JP6711240B2 - Current detector - Google Patents

Description

The present disclosure relates to a technique of detecting each phase current to a motor that is driven by being supplied with a three-phase current.

Various vehicles, such as electric vehicles and hybrid vehicles, equipped with a motor driven by a three-phase current (hereinafter referred to as a three-phase motor) as a driving source for traveling have become widespread. This type of vehicle is usually equipped with an inverter that generates U, V, and W phase currents, and the three-phase motor is driven by supplying the three-phase current from the inverter to the three-phase motor.

Further, as a method of controlling each phase current from the inverter to the three-phase motor, a method of individually detecting the current of each phase and controlling each phase current based on each detected value is widely known. In this control method, a current sensor is provided for each phase current supply path, and control is performed based on each current value detected by each current sensor.

In recent years, with the increase in output of three-phase motors, the value of each phase current supplied from the inverter to the three-phase motor has tended to increase. When the current of each phase becomes large, it becomes necessary to mount a current sensor having a large rating accordingly. However, if a current sensor with a large rating is newly selected, there is a concern that the cost will increase due to the cost of selection and the cost of the current sensor itself.

On the other hand, in Patent Document 1, a bus bar through which a current to be detected flows is branched into a plurality of parts, a current sensor is provided in one of the plurality of branch bars, and the current flows to the entire bus bar from the detection result of the one current sensor. Techniques for detecting current are described.

JP, 2002-257866, A

If the method described in Patent Document 1 is used as a method of detecting each phase current to a three-phase motor, even if the phase current is large, it is not necessary to directly detect the entire phase current, and a branched part is used. It is sufficient to detect the current of. Therefore, it is not necessary to increase the rating of the current sensor according to the size of the entire phase current.

However, since the currents of all of the plurality of branch paths are not directly detected, the entire phase current may not be accurately detected. On the other hand, if the current sensors are provided on all of the plurality of branch paths and the detection values of the respective current sensors are summed up, the detection accuracy of the entire phase current can be improved.

However, if the current sensors are provided in all the branch paths, the total number of the current sensors naturally increases, and there is a concern that the failure probability of the current sensors as a whole increases accordingly. There is also a problem that when any one of a plurality of current sensors of the same phase fails, it is difficult to specify which one of the plurality of current sensors has failed.

The present disclosure, in a motor drive device configured to supply each phase current to a three-phase motor by branching into a plurality for each phase and providing a current sensor in each of the plurality of branch paths, one of the current sensors is provided. Provided is a technology that enables a failed current sensor to be detected quickly and accurately in the event of a failure.

One aspect of the present disclosure is a current detection device (13) used in a motor drive device (1). The motor drive device supplies a predetermined phase current to each of the current supply paths (21, 22, 23) for each phase from the drive circuit (12) that supplies the three-phase current to the motor (10) to the motor. A plurality of branch paths (21a, 21b, 22a, 22b, 23a, 23b) configured to branch into a plurality of branches at a branching ratio and join again are provided. Further, each phase is provided with a current sensor (UA, UB, VA, VB, WA, WB) for detecting a current flowing through the branch path of the phase.

The current detection device of the present disclosure used in the motor drive device having such a configuration has a detection current value that is a value of the current detected by each current sensor provided in each of the plurality of branch paths of each phase. The sum is detected as the current value of the phase.

Then, the current detection device of the present disclosure includes a cross phase detection unit (13, S110), a cross phase determination unit (13, S140 to S160), and a failure determination unit (13, S180, S250).

The cross-phase detector detects the magnitude relationship between the detected current value of any one of the other two phases and the magnitude relationship between the detected current values of the other two phases in the process of periodically changing the detected current value of each branch path. At least one of a plurality of cross phases that are phases is detected.

The cross phase determination unit determines, for each cross phase detected by the cross phase detection unit, whether or not the cross phase is normal based on the normal cross phase that is the cross phase when all the current sensors are normal. to decide.

When there is a cross phase which is judged to be abnormal by the cross phase judgment unit, the failure judgment unit determines which current sensor has two detected current values whose magnitude relationship is reversed at the cross phase judged not to be normal. A failure of the current sensor is determined based on whether it has been detected.

In the current detection device thus configured, when the cross phase is detected, it is determined whether or not the cross phase is normal based on the detected cross phase and the normal cross phase.
Here, when the two detected current values whose magnitude relationship is reversed are both normal values, that is, when the two current sensors that detect these detected current values are both normal, these two detected current values are detected. The value cross phase and the normal cross phase match. Note that "matching" with respect to the cross phase does not necessarily mean that they are completely matched literally, and may be regarded as matching if the difference between them is within a certain value, for example.

On the other hand, if one of the two detected current values whose magnitude relationship is reversed is an abnormal value that is different from the actual value, the cross phase of these two detected current values matches the normal cross phase. do not do.

Therefore, it is possible to determine whether or not the detected cross phase is normal based on the comparison between the detected cross phase and the normal cross phase. Then, when it is determined that it is not normal, it can be determined that one of the two current sensors that has detected the two detected current values whose magnitude relationship is reversed at the cross phase is defective. It is also possible to specify which of the two current sensors is out of order.

Therefore, according to the current detection device of the present disclosure configured as described above, when any of the current sensors fails, the failed current sensor can be detected quickly and accurately.
It should be noted that the reference numerals in parentheses described in this column and in the claims indicate the correspondence with the specific means described in the embodiments described below as one aspect, and do not indicate the technical scope of the present disclosure. It is not limited.

It is explanatory drawing which shows schematic structure of the motor drive device of embodiment. It is explanatory drawing which shows the example of a waveform of each sensor current when all the current sensors are normal in embodiment. It is explanatory drawing which shows the example of a waveform of each sensor current in case the V phase 2nd current sensor fails in embodiment. FIG. 6 is an explanatory diagram for explaining a procedure of detecting a failure of a current sensor based on a cross phase in the embodiment. It is an explanatory view for explaining various processings performed when a V phase 2nd current sensor fails in a time series in an embodiment. It is a flow chart of failure judgment processing of an embodiment. It is a flow chart of current sensor monitoring processing of an embodiment. 8 is a flowchart of the waveform information acquisition process of S320 in the current sensor monitoring process of FIG. 7. 8 is a flowchart of a restoration/discard determination process of S340 in the current sensor monitoring process of FIG. 7. It is a flow chart of phase current calculation processing of an embodiment.

Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.
[1. Embodiment]
(1-1) Configuration of Motor Drive Device The motor drive device 1 shown in FIG. 1 includes a motor 10, a battery 11, an inverter 12, and a control device 13. The motor drive device 1 of the present embodiment is mounted in, for example, a vehicle. The motor 10 is configured to rotate the drive wheels of the vehicle as a driving source for driving the vehicle.

The battery 11 is a power source for driving the motor 10. The battery 11 has a rechargeable secondary battery such as a nickel hydrogen battery or a lithium ion battery, and DC power is output from the secondary battery.

The inverter 12 converts the DC power supplied from the battery 11 into three-phase AC power of U phase, V phase, and W phase, and supplies the AC power to the motor 10. The inverter 12 supplies the motor 10 with sinusoidal three-phase alternating currents having phases different from each other by 120 degrees, that is, three-phase currents of a U-phase current Iu, a V-phase current Iv, and a W-phase current Iw. The inverter 12 of the present embodiment has, for example, an IGBT as a power conversion element.

The motor 10 is driven by the three-phase current supplied from the inverter 12. The motor 10 of this embodiment is, for example, a synchronous motor having a permanent magnet as a field.
The inverter 12 and the motor 10 are connected by a U-phase bus bar 21, a V-phase bus bar 22 and a W-phase bus bar 23 for supplying a three-phase current from the inverter 12 to the motor 10.

