Research On Multi-car Elevator Anti-collision Method

With the development of high-rise buildings and intelligent buildings,the elevator as a vertical transport is used more widely,the elevator service quality is also put forward higher requirements,the traditional single-car elevator system is difficult to meet the growing traffic demand in high buildings.A new elevator system-"One shaft,multi-cars" elevator is the new way of vertical transportation in the world.It can raise transportation efficiency,save area and energy,reduce construction cost.Multi-car elevator system can improve traffic efficiency to increase the car in the same elevator shaft.In this thesis,for safe operation problems of Multi-car elevator control system,studying the kinematics characteristic of Multi-car elevator,discussing the Safety criterion of the Multi-car elevator,proposed the Multi-car elevator anti-collision model based on the RBF neural network and safety distance measurement technology based on binocular stereoscopic vision.This research has an important practical value and a broad application prospect.This main work of this thesis is as follows:(1)The relevant information at home and abroad is searched first,then confirms the researching task,combined with the multi-car elevator control system technology development trend,summarizes the existing multi-car elevator control algorithm,last designs the research program of the RBF neural network and safety distance measurement technology based on binocular stereoscopic vision to solve avoiding collisions problems of multi-car elevator system.(2)The system structure of multi-car elevator anti-collision has been described and the working principle has been studied,the information acquisition of the motion state such as sensor application were studied deeply.(3)For the anti-collision problem of the multi-car elevator system,based on the minimum safety distance model,the paper puts forward the control method of safety discrimination criterion.In view of the different kinematics conditions of two cars,the paper sets up anti-collision model based on RBF neural network,avoiding the collision for elevators selected in the shaft.The result of the experiment shows that the algorithm of the paper can make multi-car elevator effectively prevent collision,which provides security with multi-car elevator's safe operation.(4)The binocular vision measurement technology is used to achieve the complex multi-car elevator anti-collision system is proposed in this paper.The principle of vision binocular 3D measurement is figured out,a mathematical model for the measurement system is set up,all aspects of binocular vision measurement process are determined.The deep study of the most important step of measurement system-image feature extraction and matching is made in this chapter.An improvement scheme that make high-dimensional feature description vector matching after dimension reduction by RBF algorithm is proposed,simulation experiment by MATLAB is made and the desired matching results is obtained.After acquiring the internal and external parameters of binocular vision measurement system and feature matching,in this chapter,design the experimental environment and did the distance measurement experiment.The result of the experiment show that the multi-car anti-collision system based on binocular vision safety distance measurement technology is feasibility.(5)Summarize the study of this paper,and make a prospect to the next research.