A novel quality comprehensive evaluation method based on product gene for solving the manufacturing quality tracking problem of large equipment

In order to monitor the product manufacturing quality state in large equipment manufacturing enterprises, a novel quality comprehensive evaluation method based on product gene was proposed. Based on analyzing the similarity between organisms and products, the basic concept, model structure and encoding method of product gene were described by considering the characteristics of large equipment and its manufacturing process, where processing surface was selected as the carrier of genetic unit (Gene) and a sectional encoding method was designed to describe the quality of a product flexibly. Based on the above description and definition, the product gene knowledge base was proposed, and a set of calculation rules of attribute similarity were defined for its application. Then an innovative evaluation method based on similarity analysis for single gene was presented, and it follows the following rules: 1) The more significant the resemblance between a target and the related normal reference genes is, the higher the evaluation value will be; 2) The more significant the discrepancy between a target gene and the related defect reference genes is, the lower the evaluation value will be. To determine the quality state of a product comprehensively, a weighted comprehensive evaluation method based on fuzzy Delphi method (Fuzzy Expert Evaluation Method) was developed. Finally, an experiment of grinding roller shaft was carried out to verify the method presented in this study. The results showed that the proposed product gene model and encoding method described the manufacturing quality information of the workpiece reasonably, and the quality state of the workpiece was also described accurately. Thus, the proposed quality evaluation method is effective for monitoring the manufacturing quality in large equipment manufacturing enterprises.