基于神经网络的直流微电网故障诊断方法研究

Research on fault diagnosis method of parallel microgrid based on neural network

传统诊断方法已经无法满足直流微电网的诊断需求,针对直流微电网子系统结构复杂,诊断效率低的问题,在直流微电网系统结构基础上,提出一种基于信息融合判别的直流微电网诊断拓扑结构,通过采集直流微电网系统实时数据,形成训练数据,利用神经网络算法进行直流微电网的故障诊断,在MATLAB/Simulink仿真建模平台构建了基于神经网络优化算法的直流微电网故障诊断模型,通过仿真结果,采用神经网络算法提升了故障诊断效率,验证了算法的可行性。

the traditional diagnosis method has been unable to meet the diagnostic needs of parallel microgrid.Aiming at the problems of complex structure and low diagnosis efficiency of parallel micro grid subsystem,a diagnosis topology structure of parallel microgrid based on information fusion judgment is proposed based on the structure of parallel micro grid system.Through collecting real-time data of parallel micro grid system,training data is formed and neural network is used Network algorithm is used for fault diagnosis of parallel microgrid.The fault diagnosis model of parallel microgrid based on neural network optimization algorithm is constructed in MATLAB/Simulink simulation modeling platform.Through the simulation results,the fault diagnosis efficiency is improved by using neural network algorithm,and the feasibility of the algorithm is verified.