The U-phase current Iu output from the inverter 12 is supplied to the motor 10 by the U-phase bus bar 21 and input to the U-phase current input terminal of the motor 10. The V-phase current Iv output from the inverter 12 is supplied to the motor 10 by the V-phase bus bar 22 and input to the V-phase current input terminal of the motor 10. The W-phase current Iw output from the inverter 12 is supplied to the motor 10 by the W-phase bus bar 23 and input to the W-phase current input terminal of the motor 10.

In the U-phase bus bar 21, a part of the entire path from the inverter 12 to the motor 10 is branched by two branch bars 21a and 21b. That is, the U-phase current Iu output from the inverter 12 branches and flows in two paths of the U-phase first branch bar 21 a and the U-phase second branch bar 21 b in the middle of the U-phase bus bar 21. Then, the branched currents merge again and are input to the motor 10 via the U-phase bus bar 21.

The branch ratio of the U-phase current Iu branched to each of the U-phase first branch bar 21a and the U-phase second branch bar 21b is 50%. That is, the U-phase current Iu is bisected by the two branch bars 21a and 21b at a branch ratio of 50:50.

Further, the U-phase first branch bar 21a and the U-phase second branch bar 21b are each provided with a current sensor for detecting the value of the current flowing through the branch bar. Specifically, the U-phase first branch bar 21a is provided with a U-phase first current sensor UA, and the U-phase second branch bar 21b is provided with a U-phase second current sensor UB.

The V-phase bus bar 22 and the W-phase bus bar 23 also have basically the same configuration as the U-phase bus bar 21. That is, in the V-phase bus bar 22, a part of the entire path from the inverter 12 to the motor 10 is branched by the two branch bars 22a and 22b. The V-phase current Iv output from the inverter 12 branches and flows in two paths of the V-phase first branch bar 22 a and the V-phase second branch bar 22 b in the middle of the V-phase bus bar 22. Then, the branched currents merge again and are input to the motor 10 via the V-phase bus bar 22.

In the present embodiment, the V-phase first branch bar 22a and the V-phase second branch bar 22b each have a branching ratio of the V-phase current Iv that is 60% for the V-phase first branch bar 22a and a V-phase second branch. Bar 22b is 40%. That is, the V-phase current Iv is branched into two by the two branch bars 22a and 22b at a branch ratio of 60:40.

Further, each of the V-phase first branch bar 22a and the V-phase second branch bar 22b is provided with a current sensor for detecting the value of the current flowing through the branch bar. Specifically, the V-phase first branch bar 22a is provided with a V-phase first current sensor VA, and the V-phase second branch bar 22b is provided with a V-phase second current sensor VB.

Also in the W-phase bus bar 23, a part of the entire path from the inverter 12 to the motor 10 is branched by the two branch bars 23a and 23b. The W-phase current Iw output from the inverter 12 branches into two paths of the W-phase first branch bar 23 a and the W-phase second branch bar 23 b in the middle of the W-phase bus bar 23. Then, the branched currents merge again and are input to the motor 10 via the W-phase bus bar 23.

In this embodiment, the W-phase first branch bar 23a and the W-phase second branch bar 23b each have a W-phase current Iw branching ratio of 70% and a W-phase second branch. Bar 23b is 30%. That is, the W-phase current Iw is branched into two by the two branch bars 23a and 23b at a branch ratio of 70:30.

Further, each of the W-phase first branch bar 23a and the W-phase second branch bar 23b is provided with a current sensor for detecting the value of the current flowing through the branch bar. Specifically, the W-phase first branch bar 23a is provided with a W-phase first current sensor WA, and the W-phase second branch bar 23b is provided with a W-phase second current sensor WB.

In the present embodiment, each of the current sensors UA, UB, VA, VB, WA, and WB is a so-called through-type current sensor, has a ring-shaped detection core, and branches the branch bar to be detected with respect to the branch bar. The bar is arranged so as to penetrate the detection core. Then, the detection signal corresponding to the value of the current flowing through the branch bar to be detected is output. Various types of current sensors on the penetrating side are well known, including those having a Hall element as described in Patent Document 1, for example. The description is omitted. The detection signals output from the current sensors UA, UB, VA, VB, WA, WB are input to the control device 13.

The control device 13 is mainly composed of a well-known microcomputer having a CPU 13a and a memory 13b. The memory 13b has a semiconductor memory such as a RAM, a ROM, or a flash memory. Various functions of the control device 13 are realized by the CPU 13a executing a program stored in the non-transitional physical recording medium. In this example, the memory 13b corresponds to a non-transitional substantive recording medium storing a program. By executing this program, the method corresponding to the program is executed. It should be noted that the number of microcomputers forming the control device 13 may be one or plural.

The programs stored in the memory 13b include respective programs of a failure determination process of FIG. 6, which will be described later, a current sensor monitoring process of FIGS. 7 to 9, and a phase current calculation process of FIG. Some or all of the various functions realized by the execution of the program in the control device 13 may be realized by using one or a plurality of hardware.

The control device 13 detects the value of the current flowing through the U-phase first branch bar 21a based on the detection signal from the U-phase first current sensor UA, and U based on the detection signal from the U-phase second current sensor UB. The value of the current flowing through the phase second branch bar 21b is detected. The current value detected based on the detection signal from the U-phase first current sensor UA is referred to as the U-phase first sensor value IuA, and the current value detected based on the detection signal from the U-phase second current sensor UB is It is referred to as the U-phase second sensor value IuB. The control device 13 detects the total sum of the detected U-phase first sensor value IuA and the detected U-phase second sensor value IuB as the U-phase current value Iu.

Further, the control device 13 detects the value of the current flowing through the V-phase first branch bar 22a based on the detection signal from the V-phase first current sensor VA, and based on the detection signal from the V-phase second current sensor VB. Then, the value of the current flowing through the V-phase second branch bar 22b is detected. The current value detected based on the detection signal from the V-phase first current sensor VA is referred to as the V-phase first sensor value IvA, and the current value detected based on the detection signal from the V-phase second current sensor VB is It is referred to as a V-phase second sensor value IvB. The control device 13 detects the sum of the detected V-phase first sensor value IvA and the detected V-phase second sensor value IvB as the V-phase current value Iv.

Further, the control device 13 detects the value of the current flowing through the W-phase first branch bar 23a based on the detection signal from the W-phase first current sensor WA, and based on the detection signal from the W-phase second current sensor WB. The value of the current flowing through the W-phase second branch bar 23b is detected. The current value detected based on the detection signal from the W-phase first current sensor WA is referred to as the W-phase first sensor value IwA, and the current value detected based on the detection signal from the W-phase second current sensor WB is It is referred to as the W-phase second sensor value IwB. The controller 13 detects the sum of the detected W-phase first sensor value IwA and the detected W-phase second sensor value IwB as the W-phase current value Iw.

The control device 13 controls the drive of the motor 10 by controlling the output of the three-phase current by the inverter 12 based on various information including the detected phase current values Iu, Iv, and Iw.

Further, the control device 13 monitors whether or not each current sensor UA, UB, VA, VB, WA, WB is normal based on each sensor value IuA, IuB, IvA, IvB, IwA, IwB.

(1-2) Operation of Control Device when Current Sensor Fails Specific contents of monitoring of each current sensor by the control device 13 and an outline of operation when a failure of any current sensor is detected will be described with reference to FIGS. This will be described with reference to FIG.

The control device 13 of the present embodiment fails when any one of the six current sensors UA, UB, VA, VB, WA, WB fails and a normal detection signal is not output. One current sensor can be specified.

As shown in FIG. 2, the U, V, and W three-phase currents Iu, Iv, and Iw supplied to the motor 10 from the inverter 12 all change in a sinusoidal cycle and their phases are mutually different. 120 degrees different. Therefore, the U-phase sensor values IuA and IuB, the V-phase sensor values IvA and IvB, and the W-phase sensor values IwA and IwB all change in a sinusoidal cycle, and each phase changes. The phases are 120 degrees different.

