基于遗传算法的直接再利用网络设施定位方法

针对逆向物流网络中同时包含正向流和逆向流的设施定位问题,构建一种逆向物流直接再利用网络能力受限设施定位模型,以增加生产能力约束。利用遗传算法的全局搜索能力和新颖的约束处理避免了传统拉格朗日启发式算法中求上下限的繁琐计算过程。在不同的遗传算法参数下进行多次实验评定,均能收敛到全局最优解,且计算稳定、收敛速度快。     

基于遗传算法的机器人运动学逆解

在分析以往逆解方法的基础上，提出了用遗传算法求解机器人运动学逆解的方法，给出了用于优化求解的适合度函数，并提出用二次编码法提高解的精度．计算机模拟证明：该方法能快速收敛于全局最优解，能给出机器人的可能解，并能计算冗余度机器人的逆解．     

保留精英遗传算法收敛性和收敛速度的鞅方法分析

论文引入鞅方法取代传统的马尔科夫链理论,研究保留精英遗传算法(EGA)的收敛条件和收敛速度.通过把EGA的最大适应值函数过程描述为下鞅,基于下鞅收敛定理构造使算法满足几乎处处收敛的充分条件,分析了概率1收敛充分条件与算法操作参数的关系,并计算了EGA获得全局最优解所需的最大进化代数.使用鞅方法分析遗传算法收敛性具有独特的优势,成为分析遗传算法收敛性及其性能的新方法.     

基于遗传模糊C-均值聚类算法的地图分割

图像分割和对象提取是从图像处理到图像分析的关键步骤。经典的模糊C-均值聚类算法(FCMA)是将图像分割成C类的常用方法,但依赖于初始聚类中心的选择。该算法通常得到的是局部最优解而非全局最优解。遗传算法是一类全局优化搜索算法。通过将遗传算法(GA)与FCMA相结合,对彩色地图直接按红绿蓝(RGB)三色空间进行聚类,用遗传算法搜索全局最优解,有效地避免了模糊C-均值聚类算法收敛到局部最优的问题,并在此基础上实现了对彩色地图的分割,得到了比较满意的效果。     

基于一种改进的遗传算法的空间聚类分析

空间数据挖掘是数据挖掘的一个研究分支。空间聚类分析是空间数据挖掘的一个重要的研究领域。传统的K-均值方法用于聚类具有收敛速度快、算法实现简单等特点,但容易陷入局部最优,并对初始解敏感。遗传算法是一种全局搜索算法,但是收敛速度较慢。提出一种改进的遗传算法进行聚类,该算法通过全局搜索与局部搜索相结合,取得较好效果。实验表明:文中提出的算法在聚类分析中搜索到全局最优解(或近似全局最优解)的能力要优于经典的K-均值聚类算法,且局部收敛速度和全局收敛性能较好。     

基于复合粒子群算法的几何约束求解技术研究

在将几何约束问题的约束方程组转化为优化模型的时候,需要找到一种方法来跳出局部最优解,进而找到全局最优解。为了兼顾算法的快速性和全局性,几何约束求解时,考虑使用复合粒子群算法。这种粒子群算法是一种基于群智能方法的演化计算技术,不仅在所有的进化算法中都包括控制其自身特性的启发式参数,而且这些参数通常是与特定的问题相关,并可由用户自己定义。虽然合适的参数选择需要用户丰富的经验和对研究问题所提供信息的正确判断,更重要的是,这些启发式参数会影响到算法的收敛特性,但是即便是很有经验的用户也可能选择不恰当的参数,从而使问题得不到有效地解决,这就越来越需要对这些参数进行研究。为此可将将粒子群算法中的控制参数的选取作为一个优化问题,以便用常规遗传算法来控制粒子群算法中的启发式参数,即形成了复合粒子群优化算法,通过把复合粒子群算法成功地应用到几何约束求解技术的实验表明,该方法可以在很短的时间内找到最优解。     

基于一种改进的遗传算法的空间聚类分析

空间数据挖掘是数据挖掘的一个研究分支。空间聚类分析是空间数据挖掘的一个重要的研究领域。传统的K-均值方法用于聚类具有收敛速度快、算法实现简单等特点，但容易陷入局部最优，并对初始解敏感。遗传算法是一种全局搜索算法，但是收敛速度较慢。提出一种改进的遗传算法进行聚类，该算法通过全局搜索与局部搜索相结合，取得较好效果。实验表明：文中提出的算法在聚类分析中搜索到全局最优解（或近似全局最优解）的能力要优于经典的K-均值聚类算法，且局部收敛速度和全局收敛性能较好。     

