出租车驾驶员对城市道路交通状况较为熟悉，他们选择的路径具有代表性。为了挖掘出租 车驾驶员的路径选择经验，最大程度地保留经验选择并融合到路径规划算法中，建立了基于典型 经验路径库的路径规划算法。首先，通过对出租车GPS数据进行预处理、地图匹配、载客轨迹信 息提取，建立出租车历史载客轨迹集；然后，基于网格划分建立不同时段、不同网格之间的经验 路径库；最后，提出了基于典型经验路径库的路径规划算法，并以广州市交通网络和浮动车为实 验数据。实验结果表明，该算法得到的路径其行程时间较少，平均速度较高，能够保持道路等级 的连贯性，且能适应不同时段的交通情况，更符合个人认知。

Taxi drivers are more familiar with the traffic conditions of urban roads, and the routes they choose are of representativeness. In order to explore the taxi drivers′ route choice experience and remain the route choice as much as possible, a route planning algorithm based on typical experience route data⁃ base was established. First, the taxi route data set was built after preprocessing the GPS data, matching map, and mining operation data from floating car data. Then, experience route database between differ⁃ ent time periods and grids was set up based on grid division. Finally, a new route planning algorithm was proposed based on the typical experience route database. Taking experimental data of traffic network and floating car in Guangzhou as experimental data, the experimental results demonstrate the travel time of the route calculated by the new algorithm reduced, the average speed of it was faster, the consistency of road grade could be maintained, and could adapt to traffic in different time periods. The routes with taxi drivers′ experience are more accord with the individual cognition.