The Research On Advanced Control Methods For High Power Shunt APF

Active power filter has been widely studied since 1980s. The shunt active power filter is the popularly used one due to its easy connection and protection. Several advanced control strategies are studied in this paper in order to enhance the compensation performances of SAPRThe paper analyzed the mathematical model of the SAPF under the three-phase and d-q coordinates, and proved that the transfer function is the same under the two coordinates while the disturbance is different. The LCL-filter largely reduces the volume of the system and has good filtering performance, but the system is third order and hard to control. The L-filter is simple and easy to control, but the volume of the system and the compensation performance is not satisfying. The repetitive control, the double-loop control and the selective harmonic control pointed to the LCL-filter, while the multi-sampling based on the L-filter.The repetitive control strategy which could acquire excellent steady-state performance is proposed based on internal model even though the dynamic performances of PI control strategy is superior to that of repetitive control strategy. The strategy has been verified by simulation and experiment through the paper. Moreover, this paper puts forward a double-loop composite controller composed by a PI inner loop and repetitive controller external loop to improve the compensation precision of SAPF, which could not only reduce the THD of grid current effectively but also guarantee excellent dynamic response performance.Selective harmonic control just compensates the lower frequency harmonics and neglects the higher ones, or some of the specified harmonics. So it can also avoid the harmonics that cause the resonant of the grid system, which guarantees the stability. The simulation results proved that the selective harmonic control can compensate the specified harmonics and the experiment results proved the Discrete Fourier Transformation (DFT) can extract the harmonic signals correctly by DSP.The discretization of the controller will impact its original performance and even influence the stability. Raising the sampling rate makes the digital controller's performance closer to the corresponding analog controller. But the sampling rate is once or twice of the switch rate concerned about the traditional sampling method. Since the switching rate is constrained by the transistor especially in the high power application, the sampling rate can't be raised without confine. The multi-sampling can resolve the contradiction. The paper has done research about it and resolved the pulse loss problem while realized by DSP. So this method is fit for the high power domain.Through the study above, some results have been achieved. Firstly, the repetitive control and the double-loop control strategy could deal with the contradiction between compensation precision and dynamic response of the SAPF effectively. Secondly, the selective harmonic control makes the SAPF applicable for special circumstance of harmonics compensate. Lastly, the multi-sampling deletes the contradiction between the sampling rate and the device in high power domain.