Track Section Occupancy Detection Model Based on Infrared Ray Sensor and Time-Series Change Rate Matching

Track section occupancy detection is one of the key problems in the field of railway transportation and train dispatching, which has strong link with the safety of railway traffic. In order to reduce the cost and improve the accuracy of track section occupancy detection, this paper presents an algorithm based on infrared ray sensors and the time-series change rate matching. Once a train passes, the infrared ray sensors installed on both sides of the track collect the signal time-series of a wheel set where various noises exist. This paper makes use of the similarity between the change speed and the acceleration of the standard time-series and the time-series collected to realize the similarity matching and noise recognition, which can infer the direction of the running train and the number of its wheel sets by analyzing and judging the signal time-series and finally achieve the task of track section occupancy detection. The simulation data and field test results demonstrate that the proposed algorithm improves the accuracy, availability, and reliability of track section occupancy detection based on infrared ray sensors, with a correct rate up to 99.999%.