Fault Diagnosis Approach Based on Operational Quality Data for Manufacturing System

Abstract

Accurate fault diagnosis is the indispensable basis for reliable operation of manufacturing systems, which could play a vital role in production task scheduling and production equipment scientific maintenance decision. Therefore, it has become a research hotspot of manufacturing systems to improve the accuracy of fault diagnosis and reduce downtime. However, the current studies are aimed at equipment and do not make an integrated diagnosis of the manufacturing system from its operational mechanism. Therefore, this paper proposes a manufacturing system fault diagnosis method based on operational quality data. Firstly, in order to collect operational quality data, the relationship between manufacturing systems, production tasks, and produced products were established based on Axiomatic design theory. Secondly, the integrated health concept of the manufacturing system is proposed, that is, a healthy manufacturing system should have conditions in which the manufacturing process is stable and the produced product quality is high. Thirdly, a fuzzy comprehensive evaluation model based on entropy weight method is proposed and a manufacturing system fault diagnosis model is established to estimate the possible fault of the operational manufacturing system. Finally, a case study of the cylinder head manufacturing system is conducted to verify the effectiveness of the proposed method.