基于TSK的车辆跟驰模型

为解决基于模糊推理的车辆跟驰模型在参数校准方面存在的问题，提出了基于TSK模型的车辆跟驰模型。对TSK模型在现实驾驶行为中的意义进行了分析，解释了子系统的工作点和线性方程的物理意义，为TSK模型在车辆跟驰中的深入应用奠定了基础。采用改进的基于遗传算法的TSK模型辨识方法，对分组中数据点不足的情况作了特别处理，通过GPS采集跑车实验数据，从实测数据构建并验证TSK模型。实验结果和理论分析吻合较好，表明TSK模型用于车辆跟驰模型的可行性。

TSK based Car-following Model

A TSK based car-following model was proposed for solving the calibration problem of fuzzy car-following model. This paper, which analyzes the meaning of TSK model in real driving behavior, including operation point and linear function of sub-system, lays the foundation of using TSK model in driving behavior research. An improved GA based learning algorithm was adopted for the identification of TSK model. Emphasis of improvement is placed on special condition handling for lack of sub-group data point. The TSK model was established, calibrated and validated based on the data collected from real car-following experiment, which used cars equipped with GPS. The agreement of the theoretical analysis with the experimental result shows that the TSK model is suitable for car-following modeling.