摘要
针对水电机组振动故障诊断准确率低的问题,提出一种基于粒子群优化(PSO)算法优化长短期记忆(LSTM)网络的水电机组振动故障诊断方法。首先采用PSO算法对LSTM网络层节点数与dropout值进行优化,然后采用优化的LSTM网络对水电机组振动故障诊断。结果表明,该方法可准确诊断不同类型的水电机组故障,平均诊断准确率达99.55%,相较于标准的LSTM网络和PSO-SVM的故障诊断方法,该方法具有更快的收敛速度和更高的识别准确率,可用于实际水电机组振动的故障诊断。
A vibration fault diagnosis model for hydroelectric units is proposed based on particle swarm optimization(PSO)algorithm to optimize long short term memory(LSTM)networks in response to the problem of low accuracy in vibration diagnosis and recognition of hydroelectric units.This model uses PSO to optimize the number of nodes and dropout values in the LSTM network layer,and builds a vibration fault diagnosis model for hydroelectric units using the optimized LSTM network for fault diagnosis.The results show that the model can accurately diagnose faults of different types of hydropower units,with an average diagnostic accuracy of 99.55%.Compared with the fault diagnosis models built by LSTM network and PSO-SVM,it has faster convergence speed and higher recognition accuracy,and can be used for fault diagnosis of actual hydropower unit vibration.
作者
罗玮
陈媛
Luo Wei;Chen Yuan(School of Civil Engineering,Tsinghua University,Beijing,100084,China;CHN Energy Dadu River Big Data Services Co.,Ltd.,Sichuan Chengdu,610016,China)
出处
《机械设计与制造工程》
2024年第9期94-98,共5页
Machine Design and Manufacturing Engineering
