Research On Typical Fault Diagnosis And Accident Prevention Technology Of Elevator Based On Machine Learning

Elevators are widely used in high-rise buildings because of their convenience and speed.However,elevator accidents occur frequently.According to statistics,elevator faults and improper behaviors of passengers are important reasons for frequent elevator accidents.Therefore,it is of great significance to study the elevator accident prevention technology based on elevator fault diagnosis and passenger behavior recognition.In this work,a typical elevator fault diagnosis model based on machine learning is proposed,an elevator passenger behavior recognition method based on two-stream CNN is proposed,and an elevator accident prevention system is established.The main research work of this paper is as follows:(1)Elevator operation data analysis and mining.Based on the principle of elevator typical fault monitoring,the elevator operation data collected from an elevator Io T system are cleaned and processed by feature engineering,and the elevator operation dataset which can be used for machine learning is obtained.(2)Typical elevator fault diagnosis model.After analyzing the principle and applicability of the machine learning classification algorithm,SVM,RF,and GBDT algorithms are selected to establish a typical elevator fault diagnosis model respectively.Furthermore,four fusion models based on these three models are proposed.These models are trained and evaluated on the elevator operation dataset.(3)Elevator passenger behavior recognition model.An elevator passenger behavior recognition model based on two-stream CNN is proposed after analyzing existing elevator passenger behavior recognition methods.In this work,we collect and make a video dataset of elevator passenger behavior.And the elevator passenger behavior recognition model is trained and evaluated on the video dataset.(4)Elevator safety management information system.Based on the typical elevator fault diagnosis model and elevator passenger behavior recognition model,an intelligent elevator safety management system is designed.Integrating the supervision of passengers and elevators,the system can effectively prevent the occurrence of elevator accidents.