Position Sensorless Control of Permanent Magnet Synchronous Motor Based on Rotating High Frequency Injection Method
1. Principle explanation
PMSM position sensorless control is mainly divided into two categories: one is to use the relationship between back electromotive force and electrical angular velocity in the medium and high speed range, and obtain rotor position information by calculating back electromotive force, such as flux observer, model reference adaptive method, extended Kalman filter and sliding mode observer. The other is the high-frequency injection method using the salient pole effect of the motor, including the pulse vibration high-frequency current injection method, the rotating high-frequency voltage injection method and the pulse vibration high-frequency voltage injection method. The latter two methods have nothing to do with the basic mathematical equations. It is insensitive to motor parameters and has better robustness.
In the zero-low speed environment, the back electromotive force signal of the relevant position in the fundamental wave model is very weak, and the signal-to-noise ratio is too low when extracting. At this time, the high-frequency excitation model of the motor can be used to realize position sensorless control, and then extract by injecting high-frequency signals The position signal in the high-frequency response is sufficient, and the high-frequency injection method mainly includes the rotating high-frequency voltage injection method and the pulse vibration high-frequency voltage injection method.
The rotating high-frequency voltage injection method injects high-frequency voltage signals into the α-axis and β-axis respectively under the two-phase stationary coordinate system (a-β), and then detects the high-frequency current response of the motor through a band-pass filter (BPF). The frequency current response contains positive sequence components and negative sequence components. Only the phase of the negative sequence components can extract the position signal, and then use the synchronous shafting high-pass filter (SFF) to filter out the positive sequence components, and finally obtain the position and speed information of the rotor. , this method is easier to implement and has better performance at low speeds [. The rotating high-frequency voltage injection method has better immunity to interference at low speeds, but it is also affected by filter delay and DSP signal acquisition and control delay, and the estimated phase of the rotor will be delayed. For the delay problem, a compensation algorithm is usually used to extract the error amount from the high-frequency positive sequence current component to compensate the rotor position identification error to improve the position identification accuracy. The permanent magnet synchronous motor based on the rotating high-frequency injection method has
no The block diagram of the position sensor control system is as follows:
This block diagram is built in matlab/simulink as shown in the following figure:
In order to extract the rotor position information in the negative phase sequence high frequency current response, the fundamental frequency current, low order harmonic current, PWM switching frequency harmonic current and positive phase sequence high frequency current in the motor terminal current must be well filtered out. Signal. The amplitude difference between the fundamental current and the high-frequency current is very large, and the carrier frequency is much higher than the injected high-frequency frequency, both of which can be filtered out by a conventional band-pass filter (BPF). The rotation direction of the positive phase sequence component of the carrier current is opposite to that of the negative phase sequence component, so the positive sequence current component can be filtered out by the synchronous shafting high-pass filter (Synchronous Frame Filter, SFF).
The high-pass filter of the synchronous shaft system transforms the high-frequency current vector into a reference coordinate system that rotates synchronously with the injected high-frequency voltage vector through coordinate transformation. A high pass filter removes it.
After filtering, the remaining signal is the negative phase sequence high frequency current component, which is a useful signal that can be used to track salient poles. Its vector expression is: In order to extract the rotor position information
from the negative sequence high frequency current component , the rotor position tracking observer method is commonly used at present. Its implementation block diagram is as follows:
Two, simulink block diagram module introduction
2.1. Speed loop module
2.2. Current loop module (Iq is the same as Id)
2.3. Inverter module
The subsystem2 module is:
2.4, PMSM module (general)
2.5. SFF module (synchronous shafting high-pass filter)
in
2.6. Heterodyne module (the name in the block diagram is method)
in
2.7. Observer module (observer)
2.8. Injected high-frequency voltage module
One of them is sin and the other is cos
cos parameter:
sin parameter:
2.9. Filter parameters of the block diagram
3. Simulation waveform diagram
3.1. Speed error diagram
3.2. Actual speed and estimated speed
3.3. Rotor error
3.4. Actual rotor and estimated rotor
Four. Summary
Generally speaking, it is still good, but the error is relatively large.