Research on Degradation State Assessment Method of Switch Machine Based on MSET Reconstruction Model

Aiming at the problem that the maintainers are unable to accurately evaluate the degradation state of the equipment due to the high frequency of action of the switch machine on the metro turn-back lines, a method for assessing the degradation state of the switch machine based on Multivariate State Estimation Technique (MSET) reconstruction model is proposed. Firstly, the feature vector set is constructed based on the time domain, frequency domain, and wavelet packet Renyi entropy features of the power data of the ZDJ9 switch machine, which are mainly equipped in the metro turn-back lines. Secondly, the feature vectors in the healthy state of the switch machines are used to construct the history memory matrix in the MSET model, to establish the standard space of the equipment health state. Finally, the MSET model is used to reconstruct the feature vectors of the waiting measurement data to obtain the estimation vector. The cosine similarity between the two vectors before and after the reconstruction is calculated as the degradation indicator (DI) to construct the switch machine degradation curve, to reflect the degradation degree of the switch machine in real-time. The results show that the MSET reconstruction model can effectively capture the change rule of the switch machine’s health state and realize the accurate calculation of the degradation indicator.