Study Of The Control Strategy Of Active Power Filter With Reactive Power Compensation Function

Active power filter is a kind of power electronic devices that can achieve harmonic dynamic compensation, it has become an important means of current harmonic suppression. This paper study the key technology of APF.In order to fully employ the capacity of APF, realise the regulation of reactive power in the grid, a deep analysis of a reactive power compensation capacity tuning method based on APF has been presented. The amount of reactive power compensation is determined by the rated capacity of APF, in this way, equipment utilization has maximize improved. Simulation results show that the tuning method can realise dynamic reactive power compensation.The filter inductor parameter constantly changing will reduce the performance or even directly affect the stability of the system. On the basic principle of the least square methd, this paper propose an online parameter identification method in dq coordinate system. Simulation results show that the proposed method can eliminate the glitch in traditional identification method and improve the accuracy of identification.Based on traditional repetitive control, a novel fast repetitive control is proposed to improve the dynamic response speed and exactly track the changed harmonic current. This paper treats the harmonic current composition as a point of penetration, adopts one-sixth of the fundamental period as the delay time, as a result, the response time has been cut down from 1.2 fundamental period to one-fourth fundamental period. Simulation results show that the fast repetitive control significantly outperforms the traditional repetitive control in terms of the dynamic response speed. The proposed fast repetitive control has extensive application area, stronger robustness and better filtering property.At last, this paper treats the chosen of quality function as a point of penetration, proposes a new quality function which at the basic of outputvoltage tracking error. The proposed method converts indirect optimization to direct optimization, by this way, the harmonic current tracking effort has been improved. At the same time, a novel method to decrease mean switching frequence is proposed, by judging the current tracking error, the method can improve the chosen probability of adjacent vectors.