Research On Fault Diagnosis Of Heavy Haul Locomotive Rolling Bearing Based On Variational Mode Decomposition

The axle box bearing of heavy haul locomotive is one of the key components of the bogie.In the long-term service process,its load conditions are complex,and it is prone to pitting,scuffing,wear and other failure modes,which directly affects the operating performance of the vehicle.Wheel-rail excitation directly affects the safety and stability of axle box bearings.Since the early periodic impact signals of bearings are difficult to obtain,how to effectively extract bearing fault features for identification is the key to fault diagnosis.Therefore,this thesis adopts the GA-VMD fault diagnosis method with kurtosis and envelope entropy as the fitness function.The feasibility of the GA-VMD method is verified by the simulation signal of the bearing outer ring,the experimental data of Case Western Reserve and the simulation results of the 5-degree-of-freedom rolling bearing dynamic model.Finally,combined with the locomotive-track dynamic model,the vibration response of the axle box bearing under different wheel-rail excitations is simulated.The GA-VMD method is used to identify different faults of the axle box bearing under wheel-rail excitation,which provides a basis for effective identification of locomotive axle box bearing faults.The main contents of this thesis are as follows:(1)GA-VMD fault diagnosis method with kurtosis and envelope entropy as fitness functions is proposed and its feasibility is verified.Aiming at the selection of modal decomposition number and penalty factor in VMD algorithm,genetic algorithm is used to optimize the modal decomposition number and penalty factor of VMD input parameters.Taking kurtosis and envelope entropy as the fitness function,the GA is used to calculate the minimum value of the fitness function to obtain the best number of modal decomposition and penalty factor.The effectiveness of the GA-VMD method is verified by using the bearing outer ring simulation signal and the bearing data of Case Western Reserve.The results show that the method can effectively identify fault information.(2)Five-degree-of-freedom rolling bearing dynamic model is established to analyze the dynamic response of different faults of the bearing,and the GA-VMD method is used for fault identification.The bearing model is simplified.According to the bearing dynamics theory,the dynamic fault models of rolling bearing with different faults are established.The dynamic responses of single-point and multi-point faults of bearing are simulated and analyzed.The influence of rotational speed and radial force on the dynamic characteristics of rolling bearing is studied.Based on the vibration response of different fault bearings,GA-VMD method is used to identify different faults of inner and outer rings of rolling bearings.The results further verify the effectiveness of the method.(3)The dynamic model of HX_D2 heavy haul locomotive-track is established by SIMPACK software.The vibration response of locomotive bearing is simulated and analyzed by combining with the dynamic model of five-degree-of-freedom rolling bearing.The fault type of locomotive bearing is identified by GA-VMD method.The primary force at the axle box under different wheel-rail excitations is simulated and analyzed.The primary force is used as the radial load of the axle box bearing to obtain the vibration response of the axle box bearing under different wheel-rail excitations.The GA-VMD method is used to extract the fault feature information from the simulation results.The envelope spectrum of the best IMF component obviously contains the fault feature frequency.The algorithm can well identify the fault type of locomotive axle box bearing.