JP3866982B2 - Rotation angle detection device, detection method, and noise removal device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a rotation angle detection device, a detection method, and a noise removal device, and in particular, detects a signal from a sensor connected to a rotation body that is a detection object, and calculates and outputs the rotation angle of the rotation body based on the signal. The present invention relates to a rotation angle detection device, a detection method, and a noise removal device.
[0002]
[Prior art]
When detecting the rotation angle of a rotating body, generally, a sensor such as a resolver is connected to the rotating body, and an output signal of this sensor is detected and processed by a controller to obtain the rotation angle of the rotating body.
[0003]
However, the rotating body, which is the detection target, and the controller are often arranged at a distance from each other, and the signal line connecting the rotating body and the controller is relatively long. If the signal line is long, noise is likely to be mixed into the output signal of the sensor, and the controller malfunctions as a result of the noise being mixed. When the rotating body is controlled by the controller, the rotating operation of the rotating body is also adversely affected.
[0004]
In order to reduce the influence on noise, that is, to improve immunity performance, there are a method of shielding a signal line and a method of improving the noise removal performance of the controller itself.
[0005]
A method of shielding a signal line is widely used, but there are many cases where noise cannot be completely removed only by the shielding process, and efforts are being made to improve the noise removal performance of the controller itself.
[0006]
[Problems to be solved by the invention]
In order to improve the noise removal performance of the controller itself, for example, there is a method of inserting a low-pass filter for the output signal of the sensor. However, when using a low-pass filter, it is necessary to set the cutoff frequency low in order to sufficiently remove noise. As the cut-off frequency is lowered, there arises a disadvantage that the response of the signal to be detected is lowered. Therefore, the low-pass filter is not suitable for an application for accurately detecting the rotation angle of a rotating body whose behavior changes drastically in real time.
[0007]
Further, with a low-pass filter, when the noise level is extremely high, it is difficult to remove the noise.
[0008]
As a technique for improving the noise removal performance, the presence / absence of noise is judged from the angular velocity (hereinafter also simply referred to as speed) of the rotating body, and when noise is mixed, the sensor output signal at that time is ignored. A technique has been proposed (see, for example, JP-A-5-10781). According to this technique, when an abnormal speed exceeding a certain threshold is detected, it can be determined that noise is mixed in the output signal of the sensor at that time. Therefore, the influence of noise is reduced by ignoring the output signal and performing appropriate interpolation. This technique is effective when the speed of the rotating body is substantially constant.
[0009]
However, for example, when detecting the axle rotation angle or the motor rotation angle of an electric vehicle, it is difficult to set an abnormal speed as a threshold because the speed range of the rotating body that is the detection target is very wide. It is. That is, it is inappropriate to adopt a certain threshold value from when the electric vehicle is stopped until the maximum speed, and noise cannot be sufficiently removed.
[0010]
The present invention has been made in consideration of such a problem. When detecting the rotation angle of the rotating body, the response delay with respect to the rotation angle signal of the sensor is eliminated, and noise is mixed in the rotation angle signal. It is an object of the present invention to provide a rotation angle detecting device and a detecting method that can suppress the influence of the noise regardless of the noise level.
[0011]
Further, the present invention eliminates a response delay with respect to a signal when detecting a continuous arbitrary signal, and suppresses the influence of the noise regardless of the noise level when noise is mixed in the signal. An object of the present invention is to provide a noise removing device that makes it possible.
[0012]
[Means for Solving the Problems]
An apparatus for detecting a rotation angle according to the present invention receives a rotation angle signal of a rotating body, obtains an acceleration signal by second-order differentiation of the rotation angle signal, and an acceleration limiting unit that limits the acceleration signal with a threshold value And a rotation angle calculation unit that integrates the limited acceleration signal to obtain the correction speed and integrates the correction speed to obtain the correction angle.
[0013]
By doing so, it is possible to eliminate a delay in response to the rotation angle signal of the sensor and to suppress the influence of the noise regardless of the noise level when noise is mixed in the rotation angle signal. Further, the present invention can be applied to a rotating body whose rotation speed changes.
