This blog is PMSM position sensorless control for suppressing current harmonics to improve motor system performance
Flux linkage harmonics and dead zone effects are important factors affecting current harmonics, and the resulting voltage error harmonics are periodic disturbances of 6 times the fundamental frequency in permanent magnet synchronous motor systems, which will cause obvious 5th and 7th order Sub-phase current harmonics, there are a variety of control strategies to suppress the current harmonics, among them, the inverter dead zone effect compensation
Compensation is an effective means to suppress current harmonics. Literature [13] proposes an adaptive filter voltage compensation method, which uses the filter to obtain the 6th harmonic sequence information from the d-axis and q-axis currents, and calculates the dead zone effect compensation voltage from the voltage equation. The dead zone compensation method combining the neural network bandpass filter and the extended Kalman filter uses the extended Kalman filter to improve the traditional dead zone compensation algorithm, and uses the neural network bandpass filter to extract the d-axis and q-axis current harmonics. The dead zone compensation algorithm calculates the compensation voltage to further improve the dead zone compensation performance. However, the above dead zone effect compensation method based on voltage calculation is not suitable for current harmonic suppression caused by factors other than the dead zone. To suppress current harmonics caused by periodic disturbances, repetitive control, iterative learning, and harmonic injection methods are all effective to a certain extent. Using repetitive control and introducing a repetitive model to realize asymptotic tracking of periodic disturbances can suppress the harmonic components of periodic disturbances, but the number of samples is large and the convergence speed is slow. Iterative learning is combined with traditional proportional integral (PI) control to suppress current harmonics, and the error signal of the previous cycle is used to iteratively adjust the current cycle control signal, but it is sensitive to system disturbance and has poor robustness. The harmonic injection method has a more obvious effect of suppressing specific frequency harmonics. Through harmonic current extraction and harmonic voltage injection, the 6th torque current harmonic can be significantly reduced. The d-axis and q-axis currents are obtained by using multiple synchronous rotating coordinate transformations. After extracting the current harmonics, the compensation voltage is calculated based on the PMSM model containing harmonics, and injected into the output terminals of the d-axis and q-axis current controllers respectively. The harmonic suppression The effect largely depends on the accuracy of current harmonic extraction and motor model. Resonant control utilizes the infinite gain characteristic at the resonant frequency to suppress current harmonics under periodic disturbances. Compared with the ideal resonant controller, the quasi-resonant (QR) controller increases the bandwidth at the resonant frequency, reduces the frequency sensitivity of the system and improves the stability, and has better 5th and 7th phase current resonance Wave suppression effect, so it is more suitable for motor systems with speed fluctuations. The combination of resonance control and predictive control can obtain better overall performance of the system[18]. The typical structure is a proportional integral resonant (PIR) controller, because the PI
The permanent magnet flux linkage expression considering harmonic disturbance is as follows. The
voltage error caused by the inverter dead zone effect disturbance is:
where: Δud, Δuq are the voltage errors of the d and q axes respectively; Td is the dead zone time; Udc is DC bus voltage; Ts is the sampling time.
In equations (2) and (3), the flux linkage and voltage harmonic disturbance with a frequency of 6nω will lead to current harmonics. Since the harmonic amplitude decreases with the increase of the harmonic order, this paper focuses on the dq coordinate system 6th harmonic.
The phase currents ia, ib and ic considering flux linkage and voltage harmonic disturbance are:
From the above formula, the frequencies of the 5th and 7th harmonics of the phase current are 5ω and 7ω respectively, and the corresponding frequency of the 6th harmonic in the dq coordinate system is 6ω.
Start designing QRADRC
According to the formula (1), the q-axis ideal current state equation is:
set the state variables x1 and x2, x1=iq, x2=f,
in order to effectively suppress the harmonics in the bandwidth range near the resonant frequency, this paper applies the QR controller, its The transfer function is:
the most important control control block diagram:
Then use QRADRC to positionless sensor: (SMO based on arctangent function)
Where QRADRC:
GR(s):
actual position vs. estimated position:
error between actual position and estimated position:
actual speed vs. estimated speed:
error between actual speed and estimated speed: