The Method Of Relay Life Prediction Based On The Initial Life Information

Electrical connections between the components or integers even systems are requiredin modern systems. If there is a connection failure, the reliable operation of the wholesystem may be affected. Relay is an important basic component in all kinds of automaticcontrol equipments. When the workplace of relays has very high demands for the reliability,the life test should be done to strictly examine the quality of them. The life prediction ofrelays has important implications for screening the relays of high quality and ensuring thesafe operation of the workplace of relays. However, the life prediction of relays relieslargely on the whole life information and mainly focuses on the prediction of batchedprediction. How to achieve the life prediction of single relay under the condition ofincomplete life information is the objective of this article.Firstly, the failure reason of relay is ascertained through analyzing the failure modeland the failure mechanism of relay. The research shows that the contact failure is the mainfailure reason. The contact process of relay includes bounce process and dynamic contactprocess. After analyzing sufficiently the two processes, we can know that the selectedperformance parameters should be closely associated with the two processes. Inconsideration of the relay working condition, the selected performance parameters in relaylife test include temperature, bounce time, dynamic fluctuation time,dynamic peak voltagedrop and static contact voltage drop.Secondly, after getting the12sets of life test data of the HH52P model relays, the8sets of top10%relays life test data from the12sets are selected as relays initial life data.The decision table of performance parameters life and the decision table of performanceparameters time-varying characteristics are developed based on rough set theory.The formerreflects the relationship between the performance parameters of relays initial life and therelays life; the latter reflects the relationship between the performance parameters changingrules of relays initial life and the relays life.The simplified performance parameters set andthe simplified performance parameters time-varying characteristics set can be got by usingthe attribute reduction algorithm. The validity of the reduction results can be proved,and theextracting priority of association rules can be determined, by computing the attributeweights. And then the performance parameters-life association rules and the performanceparameters time-varying characteristics-life association rules can be got.Finally, the rest4sets of life test data are selected to determine the support-degree ofeach association rules through the support-degree algorithm.The purpose of life predictioncan be got by transforming the association rules into words. This method is able to minepotential information from the existing relays life test data. The conclution of single relay life can be got by the only need of initial life performance parameters values andperformance parameters time-varying characteristics.