基于混沌遗传算法的遥感影像分类

为提高遥感影像分类精度,本文提出基于混沌遗传算法(Chaos Genetic Algorithm)的遥感影像分类方法.首先应用混沌遗传算法对样本进行自学习得到全局最优的聚类中心,然后通过得到的聚类中心对整幅影像进行分类.该方法利用混沌变量的遍历性,进行粗粒搜索,优化遗传算法的初始种群,从而提高收敛速度;对经过选择算子、...     

改进量子粒子群算法在水电站群优化调度中的应用

针对量子粒子群算法求解水电站群优化调度问题存在的早熟收敛、寻优能力欠佳等缺陷,从种群初始化、进化和变异等方面提出了改进量子粒子群算法。该方法引入混沌搜索增强初始种群质量;通过加权更新种群最优位置中心改善种群进化模式并提升收敛速度;利用邻域变异搜索增加种群多样性避免早熟收敛。同时依据问题特点设计了矩阵实数编码方式与复杂约束处理方法。乌江梯级综合对比分析表明所提方法能切实保证快速获得高质量优化调度结果,有效提高梯级水能利用率,如长序列模拟调度较逐步优化算法分别减少8.9%的弃水和72.3%的耗时,是一种适用于大规模水电站群优化调度的高效实用方法。     

Study of resonances for the restricted 3‐body problem

Our aim is to identify and classify mean‐motion resonances (MMRs) for the coplanar circular restricted three‐body problem (CR3BP) for mass ratios between 0.10 and 0.50. Our methods include the maximum Lyapunov exponent, which is used as an indicator for the location of the resonances, the Fast Fourier Transform (FFT) used for determining what kind of resonances are present, and the inspection of the orbital elements to classify the periodicity. We show that the 2:1 resonance occurs the most frequently. Among other resonances, the 3:1 resonance is the second most common, and furthermore both 3:2 and 5:3 resonances occur more often than the 4:1 resonance. Moreover, the resonances in the coplanar CR3BP are classified based on the behaviour of the orbits. We show that orbital stability is ensured for high values of resonance (i.e., high ratios) where only a single resonance is present. The resonances attained are consistent with the previously established resonances for the solar system, i.e., specifically, in regards to the asteroid belt. Previous work employed digital filtering and Lyapunov characteristic exponents to determine stochasticity of the eccentricity, which is found to be consistent with our usage of Lyapunov exponents as an alternate approach based on varying the mass ratio instead of the eccentricity. Our results are expected to be of principal interest to future studies, including augmentations to observed or proposed resonances, of extra‐solar planets in binary stellar systems (© 2012 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

复杂性科学在页岩油气勘探和研究中的应用

复杂性科学是当今世界科学重要的前沿领域之一。为了促进复杂性科学的基本理论和方法在页岩油气勘探和研究领域的广泛应用，笔者等重点围绕地质系统复杂性的基本内涵，从地质系统构成单元的多重耦合和相互作用、地质作用的复杂时空结构、地质作用过程中的自组织临界过程动力学以及矿产资源形成在混沌边缘等方面，介绍了地质系统复杂性研究的基本理论和研究方法。将上述地质系统复杂性理论运用于页岩油气系统，探讨了利用复杂系统的总体特征通过局域组成单元相互作用的“涌现”机制，来解决目前页岩储层中存在的如何利用微观局域精细表征进行客观刻画宏观储层特征问题；分析了利用地质作用的时空结构理论，建立页岩储层非均质性的静态和分形时空结构的方法；在整体论思想指导下，根据地质系统在混沌边缘分形生长这一规律，探讨了页岩油气在自组织临界态的混沌边缘富集预测方法。复杂性科学的理论和方法，必将在页岩油气的勘探研究中发挥越来越重要的作用。