中文摘要：

为了精确获得车辆跟驰模糊推理系统的隶属度函数,避免因采用专家法而使模糊推理结果的误差增大,提出采用模糊聚类分析的方法,考虑车辆跟驰数据内部的关联性,并根据高斯函数中参数的统计学意义进行车辆跟驰模糊集的划分和隶属度函数的确定。利用NGSIM数据,将后车速度、前后车相对速度、前后车间距作为输入变量,后车加速度作为输出变量建立车辆跟驰模糊推理系统,对提出的基于模糊聚类的车辆跟驰隶属度函数确定方法进行评价。结果表明,提出的新方法能真实反映数据本身的特征和驾驶员的心理生理特性,其推理结果与真实数据误差较小,可为车辆跟驰模糊推理系统的建立提供参考。

英文摘要：

In order to obtain membership functions of fuzzy inference system of car-following scientifically and accurately, this paper used fuzzy clustering analysis method to avoid increasing the error of fuzzy inference system by expert experience method. It established Gaussian membership function of car-following, based on statistical significance of parameters in Gaussian function by taking data＇ s internal correlation into account for the first time. It used the real NGSIM data to establish a fuzzy in- ference system, which took the velocity of following vehicle, the relative velocity, the gap of two vehicles as input variables, and acceleration of the following vehicle as output variable. Then it achieved and evaluated the fuzzy inference results with the proposed fuzzy inference clustering method to establish membership function of car-following. Analysis results show that, the proposed method can truly reflect the characteristics of the data itself and the driver＇ s psychological physiological features. The fuzzy inference results＇ errors based on the membership function proposed above are small compared with the real data. So it can be used to found the fuzzy inference system of car-following.