Estimation Method Of Bus Transit Network Travel Speed

With the accelerating process of urbanization, city size continues to expand, the number of motor vehicles is also increasing and traffic congestion turns to be increasingly serious. To solve this problem, many cities have put forward the policy of bus priority, taken a series of measures to develop public transit. The bus transit network travel speed is one important indicator of the level of transit service, this paper studies on estimating of bus transit network travel speed problem.Currently a large number of the buses have installed GPS, we can easily receive road vehicle operation information, such as real-time bus travel speed, location, time and other data. With the use of GPS data information, this paper establishes three models of estimating bus transit travel speed based on weighted data fusion, based on fuzzy C-means clustering and based on curve fitting.In the model for estimating bus transit network travel speed based on weighted data fusion, first using the frequency weighted method to calculate line instantaneous speed, using the multi point interregional speed estimation method to calculate line interregional speed, and then using line instantaneous speed and line interregional speed weighted data integration method to calculate line travel speed, at last estimating the bus transit network travel speed using line travel speed integration.In the model for estimating bus transit network travel speed based on fuzzy C-means clustering, first calculating the travel speed of vehicle positioning sections, calculating individual vehicle travel speed after the clustering, and then calculating individual line travel speed after the clustering of vehicle travel speed, at last estimating the bus transit network travel speed after the clustering of line travel speed.In the model for estimating bus transit network travel speed based on curve fitting, first calculating the average speed of transit network per second in the period of time, after sorting with speed, using Matlab software to fit these discrete points, at last estimating the bus transit network travel speed after function integral.At the end of this paper starts an empirical analysis using the real-time GPS data of the Beijing Public Transport on March21,2012and April5,2012, compared with travel speed of the actual road network, verifying the validity of the three models.