Research On Characteristic Parameters Testing And Degradation Regularity Of Electromagnetic Relay

Electromagnetic relays, as a kind of typical low-voltage electrical components, have been widely applied in automation, power system, aerospace and other fields. The reliability of relay affects the stability of the entire device or system. The performance of relay will degrade with working duration grows, and the degradation data of performance parameters contains a large amount of information related to reliability and remaining useful life. Hence, testing characteristic parameter and researching on reliability assessment, life prediction as well as degradation regularity of electromagnetic relay is of great significance both on theoretical and practical value.Determine the key characteristic parameters which can reflect the degradation process of the relay based on the working principle analyzing, failure reason and dynamic curve of the relay. Design and build the hardware system of testing device and write the acquisition, calculation and storage program for the characteristic parameters of the relay. In the action process of relay, the testing system is used to collect the waveform data of contact voltage and the coil current in real-time, calculate and display the characteristic parameters(pick-up time, super-path time, contact resistance, thermal time).After processing the degradation data of key characteristic parameters, investigate and analyze their degradation regularity combined with theoretical research and the trend of the data. In view of the degradation process of performance and the characteristics of degradation data comprehensively, establish several models for reliability assessment and life prediction, which are based on random Wiener process, random Gamma process and Weibull distribution respectively. And the parameters involved in the models are estimated according to the experimental data. T.he remaining useful life of relay is predicted by the models. In order to improve the accuracy of life prediction, construct a combined forecasting model based on the single model. The case of application indicates that the approach is effective and feasible.