关键词: 城市交通, 短时客流预测, 卡尔曼滤波

Abstract: Short-term passenger flow forecasting on bus stop is an important base and technical support for decision-making on the intelligent bus dispatch system. A Kalman filter method is developed to forecast short-term passenger flow based on the characteristic analysis, and the solving algorism is presented. In the real bus line, a typical bus station with large and significantly changed passenger flow is chosen to be an example. The prediction method and the artificial neural network had been compared with the results, which show that the Kalman filter-based model is more accurate. The calculated average absolute error is 5.1771 and mean square error is 0.7961 using the Kalman filter and the average absolute error is 10.4770 and mean square error is 1.6724 using the artificial neural network, which indicates that the prediction error is small, and the method is meaningful in practical application.

Key words: urban traffic, short-term passenger flow forecasting, Kalman filter