Abstract: It is believed that the routes chosen by taxi drivers are more reasonable and passable, and it is important to plan the route using the taxi drivers’ experience. This paper establishes the taxi drivers’ experience knowledge model based on analyzing the historical floating car data (FCD) and the taxi drivers’ route planning rules. The road network is classified into different experience levels, and a new route planning algorithm is proposed based on the road network of experience levels. In the end, this paper finishes 350 experiments with the road network and historical FCD of Wuhan, and compares the new algorithm with the traditional shortest path algorithm, the results show us that the routes planed by the taxi’ experience and knowledge model are more coherent and consistent, and that the traveling time can be reduced.