Research on Optimization Method of Repetitive Controller for Core Loss Testing
摘要
Core loss test system for electric machines need to obtain sinusoidal excitation current in over wide range of amplitude and frequency, and the load has the characteristic of saturation nonlinearity that makes the model of the plant vary with the reference current. Therefore, AC current tracking and harmonic suppression algorithms have become the key of related research. In this paper, the control structure of proportional-integral (PI) controller in parallel with repetitive controller (RC) is applied. On this basis, a finite impulse response (FIR) filter designed by Lagrange interpolation is used to equate the fractional-order delay term to improve the controller performance at particular frequencies, and a weighting factor is introduced to optimize the filter coefficients. Meanwhile, an integral control algorithm based on discrete Fourier transform (DFT) is proposed. The DFT is used to extract the current fundamental amplitude in real time, and the integral unit is used to eliminate the steady-state error. The experimental results show that the modified FIR filter can reduce the total harmonic distortion (THD) of current in a wide frequency range, and the DFT-based integral control structure realizes the stable control of the nonlinear plant in a wide amplitude range. Thereby, the frequency adaptive operation is achieved and the robustness of the system is improved.