Square wave injection sensorless control method for single current sensor system

Abstract

A square wave injection position sensorless control method for a single current sensor system is disclosed. By designing an edge-aligned pulse width modulation method, the phase current under square wave voltage injection is accurately sampled, and the differential current is calculated and input into a phase-locked loop to estimate the rotor position and speed. At the same time, the polarity of the rotor magnetic pole can be identified by using positive and negative pulses and the proposed edge-aligned pulse width modulation current sampling method, thereby achieving reliable position sensorless closed-loop control. The single current sensor square wave injection position sensorless control method provided by the present invention has excellent control performance, simple implementation, small calculation amount, and clear logic, providing a practical method for achieving low-cost, high-performance, low-speed position sensorless control of embedded permanent magnet synchronous motors.

Description

Technical Field

[0001] The present invention belongs to the field of position sensorless control of permanent magnet synchronous motors, and relates to a square wave injection position sensorless control method based on an edge-aligned PWM modulation method applied to a single current sensor system. Background Art

[0002] The advent of permanent magnet synchronous motors (PMSMs) has improved the efficiency, power density, and dynamic performance of modern electric drive systems. To implement vector control, the drive requires information about the motor's rotor position and current. Conventional methods for obtaining this information use specialized sensors, but this inevitably increases system cost and instability.

[0003] To further reduce costs, enhance robustness, and expand the application areas of PMSM, position sensorless control of PMSM has become a research hotspot in the field of motor control. Position sensorless control technology typically requires acquiring two or three-phase currents from the motor. However, small and medium-sized inverters have high requirements for cost-performance, and the current sampling circuit accounts for a significant proportion of the overall machine cost. A more conventional current sampling method uses two current sensors to measure current, but this introduces errors caused by inconsistent accuracy and amplification factors between the two current sensors, and also increases costs. Using a single resistor to sample the bus current and reconstruct the motor's phase current avoids errors caused by inconsistent sensor performance and reduces system costs. Therefore, position sensorless control methods for permanent magnet synchronous motors based on single-resistor sampling systems have become a research hotspot.

[0004] The document "Sensorless Control Strategy for Low-Speed ​​Operation of Permanent Magnet Synchronous Motor with Single Resistor Sampling" (Chen Feifan. Harbin Institute of Technology, 2021.) proposes a backtracking-predictive reconstruction error compensation method, which effectively reduces the phase current reconstruction error. The compensated current is equivalent to the current value at the midpoint of the zero vector. Finally, the low-speed operation control of the air-conditioning compressor is realized by a position sensorless control strategy based on orthogonal voltage injection. The document "A High Frequency Injection Technique With Modified Current Reconstruction for LowSpeed ​​Sensorless Control of IPMSMs With a Single DC-Link Current Sensor" (Jing Zhao. Etc, McMaster University, 2019) uses a 6-way square wave injection method to force the synthetic vector to leave the unobservable area, ensuring smooth sampling of the current, and relying on high-frequency response current demodulation to obtain the rotor position to achieve motor sensorless control. The paper "Dynamic Performance Evaluation of Sensorless Permanent-Magnet Synchronous Motor Drives With Reduced Current Sensors" (Matteo Carpaneto. Etc., University of Genova) uses the form of adding test pulses to collect the corresponding phase currents and then uses a back-electromotive force observer to perform sensorless control of the motor.

[0005] All of the above methods can achieve sensorless motor control in a single-resistance system, but there are some problems:

[0006] (1) Some methods based on modulation improvement have problems such as increased switching times, reduced modulation ratio, and increased sampling times;

[0007] (2) The method of high-frequency square wave injection by reconstructing three-phase current through an observer has the problem of large computational complexity and is not suitable for low-performance processors that are limited by cost.

[0008] (3) The method of using a single current sensor to achieve three-phase current reconstruction by modifying the hardware topology generally requires changing the hardware circuit topology and using Hall elements for sampling on the phase line, which also limits the use of this method in conventional low-cost solutions.

[0009] (4) The method of using back electromotive force method for motor sensorless control has the problem of poor control performance at low speed, and there is a certain gap with the high-frequency signal injection method. Summary of the Invention

[0010] To address the low precision, complex methods, and dependence on motor parameters of existing high-frequency injection sensorless control for single current sensor systems, as well as the poor performance of sensorless control based on the back-electromotive force method at low speeds, the present invention provides a high-frequency square-wave injection sensorless control method for permanent magnet synchronous motors based on edge-aligned PWM modulation. This method for high-frequency square-wave injection sensorless control of permanent magnet synchronous motors based on edge-aligned PWM modulation is simple, does not increase the switching frequency of power devices, is independent of the electrical parameters of the controlled motor, and offers high control accuracy at low speeds, making it ideally suited for low-cost controllers using a single current sensor.

