Permanent magnet synchronous motor temperature estimation method based on high-frequency signal injection

Abstract

The invention discloses a permanent magnet synchronous motor temperature estimation method based on high-frequency signal injection. The method comprises a permanent magnet synchronous motor control system, a high-frequency signal injection mechanism, a high-frequency signal processing mechanism and a temperature estimation mechanism. The high-frequency signal injection mechanism processes a current signal and an electrical angle signal of the permanent magnet synchronous motor, injects a high-frequency signal into the permanent magnet synchronous motor, and feeds back a fundamental frequency current and an electrical angle; the high-frequency signal processing mechanism calculates a high-frequency error voltage caused by an inverter dead-zone effect and device voltage drop through a high-frequency current and an electrical angle, and identifies impedance through an expected high-frequency voltage and an expected high-frequency current; the drive temperature estimation mechanism estimates the permanent magnet temperature and the stator winding temperature from the fundamental current, the electrical angle, and the impedance. The permanent magnet synchronous motor temperature estimation method has the beneficial effects that the permanent magnet synchronous motor temperature estimation method with high universality is provided, the permanent magnet temperature and the stator winding temperature are accurately monitored, and the control performance and the safety of a permanent magnet synchronous motor system are improved.

Description

Technical Field

[0001] The present invention relates to temperature estimation, and in particular to a method for estimating the temperature of a permanent magnet and a stator winding of a permanent magnet synchronous motor. Background Art

[0002] In the past three decades, permanent magnet synchronous motors have been increasingly widely used due to their attractive features such as high power, high torque density and high efficiency. However, the temperature of the permanent magnet has a great influence on the performance of the permanent magnet synchronous motor. The increase in the temperature of the permanent magnet leads to a decrease in the magnetic field strength of the permanent magnet, which leads to a decrease in the torque density of the permanent magnet synchronous motor. When the temperature rises beyond a certain limit, this decrease in torque density may be permanent. The temperature of the stator winding is also crucial to the safety of the permanent magnet synchronous motor. The increase in the temperature of the stator winding will cause the winding resistance to increase, affecting the control performance. When the temperature of the stator winding is too high, it may even cause the winding to fuse, destroying production safety. In order to deal with various problems caused by the increase in temperature during the operation of the permanent magnet synchronous motor, accurate monitoring of the permanent magnet temperature and the stator winding temperature is necessary for both control and safety. The traditional method requires the installation of complex devices such as wireless temperature sensors on the permanent magnet and stator winding for direct temperature measurement, which is costly. In addition, due to direct contact with the permanent magnet synchronous motor, these sensors will also reduce the robustness of the permanent magnet synchronous motor.

[0003] The industry has conducted research on methods for estimating the temperature of permanent magnets and stator windings. The literature [B.-H.Lee, K.-S.Kim, J.-W.Jung, J.-P.Hong and Y.-K.Kim, "Temperature Estimation of IPMSM Using Thermal Equivalent Circuit," in IEEE Transactions on Magnetics, vol.48, no.11, pp.2949-2952, Nov.2012, doi:10.1109 / TMAG.2012.2196503.] proposed an IPMSM thermal equivalent circuit that takes into account PM eddy current loss and rotor core loss to estimate the temperature of permanent magnets and stator windings.

[0004] The above literature provides valuable research ideas. However, the above literature based on thermal models uses thermal networks to model the heat transfer characteristics of different machine components (e.g., rotor and stator lamination, shaft, stator cover, radiator, etc.), which requires understanding the machine geometry, materials, and cooling system to create the model, so each machine needs to design a specific model. Therefore, proposing a highly universal permanent magnet synchronous motor temperature estimation method is a key technical issue in permanent magnet synchronous motor temperature estimation. Summary of the invention

[0005] The technical problem to be solved by the present invention is that there is a lack of a highly universal permanent magnet synchronous motor temperature estimation method in the field of permanent magnet synchronous motor temperature estimation, and a permanent magnet synchronous motor temperature estimation method based on high-frequency signal injection is proposed.

[0006] In order to achieve the above objectives, according to one aspect of a specific embodiment of the present invention, a permanent magnet synchronous motor control system, a high-frequency signal injection mechanism, a high-frequency signal processing mechanism and a data-driven temperature estimation mechanism are provided.

