为进一步探索多智能体复杂的群集运动行为,给出了两种改进的智能体模型,即改进的有限感知范围的Boid模型和改进的具有优先方向的Leader-Follower模型。以改进的有限感知范围的Boid模型为基础,研究了群集趋同和危险躲避行为。并以改进的具有优先方向的Leader-Follower模型为基础,进一步研究了群集趋同和危险躲避行为,与Boid模型不同的是,由于引进了具有对环境先验信息已知的智能体Leader,后者的行为具有更丰富的动力学行为和更高效的群集趋同和危险躲避效果。结果表明,基于领域知识或结合工程实际背景对智能体模型进行有目的的改进是进一步推动智能体群集运动行为研究的有效途径。
In order to explore the swarm behavior of multi-agent systems further,two new reformulated agent models are introduced,respectively. They are the reformulated Boid model with a limited sensing range,and the reformulated Leader-Follower model with an added priority direction. And then,swarm consensus together with danger avoidance were touched upon with the two reformulated agent models,and especially,the reformulated Leader-Follower model with an added priority direction,which is deployed with much more dynamics and much more efficiency in swarm consensus and danger avoidance,based on account of the introduction of Leaders,which are essentially important with the prior information of the environments. It is demonstrated that,the reformulated agent model,which is put forward inspired from the field knowledge or the real engineering,is vital for the development swarm behavior research of the multi-agents.