[0011] The technical solution adopted by the present invention to solve the above technical problems is:

[0012] A square wave injection position sensorless control method for a single resistor current sampling system comprises the following steps:

[0013] Step 1) injecting a high-frequency square wave with a frequency half the PWM frequency into the estimated d-axis of the motor. The voltage signal of the d-axis is subjected to coordinate transformation and edge-aligned PWM modulation, and then amplified by the power circuit and sent to the motor;

[0014] Step 2) A single current sensor collects bus current at an appropriate sampling point, reconstructs the current to obtain the three-phase current of the motor, and then transforms the three-phase current to obtain the current under the dq axis;

[0015] Step 3) The positive and negative q-axis current is corrected and then differentiated from the corrected q-axis current of the previous cycle, and the change in the differential q-axis current is used as the input of the phase-locked loop;

[0016] Step 4) Repeat steps 1 to 3. Finally, the phase-locked loop will output the current motor speed and the angle of the motor rotor, thus completing the identification of the motor's initial angle.

[0017] Step 5) Identify the motor polarity by injecting a positive and negative voltage into the estimated d-axis and determining whether the initial angle needs to be adjusted based on the magnitude of the positive and negative response currents and the principle of magnetic saturation;

[0018] Step 6) After obtaining the accurate initial angle of the motor, the conventional FOC control method can be used to repeat steps 1 to 4 to achieve sensorless control of the motor.

[0019] Furthermore, in step 1), the motor estimates that the angle between the d-axis and the A-axis is The angle with the true d-axis is The voltage and current relationship between the estimated dq axis coordinate system and the real dq axis coordinate system is:

[0020]

[0021]

[0022] in,

[0023]

[0024] The square wave voltage injected into the d-axis is estimated to be:

[0025]

[0026] In step 3), the voltage-current relationship in the actual dq coordinate system is expressed as:

[0027]

[0028] In the estimated dq coordinate system, the voltage and current equations become:

[0029]

[0030] when When , the estimated high-frequency response current in the dq coordinate system becomes:

[0031]

[0032] when Approaching 0,

[0033]

[0034] That is, the angle difference between the estimated dq coordinate system and the real dq coordinate system will also tend to 0, and the estimated dq coordinate system coincides with the real dq coordinate system, so as long as As the input of the phase-locked loop, the estimated speed output by the phase-locked loop is integrated to obtain the estimated dq-axis coordinate system. Then, steps 1 to 3 are repeated to estimate the motor rotor angle.

[0035] In step 5), the positive and negative voltages are injected into the estimated d-axis respectively, and then the polarity is judged according to the amplitude of the positive and negative response current on the estimated d-axis. The specific reason is that steps 1 to 3 can only achieve tends to θ e or θ e +π, it is impossible to determine whether the estimated angle converges correctly; by applying a square wave pulse of equal duration and consistent amplitude, the polarity can be identified according to the amplitude of the positive and negative response currents: if the amplitude of the positive current is greater than the amplitude of the negative current, then Converges to θ e , otherwise it converges to θ e +π.

[0036] In step 6), the specific operation of realizing the sensorless control of the motor by repeating steps 1 to 4 according to the conventional FOC control method is as follows: the dq axis given current output by the speed loop is subtracted from the dq axis feedback current after the reconstruction transformation, and the dq axis given voltage is obtained by the PI controller; then the dq axis given voltage is superimposed on the high-frequency square wave voltage and the EAPWM modulation is performed through the coordinate transformation, and finally the corresponding PWM drive motor is obtained; at the same time, steps 1 to 3 are continued to be repeated to extract the high-frequency component in the reconstructed three-phase current for demodulation of the rotor position.

[0037] The beneficial effects of the present invention are mainly manifested in:

[0038] (1) In a single current sensor system, sensorless control of high-frequency signal injection is achieved with good control performance;

[0039] (2) The method is simple to implement, does not increase the computational burden of the controller, and is suitable for low-cost controllers;

[0040] (3) The method is applicable to the existing single current sensor control system without any hardware modification and has strong adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is the relationship between the motor estimated dq coordinate system, the actual dq coordinate system and the three-phase coordinate system described in the present invention.