[0007] The high-frequency signal injection mechanism processes the current signal and electrical angle signal of the permanent magnet synchronous motor, injects high-frequency signal into the permanent magnet synchronous motor, and feeds back the baseband current and electrical angle; the high-frequency signal processing mechanism calculates the high-frequency error voltage caused by the inverter dead zone effect and the device voltage drop through the high-frequency current and electrical angle, and then identifies the impedance through the expected high-frequency voltage and expected high-frequency current; the data-driven temperature estimation mechanism estimates the permanent magnet temperature and stator winding temperature through the baseband current, electrical angle and impedance.

[0008] Specifically, the permanent magnet synchronous motor control system is composed of a space vector controller, an inverter and a permanent magnet synchronous motor. The space vector controller controls the inverter to generate the desired voltage vector in the natural coordinate system. Due to the dead zone effect and device voltage drop of the inverter, the actual voltage vector u in the natural coordinate system is generated. abc To drive a permanent magnet synchronous motor.

[0009] Specifically, the high-frequency signal injection mechanism is composed of a signal processing module and a high-frequency signal injection module. The signal processing module monitors the electrical angle sensor signal θ e And process the sensor current signal i in the natural coordinate system abcm , output electrical angle θ e and the fundamental frequency current vector i in the natural coordinate system abc To the space vector controller, output electrical angle θ e and the high-frequency current vector i in the natural coordinate system abchf To the signal injection module and error voltage calculation module. The high-frequency signal injection module is based on the electrical angle θ e and the high-frequency current vector i in the natural coordinate system abchf Output the desired high-frequency voltage vector in the synchronous rotating coordinate system To the space vector controller, output the desired high-frequency voltage vector in the synchronous rotating coordinate system And the expected high-frequency current vector in the synchronous rotating coordinate system To the impedance online identification module.

[0010] Specifically, the high-frequency signal processing mechanism is composed of a high-frequency error voltage calculation module and an impedance identification module. The high-frequency error voltage calculation module is based on the electrical angle θ e and the high-frequency current vector i in the natural coordinate system abchf Output high-frequency error voltage vector Δv in synchronous rotating coordinate system dqhf To the impedance identification module, the impedance identification module combines the high-frequency error voltage vector Δv in the synchronous rotating coordinate system dqhf and the expected high-frequency voltage vector in the synchronous rotating coordinate system And the expected high-frequency current vector in the synchronous rotating coordinate system Identify the impedance Z d .

[0011] Specifically, the temperature estimation mechanism uses the permanent magnet synchronous motor data sample, applies correlation analysis to determine the temperature model structure, determines the temperature model parameters through parameter identification, obtains the temperature estimation model, and transforms the fundamental frequency current vector i in the natural coordinate system abc 、Electrical angle θ e and impedance Z d Input the temperature model to estimate the permanent magnet temperature T r and stator winding temperature T s .

[0012] The beneficial effect of the present invention is that a highly universal permanent magnet synchronous motor temperature estimation method is proposed, which can achieve accurate monitoring of permanent magnet temperature and stator winding temperature and improve the control performance and safety of the permanent magnet synchronous motor system.

[0013] The content of the present invention is further explained below in conjunction with the accompanying drawings and specific implementation methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a brief structural schematic diagram of the present invention; DETAILED DESCRIPTION

[0015] The following takes a permanent magnet synchronous motor temperature estimation as an example and describes the technical solution of the present invention in detail, clearly and completely in combination with the accompanying drawings to facilitate professionals and technicians in this field to better understand the present invention.

[0016] Example

[0017] like Figure 1 As shown in the schematic diagram of the structure of the present invention, the specific steps of the permanent magnet synchronous motor temperature estimation method based on high-frequency signal injection in this example include:

[0018] In this example, a high-frequency signal injection mechanism is first designed to process the current signal and electrical angle signal of the permanent magnet synchronous motor, inject high-frequency signals into the permanent magnet synchronous motor, and feed back the baseband current and electrical angle; then a high-frequency signal processing mechanism is designed to calculate the high-frequency error voltage caused by the inverter dead zone effect and the device voltage drop through the high-frequency current and electrical angle, and then the impedance is identified through the expected high-frequency voltage and expected high-frequency current; finally, a data-driven temperature estimation mechanism is designed to estimate the permanent magnet temperature and stator winding temperature through the baseband current, electrical angle and impedance.

