Urban rail transit station sudden large passenger flow early warning method based on data driving

Abstract

The invention provides an urban rail transit station sudden large passenger flow early warning method based on data driving. The method comprises the steps that minute-level time domain signals of station passenger flow volume are extracted from multi-source automatic fare collection data, and sliding time window characteristic parameters conforming to passenger flow signal time-varying characteristics are counted through the minute-level time domain signals; training a signal classifier based on a deep belief network, extracting high-order implicit features of the passenger flow signals, andperforming adaptive mode division on the passenger flow signals; and calculating a local abnormal factor value of the passenger flow signal in the sliding time window according to the sliding time window characteristic parameter and the self-adaptive mode division result of the passenger flow signal, and when the local abnormal factor value is greater than a set large passenger flow detection limit, carrying out large passenger flow risk early warning. The method is driven by the multi-source automatic fare collection data, can adapt to a dynamic incremental data environment, has the technicaladvantages of being high in feasibility and stable in detection result, and provides scientific method support for urban rail transit passenger flow early warning work.

Description

technical field [0001] The invention relates to the technical field of urban rail transit. More specifically, it relates to a data-driven early warning method for sudden large passenger flow at urban rail transit stations. Background technique [0002] With the continuous expansion of urban rail transit networks in major cities in my country, rail transit has become the backbone of urban comprehensive transportation systems, undertaking the task of meeting the diverse travel needs of passengers. The passenger flow demand of the urban rail transit system is significantly affected by external factors of the city, especially when large-scale cultural and sports activities, trade exhibitions, ground traffic control, and severe weather occur. When a sudden large passenger flow occurs, a large number of passengers flood into the station in a short period of time. If the warning is not given in time, it will cause overcrowding in the payment area of ​​the station, a sudden increas...

AI Technical Summary

Problems solved by technology

[0005] At present, there is no early warning method for sudden large passenger flow based on the passenger flow characteristics of urban rail transit and adapting to the dynamic incremental data environment in the prior art

Images