JP7804484B2 - Rotor magnet temperature estimation method and rotor magnet temperature estimation device - Google Patents

Description

Article 30, Paragraph 2 of the Patent Act applies. August 18, 2021, IEEJ Industrial Applications Division Conference Proceedings https://www.gakkai-web.net/gakkai/jiasc/hp21/

The present invention relates to a rotor magnet temperature estimation method and a rotor magnet temperature estimation device.

Patent Document 1 discloses a device that estimates the temperature of a permanent magnet using a magnetized element made of a temperature-sensitive magnetic material, arranged on the end plate of a rotor, and magnetized by leakage flux from the rotor core, which is generated by a permanent magnet placed in the rotor core, and a Hall element that detects the strength of the magnetic field generated by the magnetized element itself. Specifically, the magnetized element is made of a temperature-sensitive magnetic material, and its magnetic properties change with changes in its own temperature. Therefore, the temperature of the magnetized element is estimated from the strength of the magnetic field detected by the Hall element, and the temperature of the permanent magnet is indirectly estimated from that.

Patent No. 4165229

It is known that the temperature of the permanent magnets placed on the rotor varies depending on their position in the axial direction of the rotor, with the temperature increasing the closer to the center of the rotor in the axial direction. In the device described in the above document, the temperature of the permanent magnets is estimated indirectly from the temperature of the magnetizing elements placed on the end plates of the rotor, which means that the temperature is estimated at the end of the rotor in the axial direction, making it difficult to accurately estimate the temperature of parts closer to the center.

The present invention therefore aims to estimate with greater accuracy the temperature of the permanent magnets (hereinafter also referred to as rotor magnets) arranged in the rotor, including those near the center of the rotor's axial direction.

According to one aspect of the present invention, there is provided a method for estimating a rotor magnet temperature of a permanent magnet synchronous motor in which a rotor core is divided into three or more divided cores in the axial direction, and each divided core is stacked in the axial direction with a skew angle. In this method, the magnetic flux linkage numbers of the d-axis and q-axis in the rotor synchronous coordinate system or the current value of inductance are calculated from the current and voltage applied to the motor; the number of divided cores with different skew angles is designated as n; the magnetic flux linkage number λ _{d_core-n} observed on the d-axis of each divided core is calculated from the magnetic flux linkage numbers of the d-axis and q-axis in a no-load state; and the d-axis magnetic flux linkage number λ _d-n , which is the magnetic flux linkage number on the d-axis generated by the magnets of each divided core, is calculated separately using the magnetic flux linkage number λ d_core _-n of each divided core and the electrical angle θ _nn formed by the d-axes of the divided cores, using the following equation (1) ; and the rotor magnet average temperature of each divided core is estimated based on the d-axis magnetic flux linkage number λ _d-n of each coil.

According to another aspect of the present invention, there is provided a rotor magnet temperature estimation device for a permanent magnet synchronous motor in which a rotor core is divided into three or more split cores in the axial direction, and each split core is stacked in the axial direction with a skew angle. This device is equipped with a state estimator that calculates the magnetic flux linkage of the d-axis and q-axis in a rotor synchronous coordinate system from the current and voltage applied to the motor, or a current value of the inductance; and a magnet temperature calculator that calculates the magnetic flux linkage λ _{d_core-n} observed on the d-axis of each divided core from the magnetic flux linkage of the d-axis and q-axis in an unloaded state, where n is the number of divided cores with different skew angles, and separately calculates the d-axis magnetic flux linkage λ _{d-n ,} which is the magnetic flux linkage on the d-axis generated by the magnets in each divided core, using the magnetic flux linkage λ d_core _{-n of each divided core and the electrical angle θ nn} formed by the d-axes of the divided cores, using the following equation (1), and estimates the rotor magnet average temperature of each divided core based on the d-axis magnetic flux linkage λ _d-n of each coil.

The above aspect allows for more accurate estimation of rotor magnet temperature, including in areas closer to the center of the rotor axial direction.

