基于改进蚁群算法的独立分量谐波检测方法  

Independent Component Analysis Based on Improved Ant Colony Optimization Algorithm for Harmonic Detection

作　　者：张立臣[1] 金运策[1] 徐步权[1] 亢朋朋[1] 耿书超[1] 

机构地区：[1]河海大学能源与电气学院,江苏南京211100

出　　处：《水电能源科学》2013年第6期218-221,共4页Water Resources and Power

摘　　要：针对电网谐波检测高精度、实时性、高效率的要求,提出了一种基于改进蚁群算法与独立分量相结合的检测算法,采用数字低通滤波器(FIR)-拉格朗日反插值获得精确基波频率,应用独立分量法分析处理源信号,并建立基于峭度的非高斯最大化目标函数,通过改进蚁群算法求得最优解,很好地解决了电网频率非固定值及传统蚁群算法易局部收敛、收敛速度慢等难题。该方法还可检测间谐波,模拟仿真验证了方法的有效性和优越性。In view of high precision,real-time and high efficiency demands of harmonic detection,a new detection algorithm is proposed by combining improved ant colony optimization algorithm with independent component analysis.The method gains accurate fundamental wave frequency with FIR-Lagrange anti-interpolation.And then independent component analysis is used to handle the source signal.At the same time,the objective function is maximized based on the kurtosis non-Gaussian.Finally,the global optimal solution can be obtained by an improved ACO algorithm.This algorithm solves two difficult problems.One is that actual working frequency of grid is a fixed value;the other is that the traditional ACO algorithm is easy to fall in local optima and the convergence rate is too slow.Furthermore,independent component analysis can detect inter-harmonics.The superiority of this algorithm is verified by the simulation results.

关 键 词：改进蚁群算法 非高斯最大化 拉格朗日反插值 独立分量 间谐波 

分 类 号：TM93[电气工程—电力电子与电力传动]