Therefore, in the process of changing the current of each phase, a timing at which the magnitude relationship between the sensor values of any two phases is reversed periodically occurs. It should be noted that the inversion of the magnitude relation between the two sensor values, that is, the intersection of the waveforms of the two sensor values in the waveform diagram is also referred to as a cross.

As shown in FIG. 2, there are six electrical angle ranges in which two sensor value crosses occur during one electrical angle cycle. Specifically, the first cross range α1, the second cross range α2, the third cross range α3, the fourth cross range α4, the fifth cross range α5, and the sixth cross range α6 occur during one electrical angle cycle. ..

The first cross range α1 and the fourth cross range α4 are ranges in which sensor values IuA, IuB of the U phase and sensor values IwA, IwB of the W phase occur. The second cross range α2 and the fifth cross range α5 are ranges in which the V-phase sensor values IvA and IvB and the W-phase sensor values IwA and IwB cross. The third cross range α3 and the sixth cross range α6 are ranges in which the U-phase sensor values IuA, IuB and the V-phase sensor values IvA, IvB cross.

In addition, in the present embodiment, since each phase current is branched into two by the two branch bars, a total of four crosses are generated in each of the above ranges in which crosses are generated. For example, in the first cross range α1, the W-phase second sensor value IwB and the U-phase first sensor value IuA cross, the W-phase second sensor value IwB and the U-phase second sensor value IuB, the W-phase first A cross between the 1-sensor value IwA and the U-phase first sensor value IuA and a cross between the W-phase first sensor value IwA and the U-phase second sensor value IuB occur.

Further, for example, in the fifth cross range α5, as shown in an enlarged view in FIG. 2, the cross between the V-phase second sensor value IvB and the W-phase first sensor value IwA, and the V-phase first sensor value IvA and W A cross with the phase first sensor value IwA, a cross with the V phase second sensor value IvB and the W phase second sensor value IwB, and a cross with the V phase first sensor value IvA and the W phase second sensor value IwB occur. To do.

Then, each cross phase, which is each electric angle at which a cross occurs in one cycle of the electric angle, always has the same value if all the current sensors UA, UB, VA, VB, WA, and WB are normal. That is, the value of the cross phase (hereinafter, also referred to as a normal value) in one cycle of the electrical angle when the current sensors UA, UB, VA, VB, WA, and WB are all normal is uniquely determined.

For example, in the fifth cross range α5, the normal value a of the cross phase X1 at which the V-phase second sensor value IvB and the W-phase first sensor value IwA cross, the V-phase first sensor value IvA, and the W-phase first sensor value The normal value b of the cross phase X2 that crosses IwA, the normal value c of the cross phase X3 that crosses the V-phase second sensor value IvB and the W-phase second sensor value IwB, and the V-phase first sensor value IvA and W The normal value d of the cross phase X4 where the phase second sensor value IwB crosses is uniquely determined.

On the other hand, if any of the current sensors UA, UB, VA, VB, WA, and WB fails, and the sensor value from the failed current sensor shows a value different from the actual current value, the sensor value And the cross phase at which the other sensor value crosses becomes a value different from the normal value.

For example, as illustrated in FIG. 3, the V-phase second current sensor VB has failed and the V-phase second sensor value IvB is lower than the actual value, that is, the absolute value of both the positive and negative values is the absolute value of the actual current value. It is assumed that the value is smaller than the above and is output. As is clear from comparison with FIG. 2, FIG. 3 shows a state in which the amplitude of the V-phase second sensor value IvB is smaller than that in FIG.

As illustrated in FIG. 3, if the V-phase second current sensor VB fails, the cross phase at which the V-phase second sensor value IvB crosses other sensor values deviates from the normal value. For example, in the fifth cross range α5, as shown in an enlarged view in FIG. 3, the cross phase X1 at which the V-phase second sensor value IvB and the W-phase first sensor value IwA cross is not the normal value a. It becomes as. Further, the cross phase X3 at which the V-phase second sensor value IvB and the W-phase second sensor value IwB cross does not become the normal value c but becomes cs.

Therefore, in the present embodiment, the control device 13 detects a cross phase when two different sensor values cross each other, and compares the detected cross phase with a normal value to detect a failure of the current sensor. ..

When any one current sensor fails and a normal sensor value is no longer output from the current sensor, the control device 13 indicates that the current sensor has failed in the six cross ranges α1 in one electrical angle cycle. It is possible to detect in four ranges of α6. For example, when the V-phase second current sensor VB fails as illustrated in FIG. 3, the control device 13 controls the second cross range α2, the third cross range α3, the fifth cross range α5, and the sixth cross range α6. The failure of the V-phase second current sensor VB can be detected in each of the four areas. On the contrary, in the first cross range α1 and the fourth cross range α4, it can be detected that all the four current sensors UA, UB, WA, WB of U phase and W phase are normal.

The normal value of each cross phase in one electrical angle cycle is stored in the memory 13b. Various methods can be adopted as to how to obtain each normal value and store it in the memory 13b. For example, when the motor drive device 1 is shipped or when a vehicle equipped with the motor drive device 1 is shipped, an inspection including energization from the inverter 12 to the motor 10 is performed, and each cross phase is measured at that time. Each measured value may be stored in the memory 13b as a normal value. Further, for example, theoretical calculation of each cross phase may be performed at the time of designing, and the calculation result may be stored in the memory 13b as a normal value.

Further, as each cross phase, an absolute value in one electrical angle cycle may be used, or a phase when the current value becomes a specific state is set as a reference phase and a relative phase with respect to the reference phase is used. May be.

When the absolute value in one cycle of the electrical angle is used as each cross phase, for example, the zero cross timing of the U-phase current shown at timing t1 in FIGS. You may make it represent the value of a cross phase. When each cross phase is represented by this method, for example, the first normal value X1 in FIG. 2 is about 260 deg.

Also, for example, each time at least one of the phase currents becomes a specific state, such as the zero-cross timing of each phase current or the timing when each phase current reaches a peak value, the timing is taken as the reference phase, and the reference phase is taken as the reference phase. The cross phase may be represented by the relative value.

Specifically, for example, a plurality of zero-cross timings t1, t2, t3, t4, t5, t6 at which zero-cross occurs during one electrical angle cycle are used as reference phases, and a cross between a certain reference phase φ1 and the next reference phase φ2 is performed. The phase may be represented by a relative value based on the reference phase φ1. In this case, the four normal values a, b, c, d within the fifth cross range α5 are represented by relative values with the cross timing t5 as the reference phase.

Note that it is not essential to set all zero-cross timings as reference phases. For example, which zero-cross timing among the zero-cross timings in one electrical angle cycle is used as the reference phase, for example, the zero-cross timing when one of the phase currents changes from negative to positive is used as the reference phase. May be appropriately determined. Further, the adoption of the zero-cross timing as the reference phase is merely an example, and another specific timing may be set as the reference phase. For example, all or part of the timing at which each phase current reaches the peak value may be set to the reference phase.

In the motor drive device 1 of the present embodiment, the absolute position of the cross phase changes depending on the phase command value θ. Since the phase command value θ is a known parameter generally used in inverter control, detailed description thereof is omitted here. Therefore, in consideration of the change in the phase command value θ, the cross phase may be expressed not as an absolute position but as a relative value with a specific timing such as zero cross timing or a peak as a reference phase.

In the example of FIG. 3, a specific procedure for detecting a failure of the current sensor based on the detection result of each cross phase in the fifth cross range α5 will be described with reference to FIG.
When the control device 13 detects a total of four cross phases X1 to X4 in the fifth cross range α5, each detected cross phase is stored in the memory 13b as shown in the cross phase evaluation table in FIG. Compare with the normal value.

