Intelligent Prediction Method for Small-Batch Producing Quality Based on Fuzzy Least Square SVM

A quality intelligent prediction model for small-batch producing process was proposed in the article, after comparing with the common used approaches of procedure intelligent prediction and their characteristics. The prediction process and algorithm were presented too. Fuzzy least square support vector machine (FLS-SVM) is taken as the intelligent kernel for the model. On one hand, it can solve the small-batch learning better and avoid the disadvantages, such as over-training, weak normalization capability, etc., of artificial neural networks prediction. On the other hand, it makes samples fuzzy by membership function to choose optimum samples and make history data ‘nearer is more weight’. After doing lots of prediction experiments and comparing with other common prediction methods, the method proposed in the article proved to be good normalization capability, more rapidly built, and more easily realized. It offers feasibility to predict and control small-batch machining process online.