Multi-objective Optimization Research Of Precision Turning Process Parameters

The selection of process parameters is not only directly related to the quality of processing,but also has a significant impact on the processing accuracy and efficiency,as well as the performance of the machine tools and cutters.In traditional NC machining process,it is often difficult to get the ideal result through the processing handbooks and the experience of the workers.At present,although there were some researches on optimization of process parameters,but mostly aiming at single objective optimization.Moreover,the optimized results were often achieved at the expense of sacrificing the performance on other aspects,which could not at the same time satify multiple objectives optimization.Improved non-dominated sorting genetic algorithm(NSGA-II)becomes one of the benchmark of multi-objective optimization algorithm due to its rapid running speed,low computational complexity and ease of implementation,etc.Therefore,it is widely used in the field of scientific research and engineering practice field for solving multi-objective optimization problem.However,the effect is not ideal when addressing some particular multi-objective optimization problems,which are caused by uneven distribution of population convergence,poor global search ability and easily fallen to local optimum.Aiming at the problems above,this paper conducts in-depth research on process parameters optimization during precision turning processing,which takes both surface roughness and material removal rate as optimal objectives.Firstly,the actual data of the surface roughness and material removal rate were obtained by desigining orthogonal experiments,on the basis of which,the multi-objective optimization model of surface roughness and material removal rate were separately builded by genetic regression neural network and orthogonal regression method.Secondly,considering the deficiency of NSGA-II,the orthogonal crossover strategy and adaptive hybrid mutation operator were introduced in the NSGA-II to improve.Thirdly,it is innovative to solve the multi-objective optimization model of precision turning process parameters with improved NSGA-II algorithm.The concrete research contents are as follows:(1)Taking the precision turning as the research object,the tool nose radius,feed of every rotation,depth of cutting edge and spindle speed as the design variable,the surface roughness and material removal rate are optimizing index,the actual data of the surface roughness and material removal rate have been obtained by using orthogonal experiment.Then these data were analysed by adoptting the range analysis and analysis of variance,which can show the influence of the degree and regularity on surface roughness and material removal rate by the various process parameters.(2)According to the results of orthogonal experiments,a second order polynomial regression predictive model of surface roughness has been built based on the orthogonal regression method.Then this paper has compared and analysed the prediction effect of the above model combining with the index prediction model and linear prediction model.Subsequently,the generalized regression neural network has been introduced to build the surface roughness prediction model with applying improved adaptive genetic algorithm(IAGA)to optimize the extremum of the smooth factor included in the generalized regression neural network(GRNN).On this basis,the prediction effect of IAGA-GRNN model is significantly higher than the three regression prediction models above with a verification experiment.In this research,the material removal rate orthogonal regression index prediction model has been established,the result of which indicates that it could be utilized to predict the material removal rate with high precision and excellent effect.Taking the tool nose radius,feed of every rotation,depth of cutting edge and spindle speed as the design variable,the IAGA-GRNN surface roughness prediction model and material removal rate index of orthogonal regression model as objective function,the multi-objective optimization model of precision turning process parameters of 6061 aluminum alloy has been established under the constraints of process parameters,cutting force,cutting power and cutting temperature,as well as cutter life.(3)As some deficiencies such as uneven distribution of convergence in population,poorly global searching ability,easily to run into partial optimization appeared in classic non-dominated sorting genetic algorithm ?(NSGA-?),this paper proposed an improved algorithm by introducing orthogonal crossover strategy and hybrid mutation operator into the NSGA-?.The experiments on a series of test functions show that the improved NSGA-? has better performance than NSGA-? both in convergence and diversity.(4)The research about optimization of processing parameters of 6061 aluminum alloy in precision turning based on improved algorithm and classic NSGA-?,reveals that the improved algorithm could have better convergence rate and accuracy than the classic NSGA-?,which means the improved algorithm is more effective in solving the multi-objective optimization problems of machining parameters.In addition,the reliability and rationality of the above method have been proved by a verification experiment.The result shows that this method could precisely and efficiently solve the multi-objective optimization problems on process parameters.