基于TPE-Informer模型的铁路继电器寿命预测

Life Prediction of Railway Relays Based on TPE-Informer Model

针对如何准确、高效地预测铁路继电器寿命问题,提出一种基于TPE-Informer模型的铁路继电器寿命预测方法。首先,搭建铁路继电器电寿命试验平台,获取其整个生命周期的退化数据,从中提取出能够反映其运行状态的特征参数;其次,将提取的特征参数输入Informer模型进行训练,利用多头稀疏自注意力机制挖掘特征信息前后状态的关联性;最后,利用非标准贝叶斯优化算法(Tree-structured parzen estimator,TPE)优化Informer模型超参数,以获得更好的预测性能。采用试验平台采集数据对模型验证,并与其他三种深度学习算法进行结果对比。试验结果表明,所提预测模型比RNN、LSTM和Informer模型预测精度高,平均精度达到96.52%,误差率小,稳定性好,证明了该预测模型应用的可行性。

Aiming at the problem of how to predict the remaining life of railway relay accurately and efficiently,a life prediction method of railway relay based on TPE-Informer model is proposed.Firstly,the electric life test platform of railway relay is built to obtain the degradation data of its whole life cycle,and the characteristic parameters that can reflect its operation state are extracted from it.Secondly,the extracted feature parameters are input into the Informer model for training,and the multi head sparse self-attention mechanism is used to mine the correlation between the states before and after the feature information.Finally,the tree-structured parzen estimator(TPE)is used to optimize the hyper parameters of Informer model to obtain better prediction performance.The experimental platform is used to collect data to verify the model,and the results are compared with the other three deep learning algorithms.The experimental results show that the proposed prediction model has higher prediction accuracy than RNN,LSTM and Informer models,with an average accuracy of 96.52%,small error rate and good stability,which proves the feasibility of the application of the prediction model.