Study On Comprehensive Intelligent Reasoning Methods Of Cutting Data For Compacted Graphite Cast Iron

The continuous emergence of refractory materials promotes the progress of cutting technology.Due to the disadvantages of conventional metal cutting technology such as low efficiency,high energy consumption,unstable surface quality and low level of automation,new processing technology and intelligent manufacturing system are urgently needed to improve the intelligent level of machining process.The combination of high-speed machining and intelligent reasoning system can not only greatly improve the cutting efficiency and workpiece surface quality,but also reduce the cutting energy consumption and further improve the automation level of cutting.Therefore,this research takes compacted graphite cast iron as experimental material,obtains the cutting data of compacted graphite cast iron in rough machining,semi-finish machining and finish machining through high-speed milling experiment Then,based on the obtained cutting data,a comprehensive intelligent reasoning system was established to predict the tool wear,tool life,cutting force,cutting power and surface quality,and to obtain the optimal combination of cutting parameters.Two improved particle swarm optimization algorithms,VPSO and VCPSO,are proposed.These two particle swarm optimization algorithms have strong global convergence ability and are superior to standard particle swarm optimization algorithm,genetic algorithm and other optimization algorithms.An adaptive neural fuzzy inference system based on VCPSO algorithm was established to reliably predict the tool flank wear,tool life and cutting force,and their reasoning accuracy reached 93.5%,91.8%and 95.6%,respectively,which verified the validity of the reasoning model.An improved adaptive neuro-fuzzy inference system,similar reasoning model of cutting power and surface roughness was established.The effects of cutting parameters on cutting power and surface roughness were analyzed by ANOVA method.The results show that the reasoning accuracy of cutting power and surface roughness is 93.1%and 93.8%,respectively.When milling compacted graphite cast iron,the cutting speed has the greatest influence on cutting power,the second is the depth of cut,and the least influence is the feed rate.The influence of cutting speed on surface roughness is the largest,followed by feed rate,and the influence of depth of cut is negligible.A similar reasoning model of cutting power and surface roughness was established,which could predict the cutting power and surface roughness of different cast iron materials under the same cutting parameters.The reasoning accuracy of cutting power and surface roughness was 92.1%and 89.4%,respectively.An intelligent reasoning and optimization system is proposed,which includes an improved case reasoning(ICBR)method,an optimization method based on ANIFS model and VPSO algorithm.The ANN model was used to determine the weight of cutting parameters and tool wear,and the Gaussian fuzzy grey correlation method was used to solve the reuse model.ICBR method can predict cutting power and cutting vibration amplitude based on cutting parameters and tool wear state,and VPSO algorithm is used to optimize cutting parameters of adaptive neuro-fuzzy reasoning system established.The results show that the cutting speed has the greatest influence on cutting power,followed by depth of cut and tool wear,and the feed rate has the least influence.Cutting speed has the greatest influence on the amplitude of cutting vibration,followed by feed rate and depth of cut,and tool wear has the least influence.The accuracy of the improved case reasoning method in predicting cutting power and cutting vibration amplitude is 91.7%and 95.7%,respectively.The reuse models of Gaussian process regression and support vector regression were established to predict the surface roughness and residual stress under different cutting parameters and tool wear.The results show that the regression reuse model of Gaussian process has higher reasoning accuracy.The influence of cutting speed on surface roughness is the largest,followed by tool wear and feed rate,and the influence of depth of cut is the least.The feed rate has the greatest effect on the residual stress,followed by the cutting speed,and the depth of cut has the least effect.Compared with other intelligent inference models,the proposed ICBR-G inference method has higher inference effect and can realize the inference prediction of the surface quality of workpiece materials.Based on B/S model,an integrated intelligent reasoning system is constructed.The system realizes the cutting data reasoning and effective utilization by means of the adaptive neuro-fuzzy reasoning system and case-based reasoning.The system not only has the function of reasoning,but also can provide optimized cutting parameters for users.Integrated intelligent reasoning system can improve cutting efficiency,surface quality and reduce cutting energy consumption.