Multi-objective Optimization Of NC Turning Process Parameters

In the manufacturing industry,the key factor affecting the cutting performance is the selection of cutting parameters.The reasonable selection of cutting parameters is conducive to reduce the wear of tools and prolong the service life of tools.At the same time,it can effectively improve the quality of machined parts,improve the machining efficiency of machine tools,save manpower and material resources for enterprises,and improve the competitiveness of enterprises.Due to the variety and quantity of machine tools in the manufacturing industry,optimizing the cutting parameters of machine tools has become the main research trend.For the optimization of NC turning process parameters,the single objective optimization method can not meet the actual needs.Explore the main research direction of multi-objective optimization method.Taking No.25 steel as the test material and cak6136 c CNC lathe as the test equipment,this paper expounds the machining process of the workpiece in detail.By comparing various experimental design methods,cutting speed,feed rate and back draft are determined as independent variables,and cutting force,surface roughness and machining time are taken as objective functions.Two optimization schemes of NC turning process parameters are proposed and compared.The first traditional optimization scheme is multi-objective optimization of cutting parameters by response surface method.The specific process is as follows: the three factor and three-level test of cutting parameters is designed based on the design expert software,the box Behnken design(BBD)method is selected to divide the cutting parameters into 17 groups,then the NC turning test is carried out,the test data is recorded and input into the design expert software,the regression model is established,the analysis of variance is carried out,and the reliability of the model is verified;After determining that the model can be used for fitting,analyze the influence degree of each single variable and its interaction on the objective function,and finally select the best parameter combination according to the actual demand.The optimal parameter combination obtained by the response surface method of the traditional optimization scheme is as follows: the cutting speed is 110 m/min,the feed rate is 0.13 mm/r,and the back feed rate is 0.5mm;Based on this parameter combination,the cutting force is 165.95 N and the surface roughness is 0.918?m.The processing time is 5.44 s.Secondly,the intelligent multi-objective optimization scheme is used to combine the approximate model with genetic algorithm.The full factor experiment is designed by Minitab software,and then the NC turning is carried out.After obtaining the test data,the RBF approximate model and BP approximate model are established respectively based on MATLAB platform.After comparing the accuracy indexes of the two models,it is found that the accuracy index of RBF approximate model is higher and more suitable for the prediction model of NC turning process parameters based on 25 steel.Next,the NSGA-II algorithm with good convergence is selected to optimize the RBF approximate model.After setting the relevant parameters,the predicted value of the RBF approximate model is trained,and the Pareto optimal solution set is obtained.The optimal parameter combination obtained by multi-objective optimization using approximate model genetic algorithm is as follows: the cutting speed is107.45m/min,the feed rate is 0.12mm/r and the back feed rate is 1.1mm;Based on this parameter combination,the cutting force is 151.46 N,the surface roughness is0.604?m and the machining time is 4.62 s.Finally,the optimization results of traditional optimization scheme and intelligent optimization scheme are tested to verify the feasibility of the scheme.The comparison results show that the relative error of cutting force optimized by response surface method is 1.2%,and the relative error of cutting force optimized by approximate model genetic algorithm is 0.35%;The relative error of surface roughness optimized by response surface method is 1.3%,and the relative error of surface roughness optimized by approximate model genetic algorithm is 0.21%;The relative error of machining time optimized by response surface method is 8.8%,and the relative error of machining time optimized by approximate model genetic algorithm is 0.58%.;The results show that the combination of approximate model and genetic algorithm can significantly reduce tool wear and improve workpiece surface quality and machining efficiency.