基于改进的选择算子和交叉算子的遗传算法

为了有效解决遗传算法中收敛速度与局部最优解的矛盾,文中提出了一种具有改进的选择算子和改进的交叉算子的遗传算法。使用文中改进的选择算子,能够增加算法收敛于全局最优解的概率,从而不容易陷入局部最优,也就增加了找到最优解的概率,使用文中改进的交叉算子可以加快算法的收敛速度,从而缩短寻找最优解的时间。实验证明,这两种改进算子的结合能以较快速度收敛于全局最优解,因此能很好地解决遗传算法中收敛速度与局部最优解之间的矛盾。     

基于改进的选择算子和交叉算子的遗传算法

为了有效解决遗传算法中收敛速度与局部最优解的矛盾,文中提出了一种具有改进的选择算子和改进的交叉算子的遗传算法。使用文中改进的选择算子,能够增加算法收敛于全局最优解的概率,从而不容易陷入局部最优,也就增加了找到最优解的概率,使用文中改进的交叉算子可以加快算法的收敛速度,从而缩短寻找最优解的时间。实验证明,这两种改进算子的结合能以较快速度收敛于全局最优解,因此能很好地解决遗传算法中收敛速度与局部最优解之间的矛盾。     

一种融合改进模拟退火技术的新型遗传算法

通过对传统遗传算法和模拟退火算法各自优缺点的分析,提出了一种新型遗传算法。算法以最优保留策略的遗传算法作为主体流程,在主体流程过程中融入改进的模拟退火技术,即设置双阈值和保留中间最优解来减少计算量,加快算法的收敛速度,从而为求得全局最优解提供了保障。通过对F6函数仿真试验,新型遗传算法的收敛速度和跳出局部最优的能力有很大提高。     

Enhancing genetic algorithms for dependent job scheduling in grid computing environments

Genetic Algorithms (GAs) are stochastic search techniques based on principles of natural selection and recombination that attempt to find optimal solutions in polynomial time by manipulating a population of candidate solutions. GAs have been widely used for job scheduling optimisation in both homogeneous and heterogeneous computing environments. When compared with list scheduling heuristics, GAs can potentially provide better solutions but require much longer processing time and significant experimentation to determine GA parameters. This paper presents a GA for scheduling dependent jobs in grid computing environments. A?number of selection and pre-selection criteria for the GA are evaluated with an aim to improve GA performance in job scheduling optimization. A?Task Matching with Data scheme is proposed as a GA mutation operator. Furthermore, the effect of the choice of heuristics for seeding the GA is investigated.     

融入IBN-NET的轻量网络在金属圆柱工件缺陷识别中的应用

金属圆柱工件的缺陷特征容易受到环境光影响,而使用传统卷积神经网络检测金属圆柱工件缺陷,存在网络参数多,运算量大和泛化能力低等问题,难以满足工业现场检测的实时性和高精度要求.针对这些问题,提出一种融入实例-批归一化网络(IBN-NET)的轻量网络模型.在轻量卷积神经网络SqueezeNext的基础上,加入增强泛化能力的IBN-NET,将浅层卷积层后的批标准化(BN)用一定比例的实例标准化(IN)替代,形成网络模型的基础模块;通过组合基础模块,形成改进的网络模型.实验采用具有5类金属圆柱工件缺陷的图像进行对比测试,结果表明,融入IBN-NET的改进网络模型拥有更高的泛化能力,在GTX1080显卡上,改进网络模型仅需0.58 M参数量和5.54 ms的识别时间就能达到95.8%的识别精度.     

Design of highly effective multilayer feedforward neural network by using genetic algorithm

This paper presents a highly effective and precise neural network method for choosing the activation functions (AFs) and tuning the learning parameters (LPs) of a multilayer feedforward neural network by using a genetic algorithm (GA). The performance of the neural network mainly depends on the learning algorithms and the network structure. The backpropagation learning algorithm is used for tuning the network connection weights, and the LPs are obtained by the GA to provide both fast and reliable learning. Also, the AFs of each neuron in the network are automatically chosen by a GA. The present study consists of 10 different functions to accomplish a better convergence of the desired input–output mapping. Test studies are performed to solve a set of two-dimensional regression problems for the proposed genetic-based neural network (GNN) and conventional neural network having sigmoid AFs and constant learning parameters. The proposed GNN has also been tested by applying it to three real problems in the fields of environment, medicine, and economics. Obtained results prove that the proposed GNN is more effective and reliable when compared with the classical neural network structure.     