In this example, since the V-phase second sensor value IvB is an abnormal value, as shown in FIG. 3, it is not possible that X1 and X3 of the four cross phases X1 to X4 are normal values. It has a different value. Therefore, as the evaluation result of the cross phase in the fifth cross range α5, X1 and X3 are evaluated as “NG”, that is, abnormal, and X2 and X4 are evaluated as “OK”, that is, normal.

Then, based on the evaluation results of the four cross phases X1 to X4, which of the four sensor values is abnormal is specified based on the cross phase-sensor value correlation table shown in FIG. The cross phase-sensor value correlation table represents, for each combination of any two sensor values of the four sensor values, whether or not the cross phase of the two sensor values is normal in a matrix form.

In the cross phase-sensor value correlation table shown in FIG. 4, the V phase second sensor value IvB is determined to be “NG” for all the cross phases with the other two sensor values. As described above, when a certain sensor value has an abnormal cross phase with respect to a plurality of other sensor values, it can be determined that the sensor value is abnormal.

As a result, in the examples of FIGS. 3 and 4, as shown in the sensor value evaluation table in FIG. 4, as the evaluation result in the fifth cross range α5, V of each sensor value of V and W phases is V. It is obtained as an evaluation result that the second phase sensor value IvB is abnormal, that is, the second current sensor VB of V phase is out of order.

Note that the tables shown in FIG. 4 are shown for the purpose of explaining the evaluation method, and the control device 13 does not actually create these tables or matrix data corresponding thereto. However, the control device 13 may actually perform the evaluation by creating a table as shown in FIG. 4 during the evaluation process.

Here, regarding the operation of the control device 13 especially after the failure of one of the current sensors is detected, the case where the failure of the V-phase second current sensor VB is detected is used as an example and FIG. 5 is used. Explain. FIG. 5 shows an example of changes in the V-phase sensor values IvA and IvB.

FIG. 5 shows that the V-phase first current sensor VA is normal and the V-phase first sensor value IvA is normal, while the V-phase second current sensor VB has a failure at timing t11 and timing t1. The following shows an example in which the V-phase second sensor value IvB has a smaller amplitude than the normal value.

As illustrated in FIG. 5, since the two current sensors VA and VB are both normal until immediately before the timing t11, the sensor values IvA and IvB of both sensors are the actual V-phase current values at the branching ratio 60:40. The value is branched according to.

On the other hand, when the V-phase second current sensor VB fails at the timing t11 and the V-phase second sensor value IvB becomes a value having a smaller amplitude than the actual value, the V-phase second sensor value IvB is detected. And the cross phase between other sensor values deviates from the normal value.

The control device 13 continues the failure determination process of the current sensor based on the comparison between the cross phase and the normal value. Specifically, in this embodiment, each time the reference phase is reached, the failure determination process is performed for the cross range from that reference phase to the next reference phase.

When the V-phase second sensor value IvB becomes an abnormal value at the timing t11, an abnormality of the V-phase second current sensor VB is detected in the failure determination process for the V-phase second sensor value IvB that is first performed thereafter. To be done.

When detecting an abnormality in the V-phase second current sensor VB, the control device 13 makes a tentative tentative determination that the V-phase second current sensor VB is out of order to determine the state of the V-phase second current sensor VB. Is set to "during failure determination". Then, the failure determination process is performed each time the reference phase arrives, and the V-phase second current sensor VB is detected at the timing t12 when the abnormality is continuously detected at the failure determination threshold or more. A formal determination is made that VB is out of order, and the state of the V-phase second current sensor VB is set to "out of order".

The control device 13 periodically calculates the phase current values Iv, Iv, Iw by a phase current calculation process described later, and uses the calculated phase current values Iu, Iv, Iw in various controls. .. For example, the V-phase current value Iv is calculated by adding the V-phase first sensor value IvA and the V-phase second sensor value IvB if both the V-phase current sensors VA and VB are normal.

At timing t11, a tentative determination is made that the V-phase second current sensor VB has failed, but the V-phase current value Iv continues to be the V-phase first value until it is officially determined to be failed at timing t12. It is calculated by adding the sensor value IvA and the V-phase second sensor value IvB.

When the failure of the V-phase second current sensor VB is officially determined at the timing t12, the control device 13 calculates the V-phase current value Iv by the interpolation calculation using the V-phase first sensor value IvA. To do.

That is, the V-phase second sensor value IvB is invalidated, and the use of the V-phase second sensor value IvB for the calculation of the V-phase current value Iv is stopped. Then, the V-phase first current value IvA is calculated by interpolation calculation based on the V-phase first sensor value IvA and the branch ratio of the V-phase first branch bar 22a.

The interpolation calculation is performed as follows. That is, the V-phase first sensor value IvA is divided by the branch ratio 0.6 of the V-phase first branch bar 22a provided with the V-phase first current sensor VA. The result of this division is regarded as the V phase current value. This V-phase deemed current value is a value that should coincide with the actual value of the V-phase current Iv if the V-phase first sensor value IvA is a normal value. In this way, the calculation for calculating the deemed current value using the sensor value of the current sensor in which no abnormality is detected is the interpolation calculation. In the case of the example in FIG. 5, the control device 13 calculates the V-phase deemed current value by interpolation calculation using the V-phase first sensor value IvA after timing t12, and uses this as the V-phase current value Iv for various controls. ..

When the controller 13 determines the failure of the V-phase second current sensor VB at the timing t12, the controller 13 periodically executes the waveform information acquisition process. The waveform information acquisition process is a preliminary process for correcting the V-phase second sensor value IvB that is no longer a normal value and making it available.

Specifically, after timing t12, the control device 13 uses the V-phase first sensor value IvA detected by the V-phase first current sensor VA as the normal current sensor and the V-phase second sensor VB as the fault current sensor. Each of the detected V-phase second sensor values IvB has a maximum plus peak and a minimum minus peak in the process of periodically increasing and decreasing.

In the example of FIG. 5, after the timing t12, first, at the timing t13, the negative peak IpnA of the V-phase first sensor value IvA and the negative peak IpnB of the V-phase second sensor value IvB are detected. After that, at timing t14, the plus peak IppA of the V-phase first sensor value IvA and the plus peak IppB of the V-phase second sensor value IvB are detected.

The control device 13 calculates the offset error Doff and the gain error Gd used to correct the V-phase second sensor value IvB when the positive and negative maximum and minimum values are detected at the timing t14.

The offset error Doff is calculated as follows. That is, the amplitude center value PB0 of the V-phase second sensor value IvB is calculated based on the plus peak IppB and the minus peak IpnB of the V-phase second sensor value IvB as the failure detection current value. Then, a value obtained by inverting the sign of the amplitude center value PB0 is calculated as an offset error Doff. For example, when the amplitude center value PB0 is −20 [A], the offset error Doff is calculated as 20.

The offset error Doff may be calculated more accurately by using the plus peak IppA and minus peak IpnA of the V-phase first sensor value IvA as the normal detection current value. For example, the amplitude center value is calculated based on at least one of the plus peak IppA and the minus peak IpnA of the V-phase first sensor value IvA, and the calculated amplitude center value and the amplitude center value PB0 of the V-phase second sensor value IvB. The difference between and may be calculated as an offset error Doff.

The gain error Gd is calculated as follows. First, the amplitude of the V-phase first sensor value IvA is obtained based on at least one of the plus peak IppA and the minus peak IpnA of the V-phase first sensor value IvA as the normal detection current value. Then, the branch ratio of the V-phase second branch bar 22b provided with the V-phase second current sensor VB and the V-phase first provided with the V-phase first current sensor VA are added to the amplitude of the V-phase first sensor value IvA. The normal amplitude on the fault side is calculated by multiplying the ratio with the branch ratio of the 1-branch bar 22a.

