Abstract: With the expressway construction, the road users have more options when driving in the continuously enlarged network. Meanwhile, the pluralistic investment subjects also require the reasonable toll division. Because of these reasons, the problem of driving multi-path recognition is the urgent problem in the on-line toll and management of expressway. This paper analyzes the influencing facts of toll distribution in the network charging environment. An expressway tolls split model is developed based on license plate recognition, and its corresponding algorithm is designed. Then, this method is applied to split the tolls of Fujian expressway. The result shows that the method achieves more than 70% accurate split and less than 30% fuzzy split, which showa the feasibility of the proposed method.

Key words: highway transportation, expressway, tolls, license plate recognition, split

摘要： 随着高速公路的不断建设，高速公路已形成网状结构，为车辆驾驶员提供了多条选择路径. 同时高速公路投资主体的多元化必然要求合理拆分通行费，进而产生如何按照车辆行驶的实际路径进行通行费拆分的问题. 本文分析了联网收费环境下影响通行费精确拆分的主要因素. 建立了基于车牌识别的通行费拆分模型，并给出了模型求解算法. 最后结合福建省高速公路路网结构与运营数据进行实例分析，实现了70%以上的精确拆分和不到30%的模糊拆分，验证了本文方法的有效性.

关键词: 公路运输, 高速公路, 通行费, 车牌识别, 拆分