改进蚁群算法及其在高光谱影像分类中的研究

针对高光谱影像波段数目多,易造成维数灾难的问题,结合遗传算法提供的初始启发信息和蚁群算法寻优能力的优势,提出一种基于改进二进制蚁群算法的波段选择方法。该方法通过遗传算法寻优获取几组较优解,经过计算后作为二进制蚁群算法的初始启发式信息,利用二进制蚁群算法的全局搜索获取最优解;另一方面,为充分利用影像的光谱与空间信息,将波段组合的光谱特征与改进二进制蚁群算法选择的纹理特征融合进行分类,可以获得更高的分类精度。实验结果表明,改进二进制蚁群算法与遗传算法、蚁群算法、二进制蚁群算法相比全局搜索能力更强,且该方法分类精度达到95.63%。     

Stepwise-then-intelligent algorithm (STIA) for optimizing remotely sensed image rectification

ABSTRACT

To address the problems of parameter selection and accuracy optimization of models in image rectification, this article first proposes a novel stepwise-then-intelligent algorithm (STIA) for image rectification optimization, which includes the following steps. First, stepwise regression is suggested to simultaneously solve the over-parameterization problem and select the optimum parameters of the polynomial model and rational function model according to different terrains. Second, intelligent algorithms, e.g. the genetic algorithm (GA) and particle swarm optimization (PSO), are proposed to search for better results based on an innovative search range determined by the uncertainty propagation and 3-sigma rule. The experimental results show that the proposed STIA can achieve higher accuracy than conventional methods; and in most cases, the PSO algorithm used in STIA is superior to the GA used in STIA in measures of time and accuracy. Moreover, stepwise-then-PSO algorithm exhibits the best performance of all compared methods, including least squares, stepwise regression, total least squares and partial least squares.     

Randomized line search techniques in combined GA for discrete sizing optimization of truss structures

This paper presents two randomized line search techniques, each combined with a genetic algorithm (GA), to improve the convergence and the accuracy ratio for discrete sizing optimization of truss structures. The first technique is a simple one-dimensional line search in which design variable axes are selected randomly as search directions. The second is a line search technique whose search direction is determined randomly by fitness function values and differences in the genotypes of individuals. To apply the above-mentioned line search techniques without difficulty, real coding is adopted for discrete problems. The line search techniques are applied to discrete optimization problems of minimum-weight truss structures subjected to stress and displacement constraints. The proposed techniques provide convergence to better solutions than a conventional GA.     

Entropy based region reducing genetic algorithm for reliability redundancy allocation in interval environment

This research paper presents a multi-objective reliability redundancy allocation problem for optimum system reliability and system cost with limitation on entropy of the system which is very essential for effective sustainability. Both crisp and interval-valued system parameters are considered for better realization of the model in more realistic sense. We propose that the system cost of the redundancy allocation problem depends on reliability of the components. A subpopulation and entropy based region reducing genetic algorithm (GA) with Laplace crossover and power mutation is proposed to determine the optimum number of redundant components at each stage of the system. The approach is demonstrated through the case study of a break lining manufacturing plant. A comprehensive study is conducted for comparing the performance of the proposed GA with the single-population based standard GA by evaluating the optimum system reliability and system cost with the optimum number of redundant components. Set of numerical examples are provided to illustrate the effectiveness of the redundancy allocation model based on the proposed optimization technique. We present a brief discussion on change of the system using graphical phenomenon due to the changes of parameters of the system. Comparative performance studies of the proposed GA with the standard GA demonstrate that the proposed GA is promising to solve the reliability redundancy optimization problem providing better optimum system reliability.     

改进的基于粒子群优化的支持向量机特征选择和参数联合优化算法

针对支持向量机(SVM)中特征选择和参数优化对分类精度有较大影响,提出了一种改进的基于粒子群优化(PSO)的SVM特征选择和参数联合优化算法(GPSO-SVM),使算法在提高分类精度的同时选取尽可能少的特征数目。为了解决传统粒子群算法在进行优化时易出现陷入局部最优和早熟的问题,该算法在PSO中引入遗传算法(GA)中的交叉变异算子,使粒子在每次迭代更新后进行交叉变异操作来避免这一问题。该算法通过粒子之间的不相关性指数来决定粒子之间的交叉配对,由粒子适应度值的大小决定其变异概率的大小,由此产生新的粒子进入到群体中。这样使得粒子跳出当前搜索到的局部最优位置,提高了群体的多样性,在全局范围内寻找更优值。在不同数据集上进行实验,与基于PSO和GA的特征选择和SVM参数联合优化算法相比,GPSO-SVM的分类精度平均提高了2%~3%,选择的特征数目减少了3%~15%。实验结果表明,所提算法的特征选择和参数优化效果更好。