Research And Application Of Bus Transfer Algorithm Based On Real-Time Traffic

In recent years, as the development of science and technology, people’s living standards has been greatly improved, the process of urbanization has been greatly promoted, but all these things bring more and more serious problems. As the result of the expansion of cities urban public transport system is becoming more and more developed, the complexity of the bus lines make people’s travel more and more inconvenient. The increase of population density and number of motor vehicles brings huge pressure to the bus system and makes the traffic situation more serious. In order to improve the public transport system’s attraction to people, reduce the use of private cars and reduce the pressure on urban traffic, this paper puts forward an effective, efficient and impeccable public transport transfer algorithm based on real-time traffic.First, we analyzes the domestic and foreign scholars’research results on people’s travel problems, we find that improving service quality of public transport system can make people use public transport more, so this paper proposes six different transfer plans; after the analysis of research on public transport transfer algorithm both at home and abroad, this paper puts forward the algorithm model which contains two parts: static transfer algorithm and dynamic transfer algorithm; static transfer algorithm uses the two technologies:data base and mathematical modeling which can improve the query performance of this algorithm; on the basis of the static transfer algorithm, we use artificial neural network project the time-consuming, the predicted values will be the final judgment conditions; we use lots of history data to train the artificial neural network,and input the real-time state of traffic and buses’ state into the network so that we can get the predicted value of time, according to the predicted value the static transfer plan reorder to get the dynamic transfer plans.This paper also do system tests on static transfer algorithm’s accuracy, coverage and query efficiency, the results present that the algorithm can not only provide six transfer model, but also has high accuracy, high coverage and efficient query efficiency. By contrast to static transfer plans, dynamic transfer plans which uses neural network forecast can adjust according to the time spot of the query, provide real-time transfer plans; by contrast with the baidu transfer tool, the dynamic transfer plans can provide fine projections for journey time consuming, and has better guidance for people’s travel. Finally, we summarize the work and achievements during the graduate studies.