[0014]
In this case, when the acceleration signal exceeds the threshold value, a threshold value correction unit that increases or decreases the threshold value after limiting the acceleration signal with the threshold value may be provided.
[0015]
The threshold correction unit may reset the threshold to an initial value when the acceleration signal is within the threshold.
[0016]
The rotation angle detection method according to the present invention includes a step of inputting a rotation angle signal of a rotating body, second-order differentiation of the rotation angle signal to obtain an acceleration signal, and a step of limiting the acceleration signal by a threshold value. And integrating the acceleration signal to obtain a correction speed and integrating the correction speed to obtain a correction angle.
[0017]
The noise removing apparatus according to the present invention includes a first differentiation function unit for inputting a signal and differentiating the signal to obtain a first-order derivative value, and a second derivative value for obtaining a second-order derivative value by differentiating the first-order derivative value. A differential function unit; a threshold limiting unit that limits the second-order differential value with a threshold; a first integration function unit that integrates the limited second-order differential value to obtain an intermediate value; and a correction signal that integrates the intermediate value And a second integration function unit for obtaining
[0018]
DETAILED DESCRIPTION OF THE INVENTION
First, an embodiment of a rotation angle detection device and a detection method according to the present invention will be described with reference to FIGS.
[0019]
The rotation angle detection device and detection method according to the present embodiment basically detects angular acceleration (hereinafter also simply referred to as acceleration) of a rotating body, and when the acceleration exceeds a certain threshold value, The output signal of the sensor at the time is ignored, and the acceleration is further limited by a threshold value.
[0020]
As shown in FIG. 1, the rotation angle detection device 10 according to the present embodiment is applied to an electric vehicle 100.
[0021]
The electric vehicle 100 travels using a motor 102 as a drive source, and one end of a rotating shaft 102 a of the motor 102 is connected to a differential gear 104. The rotation of the rotating shaft 102 a is transmitted to the front wheel 108 via the differential gear 104 and the axle 106. The differential gear 104 may be provided with a speed change function.
[0022]
A resolver 110 that detects the actual rotation angle θ of the rotating shaft 102a is connected to the other end of the rotating shaft 102a. The output signal θ 0 of the resolver 110 is supplied to the rotation angle detection device 10.
[0023]
The rotation angle detection device 10 has a function of obtaining the rotation angle of the rotation shaft 102 a based on the supplied output signal θ 0 and controlling the rotation of the motor 102 via the inverter 112. The inverter 112 is supplied with power from a battery 114 that can be charged from external power (for example, commercial AC 100V or 200V power supply). When the electric vehicle 100 is braked, the battery 114 can be charged by a so-called regenerative braking function.
[0024]
The resolver 110 detects a change in the rotation angle of the input shaft by converting it into a change in magnetic resistance, and outputs the rotation angle of the input shaft as a digital value.
[0025]
Specifically, as shown in FIG. 2, sin (t is applied to the pair of primary windings 110b and 110c facing each other so that the poles opposed to the three poles of the stator 110a are directly connected in reverse phase. ) And cos (t) are applied. Here, t is a time variable. Thereby, a magnetic path is formed between the pole of the stator 110a and the eccentric rotor 110d.
[0026]
Another secondary coil is wound around each of the three poles of the stator 110a, and the secondary coils are combined to form a secondary circuit 110e.
[0027]
When the rotor to be detected rotates the eccentric rotor 110d, the gap between the eccentric rotor 110d and the stator 110a changes, so that the actual rotation angle θ is changed to the secondary circuit 110e by the same principle as the transformer coupling. A corresponding induced voltage (K · sin (t−δ)) is generated. Here, K is a proportionality constant, and δ is a phase difference.
[0028]
The phase difference detection unit 110f compares the induced voltage Ksin (t−δ) generated in the secondary circuit 110e with the voltage applied to the primary windings 110b and 110c, and obtains the phase difference δ by digital counting. Furthermore, the phase difference detection unit 110f obtains the actual rotation angle θ of the eccentric rotor 110d from the phase difference δ and outputs it as an output signal θ0. The output signal θ0 has an output resolution of 10,000 pulses per rotation of the rotating shaft 102a.
