基于信息融合的驾驶行为识别技术的研究

A Research on the Technique of Driving Behavior Identification Based on Information Fusion

通过分析选定驾驶行为的组成主因子,并采用层次分析法确定其权重系数;基于所采集的各主因子的传感器信号,采用BP神经网络和D-S证据理论相结合的多传感器信息融合方法进行驾驶行为识别。仿真结果表明:提出的基于多信息融合的驾驶行为识别方法能准确地识别出常见的驾驶行为。

Main constituent factors of driving behaviors are selected by analyses and their weighting coefficients are determined by analytic hierarchy process. Based on sensors＂ signals collected of each main factor, the driving behaviors are identified by adopting multi＇sensor information fusion technique combining BP neural network with D-S evidence theory. The results of simulation show that the proposed technique for driving behavior identification based on information fusion can accurately identify common driving behaviors.