Passengers Organization And Parking Guide At Multi-stop Bus Station

In recent years, with the development of social economy, people’s living standards are improving. At the same time, more problems, such as traffic congestion, traffic safety, environmental pollution, energy shortages, are becoming increasingly prominent while vehicle number continue to escalate in cities. The problems brought by social development are increasingly serious plagued the daily lives of urban residents, however, the emergence of Intelligent Transport System (ITS) provides new methods and thoughts to solve the traffic congestion problems and traffic safety problems. As the core content of Intelligent Public Transport System, information services provides a direct communication platform for transit passengers and urben public transport system, and achieve the goal that public transport users, operators and the managers of relevant government department can real-time share of public transport information. Urben public transport information service system can improve some present situation like the low scheduling adherence, the inconvenient transfer, the poor security and so on. On this background, this paper studies the dynamic stop allocation at bus stations and intelligent guide of passengers. This paper aims to reducing delays in bus stations, to improve the efficiency of public transport effectively. It can attract more potential public transport users and achieves the goal of relieving traffic pressure.Based on previous research, this paper get and analyzed the data of No.63bus line in Jinan by the GPS data of bus management terminal. We found that under the influence of same road conditions and traffic signals, the bus running time between two sites has strongly properties of randomness. However, it also has periodic fluctuation with the departure date and time. Bus dwell stop is related to departure time and headway. Therefore, this paper presents a combination forecasting model of bus running time based on Kalman Filter and Artificial Neural Networks. For bus dwell time prediction, we propose a model based on Artificial Neural Networks. We used the data at other stations of No.63bus line to prove that these two models have high accuracy in real-time prediction.Bus running time prediction and dwell time prediction provides support for the dynamic stop allocation at bus stations. This paper respectively analyzed the behavior of the bus entering and leaving the station when bus should follow the queue or can overtake to enter the station. Then we proposed a dynamic allocation stop algorithm which based on the shortest waiting time of all passengers. Case study chose the data of north Shungenglu bus station in Jingshi Road, Jinan. We analyzed the impacts of different stop number, different error of running time prediction and different errors of dwell time prediction. The result proved that the dynamic stop allocation at bus stations can reduce passengers’ waiting time and the delay of bus entering the station when the accuracy of prediction can be ensured.