[0029]
With such a configuration, the resolver 110 can output with high responsiveness.
[0030]
As shown in FIG. 3, the rotation angle detection device 10 detects the output signal θ0 of the resolver 110, outputs the rotation angle of the rotating shaft 102a and the like to the monitor 12 based on the output signal θ0. The motor 102 is controlled via
[0031]
The rotation angle detection device 10 includes a CPU 14 as a control unit that controls the rotation angle detection device 10 as a whole, a ROM 18 that stores a program 16 and data that defines the processing of the CPU 14, and data according to the processing of the CPU 14. It has a RAM 20 as a storage unit for reading and writing, and an IF 22 (interface) for transmitting and receiving data to and from an external device. If the rotation angle detection device 10 is configured using a so-called one-chip microcomputer, the size is preferably reduced.
[0032]
The CPU 14 detects the output signal θ0 of the resolver 110 via the IF 22 and supplies a signal to the monitor 12 and the inverter 112 via the IF 22.
[0033]
As an alternative to the resolver 110, for example, an analog output potentiometer may be employed. In this case, an AD (Analog-Digital) converter may be inserted between the potentiometer and the IF 22.
[0034]
As shown in FIG. 4, the program 16 includes a signal processing unit 26 that obtains the output signal θ0 of the rotating shaft 102a based on the output signal θ0, a motor control unit 28 that controls the motor 102 based on the output signal θ0, A monitor output unit 30 that displays the correction angle θ1 on the monitor 12 is provided.
[0035]
The signal processing unit 26 performs second-order differentiation of the output signal θ0 to calculate the estimated acceleration γ of the rotating shaft 102a, and sets the estimated acceleration γ to the maximum threshold γ. _MAX And the minimum threshold γ _MIN And an acceleration limiting unit 34 that limits the number of In addition, the signal processing unit 26 calculates the estimated acceleration γ and the maximum threshold γ _MAX And the minimum threshold γ _MIN The maximum threshold value γ _MAX And the minimum threshold γ _MIN And a rotation angle calculation unit 38 that calculates the correction angle θ1 of the rotation shaft 102a based on the processing result of the acceleration limiting unit 34 limiting the estimated acceleration γ.
[0036]
Next, a rotation angle detection method using the rotation angle detection device 10 according to the present embodiment will be described with reference to FIGS.
[0037]
The execution subject in the flowchart of FIG. 5 is the CPU 14 that operates according to the program 16. While the ignition key (not shown) of the electric vehicle 100 is turned on, the CPU 14 continuously executes the process shown in FIG. 5 every 100 [μsec]. So-called real-time processing is performed.
[0038]
6 and 7 show a data list 200 when the electric vehicle 100 travels. 6 and 7 are examples when the rotating shaft 102a travels at a constant acceleration of 50000 [rpm / sec], and the CPU 14 processes the actual rotation angle θ of the rotating shaft 102a (see the “θ” column 200a). Each parameter (output signal θ 0, estimated speed ω, estimated acceleration γ, corrected acceleration γ 1, corrected speed ω 1, corrected angle θ 1) is shown for each processing time of the CPU 14. Each parameter is assumed to be “0” in the initial state. Further, in FIG. 6 and FIG. 7, in the “θ0” column 200b, the portion that is a thick line frame, that is, time 0.0021 [sec], 0.0041 [sec], 0.0061 [sec], and 0.0081 [ [sec], it is assumed that noise is mixed.
[0039]
First, in step S <b> 1 of FIG. 5, the CPU 14 acquires the output signal θ <b> 0 of the resolver 110 via the IF 22. The value of the output signal θ0 acquired by the CPU 14 is shown in the “θ0” column 200b (see FIG. 6).
[0040]
The output signal θ0 may contain noise and does not necessarily match the actual rotation angle θ of the rotating shaft 102a.
[0041]
Next, in step S2, the output speed θ0 is differentiated by the function of the acceleration calculation unit 32 to obtain the estimated speed ω of the rotating shaft 102a. The estimated speed ω is obtained by subtracting the correction angle θ1 at the previous processing from the output signal θ0 at that time. The correction angle θ1 will be described in step S10.
