针对智能车辆的结构特点和任务要求,综合体系结构和信息处理两方面设计车辆导航策略。提出了一种基于多智能体且包含了慎思和反应结构的混合体系,根据系统特性设计出多个异构智能体。在现有研究的基础上将图像处理、信息融合方法与该体系相结合。仿真与实验显示,智能车辆在多智能体结构下能充分利用多传感器信息,并获得良好的避障导航效果。
Aiming at the structural characteristics and mission requirements of the intelligent vehicle,a navigation strategy was designed with respect both to architecture and information process.A hybrid architecture containing deliberation and reaction was proposed on the basis of multi-agent.A set of agents was designed according to the structural characteristics.Based on the previous research,the methods of image proceeding and information fusion were integrated into this architecture.The simulation and experiment results showed that the intelligent vehicle with the multi-agent architecture could use the information of multi-sensor efficiently and performed well on navigation and obstacle avoidance.