摘要
在计算机生成兵力(CGF)应用中,提出一种基于遗传算法的路径跟踪自学习策略,能够有效增强CGF实体的的自治性。首先构建了基于遗传算法的CGF学习行为模型框架;其次依据该框架论述了规则中条件、动作及适应度函数的确定;最后在实验部分对各参数的设置、整体模型的泛化能力以及对圆形路径的跟踪能力进行了详尽的分析,实验结果表明了该算法的有效性和可行性。
A path tracking self-learning strategy based on Genetic Algorithms was proposed with application to Computer Generated Force （CGF）, which is able to improve the autonomous properties of CGF entities. First, a learning behavioral model framework of CGF was constructed. Second, condition, action and fitness function in the model framework were elaborated. Finally, in the experimental part, parameters design and the generalization ability were analyzed in detail. And the results of experiments show that this path tracking strategy is available and feasible in CGE .
出处
《系统仿真学报》
CAS
CSCD
北大核心
2009年第10期2925-2929,共5页
Journal of System Simulation
基金
国家高技术研究发展计划(863计划)(2004AA115130-05)
国家自然科学基金资助项目(60702076)
