The Research On Elevator Safety Assessment Based On Multi-information Fusion

Elevator is an essential part of modern daily life,and its amount of investment is increasing.But at the same time,frequent accidents of elevator safety also attract people's attention.Once the elevator accidents occur,the casualties and property losses are unimaginable,so it is very necessary for the elevator safety management.At present,the method of regular maintenance is mainly used to ensure the safety of the elevator,but the reliability and timeliness are not enough.In this paper,a safety assessment model of traction elevator(hereinafter referred to as the elevator)is built,and the dynamic elevator safety assessment based on multi information fusion is realized.This paper first analyzes the parts of the system that makes up the elevator,and summarizes the type of fault and the cause of the failure.Based on the analysis of the relationship between the elevator's index parameters and faults,a safety evaluation index system for elevator is built based on the evaluation index of elevator safety.Aiming at the problem of excessive subjective weight and imperfect information in current elevator safety evaluation,two elevator safety evaluation models based on multi information fusion are established: The model of elevator safety assessment based on grey correlation analysis algorithm and the model of elevator safety assessment based on evidence theory and cloud model.For the former take the elevator safety evaluation index by AHP weight analysis method to calculate the correlation degree between the current running state of the elevator and the best operation state using grey correlation,construct grey correlation matrix,to achieve the primary assessment of the state of the elevator safety and two grade comprehensive evaluation.In the latter part,the objective weight of system index is calculated by entropy weight method.In order to take account of the equilibrium between indexes,the variable weight of system index is derived based on equilibrium function,that is,the weight of corrected index.Then we use cloud model theory to get membership degree of system index for all safety levels.Finally,we integrate the index information based on DS evidence theory to achieve quantitative evaluation of elevator risk.The example shows that the evaluation results of the two models are in accordance with the actual situation,and the reliability of the two models is verified.At last,a safety assessment software is developed based on the above evaluation model by programming language,and the real-time online risk quantitative assessment of elevator running state based on multi information fusion is realized.