Research On External Cause Analysis And Prediction Algorithm Of Community Elevator Fault

The safety and reliability of elevator has always been a problem that human attach great importance to aiming at the field of elevator fault prediction,the basic idea of elevator fault prediction research at home and abroad is to predict the elevator equipment fault according to the elevator fault situation and maintenance records.According to the small sample nature of the elevator fault data used for community elevator fault prediction,this paper uses the small sample data analysis method to analyze the external cause of elevator fault data set and research the prediction algorithm.The main work contents are as follows:First,the external factors affecting the failure of the community elevator are studied and analyzed,and the eigenvalues of the external factors related to the failure of the community elevator are selected and calculated to preprocess the elevator fault data set,then useing clustering model based on the fusion particle swarm optimization algorithm is designed.The model uses the community elevator fault data set preprocessed in the first step to obtain the analogue elevator group fault data set through data clustering;Second,the elevator fault data set with analogue ability obtained by the above elevator fault data clustering model is utilized,and the ensemble learning Boosting algorithm is used to build a community elevator fault prediction model.Fault prediction training is conducted for each elevator group fault data set respectively.According to the fault prediction results of different elevator group fault data sets,the evaluation function algorithm is constructed and the fault prediction effect is compared and analyzed.The fault prediction evaluation index parameters are used to sort the different elevator fault prediction learning models,and finally the community elevator fault prediction is completed;Third,using the training model algorithm elevator fault external cause analysis and prediction software writing community,community elevator fault prediction model simulation,the elevator maintenance enterprises in research area and 96333 emergency response platform accumulated community elevator failure data as the foundation,community elevator failure prediction,a cloud of forecast and evaluation result is stored in the database at the same time.The results of training and learning show that the elevator fault data set clustering and prediction model constructed in this paper can improve the accuracy of community elevator fault prediction compared with the general regression prediction algorithm.At the same time,the next maintenance measures can be deployed by using the prediction and evaluation results of the elevator fault in different communities,so as to avoid casualties and equipment losses and realize the maximization of safety and economic benefits.