FIG. 1 is a schematic configuration diagram of a rotor core according to a first embodiment. FIG. 2 is a diagram showing the relationship between the d and q axes of the divided cores and the d and q axes in motor output control. FIG. 3 is a schematic configuration diagram of the control system according to the first embodiment. FIG. 4 is a block diagram showing an example of a flux linkage number estimator. FIG. 5 is a block diagram showing another example of the flux linkage number estimator. FIG. 6 is a diagram showing the relationship between magnet temperature and axial position. FIG. 7 is a schematic configuration diagram of a control system according to the second embodiment. FIG. 8 is a diagram showing the dependency of the inductance value on the change in magnet temperature. FIG. 9 is a schematic configuration diagram of a control system according to the third embodiment. FIG. 10 is a diagram showing the dependency of the interlinkage magnetic flux on the d-axis and q-axis in motor control on the magnet temperature of each divided core.

Embodiments of the present invention will be described below with reference to the accompanying drawings.

[First embodiment]
1 is a schematic diagram of a rotor core 1 of a permanent magnet synchronous motor (hereinafter simply referred to as a motor) to which a first embodiment of the present invention is applied. This motor is a radial gap type motor having an air gap in the radial direction of the rotor.

The rotor core 1 contains a permanent magnet (hereinafter simply referred to as a magnet) and is divided into multiple sections along the rotation axis (hereinafter also referred to as the rotation axis direction). In this embodiment, it is divided into three sections, with the central divided core being referred to as the central core 2 and the divided cores at both ends being referred to as the end cores 3.

If the angle in the rotational direction of the rotor core 1 is defined as the mechanical angle, the rotor core 1 has a so-called step skew structure in which the central core 2 and the end cores 3 are arranged so that they are offset by a mechanical angle of θ. In the following explanation, this relative mechanical angle offset θ between the central core 2 and the end core 3 is referred to as the skew angle θ. Note that the two end cores 3 have the same mechanical angle.

In the following description, the central core 2 may be referred to as Core 1 and the end core 3 as Core 2.

2 is a diagram showing the relationship between the d- and q-axes (d-core1, 2, q-core1, 2) of the central core 2 and the end cores 3 and the d- and q-axes (d-master, q-master) for motor output control. λ _d-1 and λ _d-2 in FIG. 2 indicate the magnetic flux linkages generated by the magnets in the central core 2 and the end cores 3, respectively.

These magnetic flux linkages λ _d-1,2 are vectors. Because the central core 2 and the end core 3 are skewed, the magnetic flux linkages λ _d-1,2 have a deviation in the rotational direction by the skew angle θ on the d, q coordinates, and also have a deviation in the rotational direction by ±θ/2 relative to the d-master.

The magnetic flux linkages λ _d-1,2 occur on the d-axis of each of the cores 2, 3, and have temperature characteristics whose magnitude changes with changes in the magnet temperature. Therefore, in this embodiment, the magnetic flux linkages λ _d-1,2 are separately detected and compared with previously acquired temperature characteristics to estimate the temperature of the magnets contained within each of the cores 2, 3.

In the no-load state, the magnetic flux linkage λ _d-1,2 can be calculated by the following formula (1) using the magnetic flux linkage λ _d-core1,2 (the magnetic flux linkage observed on the d-axis of each of cores 1 and 2) calculated by the magnetic flux linkage estimator 14 described below. Note that n in formula (1) is the number of divided cores with different skew angles, and in the case of this embodiment, the two end cores 2 have the same skew angle, so n=2.
3 is a schematic diagram showing an example of a control system 10 equipped with a temperature estimation device that estimates the magnet temperature using the above-described method. The control system 10 includes a motor 13, which is the object to be controlled. Therefore, the parts excluding the motor 13 correspond to the temperature estimation device. The magnet temperature estimated by the temperature estimation device is used to control the rotational operation of the motor 13.

The temperature estimation device includes a subtractor 11 , a PI controller 12 , a magnetic flux linkage estimator 14 and a λ _dn calculator 15 as state estimators, and a magnet temperature calculator 16 .

In the subtractor 11, the current value i ^d _,q , which is a feedback input from the motor 13, is negatively fed back to the torque current command value i* _d,q corresponding to the torque command value. Then, the torque current command value i ^* _d,q passed through the subtractor 11 is input to the PI controller 12.

The PI controller 12 performs feedback control using the current values _id,q, which are observed values in the motor 13. Specifically, the PI controller 12 generates voltage command values v ^* _{d, q} so that the current values id,q in the motor 13 follow the torque current command values i ^* _d,q , and outputs these to the motor 13.

