Temperature Force Identification Of CWR Track Based On Vehicle Track Vibration Response

The CWR eliminates the rail gap,improves the track smoothness and reduces the wheel-train and maintenance impact.However,due to the inability of the rail to freely expand and contract,when there is excessive temperature force inside the track,it will cause track buckling,so it is necessary to monitor the track temperature force in real time.The displacement observation pile method is commonly used to detect the temperature force of the CWR track,but the temperature force data cannot be obtained in real-time and accurately.The ultrasonic and laser detection methods developed in recent years are susceptible to the internal damage of the rail or the external environment.Therefore,this paper proposes a method for identifying the temperature force of CWR based on vehicle-induced vibration response.This method can detect the temperature force in real time when the train running on the line,which is convenient to implement and not affected by the environment,the main content is as follows:(1)Vibration response of CWR track under temperature force.The variation of vibration acceleration and dynamic strain characteristics with temperature force at different positions of each component of the track is analyzed,and the research object required for subsequent temperature force feature extraction is determined.(2)Extraction of temperature force identification index of each component of CWR track.Based on EMD frequency decomposition,HHT time-frequency diagram and Hibert marginal spectrum,the extraction method of temperature force characteristic index is proposed.The variation law of track acceleration characteristics with temperature force under each feature extraction method is analyzed,and three temperature force identification indexes are determined,which provides a basis for subsequent temperature force identification.(3)Optimal identification algorithm of temperature force of CWR track.The recognition effect of regression algorithm,grey prediction algorithm and neural network algorithm on temperature force is analyzed.The average recognition error of multiple regression algorithm is 5.08 %,which is 3.41 % lower than that of grey prediction algorithm and 0.41 % lower than that of BP neural network.The greater the temperature force,the smaller the recognition error.The multiple regression algorithm is determined as the identification algorithm of the temperature force characteristic index.(4)Influence of different train speed and axle load on the accuracy of temperature force identification of CWR track.Train speed and axle load have a certain influence on the identification accuracy of track temperature force.The train speed is within 60-80 km / h,and the train axle load is within 21～22 t,which has a good effect on the temperature force identification of CWR.The suitable vehicle type for the temperature force identification of CWR based on vibration method is given.(5)Temperature force identification method of CWR track based on vehicleinduced vibration acceleration.The method of temperature force identification of CWR track based on vehicle-induced vibration is proposed,and the concrete vibration acquisition scheme and feature extraction method are given.The characteristic regression function is determined to realize temperature force identification,which has the advantages of convenient signal acquisition and no interruption of line operation.