Prediction And Optimization Of Gear Ring Gear Cutting Process Parameters Based On BP Neural Network

The thin-walled ring gear on the planetary reducer used by a special vehicle has a large radial size,a large difference in wall thickness and radial size,poor rigidity,high hardness,high accuracy requirements,and difficult machining.The tooth shape accuracy after processing is not high.High,the transmission is not in place,the noise is large,the vibration is violent,and the transmission efficiency is low,which affects the stability and service life of the high-speed meshing movement of the ring gear.This paper takes thin-walled helical gears as the research object,aiming at the clamping and machining process that affects the precision of thin-walled helical gears during actual processing,through Abaqus finite element simulation analysis,BP neural network technology and genetic algorithm The prediction and optimization of related process parameters are studied,and then the deformation control is realized,and a feasible solution is proposed to improve the tooth profile accuracy of the ring gear.Aiming at the clamping process of thin-walled ring gear,the Abaqus finite element simulation model of ring gear clamping was established according to the actual processing site,and the simulation data was used as sample data.The ring gear clamping based on BP neural network was established through the Matlab neural network toolbox The digital model of deformation prediction realizes accurate prediction of clamping deformation under a given clamping force,and provides a data basis for the subsequent optimization of clamping parameters and the exploration of the relationship between clamping force and clamping deformation.Aiming at the gear cutting process of thin-walled gear ring,firstly,the Abaqus finite element simulation model of gear ring gear cutting was established according to the actual gear cutting process,and the corresponding relationship between the cutting parameters and the cutting force was established.The simulation data was used as a sample Data,through the Matlab neural network toolbox,a digital model for predicting the cutting force of gear cutting based on BP neural network was established,which realized the accurate prediction of the cutting force under the given cutting parameters,for the optimization of subsequent cutting parameters and Reasonable selection of interpolation parameters provides a data basis.Aiming at the optimization of interpolation parameters of thin-walled gear rings,based on the prediction model of interpolation force based on BP neural network,the function relationship between interpolation parameters and interpolation force is fitted using BP neural network as the fitness Function,and then use the genetic algorithm to optimize the function to obtain the value of the interpolation parameter when the minimum interpolation force is obtained,and verify it through the Abaqus finite element simulation model of the gear cutting of the ring gear Optimization.The results show that the combination of modern metal cutting theory,Abaqus finite element simulation technology,BP neural network technology and genetic algorithm parameter optimization technology is used to predict and optimize the process parameters,seek the optimal combination of process parameters,and explore the thin-walled ring gear The new method for predicting and optimizing the shape parameter of the gear shaping is of great practical value for the analysis and research of the factors influencing the accuracy of the tooth profile of thin-wall helical gear shaping and the optimization control of deformation prediction.