Font Size: a A A

Design And Implementation Of Four Locomotive Safety Control Systems Based On Computer Vision

Posted on:2020-07-06Degree:MasterType:Thesis
Country:ChinaCandidate:Y J WangFull Text:PDF
GTID:2481306353452274Subject:Control Engineering
Abstract/Summary:PDF Full Text Request
Coke is an important raw material for the steel industry.Increasing the speed of the four large locomotives is the most direct and effective means for coking enterprises to increase production.In the background of speed increase of four locomotives,this paper proposes a method to replace the pedestrians in the locomotive running area by computer instead of the locomotive driver in order to ensure the pedestrian safety in the moving area of the four locomotives.This method mainly aims at pedestrian detection.The SSD(Single Shot MultiBox Detector)algorithm is used to design and implement the four-locomotive safety control system for coke oven based on computer vision,which solves the problem of pedestrian safety in the locomotive running area.The specific work is as follows:(1)A closed-loop method commonly used by coking enterprises at home and abroad to ensure pedestrian safety in the locomotive running area is reviewed.Since the four locomotives often require overhaul and cleaning,passively closing the running area is not the best solution.In response to this problem,this paper proposes a computer to correctly judge the sudden safety incidents of the four major locomotives during the running and operation,and provides a new strategy for solving the safety problem of locomotive operations.(2)Briefly introduce the current coking production process,process and the role of the four locomotives,analyze the safety hazards of the existing four locomotive automatic control systems,and explain the need to increase the visual recognition and interlock control system of the four locomotives.(3)In-depth study of the current advanced SSD target detection method,based on the method,designed and implemented the visual recognition system of the four locomotives.To verify the effectiveness of the system,a pedestrian data set was constructed.The results show that the accuracy of the system in the locomotive on-site operating environment is more than 90%,which meets the site requirements.(4)Taking the coke car with the fastest speed among the four locomotives as an example,the scheme of signal transmission and interlock between pedestrian detection system and locomotive running control system is designed.Based on this,a computer vision based locomotive is proposed and realized.The safety control system provides a new idea for solving the pedestrian safety problem in the locomotive running area.
Keywords/Search Tags:coke oven, four locomotives, computer vision, pedestrian detection
PDF Full Text Request
Related items