摘要
基于Robocode机器人战斗仿真引擎,从设计原理和实际实现两方面阐述了运用差分进化算法来设计机器人行动策略。重点介绍差分进化算法理论如何与实际应用相结合:预先对机器人各种行为进行编程,行为选择逻辑应用差分进化算法实现,让机器人具有根据不同的场景和对手,自适应的选择最优的响应策略的能力。运用设计好的机器人与其它机器人战斗,对战斗结果进行分析。最后,提出可改进的若干方向。
Give the detail description of how to apply the differential evolution(DE)algorithm in robot action strategy design based on platform "Robocode"from the aspects of both design principle and actual implementation. Focus on introduction of how the DE algorithm principle is applied in the practical application. Programming of robot actions is the preparation. And then use DE algo- rithm to implement the choice logical design which makes the robotbe capable of choosing the optimal strategy adaptivelyaccord- ing to the different scenarios and rivals. Use the designed robot to combat with other robots and analyze the results of battles. Fi- nallysuggest several aspects of improvement.
出处
《微型电脑应用》
2015年第2期39-41,共3页
Microcomputer Applications