一种基于认知引擎的 t分布变异萤火虫算法

摘要

针对认知引擎中智能优化的问题，提出了一种基于t分布变异萤火虫算法的认知无线电决策引擎技术，分析了萤火虫算法的仿生原理，从数学角度对算法实现优化过程进行了定义，并设计了一种基于 t分布随机扰动的变异方案。最后，在多载波环境下对该算法进行了仿真实验，结果表明与普通基于遗传算法的认知引擎相比，该算法收敛速度快且具有较高的目标函数值，并且在加入 t分布变异之后，算法的全局寻优能力进一步加强。%Considering the problem of cognitive radio intelligence optimization ,a scheme of cognitive en‐gine is proposed based on the glowworm swarm optimization with t distribution mutation .The bionic principle of firefly algorithm is analyzed .The optimization mechanism is defined by mathematics and a mutation plan is also designed with t distribution .Finally ,the simulations are performed in the multi‐carrier system .The results show that the algorithm has higher convergence speed and larger objective function value than the basic GA algorithm .Furthermore ,the improved algorithm with t distribution mutation has the higher capability of the global optimization than that of the basic GSO algorithm .