[0042] Figure 2 This is the timing relationship between the PWM period, sampling time and high-frequency square wave in the method of the present invention.

[0043] Figure 3 It is the overall structure and flow chart of the method of the present invention.

[0044] Figure 4 Schematic diagram of positive and negative response currents during polarity identification according to the present invention.

[0045] Figure 5 This is a schematic diagram of six conduction sequence PWMs generated by the EAPWM modulation method adopted in the present invention.

[0046] Figure 6 It is a schematic diagram of the relationship between the fundamental wave vector, high frequency vector and synthetic vector described in the present invention. Specific implementation methods

[0047] The present invention will be further described below with reference to the accompanying drawings.

[0048] Reference Figures 1 to 6 A square wave injection position sensorless control method for a single resistor current sampling system comprises the following steps:

[0049] Step 1) A high-frequency square wave with a frequency half the PWM frequency is injected into the estimated d-axis of the motor. The voltage signal of the d-axis is transformed into coordinates and modulated using edge-aligned PWM, then amplified by the power circuit and sent to the motor. Figure 1 , the motor estimates that the angle between the d-axis and the A-axis is The angle with the true d-axis is The voltage and current relationship between the estimated dq axis coordinate system and the real dq axis coordinate system is:

[0050]

[0051]

[0052] in,

[0053]

[0054] Reference Figure 2 , it is estimated that the frequency of the square wave voltage injected into the d-axis is half of the PWM period, and the amplitude is:

[0055]

[0056] Step 2) Refer to Figure 3 , a single current sensor collects the bus current at a suitable sampling point, and the three-phase current of the motor is obtained after current reconstruction. The three-phase current is then transformed into the current under the dq axis after coordinate transformation;

[0057] Step 3) Refer to Figure 3 After the positive and negative signs of the q-axis current are corrected, the difference is made between the corrected q-axis current of the previous beat, and the change in the differential q-axis current is used as the input of the phase-locked loop.

[0058] The voltage-current relationship in the actual dq coordinate system is expressed as:

[0059]

[0060] In the estimated dq coordinate system, the voltage and current equations become:

[0061]

[0062] when When , the estimated high-frequency response current in the dq coordinate system becomes:

[0063]

[0064] when Approaching 0,

[0065]

[0066] That is, the angle difference between the estimated dq coordinate system and the real dq coordinate system will also tend to 0, and the estimated dq coordinate system coincides with the real dq coordinate system. As the input of the phase-locked loop, the estimated speed output by the phase-locked loop is integrated to obtain the estimated dq axis coordinate system, and then steps 1 to 3 are repeated to estimate the motor rotor angle;

[0067] Step 4) Repeat steps 1 to 3. Finally, the phase-locked loop will output the current motor speed and the angle of the motor rotor, thus completing the identification of the motor's initial angle.

[0068] Step 5) Refer to Figure 4 , identification of motor polarity, by injecting a positive and negative voltage into the estimated d-axis, and judging whether the initial angle needs to be adjusted based on the amplitude of the positive and negative response current and the principle of magnetic saturation. The specific reason is that steps 1 to 3 can only be achieved tends to θ e or θ e +π, it is impossible to determine whether the estimated angle converges correctly; by applying a square wave pulse of equal duration and consistent amplitude, the polarity can be identified according to the amplitude of the positive and negative response currents: if the amplitude of the positive current is greater than the amplitude of the negative current, then Converges to θ e , otherwise it converges to θ e +π.

[0069] Step 6) Refer to Figure 3 、 5 After obtaining the accurate initial motor angle, the conventional FOC control method is repeated by steps 1 through 4 to achieve sensorless motor control. Specifically, the dq-axis set current output by the velocity loop is subtracted from the reconstructed dq-axis feedback current, which is then passed through a PI controller to obtain the dq-axis set voltage. The dq-axis set voltage is then superimposed on a high-frequency square wave voltage and subjected to coordinate transformation for EAPWM modulation, ultimately resulting in the corresponding PWM drive motor. Simultaneously, steps 1 through 3 are repeated to extract the high-frequency components of the reconstructed three-phase current for rotor position demodulation. Figure 5 The PWM forms under the six conduction sequences of EAPWM adopted by the present invention are shown. Figure 6 The relationship between the fundamental wave vector, high-frequency vector and synthetic vector when the present invention performs sensorless motor control is demonstrated; further, the EAPWM modulation process adopted by the present invention is explained as follows: a corresponding conduction form of EAPWM is selected according to the sector where the fundamental wave vector is located, and then the duration of each phase PWM is calculated by the synthetic vector to complete the PWM modulation.

