Safety Assessment Method Of Traction Elevator Running State For Improving SVM

With the rapid development of my country’s economy,the number of elevators continues to grow.Due to the catastrophic and social influence of elevator accidents,the current evaluation methods for the operation of traction elevators were relatively primitive and backward,Low and mostly subjective methods,so it is necessary to study the objective evaluation method of elevators.The research contents of this paper are as follows:(1)Traction elevator safety assessment system.With the development of sensors and the rise of the concept of elevator Io T,the evaluation indicators of elevators tend to be more quantitative.This paper of a more scientific evaluation system based on the reference to national standards and the research of previous researchers was established,combined with the analysis of elevator failure mechanism.(2)Multi-level variable fuzzy comprehensive evaluation method.In order to determine the index weight more accurately,in the method selection,taking into account the interconnectedness of each index of the elevator,the CRITIC method was used as the objective weighting method;the AHP method was used as the subjective weighting;and the distance function was introduced to make the above two weightings.The degree of difference between the weight values of the weight method was consistent with the degree of difference between the corresponding distribution coefficients,so the determination of the combined weighting was more reasonable.And its evaluation results will be used as the output of machine learning training and the discrimination of accuracy during testing.(3)Optimization of the index system.Since the index system established by relying on previous research,mathematical statistics and experience may have some redundancy,for the simplicity and accuracy of the calculation,the MIC method to more comprehensively analyze the correlation between the indicators of the traction elevator was adopted this paper,find the index with strong correlation and the elevator subsystem with redundancy;then,select the features of the system with redundancy by MIC-CFS method to get the optimal index subset,thus constructing a scientific and concise elevator system Pull-in elevator operation status system.(4)Improve the realization of the safety assessment of elevator operation status by SVM method.The established safety evaluation system index was used as the input of SVM,the result of fuzzy comprehensive evaluation method was used as the output of machine learning,and then machine learning was carried out through the optimization of SVM based on GE,GWO and IGWO methods.The results of the safety evaluation of the elevator operation state show that: IGWO-SVM has better accuracy in the safety evaluation results of the elevator,which verifies the effectiveness of this method.