A method and system for predicting the arrival time of a train

Abstract

The invention discloses a method and system for predicting the arrival time of trains on the whole line. By statistically processing all running records of trains at different time periods between stations, the time when operating trains are located on the whole line is accurately predicted in real time, and the accuracy is high and the real-time performance is strong. The technical solution is: discretize the track lines, record the running time of the train on the track, and then convert the running time into the time-consuming value of arrival at the station. Average the time-consuming values ​​of multiple trains in multiple time periods, and use the segmental interpolation method to fill in the points without time-consuming values ​​on the track. Perform mean filtering and smoothing on the obtained line data of the whole line, and ensure that the time-consuming data in each station conforms to the law of monotonically decreasing. Complete the production of train running time-consuming data for the whole line section. In the application, when the system receives the train position data, it can obtain the train arrival prediction value by looking up the time-consuming data. By making time-consuming data of trains in different periods, it is possible to more accurately predict the time-consuming of train arrivals in different periods.

Description

technical field [0001] The invention relates to a technique for predicting the arrival time of a train, in particular to a method and a system for accurately predicting the arrival time of a train at each platform by means of statistical processing. Background technique [0002] At present, vehicle arrival prediction methods are concentrated in the field of public transportation, and vehicle GPS data is mostly used for vehicle positioning. The main forecasting methods include time series analysis, Kalman filter, artificial neural network, etc. However, due to the complex road conditions of the bus vehicles, the large number of highways involved in the traffic flow, and the deviation between GPS and road positioning, the patents for bus arrival prediction focus on the time prediction between multiple stops and the prediction of the vehicle arrival distance. [0003] In addition, in terms of subway applications, due to the high density of subway departures, there are fewer pa...

AI Technical Summary

Problems solved by technology

However, most of the current arrival predictions in the existing technology are inter-station arrival predictions, so the obtained prediction results are not real-time, and it is difficult to provide a solution that can provide accurate results

The existing prediction method treats the lines between two stations equally, and there is a large error in the prediction accuracy

Images