An edge detection method based on a good point set genetic algorithm

In order to improve the convergence rate of genetic algorithms based on edge detection, a novel edge detection method based on a good point set genetic algorithm (GGA) was proposed. The proposed method designed the crossover operation with the theory of good point set in which the progeny inherits the common genes of the parents which represent its family so as to improve the convergence rate of the genetic algorithm. Furthermore, before the algorithm was used for edge detection, the feature space of the image grey level was transformed into the feature space of the fuzzy entropy. Dissimilarity enhancement processing next was applied to the image by using a fuzzy entropy theory to filter the non edge pixels so as to reduce the scale of the solution domain. This approach offered another efficient way to improve the convergence rate. Experimental results show the proposed algorithm performs very well in terms of convergence rate. The detected edge image is well localized, thin, and robustly resistant to noise.     

Hybrid Genetic Algorithms for Multimodal Optimization

Hybrid genetic algorithms, which are based on steepest descent algorithm and genetic algorithm, are investigated for the purpose of multimodal optimization. The performances of the hybrid genetic algorithms are evaluated with criteria such as convergence probability, average convergence time and average convergence value of the function in the case of solving global optimization for Schaffer function. It is shown that the performances of the hybrid genetic algorithms are better than steepest decent algorithm or genetic algorithm, and the hybrid genetic algorithm, in which the individuals used for local optimization by steepest decent method are chosen by chance in each generation population, is more efficient than that in which the individuals used for local optimization by steepest descent method are selected from excellent individuals.     

神经网络自变量输入模式的视电阻率求解算法

根据瞬变电磁场理论公式中的响应和自变量之间的关系特点,提出用非线性方程模式的BP神经网络求解电阻率。通过构造单输入单输出网络结构,建立以不同时间点上的电流归一化的感应电压值为输入、视电阻率值为输出的神经网络,来拟合瞬变电磁场的二次涡流曲线。利用数值方法计算出的数据验证该训练网络的精确性,比较了不同算法对训练精度和收敛速度产生的影响。以重庆大学某处的防空洞探测实验为例验证了该算法的有效性,该算法避开具体的复杂电磁场计算或数值反问题计算,从而实现电阻率快速计算,为快速成像准备必要条件。     

A Design of RBF Neural Networks Based on Genetic Algorithms

Genetic algorithms are proposed to automatically configure RBF neural networks.The network configuration is formed as a subset selection problem.Using new genetic operators,the genetic algorithms are improved in convergence and applicatility.The configuration and optimal principle of genetic algorithms are introduced.The method of generating the RBF networks is given.Finally,the simulated result on practical data is used to demonstrate the performance of the new algorithms.     

TRUST REGION METHODS FOR NONLINEAR EQUALITY CONSTRAINED OPTIMIZATION

This paper presents two trust region algorithms for solving general nonlinear equality constrained optimization problems. The different techniques are used in minimizing a quadrdtic approximate model of the original optimization problems in trust region. By the convergence analysis, the global convergences and local superlinear convergences of the algorithms are obtained. Finally,the implementations of the algorithms are given.     

A method for evolving the architecture and weights of neural networks via gene expression programming

An algorithm for automatic designation of the architecture and the weights of neural networks using gene expression programming (GEP) was presented. The fundamental ideas and procedures of the algorithm were discussed. The algorithm was improved to solve the problems of prematurity and lower variance rate. An application for neural networks designation was given. The experimental results indicate that the proposed GEP approach may evolve the architecture of neural network, and can obtain the weights more precisely. Compared to other conventional evolutional algorithms, GEP shows faster convergence.     

THE STUDY OF SYMMETRICAL COMPONENT ALGORITHMS

Three symmetrical component algorithms are presented in this paper. Walsh algorithm, half-cycle Fourier algorithm and half-cycle Fourier alogrithm of changing correlating functions on the (l,o) plane. And as these algorithms have the advantages of fast respondence, high accuracy and less calculation amount, calculation can be done with a cheap microcomputer.     

Fuzzy Backpropagation Algorithms and Their Convergence

The equivalence between fuzzy neural networks model for max-min fuzzy operator and S.Stoeva's is proved by studying the fuzzy neural networks model for max-min fuzzy based on S.Stoeva's.Then the paper proposes the fuzzy backpropagation learning algorithms for changing fuzzy power and probes their convergence properties.Finally,it simulates experiment such as state monitoring of turbo-generator set.The results show that the fuzzy backpropagation learning algorithms presented are convergent on condition that the output of training sample is between maximum and minimum of its input.     