[0042]
The values of the estimated speed ω and the correction angle θ1 that are obtained are shown in the “ω” column 200c and the “θ1” column 200g (see FIG. 6).
[0043]
Next, in step S3, the estimated speed ω of the rotating shaft 102a is obtained by differentiating the estimated speed ω by the function of the acceleration calculating unit 32. The estimated acceleration γ is obtained by subtracting the corrected speed at the previous processing from the estimated speed ω at that time. The value of the estimated acceleration γ is shown in the “γ” column 200d (see FIG. 6). The correction speed ω1 will be described in step S9.
[0044]
Next, in step S4, the estimated acceleration γ is set to the maximum threshold γ. _MAX And the minimum threshold γ _MIN Compare with Maximum threshold γ _MAX And the minimum threshold γ _MIN Is the same absolute value in the initial state, and the maximum threshold γ _MAX Is a positive value, minimum threshold γ _MIN Is a negative value.
[0045]
Maximum threshold γ _MAX And the minimum threshold γ _MIN Is set with an appropriate margin for the maximum acceleration and the minimum acceleration that can be generated by the rotating shaft 102a. In the example shown in FIGS. 6 and 7, in the initial state, the maximum threshold γ _MAX "+3", minimum threshold γ _MIN Is "-3".
[0046]
Estimated acceleration γ is the maximum threshold γ _MAX If larger, the process proceeds to step S5. The estimated acceleration γ is the minimum threshold γ _MIN When it is smaller, the process proceeds to step S7. Estimated acceleration γ is the maximum threshold γ _MAX ~ Minimum threshold γ _MIN If it is within the range, the process proceeds to step S11.
[0047]
In step S5, the function of the acceleration limiting unit 34 sets the corrected acceleration γ1 to γ1 ← γ. _MAX And set the corrected acceleration γ1 to the maximum threshold γ with respect to the estimated acceleration γ. _MAX The value restricted by.
[0048]
Next, in step S6, the threshold value correcting unit 36 functions to increase the maximum threshold value γ. _MAX Increase. That is, the estimated acceleration γ is the maximum threshold γ _MAX The larger value is considered that the possibility that noise is mixed in the output signal θ0 is very high, but it may be a normal signal. Therefore, the estimated acceleration γ continues to be the maximum threshold γ during the next processing. _MAX If it is larger, it is likely that the signal is normal without noise, so the maximum threshold γ _MAX And increase the limit range of the estimated acceleration γ. For example, the maximum threshold γ _MAX Γ _MAX ← γ _MAX Set to +1. Thereafter, the process proceeds to step S9.
[0049]
Step S7 (the estimated acceleration γ is the minimum threshold γ _MIN Is smaller), the function of the acceleration limiting unit 34 sets the corrected acceleration γ1 to γ1 ← γ. _MIN The corrected acceleration γ1 is set to the minimum threshold γ with respect to the estimated acceleration γ. _MIN The value restricted by.
[0050]
Next, in step S8, the minimum threshold γ is obtained by the function of the threshold correction unit 36. _MIN Decrease (increase absolute value). That is, for the same reason as the process of step S6, the minimum threshold γ is prepared for the next process. _MIN To increase the limit range. For example, the minimum threshold γ _MIN Γ _MIN ← γ _MIN Set to -1.
[0051]
After step S6 and step S8, in step S9, the function of the rotation angle calculation unit 38 integrates the corrected acceleration γ1 to obtain the corrected speed ω1. The correction speed ω1 is obtained by adding the correction acceleration γ1 at that time and the correction speed ω1 at the time of the previous process.
[0052]
Next, in step S10, the correction angle θ1 is obtained by integrating the correction speed ω1 by the function of the rotation angle calculation unit 38. The correction angle θ1 is obtained by adding the correction speed ω1 at that time and the correction angle θ1 at the previous process.
