Method and system for bus real-time arrival time prediction

• Main Authors: YU ZHIGUO, RUAN FANG, ZHANG HUI, SUN MEIYU, YANG KAI, MIAO JIAN, CHENG HEBIN, MENG LINGGUO, WU XUESONG
• Format: Patent
• Language: Chinese; English
• Published: 01.03.2019
• Subjects: CALCULATING; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; PHYSICS; SIGNALLING; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR; TRAFFIC CONTROL SYSTEMS

The invention discloses a method and a system for bus real-time arrival time prediction. The method comprises: using an offline learning model, performing multi-variable neural network learning on arrival time of a bus in a certain line in historical time, and storing a three-dimensional multivariate parameter model of the offline learning model in allowable error, the three-dimensional multivariate parameter model having minimum error; performing online learning by using the online learning model according to actual arrival time of the bus, and dynamically adjusting the arrival time of the bus. The method predicts using a method combining the offline learning model and the online learning model, so as to effectively improve accuracy of prediction. The method is novel and unique, the design concept is clever, a prediction result is accurate, application environment is good, and market prospect is wide. 本发明公开了种公交实时到站时间预测方法及系统，利用离线学习模型将某线路公交车在历史时间中到站时间进行个多变量神经网络的学习，在允许误差内将误差最小的离线学习模型三维多变量参数模型保存；根据公交车实际的到站时间利用在线