正交试验遗传算法及其在函数优化中的应用

针对遗传算法参数多且配置困难的问题，本文提出一种利用正交试验优化选择参数的方法，它使得对于不同领域的优化问题只需用正交试验进行一次参数配置，然后用遗传算法进行具体寻优即可取得较好效果。这种正交试验遗传算法易于编程实现且在一定程度上避免了遗传算法参数配置的盲目性。     

遗传算法全局收敛性的齐次有限马尔柯夫链分析

论证了遗传算法（Ｇｅｎｅｔｉｃ Ａｌｇｏｒｉｔｈｍ ,ＧＡ）过程是一个齐次有限马尔柯夫链,通过巧妙地构造ＧＡ 的马尔柯夫链的状态空间,并对其转移概率矩阵进行极限分析,得到的简单遗传算法（Ｓｉｍ ｐｌｅ Ｇｅｎｅｔｉｃ Ａｌｇｏｒｉｔｈｍ ,ＳＧＡ）不是全局收敛的,最优保存简单遗传算法（Ｍａｉｎｔａｉｎｉｎｇ Ｏｐｔｉｍ ｕｍ Ｓｉｍ ｐｌｅ Ｇｅｎｅｔｉｃ Ａｌｇｏｒｉｔｈｍ ,ＭＯＳＧＡ）是全局收敛的结论。     

基于混沌遗传算法的自动化生产单元调度方法

针对遗传算法在求解一类带时间窗口的自动化生产单元调度问题时易出现冗余迭代、收敛缓慢等问题,将混沌搜索技术引入至遗传算法中,通过将混沌初始化、混沌扰动与遗传算法的基本操作相结合,利用混沌运动搜索精度高、遍历性好的特点来提高遗传算法的收敛速度和优化质量.本文在给出自动化生产单元调度问题的数学模型的基础上,着重讨论了混沌遗传调度算法的设计,包括编码方式、混沌初始化、交叉操作、混沌变异操作和适应度函数的计算等.最后以自动化电镀生产线为例对提出的算法进行了验证,为此类调度问题提供了有效的算法.     

元胞蚂蚁算法的收敛性

蚁群优化算法是一种新型的解决组合优化问题的仿真型算法,在许多领域中都已获得成功的应用,但却有容易陷入局部最优的缺陷.本文将元胞自动机思想引入到蚂蚁算法中,提出一种新的元胞蚂蚁算法,通过算法的元胞演化机制对信息素的二次分配,改善了对解空间的搜索性能,并从理论上证明了算法的渐近收敛性.     

遗传算法在模糊系统优化设计中的应用研究

在模糊系统的变节点自适应模糊神经网络实现的基础上,提出一种混合GA优化算法。该算法采用混合编码策略,利用GA对模糊规则和隶属函数同时优化,而对结论参数则用最小二乘法估计。算法综合了GA强大空间搜索能力和传统优化方法的快速收敛和高精度的优点,在保证全局优化能力的条件下,综合考虑了模糊控制器的复杂程度、训练速度和控制精度。仿真结果及应用表明了该算法的有效性。     

Mechanism of Universal Access in Information Network Convergence

According to the background analysis of the universal access of information network convergence, this research suggests that the auction is an effective method to maximize the social welfare and an kind of auction mechanism is designed, which can remove the obstacle of information asymmetry between regulators and enterprises, and promote the development of operation market in the era of information network convergence.     

Convergence of Distributed Gradient-Tracking-Based Optimization Algorithms with Random Graphs

This paper studies distributed convex optimization over a multi-agent system, where each agent owns only a local cost function with convexity and Lipschitz continuous gradients. The goal of the agents is to cooperatively minimize a sum of the local cost functions. The underlying communication networks are modelled by a sequence of random and balanced digraphs, which are not required to be spatially or temporally independent and have any special distributions. The authors use a distributed gradient-tracking-based optimization algorithm to solve the optimization problem. In the algorithm,each agent makes an estimate of the optimal solution and an estimate of the average of all the local gradients. The values of the estimates are updated based on a combination of a consensus method and a gradient tracking method. The authors prove that the algorithm can achieve convergence to the optimal solution at a geometric rate if the conditional graphs are uniformly strongly connected, the global cost function is strongly convex and the step-sizes don't exceed some upper bounds.     