[0053]
Step S11 (the estimated acceleration γ is the maximum threshold γ _MAX ~ Minimum threshold γ _MIN ), The corrected acceleration γ1 is set as γ1 ← γ. That is, the estimated acceleration γ is the maximum threshold γ _MAX ~ Minimum threshold γ _MIN Therefore, it is considered that the estimated acceleration γ is a value that does not include noise. Therefore, the estimated acceleration γ is directly used as the corrected acceleration γ1. The value of the corrected acceleration γ1 is shown in the “γ1” column 200e (see FIGS. 6 and 7).
[0054]
Next, in step S12, the maximum threshold γ _MAX And the minimum threshold γ _MIN To the initial value. This is because the estimated acceleration γ is a value that does not include noise, and the acceleration is within an assumed range. _MAX And the minimum threshold γ _MIN It is appropriate to return to the initial initial value.
[0055]
In the example shown in FIGS. 6 and 7, the maximum threshold γ _MAX Γ _MAX ← “3”, minimum threshold γ _MIN Γ _MIN ← Set “−3” again.
[0056]
Next, in step S13, as in step S9, the corrected acceleration γ1 is integrated to obtain the corrected speed ω1.
[0057]
Alternatively, the estimated speed ω at that time may be set as the corrected speed ω1 as it is.
[0058]
Next, in step S14, similarly to step S10, the correction speed ω1 is integrated to determine the correction angle θ1.
[0059]
Further, the output signal θ0 at that time may be set as the correction angle θ1 as it is.
[0060]
When the steps S10 and S14 are finished, the current processing of the CPU 14 is finished. The obtained correction angle θ1 is supplied to the motor control unit 28 and the monitor output unit 30 for use.
[0061]
6 and 7, as understood from the “deviation” column 200h indicating the deviation between the actual rotation angle θ and the correction angle θ1, the actual rotation angle θ and the correction angle θ1 are the time immediately before noise is mixed. The value is the same up to 0.0020 [sec], and there is no error or response delay.
[0062]
Since there is no response delay, the rotation angle detection device 10 can effectively use the resolver 110 with high responsiveness and can be applied to a high-speed rotation target.
[0063]
At the time 0.0021 [sec] when noise is mixed, the estimated acceleration γ is “51”, but the maximum threshold γ _MAX Is limited to “3”. As a result, the deviation between the actual rotation angle θ and the correction angle θ1 becomes “−2” and is suppressed to a small value. Further, the deviation at the next time 0.0022 [sec] is also “−2”, and can be suppressed to a small value. At time 0.0023 [sec], which is the second processing from time 0.0021 [sec] when noise is mixed, the deviation is 0, and the time for the deviation to converge to 0 is short. At time 0.0023 [sec], the estimated speed ω and the corrected speed ω1 are both “17”, so the deviation regarding the speed is also zero. Therefore, the convergence is fast in terms of speed.
[0064]
In the rotation angle detection device 10, the magnitude of the noise level does not affect the result, and the result does not change even if the noise is larger. Specifically, in the example of FIGS. 6 and 7, the output signal θ0 at time 0.0021 [sec] is “419” which is larger than “369” of the actual rotation angle θ, but the output signal θ0 is extremely large. Even if the values are large, the values of the correction speed ω1 and the correction angle θ1 are not affected and become “21” and “371”, respectively. This can also be confirmed from the fact that the deviation is suppressed to a small value of “−3” even when the output signal θ 0 at time 0.0041 [sec] becomes an extremely large value “5755”. .
[0065]
Further, as shown at time 0.0061 [sec], when the output signal θ0 is a value smaller than the actual rotation angle θ, the estimated acceleration γ is set to the minimum threshold γ. _MIN (Γ _MIN By limiting by = -3), the deviation can be suppressed to a small value “3”, and the deviation can be converged to “0” in a short time.
[0066]
Furthermore, as can be understood from FIGS. 6 and 7, the estimated speed ω takes a wide range of values from “16” to “25” when no noise is mixed. Therefore, it is inappropriate to set a threshold value for the estimated speed ω in order to determine the presence of noise. This is because the difference between the value “28” of the estimated speed ω at the time 0.0081 [sec] at which noise is mixed and the value “25” of the estimated speed ω at the normal time 0.0100 [sec] is small. If the acceleration traveling is further continued, the value of the estimated speed ω can be “28”. Accordingly, it is impossible to set an appropriate threshold value, and it is impossible to determine the presence or absence of noise.
