中文摘要：

农村公路急弯、陡坡、路侧险要等典型路段往往是事故易发多发路段。在现阶段农村公路安保工程资金有限的情况下，需对既有典型路段的事故风险进行定量评价和排序，为隐患路段的排查和安保工程分期实施提供依据。在现有研究成果的基础上，构建了典型路段事故风险评价指标体系，提出了一种基于贝叶斯网络的农村公路典型路段事故风险评价方法。该方法能有效集结专家知识和经验，定量评价典型路段（包括单一路段和组合路段）的事故风险，并可以随着数据的更新不断完善模型。在农村公路事故资料相对缺乏的情况下，该方法具有较强的实用性。

英文摘要：

Typical sections such as sharp turns, steep slopes and dangerous roadsides of the rural road tend to be crash prone sections. In the case of limited budget for highway safety enhancement project, in order to identify potential high risk sections and provide the basis for implement in phases of safety enhancement project, it is necessary to quantitatively evaluate and sort the crash risk of typical sections. On the basis of related researches, crash risk assessment index system for typical sections has been established. Then a crash risk assessment method for typical sections of rural roads using Bayesian networks is proposed. It is effective to evaluate the crash risk of typical sections （including single sections and combination sections） by incorporating expert knowledge and experience with the proposed method. It is also noted that the proposed model can constantly improve as related data updates. The proposed method shows a strong practicability for crash risk assessment especially when lack of accident related information.