Research On Internal Combustion Engine Fault Diagnosis Method Based On Non-linear Torsional Vibration During Multiple Operating Conditions

As the power output shaft of the internal combustion engine,the shaft system is subject to complex torsion,bending,shearing,friction and even impact force,which makes the components of the shaft system prone to failure.The torsional vibration response of the shaft system is the sum of the multi-modal response of the shaft system caused by the excitation force of each cylinder,which can fully reflect the working state and failure of each cylinder of the internal combustion engine.Therefore,the research on the monitoring and fault diagnosis of the internal combustion engine based on the torsional vibration signal has received widespread attention and continuous attention from scholars at home and abroad.However,with the development of high-speed,highload,high-power internal combustion engines and the in-depth study of torsional vibration,new problems worthy of study still continue to emerge,such as the impact of start-stop conditions on the torsional vibration characteristics of the shaft system.New control technologies are also constantly evolving,such as the frequency modulation design of nonlinear variable stiffness couplings.In view of the existing problems,combined with the nonlinear characteristics of the shaft system,the dynamic characteristics and fault diagnosis of the internal combustion engine in multiple operating conditions are studied,and an intelligent fault method based on the torsional vibration simulation model of the shaft system is proposed.Firstly,the torsional vibration calculation method of the internal combustion engine shaft system is studied: the applicability of the central difference method to the torsional vibration simulation calculation of the shaft system in start-stop conditions is studied.It has been verified that the calculation error between the central difference method and the TVCA2006 software in steady-state conditions is within 2%.The calculated value of the method in the start-stop conditions is consistent with the experimental results.Aiming at the parameter characteristics of several typical variable-parameter couplings,a torsional vibration simulation calculation algorithm suitable for the non-linear shaft system in steady-state conditions and start-stop conditions is proposed,and on this basis,the development work of the torsional vibration simulation software PVDP2019 is completed,which provides a simulation analysis tool for fault diagnosis research.Secondly,to study the method of identifying the excitation torque of the internal combustion engine during start and stop: By studying the influence of the coupling stiffness on the torsional vibration characteristics of the shaft system,it is determined that the shaft system of the coupling with large rigidity is the research object.The deficiencies of the excitation torque identification method based on average angular acceleration are analyzed,and the excitation torque identification method based on average effective pressure correction is proposed for start-up conditions,and the excitation torque identification based on intake pressure correction is proposed for stop conditions.This has laid the foundation for the subsequent simulation research on the characteristics of the shafting in the start-stop conditionsThen,using the developed simulation calculation software and excitation torque identification method to study the torsional vibration characteristics of the shaft system:The amplitude characteristics of the resonance speed of the internal combustion engine shaft system during start-stop conditions are studied.It is found that the amplitude of the coupling stiffness and resonance speed has an inverse relationship with the amplitude of the torsion angle difference between the main and passive ends of the coupling,and the existing mechanism is analyzed.The frequency characteristics of the resonance speed of the shaft system during start-stop conditions are studied.It is found that the change of the lift rate will affect the position where the resonance speed occurs and a method for quickly estimating the low-order natural frequency of a shaft system with small stiffness is proposed.This laid the foundation for the construction of fault diagnosis model.Finally,aiming at the uneven work failure of internal combustion engine due to power loss,combined with the response characteristics of the shafting during steady-state conditions,a fault diagnosis system based on the steady-state torsional vibration simulation model was built with a convolutional neural network.The fault diagnosis model has been verified,the results show that the model can accurately identify the fault degree and fault location of the fault of uneven internal combustion engine work.Aiming at the failure of the coupling and combining the response characteristics of the shafting in the start-stop conditions,a convolutional neural network is used to build a fault diagnosis system based on the start-stop conditions,and the fault diagnosis model has been verified.The results show that the model can accurately identify the failure characteristics of the coupling caused by the change of stiffness and damping,and also accurately identify whether the main and passive ends of the coupling collide.Research indicates: considering the nonlinearity of the shafting parameters,combined with the dynamics of the shafting in multiple operating conditions,a modern intelligent fault recognition method such as convolutional neural network is selected to build a internal combustion engine fault diagnosis system based on the simulation model,which can be used to adopt nonlinear.The fault diagnosis of the internal combustion engine with linear rotating elements in multiple working conditions improves the modeling efficiency and accuracy of fault diagnosis.