부산대학교 도서관

The track circuit (TC) is a crucial safety equipment in the railway signal system. It is responsible for train occupancy monitoring and train-ground communication. The information transmission model of the TC is essential for studying its transmission characteristics and working state. Hence, an information transmission model for the ZPW-2000A TC in the adjusted state is proposed by using mixed data. Compared to the four-terminal network (FTN) model of uniform transmission line, the real working state of the ZPW-2000A TC is more accurately simulated by the proposed model. To establish the model, a FTN model of the TC in the adjusted state is constructed. By utilizing practical data and simulation data as the dataset, a BP neural network (BP-NN) model is trained to represent the information transmission model of the TC. The effectiveness of the proposed model is verified by the experimental results.