The motor 13 may include an inverter and sensors. A three-phase AC voltage generated by the inverter in response to input of the voltage command values v ^* _{d, q} is applied to the bus bar of the motor 13, thereby rotating the rotor core 1 of the motor 13. A current sensor is provided on the wiring between the inverter and the motor 13, and the three-phase AC current detected by this current sensor is phase-converted to obtain the current values i _{d, q} of the d- and q-axes.

The magnetic flux linkage estimator 14 receives as input the voltage command value v ^* _d,q which is input to the motor 13, the current value i _d,q detected in the motor 13, and the electrical phase angle θe detected by a resolver provided near the rotor of the motor 13. Based on these inputs, the magnetic flux linkage estimator 14 estimates the magnetic flux linkage λ _d,q-core1,2 which is the magnetic flux linkage observed on the d-axis of Core1 and Core2. The specific estimation method is as follows.

Magnetic flux linkage is defined as the total magnetic flux linking the coils wound around the stator multiplied by the number of coil turns, and the time derivative of this is the induced voltage generated in the coil. Therefore, the magnetic flux linkage can be determined by integrating the induced voltage over time. However, when applying it to actual control, simple time integration will result in an error in the calculation of the magnetic flux linkage due to the integration constant, so it is desirable to apply a method to suppress the error, such as combining it with a high-pass filter (see Figure 4) or combining it with a model using nominal values of the motor constants (see Figure 5).

Figure 4 is a block diagram showing an example of a flux linkage estimator 14 combined with a high-pass filter. Note that the dp axis is a rotating coordinate system synchronized with the rotor, and the αβ axis is a stationary coordinate system.

In this case, the magnetic flux linkage estimator 14 includes an integrator 20, a high-pass filter 21, and an αβ/dq converter 22. The αβ-axis magnet magnetic flux linkage λ _α ,β is calculated by time-integrating the voltage command value v ^* _α,β obtained by converting the voltage command value v ^* _d,q into an αβ coordinate value in the integrator 20. This αβ-axis magnet magnetic flux linkage λ _α,β passes through the high-pass filter 21 and the αβ/dq converter 22 and is output as magnet magnetic flux linkage λ _d,q-core1,2 .

Figure 5 is a block diagram showing an example of a magnetic flux linkage estimator 14 combined with a model using nominal motor constant values.

In this case, the magnetic flux linkage estimator 14 includes an inductance calculator 30, a dq/αβ converter 31, a subtractor 32, a PI controller 33, a dq/αβ converter 34, an adder 35, an integrator 36, and an αβ/dq converter 37.

The inductance calculator 30 calculates the dq-axis magnetic flux linkages (current model) λ _{d-i, q-i that occur in response to the current load in the motor 13 by multiplying the current value i d,} q by the inductance value L _d,q of the motor 13. This process is based on the general relationship between current, inductance, and magnetic flux linkages. Note that the inductance value L _d,q is a fixed value, and specifically, a nominal value is used.

The dq/αβ converter 31 performs dq/αβ conversion on the dq-axis magnet flux linkages (current model) λ _d-i,q-i to calculate the αβ-axis magnet flux linkages (current model) λ _α-i,β-i .

The subtractor 32 calculates Δλ _α-i,β-i by subtracting the αβ-axis magnet flux linkage number λ _α,β output from the subsequent integrator 36 from the αβ-axis magnet flux linkage number (current model) λ _α-i,β-i .

In the PI controller 33, PI control is performed on the αβ-axis magnet flux linkage numbers (current model) λ α- _{i, β-} i, thereby controlling so that the difference between the αβ-axis magnet flux linkage numbers (current model) λ _{α-i, β-i} and the αβ-axis magnet flux linkage numbers λ _{α, β} becomes small.

The dq/αβ converter 34 performs dq/αβ conversion on the voltage command value v ^* _d,q .

The output from the PI controller 33 and the output from the dq/αβ converter 34 are added in an adder 35, and the addition result is time-integrated in an integrator 36 to calculate the αβ-axis magnet flux linkage number λ _α,β . Finally, the magnet flux linkage number λ _d,q-core1,2 is output via an αβ/dq converter 37.

The magnetic flux linkage λ _d,q-core1,2 calculated by any of the above methods is input to the λ _d-n calculator 15, and the magnetic flux linkage λ _d-1,2 generated by the magnets contained in Core1 and Core2 is calculated by the above-mentioned equation (1).

