基于差分进化Takagi-Sugeno模糊神经网络的电网故障诊断模型  

Fault Diagnosis Model of Power Grid Based on Differential Evolution Takagi-Sugeno Fuzzy Neural Network

作　　者：吴杨 姚刚 徐胜 杜江 陈锦龙 李长松 熊国江 WU Yang;YAO Gang;XU Sheng;DU Jiang;CHEN Jinlong;LI Changsong;XIONG Guojiang(Power Grid Dispatching and Control Center,Guizhou Power Grid Co.,Ltd.,Guiyang 550002,China;College of Electrical Engineering,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州电网有限责任公司电力调度控制中心,贵州贵阳550002 [2]贵州大学电气工程学院,贵州贵阳550025

出　　处：《机械与电子》2023年第11期10-16,共7页Machinery & Electronics

基　　金：国家自然科学基金资助项目(51907035)。

摘　　要：为有效处理电网故障中存在的不确定性,提出一种基于改进差分进化Takagi-Sugeno模糊神经网络的电网故障诊断模型。该模型基于分布式并行处理的思路,根据配置的继电保护和断路器对每个元件而非整个电网建立相应故障诊断模型。为提高诊断模型的准确性,对差分进化算法的缩放因子和交叉率进行自适应改进,并将改进算法用于优化各模型的前件参数和后件参数。仿真结果表明,与其他神经网络相比,该模型能够成功诊断存在拒动、误动的复杂故障,提高了电网故障诊断的容错性。To effectively deal with the uncertainties in grid faults,an improved differential evolution Takagi-Sugeno fuzzy neural network grid fault diagnosis method is presented.Based on the idea of distributed parallel processing,the diagnosis model is constructed for each element instead of the whole grid according to the configured protective relays and circuit breakers.To improve the model accuracy,both crossover rate and scaling factor of differential evolution are modified adaptively.Then the improved differential evolution is utilized to optimize the structure parameters and consequent parameters of the diagnosis model.Simulation results indicate that the proposed model can successfully diagnose complex faults with mal operation and refused operation and improve the fault tolerance of fault diagnosis.

关 键 词：电网故障诊断 Takagi-Sugeno模糊神经网络 差分进化 并行诊断 容错性 

分 类 号：TM76[电气工程—电力系统及自动化]