Analysis and Control of High-Pressure Die-Casting Process Parameters with Use of Data Mining Tools

This paper presents methods of improving the quality of high-pressure die-casting (HPDC) of passenger car parts in one of the Polish foundries. The process of high-pressure die-casting has been characterized in the aspect of two control methods: engineering and statistical. Focus is put on practical application of statistical process control (SPC) results, including but not limited to control charts and their interpretation as a source of data for the Data Mining models. The end result is a set of mathematical models developed in the MATLAB environment, preceded by popular statistical analyses, such as the analysis of correlation between process parameters, normality tests, and the analysis of variance (ANOVA), which examines the impact of process parameter values on the final product quality. Computational intelligence models have been developed to predict the fraction of faulty products on the basis of out-of control conditions detected by the use of the SPC and control charts, for selected statistically relevant process parameters. The approach presented in this paper can be used as a tool supporting the decision-making processes and, in the future, as a tool for direct or automated process control.