Prediction Analysis of the Railway Track State Based on PCA-RBF Neural Network

摘要

With the development of high-speed railway, its security guarantee has received more and more attention. The railway track state is safety-critical and affected by many factors. The proposed approach focuses on establishing a track-state prediction model by monitoring data and analyzing the deformation trends in track geometric dimensions. Radial basis function (RBF) neural network is widely used in many industrial prediction domains, and principal component analysis (PCA) is a kind of methods to reduce dimensions. In this paper, we present an approach for predicting track irregularity index using PCA-RBF model. It is benefit for periodical maintenance of railway systems and safeguarding of transportation. The experiments show that the proposed model is effective.