摘要
借鉴免疫算法基于浓度和适应度的抗体更新策略,提出一种改进微粒群局部搜索能力的免疫微粒群算法,并对其进行收敛性分析。该算法在满足收敛性的条件下,根据微粒浓度和适应度动态调整加速因子,保证了群的多样性和持续搜索能力。在与遗传算法、免疫遗传算法、基本微粒群等算法的仿真比较试验中,该算法不仅搜索到了最好的近优解,而且收敛速度最快。在无人潜水器仿真平台上的控制试验表明,基于免疫微粒群算法的控制器性能良好,具有较强的抗海流干扰能力。仿真结果证明了该算法的可行性。
Drawing ideas from concentration and fitness-based antibody-update strategy of immune algorithm,an Immune Particle Swarm Optimization (IPSO) algorithm is presented to enhance local searching ability,and it is analyzed about the convergency. IPSO tunes the acceleration factors dynamicly according to particle's concentration and fitness.This guarantees the diversity and sustained searching ability of the swarm.In the comparison experiments with genetic algorithm,immune genetic algorithm,basic particle swarm optimization algorithm and so on,IPSO performs well with best solution and fastest convergence speed.And the control experiments conducted on AUV platform show that IPSO-based controller works well and has strong ability against current disturbance.The simulation results show feasibility of IPSO in application to AUV.
出处
《计算机工程与应用》
CSCD
北大核心
2009年第28期215-218,239,共5页
Computer Engineering and Applications
关键词
微粒群算法
免疫算法
收敛性
无人潜水器
模糊神经网络
particle swarm optimization algorithm
immune algorithm
convergency
Autonomous Underwater Vehicle(AUV)
fuzzy neural network