The Residual Life Prediction Method Based On Support Vector Machine

In this paper,through the analysis of previous remaining life prediction model based on support vector machine(SVM)in engineering applications it exists the shortage of the prediction precision is not stable,etc,to the introduction of the principle of kalman filter residual life prediction model,is proposed based on SVM and residual life prediction model of kalman filter,provide theoretical basis for improving the residual life prediction accuracy.The main research content of this article has the following several aspects.(1)Under the assumption that support vector machine(SVM)input variable and output residuals are independent of each other's conditions,analyzes the distribution form of support vector machine(SVM)output residuals,illustrates the known output under the condition of the distribution of residual type distribution parameter estimation method.Under a certain confidence level is deduced the calculation method of support vector machine(SVM)output variable confidence interval,and the method is applied in predicting residual life range.Through example illustrates the necessity of residual life interval prediction,and this article puts forward the residual life of interval estimation method is effective(2)For past the residual life prediction model based on support vector machine(SVM)in engineering applications it exists the shortage of the prediction precision is not stable,etc,is proposed based on SVM and kalman filter residual life prediction model.The residual life prediction of the model is based on the previous full life test data and predicted degradation signal was a recession period,compared with other models,the residual life prediction of the model is based on a more comprehensive statistical data,higher credibility.And the iterative solving process of the proposed model is described in detail.Through the example is given to illustrate the proposed model can predict the residual life prediction,has strong engineering application value and versatility,is able to provide theoretical basis for the proper equipment maintenance strategies.