基于GA-ELM的电压暂降源识别研究

Research on recognition of voltage sag source based on GA-ELM

针对传统电压暂降源识别方法分类时间长、准确率不高等问题,提出了一种基于广义S变换(GST)和遗传算法(GA)优化极限学习机(ELM)的电压暂降源识别方法。先利用广义S变换的模时频矩阵有效提取出电压暂降的起止时刻、暂降深度、相位跳变等特征量,再采用遗传算法优化ELM的输入权值和隐含层阈值,构建基于GA-ELM的电压暂降源识别模型,实现对电压暂降源的识别。通过MATLAB/SIMULINK仿真,对比GA-ELM、ELM、BP神经网络对电压暂降源的识别结果,验证了采用GA-ELM的电压暂降源识别的准确率要高于采用原始ELM和BP神经网络。

Aiming at the problem that the traditional voltage sag source identification method has long classification time and low accuracy,a voltage sag source identification method based on generalized S transform(GST)and genetic algorithm(GA)to optimize the extreme learning machine is proposed.Firstly,the model matrix of generalized S transform is used to extract the feature quantities such as start and stop time,sag depth and phase jump of voltage sag.Then the genetic algorithm is used to optimize the input weight and hidden layer threshold of ELM to build voltage sag source recognition model based on GA-ELM,and realize voltage sag source recogmition.By MATLAB/SIMULINK simulation,compare the identification results of GA-ELM,ELM and BP neural networks to identify the voltage sag source.It is verified that the accuracy of voltage sag source identification using genetic algorithm to optimize ELM is higher than that of original ELM and BP neural networks.