Proposed Position Detection Method by Recognizing Object Layout via LiDAR Sensor

A train control system detects train positions using a ground devise. Railway companies maintain and renovate the system. On the other hand, if the ground devise fails, then the workers of railway companies will be dispatched to the site to perform maintenance. Hence, it is necessary to secure a certain number of workers. Given that the declining birthrate and aging population pose challenges in securing sufficient workers, it will be necessary to secure labor with respect to ground devices. In this study, we propose a new position detection method that performs position detection on-board to save labor with respect to ground devises. Specifically, first, location information, such as latitude, longitude, and kilometer, is linked to the layout of objects on the track. Second, the method detects objects using a LiDAR sensor installed on-board and then recognizes the layout of the objects. Third, position information is retrieved by matching the recognized layout with the database on-board. Finally, a train position is detected using the position information and relative position between the train and objects. The results of tests using the vehicle indicate the feasibility of the method in recognizing the layout and detecting the train position.