基于相关机会规划的多机器人觅食过程的参数优化

Parameters Optimization in Foraging Process of Multi-Robot Based on Dependent-Chance Goal Programming

针对随机环境中以觅食为任务的多机器人系统,在建立其整体数学模型基础之上,利用随机规划中的相关机会规划对影响整体系统性能的两个重要参数进行了优化分析,并采用基于随机模拟的遗传算法得到了优化结果。仿真结果表明,利用该方法可以实现随机环境下的系统优化控制,从而可为多机器人系统的行为规则优化提供理论依据。

In this paper, the mathematical model of a multi-robot system with a foraging task in random environment is firstly introduced. Based on this model, dependent-chance programming is adopted to optimize two important parameters which influence performance of the whole system. Then the optimal result is obtained by using stochastic simulation based genetic algorithm. The simulation result shows that the adopted approach can optimize the system in random environment. And also, it provides basis for further optimizations of robot behaviors.