基于模糊推理和李亚普诺夫理论的有源电力滤波器控制

Fuzzy Control and Lyapunov-Based Stability Control of the Active Power Filter

针对负载和参数变化带来有源电力滤波器(active power filter,APF)精确数学模型难以获得和系统稳定性问题,在APF的直流电压控制中引入模糊比例积分(proportional-integral,PI)控制器,在电流控制中引入基于李亚普诺夫(Lyapunov)稳定性理论的控制方法,保证了APF的控制精度和稳定性。模糊PI控制器通过模糊化、模糊推理和解模糊,根据直流电压误差和误差的变化率对控制参数进行在线调整。通过Matlab/Simulink分别对4种控制方法进行对比实验,在负载和系统参数变化的情况下,提出的方法比使用PI和滞环控制方法的源电流谐波总畸变率(total harmonic distortion,THD)明显减少;在负载时变和系统参数变化的情况下,应用PI和滞环控制的APF运行不稳定,而应用作者提出的控制方法的APF运行稳定,源电流THD最低,电压超调最小。

Active power filter（APF） usually appears unstable and it is quite difficult to determine its exact mathematical model , when the load and parameters are changing. In order to treat this problem, fuzzy proportional-integral （PI） controller and Lyapunov-based stability theory were introduced into the direct current（DC） side voltage control and current control of APF to guarantee the precision and stability of APF. Through fuzzification, fuzzy inference and defuzzifier, fuzzy PI controller adjusted the parameters of the PI con- troller according to the error and error varying rate of the DC side voltage. After testing on four different control methods with Matlab/ Simulink, the distinctive effectiveness of the proposed method was therefore confirmed. The results showed teat, when the load and pa- rameter change abruptly, compared with PI and hysteresis controllers, the total harmonic distortion （THD） of source current after ap- plying the proposed method is decreased obviously. When the load is time-varying and parameter changes abruptly, both the THD of source current and overshoot of the DC capacitor voltage are much smaller in the proposed method, while the APF runs unsteadily after applying PI and hysteresis controllers.