Study On APF Compensation Current Control Strategy Based On Nonlinear Optimal Predictive Control Algorithm

The DC side of the Shunt Active Power Filter generally uses power capacitor as its energy storage element. In the process of controlling active power filter’s compensation current tracking harmonic current soon, there is the active loss which caused by the line resistance and switch loss, which will lead to the DC bus voltage’s fluctuation or under voltage, furthermore, it will affect the compensation current tracking’s performance. Therefore, it is necessary to improve the control of both DC bus voltage and the active power filter’s compensation current, and the good cooperation between two of them will be directly related to the system for real-time tracking characteristics of its harmonic current, which plays a decisive role to improve the overall performance of active power filter.This paper focuses on the controlling method of active power filter’s compensation current, although the hysteresis current controlling is the representative of the nonlinear control, and its hardware is easily to realize, but its switching frequency is not fixed which will cause the high loss. According to the problem of the current control, this paper puts forward a method of the compensation current control which is based on predictive control firstly. The basic idea of predictive control is to predict the output state of the system in the future firstly, and then to determine control function of the current. In predictive control, the optimization is carried out repeatedly. It is the meaning of the rolling optimization. Predictive control is to predict system’s output, and then to control. So it has the foresight, and also the advantage of switching frequency which is fixed.Predictive control is the effective way to solve the problems of the system time delay and parameter drift and so on, however, it does not fully consider the nonlinear characteristics of PWM rectifier. In essence, it is still a linear control method. So how to better solve such problems is the key to study. In reference on the basis of a large number of literatures, this paper presents a new predictive control algorithm, which is the nonlinear optimal predictive control. Its essence is applied nonlinear predictive control method based on differential geometry theory to the active power filter. Its basic principle is to get the input current value of the next control cycle and its reference value by using Taylor expansion, then according to a certain amount of control performance index to calculate the current cycle, so as to effectively eliminate the negative effects of delay. In order to further improve tracking precision of the compensation current, I considered the state variables and deviation as the target instruction value of the equation, and introduce into integral equations in the performance index, draw a new kind of nonlinear optimal control law. In addition, I got a detailed analysis on the essence of the nonlinear predictive control, and explain the important reason why the compensation current tracking performance is superior.In order to verify the correctness and superiority of control strategy mentioned in this paper, I carried on the simulation and experiment. The results show that the input current is sinusoidal, high power factor. The distortion rate is small. The transient performance is superior. It can effectively compensate the influence of delay. Therefore, it has important theoretical and practical significance about a good control algorithm research for improving the active power filter’s performance of the system.