The magnet temperature calculator 16 estimates the magnet temperature T _core1,2 based on the relationship between the magnet flux linkage λ _d-1,2 and the magnet temperature. The relationship between the magnet flux linkage λ _d-1,2 and the magnet temperature is mapped using data previously obtained through experiments or the like, for example as shown in FIG. 3, and stored in the magnet temperature calculator 16. Although the magnet temperature within each divided core has a temperature gradient in the axial direction, the magnet temperature used for mapping is the average temperature in the axial direction. In other words, the magnet temperatures T _core1,2 estimated by the above procedure are the average temperatures in the axial direction of each magnet contained within Core 1 and Core 23, respectively.

This is shown in Figure 6. The horizontal axis of Figure 6 represents axial position, the vertical axis represents magnet temperature, and _Tcore1 and _Tcore2 , shown by solid lines, are magnet temperatures estimated using the above procedure. This result also shows that the magnet temperature near the center, which is relatively high, can be estimated with greater accuracy than the method described in the above-mentioned literature. However, to further improve estimation accuracy, the actual temperature gradient can be estimated from the estimation results shown in Figure 6 using, for example, linear extrapolation, and the magnet temperature distribution and maximum temperature within rotor core 1 can be estimated as shown by the dashed lines in Figure 6. Estimating the maximum temperature more accurately makes it easier to avoid thermal demagnetization of the magnets during operation.

As described above, this embodiment provides a rotor magnet temperature estimation method for a permanent magnet synchronous motor in which the rotor core 1 is divided axially into three or more split cores 2, 3, and each split core 2, 3 is stacked axially at a skew angle θ. This method calculates the magnetic flux linkage numbers for the d-axis and q-axis in the rotor synchronous coordinate system from the current and voltage applied to the motor, and estimates the rotor magnet temperature of each split core with a different skew angle based on the calculated magnetic flux linkage numbers for the d-axis and q-axis. Because the magnetic flux linkage number is an observable value and is dependent on the magnet temperature distribution, this embodiment enables accurate estimation of the magnet temperature. Furthermore, since the rotor core 1 has a so-called step skew, and the temperatures of split cores with different skew angles are estimated, magnet temperatures with a temperature distribution in the axial direction can be accurately estimated.

In this embodiment, each split core 2, 3 is stacked with a skew angle between the split core (center core 2) located approximately in the center of the axial direction and the split core (end core 3) located at the end of the axial direction. Because the center core 2 and end core 3 are stacked with different skew angles, the magnet temperature in the center, which becomes relatively hot, can be accurately estimated.

In this embodiment, the number of split cores with different skew angles is n, and the magnetic flux linkage λ _{d_core-n} observed on the d-axis of each split core is calculated from the magnetic flux linkages on the d-axis and q-axis under no load. Then, using equation (1) from the magnetic flux linkage λ _{d_core-n} of each split core and the electrical angle θ _nn formed by the d-axes of the split cores, the d-axis magnetic flux linkage λ _d-n , which is the magnetic flux linkage on the d-axis generated by the magnets possessed by each split core, is calculated separately, and the average rotor magnet temperature of each split core is estimated based on the d-axis magnetic flux linkage λ _d-n of each coil. That is, the d-axis magnetic flux linkage λ _d-n of each split core is calculated separately using equation (1), and the magnet temperature of each split core is estimated based on this, so the magnet temperature of each split core can be detected separately.

In this embodiment, the magnetic flux linkage number λ _{d_core-n} on the d-axis of each divided core is calculated by integrating the voltage applied to the motor over time. This allows the actual magnetic flux state generated in the motor to be detected with high accuracy, improving the accuracy of estimating the magnet temperature.

In this embodiment, the temperature gradient of the rotor magnet temperature in the axial direction is calculated based on the calculated average rotor magnet temperature of each divided core and the relative axial distance between the axial center positions of each divided core. This allows for more accurate magnet temperature distribution within the rotor, making it possible to estimate the maximum magnet temperature and avoid partial thermal demagnetization of the magnet.

In this embodiment, the rotor magnet temperature at the axial center of rotor core 1 is estimated using the average rotor magnet temperature of one of the divided cores, the temperature gradient, and the distance from the axial center of that divided core to the axial center of rotor core 1. This makes it possible to obtain the magnet temperature at approximately the axial center, where the temperature is highest, thereby avoiding partial thermal demagnetization of the magnet.