[0019] The permanent magnet synchronous motor control system in this example is composed of a space vector controller, an inverter and a permanent magnet synchronous motor. The space vector controller in this example is composed of a PI controller, an inverse Park transformation module, a space vector pulse width modulator svpwm with dead time protection and a Clark-Park transformation module 1, wherein the Clark-Park transformation module is based on the electrical angle θ e The fundamental frequency current vector i in the natural coordinate system abc The three components i a 、i b 、i c Transformed into the fundamental frequency current vector i in the synchronous rotating coordinate system dq , the fundamental frequency current vector i in the synchronous rotating coordinate system dq The expected current vector in the synchronous rotating coordinate system The differential input PI controller generates the desired voltage in the synchronous rotating coordinate system Expected voltage in synchronous rotating coordinate system The expected high-frequency voltage vector in the synchronously rotating coordinate system Expected synthetic voltage in the synthetic synchronous rotating coordinate system The inverse Park transformation module is based on the electrical angle θ e The expected synthetic voltage in the synchronous rotating coordinate system Transformed into the two components of the desired composite voltage in the stationary coordinate system The space vector pulse width modulator svpwm with dead time protection is based on the two components of the desired synthetic voltage in the stationary coordinate system. The signal g is sent to control the switch of the inverter to drive the permanent magnet synchronous motor. Therefore, the inverter has a dead zone effect and is also affected by the device voltage drop, which makes the permanent magnet synchronous motor control system have an error between the expected voltage and the actual voltage. A current sensor is installed between the inverter and the permanent magnet synchronous motor to measure the sensor current signal i in the natural coordinate system. abcm The three components i am 、i bm 、i cm , and an electrical angle sensor is installed on the permanent magnet synchronous motor to measure the electrical angle θ e .

[0020] The high-frequency signal injection mechanism in this example consists of a signal processing module and a high-frequency signal injection module. The signal processing module consists of a high-frequency signal frequency -w hf and w hf A band-stop filter with a stop band center and a high-frequency signal frequency -w hf and w hf The bandpass filter is composed of a passband center, where the high-frequency signal frequency -w hf and w hf The band-stop filter at the stop band center converts the sensor current signal i in the natural coordinate system into abcm The three components i am 、i bm 、i cm The fundamental frequency current vector i in the natural coordinate system abc The three components i a 、i b 、i c They are extracted and input to Clark-Park transformation module 1, with high-frequency signal frequency -w hf and w hf The bandpass filter with the passband center transforms the sensor current signal i in the natural coordinate system abcm The three components i am 、i bm 、i cm The high-frequency current vector i in the natural coordinate system abchf The three components i ahf 、i bhf 、i chf The signal is extracted and input to Clark-Park transformation module 2. The signal processing module converts the electrical angle sensor signal θ e Distributed to the inverse Park transformation module, Clark-Park transformation module 1, Clark-Park transformation module 2 and error voltage calculation module. The high-frequency signal injection module consists of a PR controller and a Clark-Park transformation module 2, wherein the Clark-Park transformation module 2 is based on the electrical angle θ e The high-frequency current vector i in the natural coordinate system abchf The three components i ahf 、i bhf 、i chf Transformed into the high-frequency current vector i in the synchronous rotating coordinate system dqhf , high-frequency current vector i in synchronous rotating coordinate system dqhf The expected high-frequency current vector in the synchronously rotating coordinate system The differential input PR controller generates the desired high frequency voltage in the synchronous rotating coordinate system Achieve tracking of desired high frequency current vector In this example, the expected high-frequency current vector in the synchronous rotating coordinate system Defined as:

[0021]

[0022] Among them I hf is the amplitude, w hf is the frequency.

[0023] The high-frequency signal processing mechanism in this example is composed of a high-frequency error voltage calculation module and an impedance identification module. The high-frequency error voltage calculation module is based on the electrical angle θ e and the high-frequency current vector i in the natural coordinate system abchf The three components i ahf 、i bhf 、i chf Calculate the high-frequency error voltage vector Δv in the synchronous rotating coordinate system dqhf , the calculation formula is:

[0024]

