Abstract: As one part of the duration time of freeway traffic incidents，clearance time is closely connected with incident duration and incidence． A new definition of clearance time which has more correlations with incident duration regarding to emergency vehicles is put forward． The prediction model based on TSK fuzzy logic is constructed，according to data training by artificial neural network ( ANN) ． The test results indicate that the proposed model can predict the clearance time of common traffic incidents quite well when the actual clearance time is no less than 10 minutes．

Key words: freeway traffic incidents, clearance time, emergency vehicles, TSK fuzzy logic, artificial neural network

摘要： 清除时间是高速公路交通事件持续时间中的一部分，与事件影响范围密切相关。通过对某高速公路事件数 据及交警执勤记录进行分析，提出了救援车辆的清除时间新定义，该定义与事件持续时间有更强的相关性。随后 采用含自适应算法的TSK 模糊逻辑推理方法，通过对基于事件数据的人工神经网络训练，建立了交通事件清除时 间的预测模型，并讨论了模型的适用范围。经检验: 该模型对于大于10 min 的常见交通事件清除时间有较好的预 测效果。

关键词: 高速公路交通事件, 清除时间, 救援车辆, TSK 模糊逻辑, 人工神经网络