APF特定次谐波智能检测方法的研究  

Specific Harmonic Detection Algorithm for Active Power Filter

作　　者：马立新[1] 王晓丹[1] 王月晓[1] 王玉珠[1] 

机构地区：[1]上海理工大学电气工程系,上海200093

出　　处：《控制工程》2013年第2期352-356,共5页Control Engineering of China

基　　金：国家科技部政府间科技合作项目(2009014);上海市高等学校高地建设项目(5209302001)

摘　　要：有源电力滤波器(APF,Active Power Filter)是当今重要的谐波治理和无功补偿装置,APF关键的环节是实时准确地检测出谐波电流。针对传统基于瞬时无功理论的ip-iq谐波检测方法实时性差、对系统的补偿能力要求很高等不足,提出了基于BP神经网络与锁相环相结合提取特定次谐波的方案,权值调整采用BFGS拟牛顿算法,该算法可检测出基波、各次谐波分量的幅值和相角,且神经网络的并行运算能力强大,方便应用于三相电路中。改进的特定次谐波检测系统较传统的LC滤波器和并联型APF的组合调节策略响应速度高,系统结构简单,且不易出现高频谐波。通过Matlab仿真实验证明了该算法的可行性和良好的控制效果。Active power filter（APF） is an important power system harmonic control and reactive power compensation device. The ip-iq algorithm, based on the instantaneous reactive power theory with real-time harmonic detection, had been successfully applied in many real-time harmonic detection systems. However, it has a poor real-time performance, dependents overly on the system compensation property. This paper proposed the implementation of BP neural network to extract specific harmonics. The weight adjustment used the BFGS quasi-Newton algorithm, which can accurately detect the fundamental, harmonic component＇s phase amplitude and angle. The specific harmonic detection system designed and implemented with simple structure, high possessing and precision, can be well used in the three-phase circuit. Matlab simulations indicate the correctness of compensated currents examination and validity performance of BP neural network algorithm.

关 键 词：有源电力滤波器 BP神经网络 特定次谐波 MATLAB仿真 

分 类 号：TP27[自动化与计算机技术—检测技术与自动化装置]