Analysis Of Chaotic Motion Characteristics And Fault Diagnosis And Prediction Of A Mechanical Mechanism With Multiple Clearances

The mathematic model of the pair with clearance itself has strong non-linear behavior.Recent years’ research shows that the small change of system parameters in the non-linear system will lead to chaotic motion phenomenon in its operation results.Combining the idea of chaotic theory analysis with the simulation analysis of dynamic equation solution will be helpful to analyze and clarify some important points in the mechanical system.What is the impact of parameters on the system? At the same time,the possibility of failure such as wear and tear of parts in mechanical system with clearance will rise in a straight line.Therefore,the research of fault diagnosis method will be of great significance to the stability of equipment.The research object of this paper is crank-slider mechanism with clearance.The research content is helpful to deepen the application of non-linear dynamics theory in mechanical engineering and expand the application scope of chaos analysis theory in mechanical engineering.At the same time,the application of neural network in bearing fault identification and diagnosis of clearance mechanism is studied,and the application of intelligent fault identification and diagnosis of mechanical system is realized.It has certain reference significance.The main contents of this paper are as follows:This paper studies the description of clearance of rotating pair,establishes its general model,establishes dynamic equation of crank-slider mechanism with single clearance and double clearance at different positions,identifies chaotic motion phenomena in the mechanism by Poincare mapping method,phase trajectory diagram and Lyapunov index with the help of MATLAB software and ADAMS software,and analyses the conditions of different clearance positions and clearance sizes.Under this condition,the chaotic motion phenomena in the moving pair are discussed,and the influence of the number of clearances on the chaotic phenomena in the rotating pair is also discussed.On the basis of spectrum analysis and wavelet decomposition of slider acceleration simulation signals,acceleration signals with different gap sizes are classified and identified.The neural network diagnosis model is built by using MATLAB,and the modelafter training is validated.Finally,a sample data selection method is proposed and its validity is verified.