A Weather-Assisted Driver Experiences Based Path Selection Method

With the rapid increasing number of vehicles in the urban area, the metropolitan traffic becomes extremely busy in particular in rush hours. In recent years, Intelligent Transport System (ITS) is applied to provide real-time path planning and navigation in order to enable the convenience travelling from one place to another. Weather condition is a critical factor for path planning while the raining or snowing can severely degrade the vehicle moving speed. For this reason, when the experienced drivers including most taxi drivers manually plan the paths, they always consider the weather and other related factors to find the optimal path during bad weather. However, most existing works do not consider the weather and experienced drivers' selection to make path planning. Therefore, in this paper, we propose a new path planning considers by analyzing the weather effect on experienced drivers' path selection. To achieve this, we study the floating car data in Beijing to discover how taxi drivers as a group of most experienced drivers select path under bad weathers. In addition, we use the Support Vector Machine (SVM) to learn a model about the taxi drivers' path selection. Then we apply this model in the A-star. The experimental results show our proposal provides shorter travel time compared to its counterpart under bad weather condition.