Research On Location Technology And Arrival Time Prediction Algorithm For Public Transport Vehicles Based On CAN BUS

With the rapid development of national economy and the dramatic speeded-up of urbanization process, the city traffic congestion has been becoming a social problem, which influences the quality of people's life. Intelligent Public Transportation System (IPTS) developed in recent years can relieve the pressure on urban transportation efficiently, which can realize reasonable dispatch for the public transport vehicles by modern and advanced electronic, information and communication technologies based on the existing public transportation. Hence, the initiative of citizen taking bus can be enhanced greatly.In this paper, a location technology and an arrival time prediction algorithm for public transport vehicles in intelligent public transportation system are studied and the corresponding algorithm is proposed respectively.As for the location for public transport vehicles, a location technology based on CAN BUS is given in this paper. The operating data of the public transport vehicle are extracted and transmitted to the bus port by CAN BUS and GPRS network, respectively. Then these data are processed by data processing module at bus port, thus the bus can be located. The corresponding hardware module on vehicle is designed for the location method.As for the arrival time prediction of public transport vehicles, an arrival time prediction algorithm based on CAN BUS is proposed in this paper. The arrival time is divided into three parts, which are traveling time without stop on road, staying time on the station and residence time on the road intersection. The three parts of the time are predicted separately by the linear regression analysis method based on the operating data extracted for the public transport vehicle. So the arrival time of bus can be predicated ultimately.The arrival time prediction algorithm based on CAN BUS for public transport vehicles is verified by experiments. The experimental results show that the proposed algorithm has higher accuracy and strong practicability.