[0025] Where Δv is the high frequency error voltage amplitude, t dead is the dead zone protection time, t on is the active switch turn-on time, t off is the active switch off time, t s is the space vector pulse width modulation period, v dc Inverter DC input voltage, v sat is the saturation voltage drop of the active switch, v d is the forward voltage drop of the freewheeling diode, Δv abchf is the high-frequency error voltage in the natural coordinate system. The impedance identification module consists of the d-axis component selection module and the high-frequency signal frequency-w hf The band-stop filter 1 is the center of the stop band, with the high-frequency signal frequency -w hf The band-stop filter 2 and the impedance calculation module at the stop band center are composed of the expected high-frequency voltage in the synchronous rotating coordinate system. The high-frequency error voltage Δv in the synchronous rotating coordinate system dqhf The real high-frequency voltage u in the synchronous rotating coordinate system is obtained by subtraction dqhf And input the d-axis component selection module, set the q-axis component to 0, and keep the d-axis component unchanged to obtain the real high-frequency voltage processing value u in the synchronous rotating coordinate system dzhf , the actual high-frequency voltage processing value u in the synchronous rotating coordinate system dzhf By using a high frequency signal frequency -w hf The positive sequence component u is obtained by taking the band-stop filter 1 at the stop band center as dzhf+ , the expected high-frequency current vector in the synchronous rotating coordinate system By using a high frequency signal frequency -whf The positive sequence component of the band-stop filter 2 at the stop band center is obtained The positive sequence component u dzhf+ and positive sequence component The input impedance calculation module calculates the impedance Z d , the calculation formula is:

[0026]

[0027] The data-driven temperature estimation mechanism in this example uses permanent magnet synchronous motor data samples, applies correlation analysis to determine the model structure, determines the model parameters through parameter identification, obtains the temperature model, and then uses the fundamental frequency current vector i in the natural coordinate system to estimate the temperature. abc 、Electrical angle θ e and impedance Z d Input the temperature model to estimate the permanent magnet temperature T r and stator winding temperature T s The temperature model consists of a resistance calculation module, an inductance calculation module, a stator winding temperature calculation module, and a permanent magnet temperature calculation module. The resistance calculation module calculates the temperature based on the input impedance Z. d And the calculation formula:

[0028]

[0029] Calculate the resistance R dhf ,in The impedance Z d The inductance calculation module is based on the input impedance Z d And the calculation formula:

[0030]

[0031] Calculate the inductance L dhf The permanent magnet temperature calculation module calculates the temperature based on the input inductance L dhf , d-axis current i d , q-axis current i q And the calculation formula:

[0032]

[0033] Calculate the permanent magnet temperature T r , where L dhf0 is the initial value of inductance, is the relationship coefficient between d-axis current and inductance, i d is the d-axis current, is the relationship coefficient between q-axis current and inductance, i q is the q-axis current, is the coefficient of relationship between the magnetic field strength and inductance of the permanent magnet, B r0 is the initial value of the permanent magnet magnetic induction intensity, αT is the relationship coefficient between permanent magnet temperature and permanent magnet magnetic induction intensity, T0 is the initial value of ambient temperature, and the stator winding temperature calculation module calculates the temperature according to the input resistance R dhf , permanent magnet temperature T r And the calculation formula:

[0034]

[0035] Calculate the rotor winding temperature T s , where R drhf0 is the initial value of the permanent magnet partial resistance, α mag is the coefficient of relationship between the resistance of the permanent magnet and the temperature of the permanent magnet, R dshf0 is the initial value of the stator winding resistance, α Cu is the relationship coefficient between the partial resistance of the stator winding and the temperature of the stator winding.

Claims

1. A permanent magnet synchronous motor temperature estimation method based on high frequency signal injection, characterized in that , including the following steps: Step 1, design a permanent magnet synchronous motor control system; Step 2, designing a high-frequency signal injection mechanism, including a high-frequency signal injection module and a signal processing module; Step 3, designing a high-frequency signal processing mechanism, including a high-frequency voltage calculation module and an impedance identification module; Step 4, design a temperature estimation method, use the permanent magnet synchronous motor data sample, apply correlation analysis to determine the temperature model structure, determine the temperature model parameters through parameter identification, and obtain the temperature estimation model; Step 5: Introduce the data of the high-frequency signal injection mechanism and the high-frequency signal processing mechanism into the temperature model of the temperature estimation mechanism to estimate the permanent magnet temperature and the stator winding temperature.

2. The high-frequency signal injection mechanism according to claim 1, characterized in that: The algorithms include rotating high-frequency signal injection method, pulse high-frequency signal injection method and fixed-frequency high-frequency signal injection method.

3. The high-frequency signal processing mechanism according to claim 1, characterized in that: The algorithms include orthogonal demodulation, bandpass filtering, phase-locked loop (PLL) technology, and signal processing based on machine learning.

4. The data sample according to claim 1, characterized in that: Includes design samples, historical data, and online data.

5. The correlation analysis according to claim 1, characterized in that: It includes partial correlation analysis, canonical correlation analysis, grey correlation analysis, principal component correlation analysis and cross-correlation function analysis.

6. The parameter identification method according to claim 1, characterized in that: Including data fitting method and probability density analysis method.

Images