自适应免疫遗传算法在认知决策引擎中的应用  

Application of adaptive immune genetic algorithm in cognitive engine

作　　者：夏龄[1] 

机构地区：[1]四川民族学院网络信息中心,四川甘孜藏族自治州康定县626001

出　　处：《太赫兹科学与电子信息学报》2013年第3期469-473,共5页Journal of Terahertz Science and Electronic Information Technology

基　　金：四川省教育厅科研基金资助项目(10ZC025);四川省教育厅自然科学重点资助项目(13ZA0134);甘孜州重点科技基金资助项目(20095025)

摘　　要：提出一种自适应免疫遗传算法,设计自适应免疫遗传算子。该算法利用交叉率和变异率自适应调整策略,既防止交叉变异中的个体退化,又保证种群的多样性,并能快速收敛到全局最优解。仿真分析表明,与遗传算法等其他算法相比,该算法具有收敛速度快、平均适应度高、稳定性好等优点,能满足认知引擎参数优化的需要。An Adaptive Immune Genetic Algorithm（AIGA） is proposed, which introduces the adaptive immune genetic operators based on the immune genetic algorithm. Crossover probability and mutation probability are employed in the proposed algorithm to not only prevent the individual backwards during crossover and mutation processes, but also ensure the diversity of population, and this algorithm can rapidly converge to the global optimal solution. Simulation results show that AIGA exhibits a good performance in convergence speed, average fitness value and stability in comparison to Genetic Algorithm（GA） and other algorithms, which greatly satisfies the demand of parameters optimization of cognitive engine.

关 键 词：认知引擎 免疫遗传算法 遗传算子 自适应策略 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]