神经网络控制的三相并联有源电力滤波器设计

Design of Three-phase Shunt Active Power Filter by Current Control

针对现有各种谐波检测方法的不足,在分析了滤波器的工作原理和传统的神经网络控制算法之后,提出了一种新的神经网络控制算法,设计了一种应用于电力系统中补偿无功和抑制谐波的三相并联有源电力滤波器(APF),该滤波器由一个三相PWM电压源逆变器及其控制电路组成。将负载电流、直流侧电容电压与系统电压输入改进的神经网络,计算出并联有源电力滤波器的参考电流,参考电流输入滞环电流控制器获得逆变器的触发脉冲,从而得到补偿电流。通过Matlab仿真表明,由改进的神经网络控制的滤波器,在负载变化及其电流变化情况下都有良好的性能。同时,通过数字信号处理器实现的系统实验验证了该方案的正确性。

This paper presents a three-phase shunt active power filter（APF）to compensate the reactive power and reduce harmonics in power system. The active filter is based on a 3-phase pulse width modulated （PWM） voltage source inverter （VSI） and a control circuit. Operating principles of APF and conventional neural network （ANN） algorithm are analyzed, and then, a modified adaptive artificial neural network is presented. By inputting sensing load current, dc capacitor voltage and source voltage, reference current of APF is computed through modifying adaptive artificial neural network which enhances the speed of estimating the harmonic components and voltage sags in the system. Reference currents are input into modified hysteretic current controller to generate the firing pulses of the voltage source inverter,and then the compensating currents are gotten. The performances of the proposed APF are verified through simulation studies using Matlab under different loads. The simulation results show that the APF with adaptive neural network control is able to adapt itself to the changes in the non-linear load currents, respectively. The system is also implemented using a high speed Digital Signal Processor （DSP）, and experimental results are presented to confirm the validity of the scheme.