In the present embodiment, the V-phase second branch bar 22b has a branching ratio of 40% and the V-phase first branch bar 22a has a branching ratio of 60%. Therefore, here, the amplitude of the V-phase first sensor value IvA is 40/. It will be multiplied by 60. The failure-side normal amplitude obtained by this multiplication is the amplitude of the V-phase second sensor current value that should be originally detected when the failed V-phase second current sensor VB is normal.

Further, the amplitude of the V-phase second sensor value IvB is calculated based on the plus peak IppB and the minus peak IpnB of the V-phase second sensor value IvB as the failure detection current value. Then, the ratio between the normal amplitude on the failure side and the amplitude of the V-phase second sensor value IvB calculated above is calculated as the gain error Gd. For example, when the amplitude of the calculated V-phase second sensor value IvB is K1 and the fault-side normal amplitude is K2, the gain error Gd is calculated by K2/K1.

In FIG. 5, the timing t15 is the timing when the offset error Doff and the gain error Gd are calculated and the preparation for correction is completed.
In this way, when the offset error Doff and the gain error Gd are obtained at the timing t15, the state of the V-phase second current sensor VB is set to “restoration determination in progress”. Then, thereafter, the correction calculation is performed on the detected V-phase second sensor value IvB. The correction calculation is stopped when the state of the V-phase second current sensor VB returns to "normal" or "discard".

Specifically, the correction calculation is performed by the following equation (1). In the following equation (1), IvBc represents a V-phase second failure correction value which is a failure correction value of the V-phase second sensor value IvB.
IvBc=(IvB+Doff)*Gd (1)
When the state of the V-phase second current sensor VB is set to “during recovery determination” at timing t15, the control device 13 periodically executes recovery/discard determination processing instead of the waveform information acquisition processing. The restoration/discard determination processing determines whether the V-phase second failure correction value IvBc obtained by the correction calculation is an appropriate value, in other words, whether the calculated offset error Doff and gain error Gd are appropriate values. This is a process for determining.

The restoration/discard determination process is performed based on whether or not the ratio between the V-phase first sensor value IvA and the V-phase second failure correction value IvBc matches the normal ratio. The normal ratio is the ratio of the corresponding branch ratios, and is 60:40 for the V-phase first sensor value IvA and the V-phase second failure correction value IvBc. It should be noted that the term "matching" here does not mean that it is a literal perfect match, but also includes the case where it is within a range that can be regarded as a match. For example, if the ratio is within the range of 65:35 to 75:25, it may be considered that they match.

Then, in the recovery/discard determination processing that is periodically performed, it is continuously determined that the ratio between the V-phase first sensor value IvA and the V-phase second failure correction value IvBc is 60:40, and the recovery threshold value continues. Then, it is determined that the V-phase second failure correction value IvBc obtained by the correction calculation is an appropriate value, and the state of the V-phase second current sensor VB is set to “recovery”. In the example of FIG. 5, it is set to "recovery" at the timing t16.

When "recovery" is set at the timing t16, the periodical execution of the recovery/discard determination processing is stopped. In the failure determination process executed in the “recovery” state, the V-phase second failure correction value IvBc is used as the sensor value of the V-phase second current sensor. That is, the cross phase between the V-phase second failure correction value IvBc and another sensor value is an evaluation target.

Further, when set to "recovery" at the timing t16, the control device 13 uses the V-phase second failure correction value IvBc obtained by the correction calculation as the value of the current flowing through the V-phase second branch bar 22b. That is, for the V-phase current value Iv, the control device 13 calculates the sum of the V-phase first sensor value IvA and the V-phase second failure correction value IvBc, and uses this as the V-phase current value Iv for various processes. .. In the restoration/discard determination processing, if it is not determined that the V-phase second failure correction value IvBc obtained by the correction calculation is an appropriate value, the V-phase current value Iv continues to be the V-phase first value. The V-phase deemed current value using the sensor value IvA is used.

(1-3) Failure Judgment Process Next, the failure judgment process executed by the control device 13 to realize the failure judgment based on the cross phase described with reference to FIGS. 2 to 4 will be described with reference to FIG. .. The controller 13 executes the failure determination process of FIG. 6 continuously after the start-up, or periodically for a predetermined period after the predetermined execution timing has arrived.

In the present embodiment, as an example, as described above, each time the reference phase is reached, the cross range between the reference phase and the next reference phase is targeted, and all the cross phases within the cross range are detected. Until that is done, the failure determination process is periodically executed. The failure determination process is actually executed in the control device 13 by the CPU 13a described above according to a program for the failure determination process stored in the memory 13b.

When starting the failure determination process, the control device 13 performs a cross phase detection process in S110. Specifically, it is determined whether or not any two sensor values are crossed, and when they are crossed, the cross phase is associated with the two sensor values, or two currents that have detected the two sensor values. It is temporarily stored in the memory 13b in association with the sensor.

The detection of the crossing of the two sensor values can be performed by various methods. For example, each sensor value at the present time is acquired for each process of S110 of the failure determination process that is periodically performed, and each sensor value acquired last time is compared with each sensor value acquired this time. Then, when the magnitude relationship between any two sensor values is inverted between the previous time and the current time, it can be detected that the two sensor values cross. Further, the value of the cross phase can be obtained by, for example, interpolation calculation based on the previous value and the current value of the two sensor values.

In S120, it is determined whether or not the cross phase is detected in S110. If the cross phase has not been detected, the failure determination process ends. When the cross phase is detected, the process proceeds to S130. In S130, it is determined whether or not all the cross phases in the evaluation target cross range have been detected. For example, when the cross range to be evaluated is the fifth cross range α5 illustrated in FIG. 2, the process of S130 is a process of determining whether or not all four cross phases X1 to X4 have been detected.

When the undetected cross phase remains, the failure determination processing ends. If all the cross phases in the evaluation target cross range are detected, the process proceeds to S140. In S140, the number of the next cross phase is set in the variable n. When the process of S140 is performed for the first time after the reference phase is reached and the failure determination process is started, the variable n is set to the initial value “1”. After being set to the initial value 1, the variable n is incremented by 1 each time the process returns from S160 to S140.

In S150, the cross phase Xn detected and temporarily stored in the memory 13b is evaluated. Specifically, it is determined whether or not the cross phase Xn is normal according to the procedure described using the cross phase evaluation table in FIG. In S160, it is determined whether or not the evaluation in S150 has been performed for all cross phases in the evaluation target cross range. When the unevaluated cross phase remains, the process returns to S140. If the evaluation in S150 is completed for all the cross phases, the process proceeds to S170.

In S170, the variable Y is set to the number of the next current sensor. When the process of S170 is performed for the first time after the current sensor monitoring process is started, the variable Y is set to the initial value “1”. The variable Y is incremented by 1 each time it returns from S260 to S170 after being set to the initial value 1.

In S180, among the current sensors corresponding to the cross range to be evaluated, that is, among the current sensors that have detected each sensor value at which a cross occurs in the cross range, the Y-th current sensor is set as the evaluation target and the Y-th current sensor is selected. Evaluate the current sensor. It should be noted that which of the current sensors to be evaluated may be numbered may be appropriately determined.

In the evaluation of S180, it is determined whether or not the Y-th current sensor to be evaluated is normal according to the procedure described using the cross phase-sensor value correlation table and the sensor value evaluation table in FIG. Referring to the example of FIG. 4, if the Y-th current sensor to be evaluated is the V-phase second current sensor VB, the evaluation result is abnormal, and the Y-th current sensor to be evaluated is the V-phase second current sensor VB. If any of the three other than the above, the evaluation result is normal.

In S190, it is determined whether the evaluation result of S180 is normal. If the evaluation result is normal, the process proceeds to S220. In S220, the counter value of the failure counter corresponding to the Y-th current sensor (hereinafter, the Y-th failure counter value) is cleared to 0. In S230, the state of the Y-th current sensor is set to "normal". However, if it is already set to "Recovery", the setting of "Recovery" is maintained. After the processing of S230, the process proceeds to S240.

