基于GM(1,1)的残差修正模型的电梯故障率预测

为研究某城市某品牌电梯故障率发展趋势,建立了该城市该品牌电梯故障率的GM(1,1)灰色预测模型,并对所建模型进行了数据检验,检验结果表明该预测模型的预测精度波动较大。为了提高GM(1,1)灰色预测模型的预测精度,利用对模型进行数据检验时得到的残差序列,建立GM(1,1)灰色预测模型的残差修正模型,利用该残差修正模型对原预测模型进行修正。利用经残差修正模型修正后的故障率预测模型对该城市A品牌电梯的故障率进行预测,结果表明:1)残差修正模型对原模型修正后的相对误差与修正前相比有升也有降,但精度有所提高且趋于稳定,表明残差修正模型有利于提高预测精度;2)利用所建立的故障率预测模型求得的预测故障率与实际故障率相比,相对误差不超过8.010%,表明该故障率预测模型的预测精度较高;3)修正模型预测值表明,在现有状态下该城市A品牌电梯的故障率呈上升趋势,应加强该品牌电梯的检维修与管理。

Prediction of the elevator failure rate based on the residual error correcting model of GM (1,1)

This paper intends to bring about our study of the failure rate predictioning model for elevators of Brand A produced in a certain city.There must be existing a series of complicated factors that affect the failure rate of the elevator or an elevator system,which are in some ways turning up from time to time because of a lot of uncertainties.It is just for such uncertainties that have pushcd us to make efforts to disclose the myth of the problem by using a grey system.In order to reduce the impact of the multi-factors coupling on the prediction of the elevator failure rate,we have built up a GM (1,1) grey prediction model of the failure rate based on the failure rate data from the 96333 Emergency Handling Platform in Jan to Aug of 2015.The prediction data system has been made validated by using the actual failure rate statistics formula.The validated results of the failure rate statistics finally indicate the similar failure incidence rate between the prediction rate and that of the actual scoring.Nevertheless,there remains a gap and fluctuation between the actual statistics and that made by the prediction model.In hoping to heighten the precision of the prediction model,we have worked out a residual correction model,which is expected to be used for correcting the results done by the original prediction model.As a result,the prediction model corrected by the supplementary residual correction model,can be applied to predict the failure rate of the brand elevator.The predicted results the eventually gained can be shown as follows:(1) The relative error of the supplementary residual correction model to the residual correction tends to rise or fall as compared with the original one,but the predictioning accuracy should be improved and stabilized.Thus,it can be conformed that the residual correction model can be really helpful to heighten the predictioning accuracy;(2) The relative error of prediction failure rate as compared with the actual failure rate turns to be less than 8.010％,which manifests that the accuracy of the prediction model corrected by the residual correction model proves to be acceptable;(3) According to the prediction failure rates,the failure rate of the elevators may turn to be higher in the future.In order to decrease such kind of elevator prediction failure rate,it is of great need to strengthen the inspection,maintenance and management of the elevator under service.