有源电力滤波器改进无差拍-重复控制策略

Improved Deadbeat-Repetitive Control Strategy for Active Power Filter

首先推导了三相有源电力滤波器(APF)的连续时间数学模型和离散数学模型,介绍了APF的整体控制结构。对传统无差拍控制进行深入分析,发现考虑AD采样和DSP数字计算延时后,传统无差拍控制并不能真正实现控制上的无差拍,控制精度不高。为此,提出一种改进无差拍控制,两个闭环极点均位于原点,理论上可以实现真正的无差拍控制。为了提高改进无差拍控制的抗扰动性能和对电感参数变化的鲁棒性,提出一种改进无差拍-重复控制策略。该策略能有效消除AD采样误差、死区时间、电感参数偏移对控制性能的影响,具有良好的谐波补偿效果。实验结果验证了该策略的正确性和有效性。

Continuous time and discrete mathematic model of active power filter(APF)are derived,and then the whole control structure is introduced.A deep analysis of traditional deadbeat control is done.It is found that traditional deadbeat control cannot achieve the real deadbeat control and its control accuracy is not high if the delay of AD sampling and DSP digital calculation are considered.In order to solve this issue,an improved deadbeat control is presented so that two poles of closed-loop transfer function both locate origin and the real deadbeat control can be achieved theoretically.To improve disturbance resistant performance and robustness to deviation of inductance parameter,an improved deadbeat-repetitive control strategy is proposed.This strategy can eliminate the negative effect of AD sampling error,dead time,deviation of inductance parameter and achieve a good harmonic compensation performance.Experimental results validate the correctness and effectiveness of the proposed control strategy.