Research On Optimizing Sizing Process Parameters Of Casual Shoes Molding Production Line

In the sizing process of the casual shoe intelligent molding production line,selecting the appropriate sizing process parameters is the key to ensuring the quality of sizing.However,due to the complexity of the sizing process,the interaction of multiple process parameters has an effect on the sizing quality.The adjustment of the existing sizing process parameters still depends on the experience value obtained by the operators in the trial and error.This situation causes the instability of the sizing quality.In response to this problem,this thesis selected the sizing width and the sizing thickness as the quantitative indicators to measure the sizing quality,and used the five process parameters that have the greatest impact on the sizing quality as the optimization objects,and applying intelligent optimization algorithms such as neural networks and genetic algorithms to the optimization of sizing process parameters,which provided a new way for the adjustment of sizing process parameters.In view of the complexity of the sizing process,it is very difficult to optimize the sizing process parameters through physical modeling.Aiming at the complex non-linear characteristics between sizing process parameters and sizing quality,this thesis used the BP neural network improved by L-M algorithm based on Bayersian regularization to model the prediction model of sizing quality.The results show that the average relative error of the 5-11-2 three-layer network structure for sizing width prediction is 1.56%,and the average relative error for the sizing thickness prediction is 5.91%.In view of the shortcomings of the neural network's sensitivity to the initial connection weights,this thesis uses genetic algorithms to optimize the initial connection weights of the BP neural network.The average relative error of the optimized BP neural network's prediction of the sizing width drops to1.17%.The average relative error of the glue thickness prediction drops to 4.04%,the accuracy of the optimized network has been improved,and it can be used to optimize the sizing process parameters.Based on the sizing quality prediction model,this thesis established a mathematical model for sizing process parameter optimization based on actual problems.The NSGA-II intelligent optimization algorithm was used to solve the multi-objective problem of sizing process parameter optimization.In the Pareto optimal solution set obtained by the solution,the optimal solution is determined according to actual needs.The sizing process parameters obtained by the optimal solution were used for spraying experiments,and compared with the results before optimization.The results show that the optimized finished shoes Sizing quality has been improved.Aiming at the sizing process parameter optimization method proposed in this thesis,a sizing process parameter optimization software had been developed.Application verification shows that the software basically meets the needs for actual sizing process parameter optimization.