A Study on a Visible Light Communication-Based Train Positioning Method Using Particle Filtering and Tracking-Learning -Detection

Addressing the limitations of traditional train positioning methods, a visible light communication train positioning method that utilizes particle filtering and TLD target tracking is proposed to achieve continuous positioning of trains in complex environments of subway tunnels, considering that optical signals are easily obscured. In this method, LED lights in the subway tunnel serve as the transmitters and the Photodetectors (PDs) installed on the front of the train act as the receiver, and the visible light channel model is established. The RSSI dataset is constructed using the LED light signal intensity, and the TLD algorithm tracks the target through the position information predicted by particle filtering to optimize the positioning results, thereby improving the accuracy of visible light communication train positioning. In order to verify the feasibility and effectiveness of the proposed method, simulation tests are carried out for trains under three basic motion models based on real line data and equipment parameters of Chengdu Metro Line 1. The simulation results demonstrate that the positioning results obtained by this method meet the requirements of the IEEE 1474.1 standard and offer a promising new option for train positioning in future CBTC systems.