An Quality Prediction Method of Injection Molding Batch Processes Based on Sub-Stage LS-SVM

Injection molding process, a typical batch process, due to multistage, nonlinear and unavailability of direct on-line quality measurements, it is difficult for on-line quality control. A sub-stage LS-SVM quality prediction method is proposed for dedicating to reveal the nonlinearly relationship between process variables and final qualities at different stages. Firstly, using an clustering arithmetic, PCA P-loading matrices of time-slice matrices are clustered and the batch process is divided into several operation stages, the most relevant stage to the quality variable is defined, and then applying correlation analysis to get irrelevant variables of un-fold stage data according to time as input of LS-SVM for end-of-batch product quality prediction. For comparison, a sub-MPLS quality prediction method is applied. The experimental results prove that the effectiveness of the proposed quality prediction method is superior to the one of the sub-MPLS method.