Abstract: The lane change decision-making model was established based on the fuzzy multi-objective theory. The predicted results of lane change models which considered and did not consider driving tendency difference were calculated according to the experiment data of road. The predicted traffic flow macroscopic parameter (lane change rate) was compared with the result of road experiment to verify the inferential results. The experimental results show that accuracy of the method used for identifying vehicle driving tendency can be improved significantly.

Key words: traffic safety, vehicle driving-assistance, free flow, driving tendency recognition, simulated experiment verification

摘要： 应用模糊多目标理论建立驾驶员换道决策模型，根据实际道路实验所得数据，计算出不考虑和考虑驾驶倾向差异的换道模型预测结果，将模拟出的交通流宏观参数(换道率)与道路实验情况相对比，进行驾驶倾向性推理效果的验证。实验结果表明，所提方法可以明显改善汽车驾驶倾向性辨识模型的准确率。

关键词: 交通安全, 汽车辅助驾驶, 自由流, 驾驶倾向性辨识, 仿真实验验证