Research On Methods And Applications Of Trend Extraction Of Rail Corrugation

As an important means of transport,railway transport has been one of the most widely used means of transport.In China,the continuous development and maturity of high-speed rail technology not only brings convenience to people,but also a higher demand to the railway trunk construction and maintenance.Theoretical research and engineering practice show that the orbital wave mill will cause the vibration of the vehicle body during the driving process,which leads to the increase of the interaction force between the wheel and the track,which is the main factor limiting the speed and safety of the train.For the existing orbital uneven detection method,whether it is static detection or dynamic detection,it will inevitably contain a vibration cycle is greater than the signal length of the low frequency trend items,must be removed.At present,there are many ways to use the signal in the trend of the items to remove,different methods of its conditions and accuracy of the results are also very different.Empirical Mode Decomposition(EMD)can decompose the signal according to the signal itself without any prior knowledge,so that the original signal is decomposed into several inherent modal functions(Intrinsic Mode Functions,IMF)and the remainder Of the combination.Since the stop criterion for EMD decomposition is whether the function converges,the decomposition term can be treated as a trend term for the signal.The advantage of EMD decomposition is that without the need for a priori knowledge,it is possible to intelligently decompose the remainder according to the original signal to obtain the signal trend term.Nevertheless,EMD also has some inherent flaws,such as endpoint effects and modal aliasing,etc.,so it is still need to be improved to improve accuracy in practice.In order to solve the problem of endpoint effect of EMD,this paper proposes an improved EMD method based on slope matching waveform continuation method,which is used to remove the trend items in the track wave data and apply it to the design of the track wave grinding system.The main contents are as follows:First,the related concepts of track wave and trend items are discussed,and the theoretical basis of detection technology is discussed.This paper introduces the detection method of orbital wave mill,and introduces the most representative method of track wave grinding:inertial datum and chord measurement method.This paper focuses on the method of removing the trend of wave and wave in China and abroad,and briefly analyzes the advantages and disadvantages of several trend items removal methods.Secondly,discussed the basic concepts and principles of empirical mode decomposition in detail,analyzed the advantages and disadvantages of its existence,and analyzed some existing improvement methods.Aiming at the problem of endpoint effect of traditional EMD,The slope of the adaptive function is used to improve the traditional EMD,and the simulation results are compared with other extension methods to verify the improvement effect.Thirdly,in this paper,Several typical trend term removal methods,including wavelet decomposition method,least squares method and traditional EMD decomposition method,are presented and analyzed for the trend items in the orbital wave data.The slope matching waveform proposed in this paper is used to extend the improved EMD Method for comparison.By analyzing the results obtained by analyzing the simulation and the measured signal,the eff-ect of removing and removing the trend items in the signal is compared with each other,and the effectiveness of the improved algorithm in the practical application is verified.Lastly,This paper introduces the development of the track wave mill detection system,including the overall structure of the track test platform and the software platform and the detailed working principle of the various functional modules,and apply the improved EMD to the software design which is used for trend item extraction.The accuracy and reliability of the track-wave detection system are verified by actual measurements on the actual orbit line.