Study On Slip And Slide Detecting Of Heavy-haul Locomotive And Its Application

As the main railway freight tool,heavy-haul locomotives have achieved heavy-load breakthroughs again and again.The increasing weight of locomotives also puts forward higher and higher requirements on the traction control of locomotives.When the adhesion condition between the wheels and rails becomes worse and the traction force of the locomotive wheels is greater than the available adhesion force between the wheels and rails,it will cause the locomotive to slip / slide.The slipping / sliding of the locomotive not only causes a waste of traction,but also causes serious accidents such as wheel and rail scratches or even derailment of the locomotive in severe cases.Therefore,fast and accurate slip / slide detecting plays an important role in locomotive traction control.Currently widely used are slip / slide detecting based on creep speed 、wheel set acceleration and wheel acceleration differential.These methods can accurately detect the running status of the locomotive,but there are certain deficiencies.In addition,most of the current slip / slide detecting methods perform signal processing on a single scale,and perform more time-domain analysis,without considering the frequency-domain characteristics.This paper proposes the idea of signal processing on multiple scales,combining time and frequency domains,and using their respective advantages to process signals together to better detect the running status of the locomotive.Firstly,this paper designs a method for slip / slide detecting based on multiscale fuzzy entropy.This method takes the wheel speed as input,processes the speed signal on multiple scales,calculates the fuzzy entropy value of the wheel speed on each scale,and then judges on the current scale based on whether the calculated fuzzy entropy value reaches the set threshold,and finally use the voting method to comprehensively judge the running status of the locomotive for each scale judgment result.Compared with the traditional slip / slide detecting method,the calculation based on the multiscale fuzzy entropy method is accurate and free from noise interference,and can effectively detect the locomotive’s slip / slide in time.Secondly,the slip / slide detecting method based on EEMD is designed.EEMD can adaptively decompose the signal according to the characteristics of the signal itself,which has a great advantage in processing non-stationary nonlinear signals such as wheel set speed.EEMD decomposition decomposes the wheelset speed signal into high-to-low frequency components,selects the appropriate components from the decomposed components to calculate the fuzzy entropy value,and judges the locomotive status according to whether the fuzzy entropy value reaches the set threshold.Comprehensive judgment of each component to determine the slip / slide status of the locomotive.Compared with traditional identification methods,analysis is usually performed in the time domain.The slip / slide detecting based on EEMD combines the time-frequency domain to more accurately detect the running status of the locomotive.Finally,taking a certain type of heavy-haul locomotive as the research object,a locomotive dynamics simulation model is built in the environment of MATLAB / Simulink,and various working conditions of the actual operation of the locomotive are simulated through the rail surface transformation.Through simulation,the effectiveness of the multiscale fuzzy entropy and slip / slide detecting method based on EEMD decomposition designed in this paper is verified.The simulation results show that these two detecting methods can quickly and accurately detect the running status of the locomotive,and have good application prospects.