Revisiting the ODE Method for Recursive Algorithms:Fast Convergence Using Quasi Stochastic Approximation

Several decades ago, Profs. Sean Meyn and Lei Guo were postdoctoral fellows at ANU,where they shared interest in recursive algorithms. It seems fitting to celebrate Lei Guo's 60 th birthday with a review of the ODE Method and its recent evolution, with focus on the following themes:The method has been regarded as a technique for algorithm analysis. It is argued that this viewpoint is backwards: The original stochastic approximation method was surely motivated by an ODE, and tools for analysis came much later(based on establishing robustness of Euler approximations). The paper presents a brief survey of recent research in machine learning that shows the power of algorithm design in continuous time, following by careful approximation to obtain a practical recursive algorithm.While these methods are usually presented in a stochastic setting, this is not a prerequisite. In fact,recent theory shows that rates of convergence can be dramatically accelerated by applying techniques inspired by quasi Monte-Carlo.Subject to conditions, the optimal rate of convergence can be obtained by applying the averaging technique of Polyak and Ruppert. The conditions are not universal, but theory suggests alternatives to achieve acceleration.The theory is illustrated with applications to gradient-free optimization, and policy gradient algorithms for reinforcement learning.     

信噪比失配对LDPC码BP译码收敛性的影响

针对信噪比失配下置信传播译码的收敛性问题,提出了一种收敛性分析方法。该方法利用改进的高斯近似理论计算译码消息的均值和方差,获得外信息,进而跟踪译码的收敛过程。可分析不同码型、信道状况、失配程度下的译码收敛性,并给出了具体步骤和例子,分析结果可为自适应地确定译码器的估计精度标准提供参考。最后给出了一种灵活的信噪比估计算法,仿真验证了算法的有效性。     

Global and Superlinear Convergence for Generalized Broyden＇s Class Methods

ＧｌｏｂａｌａｎｄＳｕｐｅｒｌｉｎｅａｒＣｏｎｖｅｒｇｅｎｃｅｆｏｒＧｅｎｅｒａｌｉｚｅｄＢｒｏｙｄｅｎ＇ｓＣｌａｓｓＭｅｔｈｏｄｓＺＨＡＯＹｕｎｂｉｎ（ＣｈｏｎｇｑｉｎｇＩｎｄｕｓｔｒｙａｎｄＭａｎａｇｅｍｅｎｔＩｎｓｔｉｔｕｔｅ．Ｃｈｏｎｇｑｉｎ...     

平衡问题解的收敛性

利用集收敛、函数收敛建立了集值映射收敛的概念及其性质 ,并依此讨论了平衡问题解的收敛性 .     

一种改进的推广卡尔曼滤波收敛特性研究

提出了一种改进的推广卡尔曼滤波算法,这一算法不仅具有良好的数值稳定性,而且计算量较小,并进一步分析研究了这一算法的收敛特性,给出了指数收敛速度,分析结果表明改进的算法得到的滤波器增益和状态估计能很好地跟踪原算法得到的滤波器增益和状态估计。     

退火进化规划算法及其收敛性

基于排序的选择方式在一定程度上会导致种群搜索范围变窄,进化规划算法过早收敛。针对此问题,将退火概率与适应度结合的选择方式引入进化规划算法的选择操作,形成了退火进化规划算法(AEP)。然后利用非时齐Markov链对退火进化规划算法进行了描述,并证明了其全局收敛性。数值实验表明,退火进化规划算法能保证种群的全局收敛性,且收敛速度较快,可较好地避免早熟收敛和局部极值。     

二阶自适应IIR陷波器复数算法的收敛性分析

针对背景噪声中复正弦信号的检测和估计问题,分析了自适应格型IIR陷波器的两种复数算法即梯度算法和简化的梯度算法的收敛性能,证明了输入白噪声方差的大小不影响收敛速度。并分析了由一阶自回归模型产生的有色噪声对算法的收敛性能的影响,为有色噪声中复正弦信号检测的自适应过程的参数的选取提供了理论依据。     

