Quality Analysis And Prediction Of Discrete Manufacturing Processes

For a long time,China’s manufacturing industry has been in a phase of rapid development.In order to adapt to the changing needs of the market and fierce competition,traditional mass production methods have gradually given way to discrete manufacturing as a means of meeting the diverse and individual order requirements of customers.The manufacturing industry is gradually being transformed and upgraded to become intelligent as information technology develops.Various factors in the production process can affect the final quality of products.Through the mining and analysis of production data,managers can predict quality problems in advance,effectively reduce costs and make efficient decisions.This dissertation studies quality forecasting in the discrete manufacturing process.Using historical data generated in the manufacturing process,an effective quality forecasting model for the production is built after data prossessing.After which,association rule techniques is used to mine the data,so as to reduce the risk that managers will need to rework the product because it is not of acceptable final quality.Firstly,the direction and content of this dissertation will be determined by domestic and international research and analysis of discrete manufacturing and quality prediction.Secondly,after analyzing the characteristics of discrete manufacturing,this dissertation introduces machine learning methods,focuses on the optimization algorithm,and proposes to use the sparrow search algorithm to optimize the relevant parameters of the least squares support vector machine.Thirdly,the quality prediction model of SSA-LSSVM is constructed,and multiple evaluation indicators are used to verify the accuracy of the model.Finally,this paper verifies the feasibility of the SSA-LSSVM quality prediction model by using examples,and concludes that the SSA-LSSVM model has high accuracy and faster algorithm detection speed,which is suitable for quality prediction in discrete manufacturing processes.The above results are combined with the association rule method,the data is further mined,and the rule results of abnormal quality data are obtained,which can be applied to the actual situation,and the production parameters can be controlled in the association rules of this feature,guiding the quality control in the discrete manufacturing process and reduce the cost.