混合蝙蝠和布谷鸟算法的认知决策引擎

Cognitive Decision Engine based on Hybrid Bat and Cuckoo Algorithms

针对认知无线电中参数配置问题,提出一种混合蝙蝠算法和布谷鸟算法的认知决策引擎(hybrid bat algorithm and cuckoo search,HBA-CS),首先将参数配置建模为多目标优化问题,然后利用布谷鸟算法优化,同时为了克服布谷鸟算法收敛速度慢、容易陷入局部最优的缺点,在经过进化之后,再利用全局寻优能力强、收敛速度快的蝙蝠算法优化,该算法解决了局部搜索和全局搜索的平衡问题,从而改善了算法收敛性、避免陷入局部最优。仿真结果表明,基于HBA-CS的认知决策引擎的收敛速度和精度优于混合粒子群和遗传的算法(HBPGA)和优于混沌量子粒子群算法(CQPSO),优化后得到的系统参数具有更好的性能。

Aiming at the parameter configuration problem in cognitive radio,a hybrid bat algorithm and cuckoo search(HBA-CS)is proposed.Firstly,the parameter configuration is modeled as a multi-objective optimization problem.Then,the cuckoo algorithm is used to optimize this problem.At the same time,in order to overcome the shortcomings of slow convergence speed and easy falling into local optimum of cuckoo algorithm,the bat algorithm with strong global search ability and fast convergence speed is used to optimize the algorithm.The algorithm solves the balanc of between local search and global search,thus improving the convergence of the algorithm and avoiding the falling into local optimum.The simulation results indicate that the convergence speed and accuracy of the cognitive decision engine based on HBA-CS are better than those of the hybrid particle swarm optimization and genetic algorithm(HBPGA)and chaotic quantum particle swarm optimization(CQPSO),and the optimized system parameters have even better performance.