Research on Reliability Prediction method of Complex Mechanical Product Based on Meta-action Unit

Abstract

Reliability prediction of complex mechanical product is helpful to determine the reliability state of products and that how to optimize products. Traditional reliability prediction methods of complex mechanical products come from electronic products, which need a certain amount of reliability data as support. In addition, the reliability of system complex mechanical product is generally predicted by the parts or component that make up it. However, failure of one part may not cause failure of the whole product, that will affect the accuracy of reliability prediction. To solve these problems and in order to accurately predict the reliability of complex mechanical product, a methodology based on meta-action unit (MAU) which is the basic units affecting reliability is proposed in this paper. The mechanical product is decomposed into many MAUs, and the conception of “unit quality entropy” is introduced to predict the reliability index of MAU. Subsequently, according to the reliability mathematical model of complex mechanical product and MAU, the reliability is predicted by combining interval analytical hierarchy method and eigenvector method. This paper also verifies that the method is applicable and is more accurate with a case study of the reliability of spindle box.