Short-term passenger flow prediction system and method for urban rail transit stations

Abstract

The invention discloses a short-term passenger flow prediction system and method for urban rail transit stations. The system consists of AFC equipment, video passenger flow statistics equipment, data screening module, data preprocessing module, federal Kalman filter module, neural network prediction module and It consists of a database module; the beneficial technical effects of the present invention are: it proposes a short-term passenger flow prediction system and method for urban rail transit stations. The present invention can improve the diversity, comprehensiveness and accuracy of data sources, and at the same time improve the quality of the data. Accuracy ultimately makes the prediction results more accurate.

Description

technical field [0001] The invention relates to a short-term passenger flow prediction technology, in particular to a short-term passenger flow prediction system and method for urban rail transit stations. Background technique [0002] my country's current urban rail transit has the characteristics of increasing total passenger traffic and steady growth of passenger flow intensity. Reasonable and accurate passenger flow forecasting is beneficial to passenger flow induction, safety management and operation organization of rail transit. [0003] According to different needs, passenger flow forecasting can be divided into medium and long-term forecasting, short-term forecasting and short-term forecasting; medium- and long-term forecasting (usually referring to the next 10 to 25 years) is mainly used to assist rail transit network development planning and station design; short-term forecasting (usually within the next week or one month) is mainly used for traffic status assessme...

AI Technical Summary

Problems solved by technology

[0005] In addition to periodicity and peaks under normal conditions (working days and weekends), the passenger flow change characteristics of urban rail transit stations also show differences and particularities due to abnormal factors such as holidays and large-scale urban activities. The resulting changes in passenger flow have a major impact on the safe operation of urban rail transit; among the existing technologies, relatively mature short-term passenger flow prediction technologies based on neural networks are mostly aimed at passenger flow prediction under normal conditions, and less involving normal and abnormal conditions. Combined with the problem, this makes the existing technology not systematic and comprehensive in predicting the short-term passenger flow of the site under abnormal conditions.

[0006] At the same time, in the existing technology, common passenger flow information collection equipment mainly includes automatic fare collection equipment (AFC equipment) and video image processing equipment (video passenger flow statistics equipment). For a certain type of equipment, the information source is relatively single, and the data is not necessarily reliable and accurate; in addition, the existing technology does not process the collected passenger flow information, but directly uses it in the neural network, which makes some abnormal data also Act on the neural network, thereby affecting the accuracy of predictions

Images