基于GWO优化SVM的小麦籽粒优劣分级研究

小麦籽粒优劣不仅是产量及品质的重要决定因素,也是育种适应性的综合指标。为了提高小麦籽粒优劣分级的准确率,同时克服神经网络中存在的收敛速度慢、容易陷入局部极值等缺陷,提出一种灰狼算法（GWO）优化支持向量机（SVM）的小麦籽粒优劣分级方法,以航麦8805为研究对象,利用图像处理技术对小麦籽粒图像进行预处理并提取小麦籽粒的形态、颜色和纹理等21个特征。然后采用灰狼算法对支持向量机的两个参数（c、σ）进行优化,建立GWO-SVM模型,从而对小麦籽粒进行优劣分级。与其他算法相比,GWO优化SVM的算法对小麦籽粒的分级准确率有明显的提高,对小麦籽粒优劣分级的准确率可达到95.08%。     

不同失稳判据下边坡稳定性的规律性

以塑性区贯通、位移增量突变、计算不收敛3种边坡失稳判据为依据,采用强度折减有限元法和重度增加有限元法对简单边坡进行了分析。结果表明：以边坡潜在滑动面上某点位移增量突变作为边坡失稳判据是准确的;对于不同土体强度参数下,以位移增量关系曲线突变为判据得到的边坡的安全系数较另外两种方法稳定;对应于塑性区贯通、位移增量曲线突变和计算不收敛的3种判据,边坡潜在滑动面依次向深层发展,边坡的安全系数依次增加。     

Self-adaptation Optimize BP Neural Network Design Based on the Genetic Algorithms

BP(Back Propagation) Neural networks is in the presence of the local optimization in the Neural networks training.The algorithm have slow convergence and the local convergence problem which impact the neural networks work performance.In order to cover these shortcomings and solves the size's hugeness and the low efficiency of the net problem in the traditional NN designing,the action principles of BP-Neural network's structure are analyzed,and a new method is formed which is confirmed from the Enhance genetic algorithms(EGA).The method can identify network configuration and network training methods.By adopting the number coding,self-adaptable multi-point variations operation,this method can effectively reduce the network size and the network convergence time,increase the network training speed.Tomatoes disease diagnosis examples illustrate the feasibility of this approach.     

Calculation Method of Hydraulic Characteristics for Direct-Return Chilled Water System

In order to optimize design and operation, it is necessary to calculate accurately hydraulic characteristics large central air-conditioning systems. Simplified Model can not be applied to hydraulic calculation of large pipe network topological structure for its large calculation error. Taking into account the regulating characteristics of terminal branch temperature regulating valve, a direct-return chilled water system is studied, an accurate mathematical model of pipe network hydraulic characteristics has been established and a computer logic algorithm with virtual flow has been given. A direct-return pipe network with ten AHU branches is used as simulation calculation object, the pipe network supply-return water pressure difference is calculated, and also the temperature regulating valve opening and actual flow of each branch are calculated in the condition of different supply-return water pressure difference. The calculated results accord with the natural characteristics of high pressure difference unbalanced for direct-return pipe network. The calculation method can ensure calculation convergence.     

Improvement and performances of DV hop localization algorithm in wireless sensor networks

In order to improve the node localization precision of Range Free based DV Hop localization algorithm in wireless sensor networks (WSNs) with randomly distributed nodes and dynamic topology, the improved algorithm is proposed. After analyzing the DV Hop algorithm, considering the obvious errors of the estimated node coordinates calculated by Polygon based method in traditional DV Hop algorithm, the numerical iterative algorithm is constructed by employing Taylor series expansion, and simulation studies for the improved DV Hop algorithm are conducted. The selection criteria for the convergent threshold of iterative step is determined, the localization performance of the improved localization algorithm is analyzed by comparing with the traditional DV Hop algorithm under the same condition of selected convergent threshold and simulation parameters, while the calculation amount and convergence rate of the improved algorithm are also measured by the statistic iterations. The simulation results show that by selecting reasonable iterative threshold values and appropriately increasing calculation amount for node localization, the improved DV Hop localization algorithm greatly improves the localization precision and the error stability, which is feasible for node localization in WSNs with both randomly distributed nodes and dynamic topology.     

