Font Size: a A A

Research On Rail Transit Foreign Matter Intrusion Detection And Train Operation Control Strategy

Posted on:2022-07-02Degree:MasterType:Thesis
Country:ChinaCandidate:Z L TanFull Text:PDF
GTID:2531307070456094Subject:Transportation engineering
Abstract/Summary:PDF Full Text Request
In recent years,with the rapid development of rail transit in China,the safety of train operation has become the focus of social concern,wherein foreign matter intrusion is one of the key factors of influencing train operation safety.Therefore,accurate detection of foreign matter intrusion is an urgent problem.The existing railway foreign matter intrusion detection technology has the problems of low automation degree,insufficient judgment of the risk degree of intrusive foreign matter,inadequate detection result application and the like.Accordingly,the research on rail transit foreign matter intrusion detection and train operation control strategy has important significance for train operation safety and efficiency.On the basis of video image captured,a rail transit foreign matter intrusion detection and train operation control strategy are studied.The specific contents are as follows:(1)Dynamic division of detection area.Simulation calculation is performed on the boundary relationship corresponding to the train speed and the aerodynamic force by using the FLUENT software,and the detection area is dynamically divided into a dynamic pedestrian danger area,a dynamic high-risk area,a potential risk area and a safe area according to the danger degree.(2)Recognition and classification of intrusive foreign matter.Background frame subtraction,three frames subtraction with fusing Gaussian Mixture Model and the Mask RCNN object detection algorithm are combined to realize identification and classification of the intrusive foreign matter,and obtain the attribute and motion state information of the intrusive foreign matter.(3)Trajectory tracking and prediction of dynamic foreign matter.Trajectory tracking is performed on the dynamic foreign matter by using CSR-DCF(Discriminative Correlation Filter with Channel and Spatial Reliabilit),and trajectory prediction is performed on the dynamic foreign matter in the safe area by using extended Kalman filtering to obtain the prediction result of the dynamic motion state of foreign matter of the safe area.(4)Danger level division of intrusive foreign matter and train operation control.According to information such as the attribute of the intrusive foreign matter,danger level division is carried out through the fuzzy logic system.The speed protection curve corresponding to each danger level is formulated by using the target-distance control mode.The train operation safety is ensured,and the train operation efficiency is improved.
Keywords/Search Tags:Rail transit, Foreign matter intrusion, Train operation control, Mask R-CNN algorithm, Speed protection
PDF Full Text Request
Related items