If the evaluation result is abnormal in S190, the process proceeds to S200. In S200, the Yth failure counter value is incremented by 1 from the current value. In S210, the state of the Y-th current sensor is set to "during failure determination".

In S240, it is determined whether the Yth failure counter value is equal to or greater than the failure determination threshold value. The failure determination threshold may be appropriately determined. If the Yth failure counter value is less than the failure determination threshold value, the process proceeds to S260. If the Yth failure counter value is equal to or greater than the failure determination threshold value, the process proceeds to S250. In S250, the state of the Y-th current sensor is set to "failure".

In S260, it is determined whether or not the evaluation has been performed for all the current sensors to be evaluated, that is, whether or not the processes of S180 and the following have been performed for all the current sensors to be evaluated. If the evaluation has already been performed for all the current sensors to be evaluated, the failure determination processing ends. If there is a current sensor that has not been evaluated yet, the process returns to S170.

For example, when the fifth cross range α5 illustrated in FIG. 3 is the evaluation target and only the first current sensor of the four current sensors to be evaluated is still evaluated, the process returns to S170. In this case, in S170, the variable Y is set from the current "1" to "2" indicating the next current sensor, which allows the second current sensor to be processed in S180 below. Will be done.

(1-4) Current Sensor Monitoring Process Next, the current sensor monitoring process executed by the control device 13 to realize the operation illustrated in FIG. 5 will be described with reference to FIG. 7. After activation, the control device 13 repeatedly executes the current sensor monitoring process of FIG. 7 in a predetermined control cycle during a predetermined monitoring period in which each current sensor should be monitored. The current sensor monitoring process is actually executed in the control device 13 by the above-described CPU 13a in accordance with the current sensor monitoring process program stored in the memory 13b.

The current sensor monitoring process is individually executed for each phase of U, V, and W. That is, the current sensor monitoring process executed with each of the U-phase current sensors UA and UB as a monitoring target, the current sensor monitoring process executed with each of the V-phase current sensors VA and VB as a monitoring target, and each of the W phase The current sensor monitoring process executed by using the current sensors WA and WB as monitoring targets is executed in parallel as, for example, a multitask.

When the control device 13 starts the current sensor monitoring process in FIG. 7, in S310, one state of the first current sensor and the second current sensor in the phase to be processed is “failure” and the other state. Is determined to be "normal" or "recovery". If an affirmative decision is made in S310, the operation proceeds to S320.

In S320, waveform information acquisition processing is executed. Details of the waveform information acquisition processing are as shown in FIG. When proceeding to the waveform information acquisition process, the control device 13 performs a peak acquisition process in S410.

This peak acquisition process is a process of acquiring the above-mentioned plus peak and minus peak for each sensor value of the phase to be processed. The control device 13 acquires each sensor value of the relevant phase at the present time, and determines whether each is the maximum value or the minimum value. If there is a current sensor whose status is "recovery" out of the two current sensors, the current sensor is either the maximum value or the minimum value for the failure correction value, not the sensor value itself. Judge whether or not

Then, if there is a sensor value that is the maximum value, that sensor value is acquired as a plus peak of the current sensor, and if there is a sensor value that is the minimum value, that sensor value is obtained from the current sensor. Get as a negative peak. The acquired plus peak or minus peak is temporarily stored in the memory 13b.

In S420, it is determined whether or not both the positive peak and the negative peak have been acquired for both the first sensor value of the first current sensor and the second sensor value of the second current sensor, that is, they are stored in the memory 13b. Determine if both are in a saved state. If there is a peak that has not been acquired yet, the current sensor monitoring process ends. If all the plus peaks and minus peaks to be acquired are in the acquired state, the process proceeds to S430.

In S430, the offset error is calculated by the same method as the method of calculating the offset error Doff described above with reference to FIG. That is, the offset error is calculated by the above-described method from the plus and minus peaks of the sensor value of the current sensor whose state is set to "failure" as the fault detection current value.

In S440, the gain error is calculated by the same method as the method of calculating the gain error Gd described above with reference to FIG. Specifically, for the sensor value of the current sensor whose status is set to "failure" as the failure detection current value, calculate the failure-side normal amplitude and calculate the ratio with the amplitude of the sensor value actually detected. It is calculated as an error Gd.

In S450, the correction process is started for the failure detection current value which is the sensor value of the current sensor in the "failure" state. That is, with respect to the sensor value of the current sensor, the failure detection current value actually detected is used until the current sensor enters the “normal” state or the “discard” state. The failure correction value is calculated according to 1). In S460, the state of the current sensor in the "failure" state is set to "restoring determination".

Returning to FIG. 7, the description will be continued. When a negative determination is made in S310, the process proceeds to S330. In S330, it is determined whether one state of the first current sensor and the second current sensor in the phase to be processed is "restoring determination" and the other state is "normal" or "restoring". To do. When a negative determination is made in S330, the current sensor monitoring process ends. If an affirmative decision is made in S330, the operation proceeds to S340.

In S340, restoration/discard determination processing is executed. Details of the restoration/discard determination processing are as shown in FIG. When the control device 13 proceeds to the restoration/discard determination processing, in S510, it is determined whether or not the ratio between the currently calculated failure correction value and the sensor value of the normal current sensor matches the normal ratio. For example, in the case of the operation example of FIG. 5, it is determined whether or not the ratio of the V-phase first sensor value IvA and the V-phase second failure correction value IvBc matches the regular ratio of 60:40.

When the ratio between the failure correction value and the sensor value of the normal current sensor matches the normal ratio, in S520, the counter value of the recovery counter (hereinafter, recovery counter value) is incremented by 1 from the current value. After the processing of S520, the process proceeds to S550.

When the ratio between the failure correction value and the sensor value of the normal current sensor does not match the normal ratio in S510, the counter value of the discard counter (hereinafter, the discard counter value) is incremented by 1 in S530. .. In S540, the recovery counter value is cleared to 0. The initial values of the recovery counter value and the discard counter value are both 0.

The process of clearing the recovery counter value in S540 may be omitted. Conversely, the discarding counter value may be cleared to 0 either before or after the recovery counter value is incremented by 1 in S520.

In S550, it is determined whether the discard counter value is greater than or equal to the discard threshold. If the discard counter value is equal to or greater than the discard threshold value, the state of the current sensor in the “recovery determination in progress” state is set to “discard” in S560.

If the discard counter value is less than the discard threshold value in S550, it is determined in S570 whether the recovery counter value is equal to or greater than the recovery threshold value. If the restoration counter value is not greater than or equal to the restoration threshold value, the current sensor monitoring process ends. If the recovery counter value is equal to or greater than the recovery threshold value, the state of the current sensor in the "recovery determination in progress" state is set to "restore" in S580.

(1-5) Phase Current Calculation Process Next, the phase current calculation process executed by the control device 13 will be described with reference to FIG. 10. After starting, the control device 13 repeatedly executes the phase current calculation process of FIG. 10 in a predetermined control cycle during a predetermined current detection period in which the current value of each phase should be detected. The phase current calculation process is actually executed in the control device 13 by the CPU 13a described above in accordance with the program of the phase current calculation process stored in the memory 13b.

The phase current calculation process is individually executed for each of the U, V, and W phases. That is, the phase current calculation process for calculating the U-phase current, the phase current calculation process for calculating the V-phase current, and the phase current calculation process for calculating the W-phase current are performed in parallel as multitasks, for example. Will be executed.

When the control device 13 starts the phase current calculation process in FIG. 10, in S610, the states of the first current sensor and the second current sensor in the phase to be processed are both “normal”, or one of them is “normal”. It is determined whether the setting is changed from "normal" to "during failure determination" in the "normal" state. The state in which the setting is changed from "normal" to "during failure determination" is currently "during failure determination", and the state before switching to "during failure determination" is "normal". It means that.