NLPNN Model in Power Systems Reliability Evaluation

The purpose of a power system is to deliver the power the customers require in real time, on demand, within acceptable voltage and frequency limit, and in a reliable and economic manner. Using NLPNN (New Linear Programming Neural Network) model, the behavior analysis of power system under fault conditions is carried through. The stability and convergence theorems are set up to obtain the resolution for the problem of the load curtailment strategy. A new way is open up to evaluate the reliability of power systems and the great progress in calculation analysis model is achieved. The new calculation technique and theory foundation of reliability evaluation are supplied.     

Application of a Particle Swarm Optimization Algorithm in Truss Structure Optimal Design

The particle swarm optimization (PSO) algorithm developed in recent years is a stochastic optimization algorithm based on swarm intelligence. It possesses advantages such as being a simple concept, ease of implementation and low resource occupation. The PSO algorithm was adopted to solve the problem of size optimization of truss structures with stress and displacement constraints. We present the basic principle of the original PSO algorithm in detail, and then introduce a constriction coefficient to modify it. Moreover, reasonable values of the coefficients are proposed for the modified PSO algorithm. Several classical problems are solved using the modified PSO algorithm, and the results compared with those using traditional optimization algorithms and genetic algorithms. Numerical results show that the modified PSO algorithm has good convergence and stability, and can be applied to the size optimization of truss structures.     

Modeling and Algorithms Implement of Traversing of BOM Production Based on Relational Database

The mono-layer BOM and multi-layer BOM trees are defined, and two of the basic algorithms of traversal of multilevel BOM are summarized as well,which are Recursive Search and Level Search. Based on the analysis of algorithms principles, and evaluating the calculating rate, percentage of using CPU and practicability of them, a kind of improved Traversal of multi-level BOM algorithms is proposed. It is applied in relational database with high calculating rate, lower percentage of using CPU and high applicability. The software implement of these algori-thms is also introduced in detail, and the results of its application show that these algorithms are feasible and effective.     

Regional Economic Convergence in Mexico: An Analysis by Industry

ABSTRACT Existing studies of convergence across jurisdictions of a nation have focused on developed economies. A key assumption underlying regional convergence is geographical factor mobility, and in a developed economy, mobility is facilitated by low transportation costs. By the same token, convergence in a less-developed economy may be impeded by the absence of a well-developed transportation infrastructure. We examine the rate and industrial composition of economic convergence in a neighboring less-developed country (LDC), Mexico, to examine how it might have differed from the U.S. experience. We find evidence of stronger convergence in Gross State Product per capita in Mexico relative to existing estimates of U.S. convergence. Further, while manufacturing activity has been found to be a primary source of convergence in the U.S., we find weaker evidence of convergence of manufacturing activity in Mexico. On the other hand, industries such as hotels and transportation were found to be significantly influential in regional convergence in the Mexican economy.     

Feature of Iteration Convergence in Concurrent Design

A parameter, proportion of all work on every task in each stage, is introduced to build a new WTM. Based on this new WTM, this paper deeply analyzes the relationships between the eigenvalue and the convergence rate of the total work, and the relationships between the eigenvector and the total work. By comparative analysis between the existing WTM and new WTM, it is investigated that this parameter causes the increasing total work and decreasing convergence rate, and the existing WTM is just a speciality when this parameter becomes the maximum. Finally, the generality of the new WTM is tested by the design process of the camera.     

Cylindricity error evaluation using artificial bee colony algorithm with tabu strategy

For cylindricity error evaluation, the Least Squares Method (LSM) is not good enough because of the big error, while the bionics algorithms such as Genetic and Ant Colony Algorithm need to set many parameters and converge slowly, an Artificial Bee Colony Optimization Algorithm is proposed to evaluate the minimum zone cylinder (MIC). This Algorithm refers the tabu strategy for tabu search algorithm to use the Tabu table to save the local optimization results. It enhances the control effect of parameter limit and improves the global convergence ability. Experiment results indicate that this method can converge to the global optimization very quickly. The average runtime is nearly 1.2 s. It is applicable to the real time processing system such as Three Coordinate Measuring Machine.     

Simulation of the constant modulus blind equilibrium algorithm with variable step size based on the mean square error

The mathematic models and basic theory of constant modulus algorithm (CMA) for blind channel equalization are introduced. An improved algorithm with variable step size is proposed based on CMA. The improved algorithm uses the mean square error (MSE) to obtain the new variable step size to solve the contradiction between the convergence rate and accuracy in traditional CMA with fixed step. To smooth the MSE learning curve for easy comparison, an ensemble average technique is used to analyze the performance of convergence with different step sizes. The theoretical analysis and the simulation results show that the improved algorithm is superior to the traditional constant modulus(CM) in terms of convergence speed and residual error.