| Subway has become an inevitable choice to alleviate the pressure of public transportation in metropolises.Communications-Based Train Control systems are the mainstream train operation control system for subway.Train positioning information obtained by the positioning technology is one of the important parameters to ensure the safe train operation.The traditional positioning methods of subway train include balise,cross-induction loop,axle counter,WLAN,LTE and so on,but the above methods are all facing certain shortcomings,which are unable to meet the needs of continuous positioning,high cost and difficult to maintain,low precision,susceptible to electromagnetic interference,limited spectrum resources,etc.With the deep research of visible light,the advantages of Visible Light Communication(VLC)technology highlight in various industries such on high transmission rate,taking into account lighting and communication,green environmental protection,and free spectrum resources.Among them,the positioning technology with VLC,using Image Sensor(IS)as the receiver,inherits the above advantages,and also has the advantages of high accuracy,low cost,excellent portability,and easy system construction.Aiming at the tunnel environment in the process of subway running,a positioning method for subway train based on Optical Camera Communication(OCC)is proposed.Firstly,according to the tunnel scene of subway,the train positioning model based on OCC,train moving model and observation model of train positioning have been constructed.Secondly,according to the inherent arrangement of LED lamps in subway tunnels,a positing algorithm with OCC based on dual LED lamps is designed,which makes use of cameras to catch LED lamps.Region of Interest(ROI)is extracted from LED lamps by image processing technology to identify the features of the striped images within the ROI.Next,ID information of the LED lamps is obtained by machine learning,and the train position information is achieved through geometric principles.Finally,in view of the possible situation that two LED lamps cannot be captured,a positioning algorithm with OCC based on single LED lamp is proposed,which uses the shape information of the LED lamp to solve the scale factor k.In addition,in order to reduce the scale factor error when the imaging is tilted,an anti-tilt algorithm is proposed to improve the accuracy of the positioning algorithm.Relying on the Gansu Provincial Artificial Intelligence and Graphics and Image Processing Engineering Research Center,an experimental platform is built to verify the positioning performance of the proposed positioning algorithm with dual-LED lamps.According to the actual line and equipment information of Chengdu Metro Line 1,MATLAB is used to conduct simulation experiments to verify the accuracy and effectiveness of the positioning algorithm with single LED lamp.The experimental results show that the average positioning error of the proposed positioning algorithm with dual-LED lamps is 2.34 cm,and the average time of the proposed positioning algorithm is 51.24 ms.The average positioning error of the proposed positioning algorithm with single LED lamp is 15.56 cm.The positioning methods proposed in the thesis can meet the requirements of the IEEE 1474.1 standard for the train positioning accuracy of CBTC systems,have real-time performance and ensure the stability during the positioning process,which have wide application prospects. |