In this embodiment, the magnetic flux linkage numbers λ _d,q-core1,2 are calculated by observing the magnetic flux linkages on the d-core1, 2, but the calculation method is not limited to this. For example, they may be calculated by coordinate transformation using the observed values of the magnetic flux linkage numbers λ _{d-master and} λ _q-master on the d- _master and q-master and the skew angle θ.

Second Embodiment
In the first embodiment, the magnet temperature is estimated using the temperature characteristics of the magnetic flux linkage of Core 1 and Core 2, but in this embodiment, instead, the characteristic that the inductance value L d,q _-master of the _d,q-master has dependency on the magnet temperature change of Core 1 and Core 2 is used.

The inductance value L _d,q-master is calculated by applying an AC voltage to any position between the d-axis and the q-axis and using the ratio between the amplitude of the applied AC voltage and the amplitude of the current generated at the same frequency as the applied AC voltage. The frequency of the applied AC voltage is set to be equal to or higher than the frequency of the current command value for motor control.

Fig. 7 is a schematic diagram showing an example of a control system 10 equipped with a temperature estimation device according to this embodiment. The differences from Fig. 3 are that a high-frequency current superimposer 40 and an adder 41 are added, and that a band-pass filter 42 and an inductance calculator 43 serving as a state estimator are provided instead of the magnetic flux linkage estimator 14 and the λ _d-n calculator 15. The calculation method performed by the magnet temperature calculator 16 is also different. The following explanation will focus on these differences.

When Trig is input in response to a magnet temperature estimation request from a higher-level device (not shown), the high-frequency current superimposer 40 generates a high-frequency current command value so that a high-frequency current Δi flows only during that period. The amplitude of the high-frequency current is I _d,q-HF and its frequency is ω. Note that the frequency ω is an angular velocity defined by the frequency f of the superimposed high-frequency current, and can be calculated by 2πf.

In the adder 41, the high frequency current command value Δi is superimposed on the torque current command value i ^* _d,q , and the superimposed torque current command value i ^* _d,q is input to the subtractor 11 described above.

The bandpass filter 42 extracts the voltage (amplitude V _d,q−HF ) required to superimpose the high-frequency current from the voltage command value v ^* _d,q , extracts the high-frequency current from the current value i _d,q, and inputs these to the inductance calculator 43 .

The inductance calculator 43 calculates the inductance value L _d,q-master using the following equation (2).
The magnet temperature calculator 16 calculates the magnet temperatures T _core1,2 of Core1 and Core2 based on the dependency of the inductance value L _d,q-master on changes in magnet temperature (see FIG. 8).

Figure 8 plots combinations of Core 1's magnet temperature _Tcore1 and Core 2's magnet temperature _Tcore2 in a coordinate system where the horizontal axis is the d-master inductance value Ld _-master and the vertical axis is the q-master inductance value Lq _-master . Rows q1 to q4 are combinations where the Core 2 magnet temperature is constant and the Core 1 magnet temperature is different, with q1 being the lowest and q4 being the highest for Core 2. Columns d1 to d4 are combinations where the Core 1 magnet temperature is constant and the Core 2 magnet temperature is different, with d1 being the lowest and d4 being the highest for Core 1.

The inductance value L _d,q-master exhibits the dependency shown in FIG. 8 on the magnet temperature T _core1,2 of Core 1 and Core 2 for each load current I _d,q. Therefore, the magnet temperatures of Core 1 and Core 2 can be estimated from the inductance value L _d,q-master calculated by equation (2) and FIG. 8, even when there is a load.

The magnet temperature estimated using the above procedure is the average temperature in the axial direction of Core 1 and Core 2, as in the first embodiment. Then, as in the first embodiment, the axial temperature gradient is calculated from the estimated magnet temperature, and the actual magnet temperature can be estimated from this and the axial length of each core.

As described above, this embodiment provides a rotor magnet temperature estimation method for a permanent magnet synchronous motor in which the rotor core is divided axially into three or more split cores 2, 3, and each split core 2, 3 is stacked axially with a skew angle. This method calculates the current values of the d-axis and q-axis inductances in the rotor synchronous coordinate system from the current and voltage applied to the motor, and estimates the rotor magnet temperature of each split core with a different skew angle based on the calculated d-axis and q-axis inductances. This allows for accurate magnet temperature estimation, similar to the first embodiment.

In this embodiment, the d-axis and q-axis inductances are calculated by applying an AC voltage to any position between the d-axis and q-axis, and using the ratio between the amplitude of the applied AC voltage and the amplitude of the current generated at the same frequency as the applied AC voltage. This reduces the time required to calculate the inductance, thereby preventing deterioration of motor controllability.

