灰色GM(1,1,k,k2)模型背景值及时间响应函数优化

本文提出了一种新的带有时间幂次项的灰色GM(1,1,k,k²)模型,给出了其灰微分方程和白化微分方程基本形式。基于最小二乘法获得了该模型参数估计值,并推导了该模型时间响应函数。鉴于GM(1,1,k,k²)模型灰微分方程与白化微分方程之间存在跳跃关系,首先对灰微分方程的背景值进行了优化,并推导了优化后的背景值计算公式。为了克服初始值的影响,根据误差平方和最小,进一步优化了GM(1,1,k,k²)模型时间响应函数。最后,该优化后的GM(1,1,k,k²)模型被应用于软土地基沉降预测,获得了较好的模拟预测效果,说明模型是可行的。     

成分数据的组合预测

成分数据是一类具有复杂性质的数据,特别是它的预测研究在管理学与经济学中占有很重要的地位.组合预测则是近年来在预测中应用比较广泛的一种方法,它能够充分利用单预测模型的信息,提高预测精度,增强预测的稳定性,且具有较高的适应能力.本文首次把组合预测方法应用到成分数据的预测分析中,基于成分数据的一些基本性质,利用组合预测得到了较好的预测结果.     

一类非线性抛物方程组解的爆破时间上下界估计

本文研究了一类非线性抛物方程组uj/t=△uj+fj(u)解的爆破时间的估计问题.通过构造恰当的辅助函数和建立一系列微分不等式,获得了此类非线性抛物方程组解的爆破时间上下界的估计.从而将单个方程的结论推广到了方程组的情形.     

Synchronization in multilayer networks through different coupling mechanisms

In recent years, most studies of complex networks have focused on a single network and ignored the interaction of multiple networks, much less the coupling mechanisms between multiplex networks. In this paper we investigate synchronization phenomena in multilayer networks with nonidentical topological structures based on three specific coupling mechanisms:assortative, disassortative, and anti-assortative couplings. We find rich and complex synchronous dynamic phenomena in coupled networks. We also study the behavior of effective frequencies for layers I and II to understand the underlying microscopic dynamics occurring under the three different coupling mechanisms. In particular, the coupling mechanisms proposed here have strong robustness and effectiveness and can produce abundant synchronization phenomena in coupled networks.     

Analytical formulas for calculating the blocking probability of a dynamic star network

Routing and wavelength assignment (RWA) has becomeone of the key problems for today's complicated wave-length division multiplexing (WDM) networks. In gen-eral, RWA can be classified either into the dynamiccase and the static case according to the wavelengthallocation[1], or into the star, ring and mesh networksaccording to the topology. Although extensive researchworks[2,3] have been done on the RWA and many al-gorithms have been developed for the case of dynamicrings[4], little has been …