The Diagnosis And Maintenance Of Planetary Gearbox Damage Status Based On The Fusion Of Swarm Intelligence Algorithm

As a complex compound gear transmission system,planetary gearbox with a compact structure,large transmission ratio,high transmission efficiency,smooth operation and other characteristics,has been widely used in automobile gearbox,wind turbine,aviation engine set.Due to the harsh working environment,gear and bearing damage and failure often occur,and due to different degrees of damage,failure mode and complex vibration signal transmission path and other reasons,so the problem of the planetary gear state identification and fault diagnosis has not been effectively solved.In recent ten years,the intelligent technology of biological community has emerged in the field of fault diagnosis with its unique advantages.This paper mainly studies the fusion strategy of shuffled leap frog algorithm(SFLA)and particle swarm optimization(PSO)in swarm intelligence algorithm to improve the performance of the algorithm,and applied it to the KPCA feature extraction of planetary gears and the modeling of vibration transmission path as well as its parameter optimization.On the basis of the experiment of planetary gearbox fault,the fusion algorithm is used to optimize the neural network to identify and diagnose the complex working conditions of planetary gear transmission.The main research contents and conclusions are as follows:1.A swarm intelligence fusion algorithm(SFLA-PSO)is proposed,which is a combination of the shuffled frog leaping algorithm(SFLA)and particle swarm optimization(PSO).The fusion strategy of two-layer optimization and internal and external circulation is given.The fusion algorithm flow is designed.And the simulation and performance test of four single-peak and multi-peak test functions are carried out for the fusion algorithm.Simulation results show that the proposed SFLA-PSO fusion algorithm has better global search ability,and is superior to SFLA and PSO in convergence time,iteration speed and convergence accuracy.2.A KPCA feature extraction method(CKKPCA)for linear combination of polynomial and radial basis(RBF)kernel functions is proposed;The mathematical model of kernel parameters and kernel weight optimization based on Fisher's criterion is established,and the SFLA-PSO fusion algorithm is applied to realize optimization.The optimized CKKPCA is applied to Iris data set for simulation research and engineering verification of planetary gearbox.The analysis results show that the intelligent optimized CKKPCA has a good recognition effect on the nonlinear behavior of mechanical equipment,such as damage and fuzzy fault boundary.3.The danamic model of the planetary gearbox is established,and a weighted and improved Hamming function is used to express the influence of the transmission path.The method of SFLA-PSO optimizing the influence function weight of the path is proposed,which is applied to the analysis of simulated vibration signals of the planetary gear under normal working conditions and planetary gear crack faults.Considering the effect of vibration source and transmission path,the total signal of planetary gearbox is analyzed and compared with the experimental test results.At last the location of the sensor is optimized4.The fault simulation experiment of planetary gearbox was carried out in the laboratory,and the vibration signals of the planetary gear's multi-wear state,the composite fault state of the sun gear and the planetary gear,the planetary gear crack and the broken tooth of the sun gear were measured,and the vibration signal processing and characteristic frequency analysis were completed.5.A fault diagnosis model based on neural network is established,and a BP neural network fault diagnosis method based on SFLA-PSO intelligent fusion algorithm is proposed,which is applied to the identification and diagnosis of planetary gear wear and the composite faults of planetary gear combined with sun gear.The test results show that,compared with before optimization,the diagnostic accuracy of all test samples of BP network optimized by SFLA-PSO is 100%,and the overall output errors are small with obvious effect.As the SFLA-PSO fusion algorithm plays a role in adjusting and optimizing the parameters of neural network,it improves the performance of the algorithm and improves the accuracy and recognition rate of diagnosis,and the accuracy and recognition rate of planetary gear wear damage status recognition and multi-mode composite fault diagnosis are improved.