If an affirmative decision is made in S610, the operation proceeds to S620. In S620, the sum of the sensor values is calculated as the current value of the phase. That is, for example, in the case of the U phase, the sum of the U phase first sensor value IuA and the U phase second sensor value IuB is calculated as the U phase current value Iu.

When a negative determination is made in S610, the process proceeds to S630. In S630, it is determined whether or not one of the states of the first current sensor and the second current sensor is "normal" and the other is "failure", "recovery determination in progress", or "discard".

If an affirmative decision is made in S630, the operation proceeds to S640. In S640, the sensor value of the current sensor whose status is "fault", "recovery determination in progress" or "discard" is invalidated, and the phase of the relevant phase is calculated by the above-described interpolation calculation using the sensor value of the current sensor in any other status. Calculate the current value. In the operation example of FIG. 5, the process of calculating a value twice the V-phase first sensor value IvA as the V-phase current value Iv between timings t12 and t16 corresponds to the process of S640.

When a negative determination is made in S630, the sensor value of the current sensor whose state is not “normal” is replaced with the failure correction value. Then, the sum of the failure correction value and the sensor value of the current sensor in the “normal” state is calculated as the current value of the phase. In the operation example of FIG. 5, the process of calculating the sum of the V-phase first sensor value IvA and the V-phase second failure correction value IvBc as the V-phase current value Iv after timing t16 corresponds to the process of S650. ..

(1-6) Effects of the Embodiment According to the embodiment described above, the following effects (1a) to (1f) are achieved.
(1a) The control device 13 detects, for each of the cross ranges α1 to α6 generated in six periods in one electrical angle cycle, a plurality of cross phases generated in the cross range and corresponds to each of the cross phases. It is determined whether or not each cross phase is normal according to the comparison with the normal value. Then, if there is an abnormal cross phase, the defective current sensor is detected based on the abnormal cross phase in the manner described with reference to FIG.

Therefore, according to the control device 13 of the present embodiment, when any one of the current sensors fails, the failed current sensor can be detected quickly and accurately based on the cross phase.

(1b) In the present embodiment, the branch ratios of the V phase and the W phase among the three phases are set to ratios different from 50:50. That is, the values of the two branched currents are different in both the V-phase and the W-phase. As a result, the V-phase sensor value and the other-phase sensor value have different cross phases, and the W-phase sensor value and the other-phase sensor values also have different cross phases. Therefore, the failure determination of the current sensor based on the cross phase can be performed with higher accuracy.

(1c) In this embodiment, a cross phase having a relative value based on a specific reference phase such as zero cross timing of each phase current is detected, and various evaluations are performed based on the cross phase. Therefore, even if the operating condition of the motor drive device 1 changes and the phase command value θ changes, it is possible to accurately determine the failure of the current sensor.

(1d) When any one of the current sensors is officially determined to be faulty, the sensor value itself of the fault current sensor is invalidated. As the current value of the branch bar provided with the fault current sensor, a fault correction value obtained by a correction calculation using the sensor value of the fault current sensor is used instead of the actual sensor value.

Further, when performing the correction calculation, the offset error and the gain error are calculated in advance. Then, after the offset error and the gain error are obtained, the correction calculation is performed using them.

Therefore, even if one of the current sensors fails, the sensor value of the failed current sensor is corrected to the failure correction value by the correction calculation and reused, so that the redundancy can be restored and the failure correction value can be restored. The phase current can be calculated with high accuracy by using the phase current.

(1e) When any one of the current sensors is officially determined to be defective, the interpolation using the sensor value of the other normal current sensor in the same phase is performed until the offset error and the gain error for correction calculation are obtained. The current value of the phase is calculated by the calculation. Therefore, until the offset error and the gain error are obtained, it is possible to isolate the sensor value of the current sensor that has been determined to be faulty and to maintain the calculation result of the current value of the relevant phase at a level that is not practically problematic.

(1f) Even if the offset error and the gain error are obtained and the failure correction value can be calculated, the failure correction value is not immediately used for calculating the current value of the phase, and the failure correction value is a practical problem. Check if there is no precision. Specifically, it is determined whether the ratio between the failure correction value and the normal sensor value matches the normal ratio. Then, when the determination that the ratio matches the regular ratio continues for the failure determination threshold times or more continuously, it is determined that the failure correction value is a value that has no practical problem, and thereafter, the failure correction value is used to Calculate the phase current. Therefore, the reliability of the phase current calculated using the failure correction value can be improved.

(1-7) Correspondence between Claims and Claims Here, the correspondence between the wording of the embodiment and the claims will be described. The control device 13 corresponds to a current detection device. The inverter 12 corresponds to a drive circuit. The bus bars 21, 22, 23 of each phase correspond to the supply path. Each branch bar 21a, 21b, 22a, 22b, 23a, 23b corresponds to a branch path. The CPU 13a of the control device 13 includes a cross phase detection unit, a cross phase determination unit, a failure determination unit, an abnormal phase current calculation unit, a failure side peak acquisition unit, a normal side peak acquisition unit, a gain error calculation unit, an offset error calculation unit, and a failure. It corresponds to the correction value calculation unit and the correction result determination unit. The memory 13b of the control device 13 corresponds to a storage unit.

S110 in FIG. 6 corresponds to the process of the cross phase detection unit. S140 to S160 of FIG. 6 correspond to the processing of the cross phase determination unit. S180 and S250 of FIG. 6 correspond to the process of the failure determination unit. S410 of FIG. 8 corresponds to the processing of the fault side peak acquisition section and the normal side peak acquisition section. S430 of FIG. 8 corresponds to the process of the offset error calculation unit. S440 of FIG. 8 corresponds to the process of the gain error calculation unit. S450 in FIG. 8 corresponds to the process of the failure correction value calculation unit. S510 of FIG. 9 corresponds to the process of the correction result determination unit. The processing of S640 and S650 in FIG. 10 corresponds to the processing of the abnormal phase current calculation unit.

[2. Other Embodiments]
Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above-described embodiments, and various modifications can be implemented.

(2-1) In the above embodiment, the branching ratio of each phase is 50:50 for the U phase, 60:40 for the V phase, and 70:30 for the W phase, but these branching ratios are merely examples. is there. The branching ratio of each of the plurality of branching paths in each phase may be appropriately determined.

For example, all the phases may be 50:50. Further, for example, all the phases may have a branching ratio different from 50:50. In that case, each phase may have the same branching ratio, or each phase may have a different branching ratio.

(2-2) The number of branches of the bus bar for each phase, that is, the number of branch bars may be appropriately determined. That is, in the above embodiment, the two-branch configuration has two branch bars, but three or more branches may be used.

(2-3) There is no need to match the number of branches for each phase. For example, the number of branches of each phase may be appropriately determined such that the U phase has two branches and the V phase and the W phase have three branches, or the number of branches is different for each phase.

(2-4) It is not essential to detect the failure of the current sensor based on the cross phase in all of the six cross ranges α1 to α6 in one electrical angle cycle. For example, failure detection may be performed in at least one specific range out of the six cross ranges α1 to α6. For example, the failure detection may be performed in the three cross ranges of the first cross range α1 to the third cross range α3, and the failure detection may not be performed in the fourth cross range α4 to the sixth cross range α6.

Further, it is not essential to perform the failure detection based on all the generated cross phases in one cross range, and it is appropriate to determine which of the generated cross phases and how many cross phases are used to perform the failure detection. You may decide. For example, in the fifth cross range α5 of FIG. 3, the failure detection may be performed based on any one, two, or three of the four cross phases.

For example, the failure detection of the W-phase first current sensor WA is possible by performing the failure detection based only on the two cross phases X1 and X2. Further, when one cross phase shift can be detected, it is difficult to specify which current sensor has failed by only the cross phase shift, but for example, two sensors corresponding to a cross phase different from the normal value can be detected. It is possible to detect which of the two sensor values is abnormal by checking the ratio of each value to the other sensor value of the same phase.