[0070] The embodiments of this specification are merely examples of implementations of the invention and are provided for illustrative purposes only. The scope of protection of the present invention should not be considered limited to the specific embodiments described in these embodiments. The scope of protection of the present invention also extends to equivalent technical means that can be conceived by a person of ordinary skill in the art based on the invention.

Claims

1. A square wave injection position sensorless control method for a single current sensor system, characterized in that: The control method comprises the following steps: Step 1) injecting a high-frequency square wave with a frequency half the PWM frequency into the estimated d-axis of the motor, and the voltage signal of the d-axis is coordinate transformed and edge-aligned PWM modulated, and then amplified by the power circuit and sent to the motor; Step 2) A single current sensor collects bus current at an appropriate sampling point, and after current reconstruction, the three-phase current of the motor is obtained, and then the three-phase current is coordinate transformed to obtain the current under the dq axis; Step 3) The positive and negative q-axis current is corrected and then differentiated from the corrected q-axis current of the previous cycle, and the change in the differential q-axis current is used as the input of the phase-locked loop; Step 4) Repeat steps 1) to 3), and finally the phase-locked loop will output the current motor speed and the angle of the motor rotor, thus completing the identification of the motor's initial angle; Step 5) Identify the motor polarity by injecting a positive or negative voltage into the estimated d-axis. Based on the magnitude of the positive or negative response current and the principle of magnetic saturation, determine whether the initial angle needs to be adjusted. Step 6) After obtaining the accurate initial angle of the motor, repeat steps 1) to 4) according to the FOC control method to achieve sensorless control of the motor; In step 1), the motor estimates the angle between the d-axis and the A-axis to be The angle with the true d-axis is The voltage and current relationship between the estimated dq axis coordinate system and the real dq axis coordinate system is: in, The square wave voltage injected into the d-axis is estimated to be:

2. The square wave injection position sensorless control method for a single current sensor system according to claim 1, characterized in that: In step 4), the voltage-current relationship in the actual dq coordinate system is expressed as: In the estimated dq coordinate system, the voltage and current equations become: when When , the estimated high-frequency response current in the dq coordinate system becomes: when Approaching 0, That is, the angle difference between the estimated dq coordinate system and the real dq coordinate system will also tend to 0, and the estimated dq coordinate system coincides with the real dq coordinate system, so as long as As the input of the phase-locked loop, the estimated speed output by the phase-locked loop is integrated to obtain the estimated dq axis coordinate system, and then steps 1) to 3) are repeated to estimate the motor rotor angle.

3. The square wave injection position sensorless control method for a single current sensor system according to claim 1 or 2, characterized in that: In step 5), the positive and negative voltages are injected into the estimated d-axis respectively, and then the polarity is judged according to the amplitude of the positive and negative response current on the estimated d-axis. The specific reason is that steps 1) to 3) can only be achieved. tends to θ e or θ e +π, it is impossible to determine whether the estimated angle converges correctly. By applying a square wave pulse of equal duration and amplitude, the polarity can be identified according to the amplitude of the positive and negative response currents: if the amplitude of the positive current is greater than the amplitude of the negative current, then Converges to θ e , otherwise it converges to θ e +π.

4. The square wave injection position sensorless control method for a single current sensor system according to claim 1 or 2, characterized in that: In step 6), the specific operation of realizing the sensorless control of the motor by repeating steps 1) to 4) according to the conventional FOC control method is as follows: the dq axis given current output by the speed loop is subtracted from the dq axis feedback current after the reconstruction transformation, and the dq axis given voltage is obtained by the PI controller; then the dq axis given voltage is superimposed on the high-frequency square wave voltage and the EAPWM modulation is performed through the coordinate transformation, and finally the corresponding PWM drive motor is obtained; at the same time, steps 1) to 3) are continued to be repeated to extract the high-frequency component in the reconstructed three-phase current for demodulation of the rotor position.

5. The square wave injection position sensorless control method for a single current sensor system according to claim 4, characterized in that: In step 6), the EAPWM modulation process is as follows: selecting a corresponding EAPWM conduction form according to the sector where the fundamental wave vector is located, and then calculating the duration of each phase PWM by synthesizing the vector to complete the PWM modulation.

Images