In this embodiment, the frequency of the applied AC voltage is equal to or greater than the frequency of the current command value for motor control. This avoids interference with motor control, thereby minimizing deterioration of motor controllability.

In this embodiment, the correlation between the average rotor magnet temperature of each divided core and the d-axis and q-axis inductance under load is calculated in advance and stored. The d-axis and q-axis inductance under load are used as input values to calculate the average rotor magnet temperature of each divided core based on this correlation. This reduces the impact of fluctuations in magnetic flux linkage due to load conditions, thereby further improving the accuracy of magnet temperature estimation.

[Third embodiment]
In the first embodiment, the magnet temperature is estimated using the temperature characteristics of the magnetic flux linkage of Core 1 and Core 2. However, in this embodiment, instead, the dependency of the d,q-master magnetic flux linkage λ _d,q-master on the magnet temperature of Core 1, 2 at the load current I _d,q (see FIG. 10 ) is utilized.

Fig. 9 is a schematic diagram showing an example of a control system 10 equipped with a temperature estimation device according to this embodiment. The difference from Fig. 3 is that the magnet temperature calculator 16 calculates the magnet temperature based on the flux linkage λ _{d,q-master of the d,q} -master, as described above. Note that the output of the flux linkage estimator 14 serving as a state estimator is λ _d,q-master , which is the same as λ _d,q-core1,2 in the first embodiment.

10 is a plot of combinations of the magnet temperature T _core1 of Core 1 and the magnet temperature T _core2 of Core 2 in a coordinate system with the horizontal axis being λ _d-master and the vertical axis being λ _q-master . Rows q1 to q4 and columns d1 to d4 are the same as in FIG.

Since the flux linkage λ _{d,q-master of the d,q-master} exhibits the dependency shown in FIG. 10 on the magnet temperature of Cores 1 and 2 at the load current I _d,q , the magnet temperatures of Cores 1 and 2 can be estimated from λ _d,q-master calculated by the flux linkage estimator 14 and FIG. 10 even when there is a load.

The magnet temperature estimated using the above procedure is the average temperature in the axial direction of Core 1 and Core 2, as in the first embodiment. Then, as in the first embodiment, the axial temperature gradient is calculated from the estimated magnet temperature, and the actual magnet temperature can be estimated from this and the axial length of each core.

As described above, this embodiment achieves the same effects as the first embodiment.

In addition, in this embodiment, the correlation between the average rotor magnet temperature of each divided core and the d-axis and q-axis magnetic flux linkages under load is calculated in advance and stored, and the d-axis and q-axis magnetic flux linkages under load are used as input values to calculate the average rotor magnet temperature of each divided core based on this correlation. This reduces the impact of magnetic flux linkage fluctuations due to load conditions, thereby further improving the accuracy of magnet temperature estimation.

The present invention is not limited to the above-described embodiment, and it goes without saying that various modifications can be made within the scope of the technical concept described in the claims. For example, while the embodiment describes a case in which the rotor core 1 is divided into three parts, this is not limited to this and can be applied equally to a case in which the rotor core is divided into four or more parts. When the number of parts is odd, the magnet temperature at the axial center, which is the hottest, can be estimated from the temperature of the divided core located in the axial center, just as in the case of a three-part division. On the other hand, even when the number of parts is even, the magnet temperature at the axial center can be estimated from the temperature of the divided core located approximately in the axial center.

1. Rotor, 2. Center core (Core 1), 3. End core (Core 2)

Claims (10)

