【作者】 高鹏；

【导师】 梁光胜； 高雪莲；

【作者基本信息】 华北电力大学（北京） ， 信号与信息处理， 2009， 硕士

【摘要】 本文借鉴免疫系统中的克隆选择原理,结合小生境技术,引入新的克隆选择机制,提出了应用于多模态函数优化问题的克隆小生境算法(CNSA)以及应用于多目标优化问题的克隆小生境算法(MCNSA)。论文通过对典型的多模态函数和公认的benchmark问题进行仿真,并与传统寻优算法的相应仿真结果进行对比,实验结果说明本文提出的克隆小生境算法具有较强的多模态函数寻优能力与多目标优化解决能力。最后,论文将解决多目标优化问题的克隆小生境算法(MCNSA)应用于QOS多播路由寻优。该方法将QOS多播路由寻优作为一个以网络代价、带宽、时延为目标的多目标优化问题来处理,并用多目标优化问题的克隆小生境算法(MCNSA)来求解该问题的Pareto最优解集,通过仿真结果表明了其有效性。更多还原

【Abstract】 New algorithms for multi-modal function and multi-objective problems optimization,the clonal niche selection algorithm(CNSA) and multi-objective clonal niche selection algorithm(MCNSA) are proposed by using the clonal selection principle of immune system,combining with the niching technology and introducing a new clone selection mechanism.Then these algorithms are applied to the optimization of typical multi-modal functions and benchmark problems.By comparing with some traditional methods,the results of those experiments show the effectiveness of the new algorithms obviously.Finally,multi-objective clonal niche selection algorithm(MCNSA) is applied to multicast routing optimization in the paper.In this algorithm,we consider the work as multi-objective optimization problem on net cost,bandwidth and delays of network,and use the multi-objective clonal niche selection algorithm (MCNSA) to obtain the Pareto optimal solution set of the problem.At last,some numerical simulation results show its validity.更多还原