Convergence rates of Markov chain approximation methods for controlled diffusions with stopping

<正> This work is concerned with rates of convergence of numerical methods using Markov chainapproximation for controlled diffusions with stopping (the first exit time from a bounded region).In lieuof considering the associated finite difference schemes for Hamilton-Jacobi-Bellman (HJB) equations,a purely probabilistic approach is used.There is an added difficulty due to the boundary condition,which requires the continuity of the first exit time with respect to the discrete parameter.To prove theconvergence of the algorithm by Markov chain approximation method,a tangency problem might arise.A common approach uses certain conditions to avoid the tangency problem.Here,by modifying thevalue function,it is demonstrated that the tangency problem will not arise in the sense of convergencein probability and in L~1.In addition,controlled diffusions with a discount factor is also treated.     

离散时间细胞神经网络的收敛性条件

神经网络的收敛性是网络各种应用的基础。主要研究了离散细胞神经网络的收敛性 ,并给出了几个新的网络收敛性条件。如果细胞网络的模板不是互补的 ,则给出一个网络在细胞格子非相互作用演化方式下的收敛性结果 ,所获结果推广了已有的结论。如果模板是互补的 ,且是行占优的 ,则网络按细胞格子行方式进行演化是收敛的。如果模板是互补的 ,且是列占优的 ,则网络按细胞格子列方式进行演化是收敛的。     

多目标进化计算收敛到Pareto最优解集的证明

多目标优化方法经历了一个从确定性搜索算法到随机搜索算法的过程 ,本质上仍是单目标优化的目标组合方法到真正意义上的向量优化方法的过程 ,至今仍在不断地发展中 ,但仍有大量未解决的问题。对多目标进化计算的研究是近年来求解多目标优化问题的重点 ,但目前仍未能证明多目标进化计算的收敛性 ,同时 ,单目标进化计算的收敛性结论不一定能推广到多目标的情况。对该问题进行了探讨 ,提出并证明了三个定理 ,并且算例说明了该理论的正确性。     

Global Convergence Property of Conjugate Gradient Methods for Optimization

１　ＩｎｔｒｏｄｕｃｔｉｏｎＷｅｃｏｎｓｉｄｅｒｔｈｅｃｏｎｖｅｒｇｅｎｃｅｐｒｏｐｅｒｔｉｅｓｏｆｔｈｅｃｏｎｊｕｇａｔｅｇｒａｄｉｅｎｔｍｅｔｈｏｄｓｆｏｒｔｈｅｕｎｃｏｎｓｔｒａｉｎｅｄｎｏｎｌｉｎｅａｒｏｐｔｉｍｉｚａｔｉｏｎｐｒｏｂｌｅｍ．ｍｉｎｆ（ｘ）ｗｈｅｒｅｆ∶Ｒｎ→Ｒ１ｉｓｃｏｎｔｉｎｕｏｕｓｌｙｄｉｆｆｅｒｅｎｔｉａｂｌｅａｎｄｉｔｓｇｒａｄｉｅｎｔｉｓｄｅｎｏｔｅｄｂｙｇ．Ｗｅｃｏｎｓｉｄｅｒｏｎｌｙｔｈｅｃａｓｅｗｈｅｒｅｔｈｅｍｅｔｈｏｄｓａｒｅｉｍｐｌｅｍｅｎｔｅ…     

基于卡尔曼滤波的二次型神经网络学习算法及收敛性分析

提出了二次型多层前馈神经网络的卡尔曼滤波学习算法,并证明了该算法的收敛性。与文献［２,３］中的学习算法和经典的误差反向传播学习算法相比,新的学习算法具有更快的学习速度、良好的泛化能力,并且对学习率有很好的鲁棒性,不容易陷入局部极小点。仿真实验结果表明了新算法的有效性。     

CONVERGENCE PROPERTIES OF THE DEPENDENT PRP CONJUGATE GRADIENT METHODS

In this paper, a new region of βk with respect to ;βk^PRP is given. With two Armijo-type line searches, the authors investigate the global convergence properties of the dependent PRP conjugate gradient methods, which extend the global convergence results of PRP conjugate gradient method proved by Grippo and Lucidi （1997） and Dai and Yuan （2002）.