A rotor magnet temperature estimation method for a permanent magnet synchronous motor in which a rotor core is divided into three or more divided cores in an axial direction, and each divided core is stacked in the axial direction at a skew angle, comprising:
Calculating the current values of the magnetic flux linkages or inductances of the d-axis and q-axis in the rotor synchronous coordinate system from the current and voltage applied to the motor;
The number of divided cores with different skew angles is n,
Calculating the magnetic flux linkage number λ _{d_core-n} observed on the d axis of each divided core from the magnetic flux linkage numbers on the d axis and q axis in a no-load state ;
Using the magnetic flux linkage number λ _{d_core-n} of each divided core and the electrical angle θ _nn formed by the d axes of each divided core, the d- axis magnetic flux linkage number λ _d-n , which is the magnetic flux linkage number on the d axis generated by the magnets of each divided core , is calculated separately using the following formula (1):
A rotor magnet temperature estimation method characterized by estimating the average rotor magnet temperature of each divided core based on the d-axis magnet flux linkage number λ _d−n of each coil.
2. The rotor magnet temperature estimation method according to claim 1,
A rotor magnet temperature estimation method, wherein each divided core is stacked with a divided core located approximately in the center of the axial direction and a divided core located at an end in the axial direction at a skew angle. 3. The rotor magnet temperature estimation method according to claim 1 , further comprising :
A rotor magnet temperature estimation method in which d-axis and q-axis inductances are calculated by applying an AC voltage to any position between the d-axis and q-axis, and using the ratio between the amplitude of the applied AC voltage and the amplitude of a current generated at the same frequency as the applied AC voltage. 4. The rotor magnet temperature estimation method according to claim 3,
A rotor magnet temperature estimation method, wherein the frequency of the applied AC voltage is equal to or higher than the frequency of a current command value for motor control. 5. The rotor magnet temperature estimation method according to claim 1,
A rotor magnet temperature estimation method that calculates the magnetic flux linkage number λ _{d_core-n} on the d axis of each divided core by integrating the voltage applied to the motor over time. A rotor magnet temperature estimation method for a permanent magnet synchronous motor in which a rotor core is divided into three or more divided cores in an axial direction, and each divided core is stacked in the axial direction at a skew angle, comprising:
The correlation between the rotor magnet average temperature of each divided core and the magnetic flux linkage number or inductance of the d-axis and q-axis magnets under load is calculated and stored in advance,
Calculating the current values of the magnetic flux linkages or inductances of the d-axis and q-axis in the rotor synchronous coordinate system from the current and voltage applied to the motor;
A rotor magnet temperature estimation method that calculates the average rotor magnet temperature of each divided core based on the correlation using the d-axis and q-axis magnetic flux linkages or inductance in a loaded state as input values. 7. The rotor magnet temperature estimation method according to claim 1,
A rotor magnet temperature estimation method for calculating a temperature gradient of the rotor magnet temperature in the axial direction based on the calculated average rotor magnet temperature of each divided core and the relative distance in the axial direction between the axial center positions of each divided core. 8. The rotor magnet temperature estimation method according to claim 7,
A rotor magnet temperature estimation method for estimating the rotor magnet temperature at the axial center position of the rotor core using the average rotor magnet temperature of any one of the divided cores, the temperature gradient, and the distance from the axial center position of the divided core to the axial center position of the rotor core. A rotor magnet temperature estimation device for a permanent magnet synchronous motor in which a rotor core is divided into three or more divided cores in the axial direction, and each divided core is stacked in the axial direction at a skew angle,
a state estimator that calculates the current values of the magnetic flux linkages of the d-axis and q-axis in the rotor synchronous coordinate system or the inductance from the current and voltage applied to the motor;
a magnet temperature calculator that calculates a magnetic flux linkage λ _{d_core-n} observed on the d-axis of each divided core from the magnetic flux linkages on the d-axis and q-axis in a no-load state, where n is the number of divided cores with different skew angles , and separately calculates a d-axis magnetic flux linkage λ _d-n , which is the magnetic flux linkage on the d-axis generated by the magnets in each divided core, using the magnetic flux linkage λ _{d_core-n} of each divided core and the electrical angle θ _nn formed by the d-axes of the divided cores , using the following formula (1), and estimates an average rotor magnet temperature of each divided core based on the d-axis magnetic flux linkage λ _d-n of each coil;
A rotor magnet temperature estimation device comprising:
A rotor magnet temperature estimation device for a permanent magnet synchronous motor in which a rotor core is divided into three or more divided cores in the axial direction, and each divided core is stacked in the axial direction at a skew angle,
a storage unit that preliminarily determines and stores the correlation between the rotor magnet average temperature of each divided core and the magnetic flux linkage number or inductance of the d-axis and q-axis magnets under load;
a state estimator that calculates the current values of the magnetic flux linkages of the d-axis and q-axis in the rotor synchronous coordinate system or the inductance from the current and voltage applied to the motor;
a magnet temperature calculator that calculates the rotor magnet average temperature of each divided core based on the correlation using the d-axis and q-axis magnetic flux linkages or inductance in a loaded state as input values;
A rotor magnet temperature estimation device comprising: