A Study On Quality Optimization Of Squeeze Casting Process Parameter Data Of Process Parameter Design

Squeeze casting technology has developed rapidly in recent years because of its advantages in forging and casting.The process parameters of squeeze casting are the key factors to the production of squeeze casting,and currently they are mainly achieved through experiments and experience,which is time-consuming,costly and inapplicable.To address this challenge,the data-driven rapid design method of squeeze casting process parameters is naturally expected based on the past existing data of squeeze casting process.According to the analysis,there are many quality problems in the existing data.Thus,this thesis focuses on optimizing the data quality of the existing squeeze casting process parameters such that lays a foundation for constructing the data-driven design method of squeeze casting process parameters.(ⅰ)Through the analysis of the generating characteristics of the existing squeeze casting process parameters data,the main quality problems were summarized,such as the data missing,the different reliability and inconsistent process parameters of the same material.(ⅱ)In view of the missing of squeeze casting process parameters data,the paper first summarizes the characteristics of data missing.Based on this,combined with the reasons for the missing of squeeze casting process parameters data,the missing features were analyzed.Then,based on the comparison of the applicable scope and advantages and disadvantages of existing missing data processing methods,a collaborative filtering missing data processing method based on user clustering is established,and its applicability and interpolation performance were verified.(ⅲ)In view of the different reliability of the data of squeeze casting process parameters,the evaluation index system and evaluation model of the reliability were established by analyzing the main ways of obtaining the squeeze casting process parameters at present,and the reliability grades were divided.(ⅳ)In view of the problem of inconsistency conflict between the parameters of the s casting process parameters,combined with the causes,the treatment methods were established in two cases:① same process parameters,different performance indexes;② different process parameters,different performance indexes.For the first case,the most reliable process parameter data is used as the processing result.For the second case,firstly,all the process parameter data with reliability level of 1 or two sets of process parameter data with highest reliability are found through the reliability evaluation,and then the Partial Least Squares(PLS)modeling method is used to judge the inconsistency of the data.If the judgement result proves to be inconsistent,the sets of process parameter data are combined to establish PLS model and the Particle Swarm Optimization(PSO)is used to obtain the final optimized result.Finally,the application and analysis of the two cases of processing are carried out separately.(v)Based on the above optimization method of process parameter data,the automatic quality optimization process of squeeze casting process parameter data was determined,and the data quality optimization system of squeeze casting process parameters was designed,and the system was developed by MATLAB GUI,which realized the automation quality optimization of the squeeze casting process parameter data.Finally,two sets of running experiments are carried out to verify the effectiveness of the method and the reliability of the system.