(2-5) When the number of branches is three or more, and when one of the current sensors of each branch bar is determined to be in failure, the deemed current value is not determined to be in failure. It may be calculated by appropriately using the sensor values of two or more current sensors. For example, the deemed current value may be calculated using only one of the sensor values of the other two or more current sensors that have not been determined to be faulty, or may be calculated using any two or more. The deemed current value may be calculated.

The calculation procedure in the case of calculating the considered current value using any two or more is as follows, for example. That is, each of the two or more sensor values used to calculate the deemed current is divided by the corresponding branch ratio. Then, the sum of the division results is taken as the deemed current value.

(2-6) The motor to which the technology of the present disclosure can be applied is not limited to the synchronous motor having the permanent magnet as the field magnet, which is exemplified in the above embodiment. The present disclosure can be applied to various three-phase motors that can operate with three-phase current.

(2-7) The through-type current sensor is merely an example, and what kind of current sensor may be used may be appropriately determined.
(2-8) By implementing a plurality of functions of one component in the above embodiment by a plurality of components, or by implementing a single function of one component by a plurality of components Good. Further, a plurality of functions of a plurality of constituent elements may be realized by one constituent element, or one function realized by a plurality of constituent elements may be realized by one constituent element. Further, a part of the configuration of the above embodiment may be omitted. Further, at least a part of the configuration of the above-described embodiment may be added or replaced with respect to the configuration of the other above-described embodiment. Note that all aspects included in the technical idea specified by the wording recited in the claims are embodiments of the present disclosure.

DESCRIPTION OF SYMBOLS 1... Motor drive device, 10... Motor, 11... Battery, 12... Inverter, 13... Control device, 13a... CPU, 13b... Memory, 21... U phase bus bar, 21a... U phase 1st branch bar, 21b... U phase Second branch bar, 22... V-phase bus bar, 22a... V-phase first branch bar, 22b... V-phase second branch bar, 23... W-phase bus bar, 23a... W-phase first branch bar, 23b... W-phase second Branch bar, UA... U-phase first current sensor, UB... U-phase second current sensor, VA... V-phase first current sensor, VB... V-phase second current sensor, WA... W-phase first current sensor, WB... W-phase second current sensor.

Claims (7)

A plurality of currents of the phase are provided at a predetermined branching ratio in each current supply path (21, 22, 23) for each phase from the drive circuit (12) that supplies the three-phase current to the motor (10) to the motor. A plurality of branch paths (21a, 21b, 22a, 22b, 23a, 23b) configured to branch and join again are provided, and a current sensor (for each branch path that detects a current flowing through the branch path). UA, UB, VA, VB, WA, WB) provided in the motor drive device (1), and each phase is detected by each of the current sensors provided in the plurality of branch paths of the phase. A current detecting device (13) configured to detect a sum of detected current values, which are current values, as a current value of the phase,
It is a phase when the magnitude relationship with the detected current value of any one of the other two phases is reversed in the process of cyclically changing the detected current value of each of the branch paths in any phase. A cross phase detection unit (13, S110) for detecting at least one of the plurality of cross phases,
For each of the cross phases detected by the cross phase detection unit, a cross that determines whether the cross phase is normal based on a normal cross phase that is the cross phase when all the current sensors are normal. A phase determination unit (13, S140 to S160),
When there is the cross phase determined to be abnormal by the cross phase determination unit, which of the current sensors has the two detected current values whose magnitude relationship is inverted at the cross phase determined to be abnormal A failure determination unit (13, S180, S250) for determining a failure of the current sensor based on whether it has been detected,
A current detection device comprising. The current detection device according to claim 1, wherein
At least one phase among the three phases is a current detection device configured such that the branching ratios of the branching paths of the phase are not all equal. The current detection device according to claim 1 or 2, wherein
A storage unit (13b) in which a relative value based on a reference phase, which is a phase when at least one of the currents flowing through the respective branch paths is in a specific state, is stored in advance as the normal cross phase. Prepare,
The cross phase detection unit is configured to detect, as the cross phase, a relative value based on the reference phase,
The cross phase determination unit, for each of the cross phases detected by the cross phase detection unit, based on the normal cross phase corresponding to the cross phase stored in the storage unit, the cross phase is Configured to determine if it is normal,
Current detection device. The current detection device according to any one of claims 1 to 3,
When the failure of the current sensor is determined by the failure determination unit, the detected current value of the current sensor determined to be the failure is invalidated, and the abnormality is a phase corresponding to the current sensor determined to be the failure. The branch provided with at least one of the detected current values other than the invalid detected current value among the detected current values in the phase, and the current sensor that detects the at least one detected current value. A current detection device comprising: an abnormal phase current calculation unit (13, S640) that calculates a current value of the abnormal phase based on the branching ratio of the path. The current detection device according to any one of claims 1 to 3 ,
When a failure of the current sensor is determined by the failure determination unit, the current sensor determined to be the failure is used as a failure current sensor, and the failure detection current value is the detection current value detected by the failure current sensor. A fault-side peak acquisition unit (13, S410) that acquires a maximum value and a minimum value in the process of cyclically increasing and decreasing,
When the failure determination unit determines a failure of the current sensor, at least one other current sensor other than the failed current sensor in the abnormal phase, which is a phase corresponding to the current sensor determined to be the failure, is normally operated. As a current sensor, a normal-side peak acquisition unit (13) that acquires at least one of a maximum value and a minimum value in the process of cyclically increasing and decreasing the normal detection current value that is the detection current value detected by the normal current sensor. , S410),
The amplitude of the normal detection current value based on at least one of the maximum value and the minimum value of the normal detection current value acquired by the normal-side peak acquisition unit, in the branch path provided with the fault current sensor A value obtained by multiplying the ratio of the branch ratio and the branch ratio of the branch path provided with the normal current sensor, as a fault side normal amplitude that is the amplitude of the current that should be originally detected by the fault current sensor, Calculate a gain error that is the ratio of the fault-side normal amplitude and the amplitude of the actual fault-detection current value based on the maximum value and the minimum value of the fault-detection current value acquired by the fault-side peak acquisition unit. A gain error calculation unit (13, S440)
Offset that is the difference between the amplitude center value of the failure detection current value based on the maximum value and the minimum value of the failure detection current value acquired by the failure side peak acquisition unit and the amplitude center value of the normal detection current value An offset error calculation unit (13, S430) for calculating an error,
The fault correction is performed by adding the offset error calculated by the offset error calculation unit to the fault detection current value, and multiplying the addition result by the gain error calculated by the gain error calculation unit. A failure correction value calculation unit (13, S450) for calculating a value,
If the fault correction value is calculated by the previous SL fault correction value calculation unit, and the fault correction value, thus calculated, and the detected current value by each of the other said current sensor other than the fault current sensor in the abnormality phase An abnormal phase current calculation unit (S650) that calculates the sum of the above as an abnormal phase current value,
A current detection device comprising. The current detection device according to claim 5, wherein
When the failure correction value is calculated by the failure correction value calculation unit, the ratio between the failure correction value and the current value detected by the current sensor other than the failure current sensor in the abnormal phase corresponds to A correction result determination unit (13, S510) for determining whether or not the ratio of the branching ratio
The abnormal phase current calculation unit, when it is determined by the correction result determination unit to match, the failure correction value and the detected current value of each of the current sensors other than the failure current sensor in the abnormal phase Calculate the sum as the current value of the abnormal phase,
Current detection device. The current detection device according to any one of claims 1 to 6,
In each of the phases, there are two branch paths,
At least two of the three phases are configured such that the branching ratios of the two branching paths of the phase are different from each other,
Current detection device.

Images