| Summary: | Accurate positioning of trains is of great importance to the safety of train operation. In some foreign subway projects, limited by the maturity of technical system and control of design cost, conventional GPS and CBTC positioning technologies cannot meet the demand of train positioning in the projects. Therefore, a nonlinear compensation based self-positioning method for rail transit trains is proposed. This method can achieve precise positioning of trains solely relying on existing onboard equipment without adding new equipment such as GPS, CBTC, etc. The proposed method features simple composition, low cost, high accuracy, and high reliability (free from environmental affection), which is worth further promotion. Firstly, the specific process used neural networks to establish a compensation model for the instantaneous speed of trains and improve the accuracy of train instantaneous speed through the compensation model, and then the real-time running distance of the train was obtained by integrating the compensated real-time speed and operating time of the train. Afterwards, the distance of the train to the next station was calculated based on the fixed train operation diagram, thus to realize positioning. The experimental results show that after compensation the positioning accuracy is less than 0.2 m, upgraded by 3-4 times of the accuracy before compensation.
|
|---|
