Research On Dynamic Multi-Objective Optimization Strategy Of Milling Parameters Based On Digital Twin

As a basic equipment.CNC machine tool is the representative of the overall manufacturing level of the society.Nowadays,every manufacturing power in the world has put forward their own decisions on revitalizing and developing high-end manufacturing.CNC machine tools must be comprehensively upgraded to adapt to the intelligent manufacturing mode.Intelligent operation optimization decision-making technology is the key to realize the intelligence of CNC machine tools.The intelligent operation optimization decision-making of CNC machine tools represented by multi-objective optimization technology can optimize the existing processing scheme and obtain the global or local optimal processing scheme under the condition of considering many processing objectives.However,the currently applied multi-objective optimization technology basically belongs to the static category,and does not incorporate the dynamic performance changes of CNC machine tools throughout the life cycle into the optimization model.Therefore,the obtained optimization results of machining scheme have great limitations,and it is difficult to meet the optimization requirements of the whole life cycle of CNC machine tools.To carry out the dynamic multi-objective optimization of milling parameters under considering the varying performance of machine tool,a strategy of dynamic multi-objective optimization of milling parameters is developed based on digital twin.The main work is as follows:This paper firstly discusses and verifies the significant impact of tool wear on machine tool processing performance.and selects the tool wear as the characteristic factor of the dynamic machine tool processing performance.Then,according to the functional requirements of dynamic optimization of milling parameters,the overall framework of dynamic multi-objective optimization strategy of milling parameters based on digital twin is designed.It combines and coordinates the functions of three major modules:fitting prediction module,parameter optimization module,and decision analysis module to achieve dynamic optimization of milling parameters.Firstly,the gradient boosting ensemble learning framework is used to construct the nonlinear mapping relationship between processing parameters and processing results by fusing gaussian process regression algorithm,random forest algorithm,and support vector regression algorithm in fitting prediction module.Then,the parameter optimization module dynamically optimizes the milling parameters by combining the digital twin technology and the dynamic non-dominated sorting genetic algorithm DNSGA-II-A considering tool wear.Finally,based on obtained pareto optimal solution,a decision analysis model is established by combining analytic hierarchy process(AHP)and technique for order preference by similarity to an ideal solution(TOPSIS),and the visual analysis ranking is realized as well.In this paper,the milling experiments of 45#steel workpiece are designed and implemented.By processing the milling data,on the one hand,the accuracy of the fitting prediction module is verified;On the other hand,it illustrates the effectiveness of the parameter optimization module and the diversity of optimization results;On the other hand,the interactive interface of milling parameter optimization decision-making is designed by Python language,and the feasibility and convenience of the decision analysis module are verified by comparing the sorting optimization scheme with the actual one.The dynamic optimization strategy proposed in this paper can provide the optimal milling parameter selection scheme in accordance with the characteristics of the current machine tool,which can help CNC machine tools to realize intelligent operation optimization decision-making function.