关键词: 交通运输工程, 智能交通, 公交出行, 客流量, 行程时间预测, k-NN, 模糊专家系统

Abstract: Bus is an important means of transportation for urban residents. Being disturbed by the uncertainties of journey time, passenger flow etc., the bus travel often causes delay of passengers. To solve this problem, a reliable bus travel service model was proposed aiming at improving the whole process of bus travel. The proposed model ensured that passengers could arrive at the destination on time and save time effectively. Firstly, the LSTM (Long Short-Term Memory) algorithm was implemented to predict bus journey time, and to provide passengers with target vehicles. Secondly, the influence of dynamic passenger flow on bus travel was taken into consideration. Then, a fuzzy expert system was established to evaluate the passenger flow of public transportation, which was used for planning reasonable departure time and avoiding delay caused by large amount of passenger flow. The passenger flow data in fuzzy expert system was obtained by k-NN algorithm. The actual case shows that the proposed model can provide reliable travel strategy for travelers. The research results have a positive role in promoting bus service level and developing smart travel.

Key words: traffic and transportation engineering, intelligent transportation, bus travel, passenger flow, travel time prediction, k-NN, fuzzy expert system