[0067]
Further, the value of the estimated speed ω at time 0.0081 [sec] has a small difference from the values “23” and “24” of the estimated speed ω at the previous times 0.0080 [sec] and 0.0079 [sec]. Therefore, an appropriate threshold value cannot be set.
[0068]
In the rotation angle detection device 10, the maximum threshold value γ with respect to the estimated acceleration γ. _MAX And the maximum threshold γ _MIN Therefore, even at this time 0.0081 [sec], the maximum threshold γ to be compared with the estimated acceleration γ of “5” _MAX (Γ _MAX = + 3), it is possible to clearly determine the presence of noise.
[0069]
In the rotation angle detection device 10, the estimated acceleration γ is set to the maximum threshold γ. _MAX And the minimum threshold γ _MIN And the presence or absence of noise is determined. Since the estimated acceleration γ has a small fluctuation range when it is normal, the estimated acceleration γ and the maximum threshold γ _MAX And the minimum threshold γ _MIN Can be accurately determined whether or not noise is present.
[0070]
FIG. 8 is a data list 201 showing each parameter for each processing time of the CPU 14 when the rotating shaft 102a travels at a constant speed of 2000 [rpm].
[0071]
As can be understood from FIG. 8, even when the electric vehicle 100 travels at a constant speed, the influence of noise can be sufficiently reduced. For example, at time 0.0021 [sec] when noise is mixed, the deviation becomes “−2” and is suppressed to a small value. Further, at the subsequent time 0.0023 [sec], the deviation becomes “0”, and the time for the deviation to converge to “0” is short.
[0072]
In addition to the electric vehicle 100, the rotation angle detection device 10 described above may be applied to, for example, a hybrid car (a vehicle having two types of power sources such as an electric motor and an internal combustion engine). The application target is applicable not only to the vehicle but also to an arbitrary rotating body, and is particularly effective for a rotating body whose rotational speed changes.
[0073]
In the acceleration calculation unit 32, the estimated acceleration γ is obtained by second-order differentiation of the output signal θ0, and the presence / absence of noise is determined based on the estimated acceleration γ. The acceleration change rate may be obtained, and the presence or absence of noise may be determined based on the estimated acceleration change rate.
[0074]
Further, in the rotation angle detection device 10, the output signal θ 0 obtained by detecting the actual rotation angle θ by the resolver 110 is processed. However, the rotation angle detection device 10 may be applied to the output signal of the speed sensor as an alternative to the resolver 110. In this case, the acceleration calculation unit 32 may obtain the rate of change of acceleration by second-order differentiation and compare and determine the obtained rate of change of acceleration and the maximum threshold value and the minimum threshold value. That is, in the above embodiment, the output signal θ0 may be regarded as a speed signal output from the speed sensor, and the rotation angle detection device 10 can be applied to the speed sensor without any change.
[0075]
Next, an embodiment of a noise removing apparatus according to the present invention will be described with reference to FIG.
[0076]
The noise removal apparatus 300 according to the present embodiment has basically the same configuration as that of the rotation angle detection apparatus 10 and removes noise from a continuous arbitrary signal.
[0077]
As shown in FIG. 9, the noise removing apparatus 300 receives a signal 302 obtained by sampling a continuous arbitrary physical quantity, differentiates the signal 302 to obtain a first-order differential value 304, A second differentiation function unit 310 that obtains a second order differential value 308 by differentiating the first order differential value 304, a threshold value limiting unit 312 that limits the second order differential value 308 with positive and negative threshold values, and a limited second order differential value A first integration function unit 318 that integrates 314 to obtain an intermediate value 316; and a second integration function unit 322 that integrates the intermediate value 316 to obtain a correction signal 320. The noise removing apparatus 300 is realized by a computer and a program (not shown), and includes a signal 302, a first-order differential value 304, a second-order differential value 308, a limited second-order differential value 314, an intermediate value 316, and a correction signal 320. Performs processing while reading from and writing to memory and registers.
