Research Of Elevator State Non Intrusive Intelligent Detection Method

China is a large country of elevator production and use,with the continuous growth of elevator ownership,the number of old elevators increases year by year,elevator trapped failure and safety accidents occur from time to time,seriously affecting the safety of people’s lives and property and the stability of economic and social development.At this stage,the supervision and maintenance model of the elevator for the regular mode,maintenance interval is relatively fixed,for the use of frequent elevators,the frequency of regular maintenance is low;and for the number of runs less elevator,the frequency of regular maintenance is too high.To this end,the General Office of the State Council issued the "General Office of the State Council on strengthening the quality and safety of elevators," according to the law to promote on-demand maintenance,to promote the "full package maintenance" and other new models.Elevator on-demand maintenance can be targeted,efficient maintenance of elevators,saving social resources.To promote on-demand maintenance,the first problem to be solved is the real-time monitoring of the elevator’s operation status.At present,the monitoring of elevator operation status is mainly realized by collecting signals from the elevator control chip,which is costly and requires invasive installation for ordinary elevator manufacturers,maintenance companies and old elevators that have been installed,and requires professional technicians to install.In this thesis,we propose a non-intrusive intelligent detection method for elevator operation status,which can monitor the operation status,position and vibration condition of elevators in real time,and design and implement a non-intrusive intelligent detection system for elevator operation status on this basis.It is of practical significance to promote the on-demand maintenance of elevators.The specific research of this thesis is as follows.(1)Designing the acceleration and velocity estimation method during the operation of elevator.The ideal state of elevator operation and the characteristics of each operation stage are analyzed,and the spatial model of elevator motion is established,while the signals measured by acceleration sensors are decomposed into vertical and horizontal directions and optimized by using the Weiszfeld algorithm;Kalman filtering and the ideal motion characteristics of the elevator are used to optimally estimate the acceleration and velocity characteristics of the elevator.The Kalman filter is used to optimally estimate the acceleration and velocity characteristics of the elevator in combination with the ideal motion characteristics.(2)The elevator real-time operation position estimation method is proposed.The information of elevator running floor is initialized by acceleration sensor,the calculation result of air pressure sensor is smoothed to reduce noise,and the automatic correction of air pressure sensor is combined with elevator motion characteristics;the state transfer function combined with floor information is designed based on the traceless Kalman filter to fuse acceleration sensor and air pressure sensor information,and the real-time elevator position is optimally estimated.(3)An on-line elevator vibration analysis and outlier detection method is designed.By recording the horizontal vibration energy of the car,the vibration baseline of the elevator in normal operation state is established,the moving average is used as the convergence criterion,and the intrinsic characteristic scale decomposition(ICD)is used to identify the abnormal values outside the baseline.According to the relevant technical conditions,the quality parameters of the elevator are calculated,and the outliers are detected by using Isolation Forest(i Forest)combined with the total running mileage and times of the elevator,the health status of the elevator is evaluated and early warning.(4)Design and implementation of a prototype elevator status monitoring and early warning system.The system includes a data collection device for monitoring status,a data transmission mechanism,a storage server supporting concurrent connections,alarm information,active pushing of elevator-related statistics and related email services,and a websocket-based web interface for data visualization.