Research On SD Identification System And Its Application In Early Fault Diagnosis Of Axle Box Bearings In Rail Vehicles

The weak signals of early failure of axle box bearings in rail vehicles are often submerged in strong noise,which is difficult to identify.If we can identify the useful weak fault signal from the strong noise as soon as possible,it will play an important role in avoiding major accidents.Based on SD(smooth and discontinuous)system,a identification system is constructed,and the identification of frequency,phase and amplitude of early fault signals of axle box bearings in rail vehicles is studied,and the correctness of the proposed method is verified through experimental research.The main research contents and conclusions are as follows:(1)Based on the stochastic variable-scale SD limit system,a variable scale-convex-peak frequency identification method is proposed.By the stochastic kind Melnikov method,the influence of the phase of the detected signal on the variable scale-convex-peak frequency identification method is studied,and the method is further optimized.Through the relationship between functions and numerical simulation,the range of parameters of the stochastic variable-scale SD limit system is determined,which improves the universality of the variable scale-convex-peak frequency identification method,that is frequency persistence.(2)Based on the stochastic variable-scale SD truncation system considering phase,that is,the stochastic variable-scale Duffing system considering phase,a variable-scale convex peak phase identification method is proposed.According to the position of periodic motion in a period,this method accurately identifies the phase of early fault signal in three situations by using the phase bifurcation diagram of reference signal.(3)Based on the stochastic variable-scale SD limit system,the amplitude of early fault signals is identified.Three methods are put forward: chaos threshold method,amplitude bifurcation diagram method and system phase diagram method.Each method can be used independently or mutually verified.The amplitude bifurcation diagram method is suitable to be extended to engineering practice because of its simple operation.(4)Based on the stochastic variable scale SD system,the circuit diagram is designed,the PCB board is made,the hardware platform is built,and the circuit experiment is carried out.Through the circuit simulation experiment,the frequency,phase and amplitude of the early fault signal are identified,which leads to the optimization of the non-smooth variable scale-convex-peak frequency identification method.The universality of the non-smooth variable scale-convex-peak frequency identification method is improved,and the feasibility of realizing the research results is confirmed by the circuit experiment.(5)The flow chart of early fault signal identification of axle box bearings in rail vehicles is constructed.The data of early fault signals of axle box bearings in rail vehicles are obtained through experiments,and the frequency,phase and amplitude of these fault signals are identified.Hence,the fault location of axle box bearings in rail vehicles is determined.The correctness of the research results in this thesis is verified by experiments.