[0078]
The signal 302 is obtained by digitally sampling an arbitrary signal that changes with time, and can be applied to a general physical quantity. For example, the present invention can be applied to a continuously changing physical quantity such as a linear displacement signal, a speed signal, an acceleration signal, and a temperature signal. Digital sampling is performed every predetermined time by an input unit (not shown), and a first signal 302 is obtained. It is supplied to the differentiation function unit 306. In addition, the signal 302 does not necessarily accurately digitally sample the physical quantity to be detected, and may contain noise. This signal 302 corresponds to the output signal θ 0 in the rotation angle detection device 10.
[0079]
The first differentiation function unit 306 differentiates the signal 302 to calculate a first-order differential value 304. The first differentiation function unit 306 corresponds to the first differentiation function of the acceleration calculation unit 32 in the rotation angle detection device 10. The first-order differential value 304 corresponds to the estimated speed ω, and represents the change speed of the physical quantity to be detected.
[0080]
The second differentiation function unit 310 further differentiates the first-order derivative value 304 to calculate a second-order derivative value 308. The second differentiation function unit 310 corresponds to the second differentiation function of the acceleration calculation unit 32 in the rotation angle detection device 10, and the second order differential value 308 corresponds to the estimated acceleration γ.
[0081]
The threshold limiting unit 312 limits the second-order differential value 308 with positive and negative thresholds. This threshold value may be set in consideration of the characteristics of the physical quantity to be detected, and is set to “± 3”, for example. Thereby, when the absolute value of the second-order differential value 308 is “3” or more, the absolute value is limited to “3” by this threshold. The threshold limiting unit 312 corresponds to the acceleration limiting unit 34 in the rotation angle detecting device 10, and the threshold is the maximum threshold γ. _MAX And the minimum threshold γ _MIN It corresponds to. The limited second-order differential value 314 corresponds to the corrected acceleration γ1.
[0082]
With the function of the threshold limiting unit 312, noise mixed in the signal 302 can be clearly distinguished from a normal value, and the noise component can be limited. Further, since the threshold is limited with respect to the second-order differential value 308, noise can be removed regardless of the change rate of the physical quantity to be detected, that is, the range of values that the first-order differential value 304 can take. Even when the noise level is extremely high, all are limited to the threshold value, so that they are not affected by the noise level.
[0083]
The first integration function unit 318 integrates the limited second-order differential value 314 to calculate an intermediate value 316. The first integration function unit 318 corresponds to the first integration function of the rotation angle calculation unit 38 in the rotation angle detector 10, and the intermediate value 316 corresponds to the correction speed ω1.
[0084]
The second integration function unit 322 calculates the correction signal 320 by integrating the intermediate value 316 again. The second integration function unit 322 corresponds to the second integration function of the rotation angle calculation unit 38 in the rotation angle detection device 10, and the correction signal 320 corresponds to the correction angle θ1.
[0085]
Thus, according to the noise removal apparatus 300, the response delay is eliminated with respect to the continuous arbitrary signal 302, and when noise is mixed in the signal 302, the influence of the noise is affected regardless of the level of the noise. Can be suppressed.
[0086]
The rotation angle detection device, the detection method, and the noise removal device according to the present invention are not limited to the embodiment using a computer and software, and may be configured by an analog circuit. Specifically, a first differentiating circuit for inputting a signal to obtain a first order differential value and a second differentiating circuit for differentiating the first order differential value to obtain a second order differential value are connected in series. A first integration circuit that integrates a second-order differential value limited by a threshold limiting circuit having a function of limiting by a threshold value to obtain an intermediate value; and a second integration circuit that integrates the intermediate value to obtain a correction signal. You can do it.
[0087]
The rotation angle detection device, the detection method, and the noise removal device according to the present invention are not limited to the above-described embodiments, and various configurations and steps can be adopted without departing from the gist of the present invention. It is.
