摘要： 针对现有汽车追尾预警系统以车间距离测量作为碰撞发生的唯一决定因素的不足，提出基于MAS和驾驶员行为的追尾预警模型(MRWMB)。采用MAS技术和基于Bayes决策的追尾预警算法，实现追尾碰撞概率的实时计算。仿真结果证明，MRWMB模型在预警的自主性和抗干扰方面比传统的预警系统有明显优势，预警平均性能提高10%左右。

关键词: Agent技术, 汽车追尾, 行为

Abstract: Aiming at the shortcomings of distance between vehicles measurement as the only determining factor in car rear-end alarm system, a Modeling of Rear-end Warning based on MAS and Behavior(MRWMB) is proposed. Bayes decision theory is utilized to prevent collision happened real time by calculating the probability of collision. The effectiveness and robustness of the model are confirmed by the simulation result and its average performance of warning is improved by around ten percent.

Key words: Agent technology, car rear-end, behavior

中图分类号: 

• TP242.6