Variation Control Technology During Manufacturing Process

Aiming at problems during the mechanical manufacturing process, such asparameters’ variation, performance inconsistency, the paper firstly defined the variationand introduced the process ability distinguishing during the manufacturing process.According to the process ability, the variation parameters can be distinguished. Thecauses of variation parameters during the manufacturing process were classified into sixtypes: Man Factor, Machine Factor, Material Factor, Method Factor, EnvironmentFactor and Measurement Factor which were concretely qualitatively analyzed and thevariation parameter control measures during the manufacturing process were putforward. Moreover, an important component belonged to a certain machining center wasacted as the study object and the manufacturing process ability was determined.To extract variation parameters need to be emphatically controlled during themanufacturing process, the article put forward the concept of variation entropycombining the entropy in Information Theory and utilized its quantified results as thestandard for diverse potential variation parameters during the mechanical producing andmanufacturing process. The paper set up a variation parameter extraction andsequencing model based on variation entropy. Firstly, the standardization distinguishingmatrix was worked out by initial distinguishing matrix. Secondly, the relativeproportions of different variation parameters and entropy values of variation parameterswere obtained. Finally, entropy weights were gained through expert scoring method.Thus, the key potential variation parameters would be get after the extraction andsequencing of various variation parameters according to different entropy weights ofdiverse variation parameters. In the same way, by the instance analysis of the importantcomponent belonged to a certain machining center, the variation parameters in practicalwere extracted and sequenced.In order to realize the determination of the relation between variation parametersduring the manufacturing process and the mechanical product performance, after theextraction of the index of reliability MTBF, the product performance estimation andprediction model was built based on GM(1,l) model to fulfill the estimation andprediction of the product performance influenced by variation parameters. The articleemployed Grey System Theory. Utilizing Grey Correlation Analysis Theory andconducted the study about the correlation extent between variation parameters and product reliability. and the sample values of variation parameters during themanufacturing process, their grey correlation degrees were calculated in accordancewith Grey Correlation Theory on the basis of variation parameters’ values. The variationparameters which influenced more on the product performance were distinguished.Eventually, focusing on the reliability of a certain horizontal machining center and thevariation parameters of the important component belonged to the machining center, theexchange tray carrier, the grey correlation analysis was conducted. The productreliability influenced by variation parameters during the manufacturing processestimation and prediction were both completed.To complete the study scope, a variation parameter model during the manufacturingprocess based on BP Neural Network was established in which the variation parametersduring the manufacturing process were as the input information and the outputinformation was the mechanical product performance. And the model is applied to fitthe mapping relation between variation parameters during the manufacturing processand the product performance.At last, orienting variation parameters during the manufacturing process and themechanical product performance, the variation parameter analysis system during themanufacturing process was established in which analysis study mentioned above wasinvolved. The system realized the automation control of variation parameters during themanufacturing process.