自适应检测器生成算法研究

基于小生境策略的否定选择算法利用在搜索空间中计算检测器之间的海明距离,构建小生境;定义适应度函数与亲合力函数相关,更客观地反映检测器的检测能力;利用进化策略,进行遗传操作,而生成多样性和通用性的最佳检测器集.同时该算法可以减少生成检测器的时间开销.     

Ensemble of niching algorithms

Although niching algorithms have been investigated for almost four decades as effective procedures to obtain several good and diverse solutions of an optimization problem, no effort has been reported on combining different niching algorithms to form an effective ensemble of niching algorithms. In this paper, we propose an ensemble of niching algorithms (ENA) and illustrate the concept by an instantiation which is realized using four different parallel populations. The offspring of each population is considered by all parallel populations. The instantiation is tested on a set of 16 real and binary problems and compared against the single niching methods with respect to searching ability and computation time. Results confirm that ENA method is as good as or better than the best single method in it on every test problem. Moreover, comparison with other state-of-the-art niching algorithms demonstrates the competitiveness of our proposed ENA.     

Combining the Filtered Evaluation Function With Coevolutionary Shared Niching in Genetic Algorithms

In this paper, we propose the combination of filtered evaluation and coevolutionary shared niching (CSN) for extending the search ability of genetic algorithms (GA). The proposed scheme can overcome the problems of the filtering GA (FGA) and the CSN. The successful optimization ability of the FGA is supported by the filtered evaluation method that can modify the landscape for escaping local optima. However, the problem of the FGA is the relatively high cost to maintain the filter. The CSN can autonomously maintain the shared distance using the coevolution between two populations (called customers and businessmen). However, the escaping ability from local optima of the CSN is still insufficient. Therefore, the combination of the filtered evaluation and the CSN is proposed, to reduce the cost of the FGA filter. The effectiveness of the proposed scheme is confirmed through test problems.     

基于小生境混沌遗传算法的水资源优化调度

水资源调度具有多目标、大规模和不确定性等特点,利用混沌遗传算法求解水资源调度问题,在一定程度上避免了局部优化并提高了求解速度,但由于损坏了种群多样性导致求解精度较低.为此提出了基于小生境的混沌遗传算法(NCGA),该算法通过小生境技术保留源中心个体的方法保护了种群多样性,同时利用混沌的随机性、遍历性及规律性与遗传算法的快速收敛性相结合,从而使该算法提高了求解速度和求解精度.将该算法应用到水资源优化调度模型中,仿真结果验证了该算法比混沌遗传算法能更合理高效地分配水资源,达到了综合效益最大化.     

基于并行小生境技术优化的智能模糊聚类图像分割算法

利用PBM模糊聚类有效性函数以图像特征空间为搜索空间,实现有效性函数的全局寻优,用并行小生境技术解决粒子群(PSO)算法的早收敛问题,优化聚类的全局收敛性能,实现有效聚类数目与聚类中心的并行寻优。通过对遥感图像分割的实验证明,与传统粒子优化群算法的分割结果相比,本文算法拥有更高的有效性且分割效果更优。     

基于改进遗传算法的可重构计算任务划分

为实现可重构计算中的软硬件任务自动划分,引入了遗传算法来搜寻最优解。为解决标准遗传算法可能出现种群早熟和种群进化后期收敛速度慢的问题,使用了小生境技术来保护种群中基因的多样性。设计了能够随适应度自动改变的自适应遗传算子（杂交算子和变异算子）。对算法进行了50次随机实验,并对结果进行分析。实验表明,改进后的遗传算法搜寻到全局最优任务划分的概率和搜寻到最优任务划分时的进化代数都要优于标准遗传算法。     

一种基于密度聚类的小生境差分进化算法

针对基本差分进化算法早熟收敛的缺陷,提出了一种基于密度聚类的小生境差分进化算法。该算法基于DE/rand/2/bin变异方式全局搜索能力强、鲁棒性好和DE/best/2/bin变异方式局部搜索能力强、收敛速度快的特点,首先初始化一个没有子种群的全局种群,再在全局种群中采用DE/rand/2/bin进行迭代搜索,并对其中的个体进行聚类,当聚类簇中的个体数目达到规定的最小规模时形成一个小生境子种群,然后在各子种群中采用改进的DE/best/2/bin进行迭代搜索并重新进行聚类,从而提高进化过程中种群的多样性,增强算法跳出局部最优的能力。仿真实验表明,该方法能显著提高算法的收敛速度和全局搜索能力,有效避免早熟收敛。     

基于两阶段搜索算法的多峰函数优化

多峰优化问题需要搜索多个最优值（全局最优／局部最优）,这给传统的优化算法带来很大程度上的挑战。本文提出了一种两阶段算法求解多峰优化问题。第一阶段采用带有邻域变异策略的排挤差分演化算法进行粗粒度搜索,在适应度景观上尽可能多的找到最优解的大概位置。搜索一定代数之后,调用DMC聚类方法把搜索种群划分成多个聚类,然后在每个聚类上调用协方差矩阵自适应演化策略算法进行精细搜索。另外,本文还提出搜索点补充策略用于平衡每个聚类的大小及增加算法初期的搜索能力。我们提出的方法和9个较新的经典算法在两个基准测试集上进行了大量对比测试,结果表明新算法是有效的,在大多数测试函数上都优于其它算法。     

多样性检测器生成算法研究

文章基于小生境策略的否定选择算法利用在搜索空间中计算检测器之间的海明距离,构建小生境;定义适应度函数与亲合力函数相关,更客观地反映检测器的检测能力;利用进化策略,进行遗传操作,而生成多样性和通用性的最佳检测器集。同时该算法可以减少生成检测器的时间开销。     

An adaptive species conservation genetic algorithm for multimodal optimization

This paper introduces an adaptive species conservation genetic algorithm (ASCGA) by defining a species with three parameters: species seed, species radius and species boundary fitness. A species is defined as a group of individuals that have similar characteristics and that are dominated by the best individual in the species, called the species seed. Species radius defines the species' upper boundary and the species boundary fitness is the lowest value of fitness in the boundary. Some heuristic algorithms have been developed to adjust these parameters and an ASCGA has been proposed to solve multimodal optimization problems. With heuristic techniques, ASCGA can automatically adjust species parameters and allow the species to adapt to an optimization problem. Experimental results presented demonstrate that the proposed algorithm is capable of finding the global and local optima of test multimodal optimization problems with a higher efficiency than the methods from the literature. ASCGA has also successfully found a significantly different solution of a 25‐bar space truss design and identified 761 local solutions of the 2‐D Shubert function. Copyright © 2009 John Wiley & Sons, Ltd.