微粒群优化求解的多机器人救援任务分配

Rescue Task Allocation by Multiple Robot Based on Particle Swarm Optimization

面向单一救援任务类型,研究灾难发生后多机器人救援任务分配问题,提出基于改进微粒群优化的问题求解方法,以生成机器人救援任务分配序列.首先,根据被困人员(即目标点)的分布情况,根据目标点之间的距离将其分成若干大类;接着,基于所分的类以救援个数最大为目标给出所求问题的数学模型;接着,设计微粒群优化方法来求取任务分配序列,并基于所求问题设计新微粒编码和解码方法.多种仿真场景验证了所提方法的有效性.

Aiming at single rescue task type, we study the rescue task allocation problem by multiple robots after disaster, and propose a modified PSO method to gain its solutions in order to generate the robots＇ rescue task allocation sequence. Firstly, the trapped men（i.e., targets） are classified into several groups according to their distribution and their distances between each other. Secondly, based on the groups, the mathematical model is established which takes the maximum rescued targets as the objective; and then, Particle Swarm Optimization（PSO） algorithm is designed to solve the above mathematical model, and based on the problem, the new encode and decode methods are proposed. Several simulations verify the effectiveness of the proposed method.