Research On Technology Of Residual Life Prediction Of Electric Water Heaters' Relay

With the rise of the concept of smart appliances,the growing demand for traditional home appliances intelligent is reflected.This concept puts forward new technical requirements to the traditional home appliance industry,at the same time,it also points out the new space for innovation and innovation direction.Life visualization prevent relay failure occurrence when electric water heater has no heating and still heating,change from passive to active warranty maintenance,reduce the losses caused by sudden faults,extend the life of the water heater,so as to enhance the core competitiveness of products,and plays a fundamental role and inspiration for the research to the water heater relay and other weak links by providing users timely and accurate feedback of residual life relay.In this paper,the Bayes update and BP neural network based on GA is applied to predict the residual life of the electric water heater.The main research contents and innovation points are summarized as follows:Firstly,based on market feedback and historical experiments the failure mechanism and the influence factors of residual life of electric water heater are analyzed.Two kinds of failure mode and the key factors on effect of the life of the relay are obtained,and screen out the characteristic quantity which can not only characterize the degradation process of the relay the relay,but also and easy measure and measurement cost is low.Secondly,the acceleration model and the acceleration factor of the electric water heater relay are obtained based on the model of the physical acceleration by using the historical experimental data.Based on this model,the design and data acquisition of the relay accelerated life test are carried out,which provide data support for the following modeling and prediction.In addition,the failure criterion of the relay is extracted according to the censored test data.Thirdly,model is build based on the Wiener process degradation.Evaluate the overall reliability of relay group on the stress condition.According to the Wiener process degradation modeling and Bayes update method to predict the single relay residual life,and compared with the linear regression model of stochastic coefficient,proving the effectiveness of the method.Fourthly,based on time series analysis,the model of BP neural network time series rolling prediction based on GA is build.It can predict real-time five attributes of the relay,and the remaining life prediction results are given according to the results of rolling relay.By comparing with the standard BP neural network prediction results,the results show that the GA_BP algorithm is better than the BP algorithm.Finally,based on the study of the residual life prediction technology of the electric water heater relay,the requirement analysis,function design and database design of the residual life prediction system are completed.