[0088]
【The invention's effect】
As described above, according to the rotation angle detection device and the detection method of the present invention, when detecting the rotation angle of the rotating body, the response delay with respect to the rotation angle signal of the sensor is eliminated, and the rotation angle signal is converted into the rotation angle signal. When noise is mixed, the effect that the influence of the noise can be suppressed regardless of the level of the noise is achieved.
[0089]
Further, according to the noise removal device of the present invention, when detecting a continuous arbitrary signal, the response delay is eliminated with respect to the signal, and when noise is mixed in the signal, the noise is not affected. Thus, the effect that the influence of the noise can be suppressed is achieved.
[0090]
Furthermore, the rotation angle detection device, the detection method, and the noise removal device according to the present invention can be applied to a signal whose differential value changes.
[Brief description of the drawings]
FIG. 1 is a schematic perspective view showing a drive system of an electric vehicle.
FIG. 2 is a schematic block diagram illustrating the principle of a resolver.
FIG. 3 is a schematic functional block diagram of a rotational position detection device according to the present embodiment.
FIG. 4 is a schematic functional block diagram of a program executed in the rotational position detection device according to the present embodiment.
FIG. 5 is a flowchart showing a procedure of a rotational position detection method according to the present embodiment.
FIG. 6 is a data list (part 1) showing changes in parameters processed by a CPU when the electric vehicle travels at a constant acceleration.
FIG. 7 is a data list (part 2) showing changes in parameters processed by the CPU when the electric vehicle travels at a constant acceleration.
FIG. 8 is a data list showing changes in parameters processed by the CPU when the electric vehicle travels at a constant speed.
FIG. 9 is a schematic block diagram of a noise removal device according to the present embodiment.
[Explanation of symbols]
10 ... Rotation angle detection device 14 ... CPU
16 ... Program 26 ... Signal processing unit
28 ... Motor control unit 30 ... Monitor output unit
32 ... Acceleration calculation unit 33 ... Acceleration limiting unit
33 ... Acceleration limiting unit 36 ... Threshold correction unit
38 ... Rotation angle calculation unit 100 ... Electric vehicle
102 ... Motor 102a ... Rotating shaft
110 ... Resolver 200, 201 ... Data list
300: Noise removing device 306: First differential function unit
310: Second differentiation function unit 312: Threshold limiting unit
318: First integration function unit 322: Second integration function unit
γ ... Estimated acceleration γ1 ... Modified acceleration
γ _MAX ... maximum threshold γ _MIN ... minimum threshold
θ: Actual rotation angle θ1: Output signal
ω ... Estimated speed ω1 ... Correction speed

Claims (5)

An acceleration calculation unit that inputs a rotation angle signal of the rotating body and obtains an acceleration signal by second-order differentiation of the rotation angle signal;
An acceleration limiting unit that limits the acceleration signal with a threshold;
A rotation angle detection device comprising: a rotation angle calculation unit that integrates the limited acceleration signal to obtain a correction speed and obtains a correction angle by integrating the correction speed. The rotation angle detection device according to claim 1,
The rotation angle detecting device further includes a threshold value correcting unit that increases or decreases the threshold value after the acceleration signal is limited by the threshold value when the acceleration signal exceeds the threshold value. The rotation angle detection device according to claim 2,
The rotation angle detection device, wherein the threshold correction unit resets the threshold to an initial value when the acceleration signal is within the threshold. Inputting a rotation angle signal of the rotating body, obtaining the acceleration signal by second-order differentiation of the rotation angle signal;
Limiting the acceleration signal with a threshold;
A rotation angle detecting method comprising: integrating a limited acceleration signal to obtain a correction speed; and integrating the correction speed to obtain a correction angle. A first differentiation function unit for inputting a signal and differentiating the signal to obtain a first-order differential value;
A second differentiation function unit for differentiating the first-order differential value to obtain a second-order differential value;
A threshold limiting unit that limits the second-order differential value with a threshold;
A first integration function unit for integrating a limited second-order differential value to obtain an intermediate value;
And a second integration function unit for integrating the intermediate value to obtain a correction signal.

Images