Research on the Iron Rail Compressive Stress Inspection Based on Improvement of Neural Networks
In order to improve the accuracy and efficiency of railway track compressive strength prediction,an improved neural network based railway track compressive strength prediction method is proposed.The rail vibration signal is extracted by empirical mode decomposition.The noise of rail vibration signal is reduced by wavelet decomposition.Combined with linear regression and fuzzy function,the neural network is improved to obtain more effective membership function,and a prediction model of railway track compressive strength is constructed.Based on FCM algorithm,attribute reduction of vibration signal is carried out through discretization processing,and the compressive strength of rail is predicted.The experimental results show that the prediction er-ror of the rail compressive strength of the proposed method is less than 2 MPa,and the prediction time is about 25 min,which im-proves the prediction accuracy and prediction efficiency,and has good practical application value.
万方数据
