Search

A grey wolf optimization algorithm with global search strategy is proposed to address the issue that the standard grey wolf optimizer (GWO) has the disadvantage of global search ability. Adaptive weight and search strategy randomly are added to strengthen the global search ability. 13 test function are used to test the performance of the proposed algorithm compared with other well-known meta-heuristics, and the results show that the proposed algorithm is feasible and effective.

Lion swarm optimization is a swarm intelligence algorithm that simulates the behaviors of lion king, lioness and cub. Lion king is the individual who occupies the best position in the lion group, but only a single position update is carried out in each iteration, so the search is not careful enough. To solve the above problem, a tabu annealing lion swarm optimization algorithm is proposed, which uses the memory tabu annealing algorithm with tabu list to enhance the lion king's optimization ability. The experimental comparisons on CEC2013 test function show that the proposed tabu annealing lion swarm optimization algorithm has faster convergence speed, higher accuracy and is more effective.

A novel local enhanced sampling strategy is proposed for microwave component optimization in this paper. A greedy-like method is employed to increase the samples near potential optimal solutions to facilitate optimization-oriented modeling. The Ackley test function and a microwave filter are used to verify the ability of the local enhanced sampling strategy in surrogate-based optimization. It is shown that our proposed sampling method promotes the ANN model to obtain better optimization results. We also investigate the effect of numbers of layers and nodes of the ANN model on the modeling accuracy.

The Ground station Scheduling Problem refers to the allocation of the communication tasks, among the ground stations and satellites. In general, the problem is formulated as a many-objective problem. The NSGA-III is an algorithm developed to solve such problems. Due to its selection operator that uses a number of reference points, the NSGA-III gives users the option to specify their own reference points. In this paper, we use this opportunity by generating distributed reference points, shifted depending on weights. A specific weight is assigned to each objective that corresponds to reference points for preferred solutions. The generation of these reference points and their effect on the final objective function values of the Pareto front are first tested on the DTLZ2 test function with 4 objectives and for a different combination of weights. Finally, various weights are applied to an instance of the Ground station Scheduling Problem, leading to Pareto fronts that favor specific objectives and feasible schedules.

The development of industry has promoted the progress of science and technology, and steel production is the basis of industrial development. With the progress of science and technology, the excessive consumption of global resources and environmental problems are becoming increasingly serious. As one of the important processes in iron and steel production, sintering has a large demand for raw materials containing iron ore and it has a high production cost. On the basis of reaching the range of technological standards, it is of great significance to study the batching scheme that can meet the requirements of technological indexes.Based on the analysis of the characteristics of sintering burdening process and the study of sintering burdening process, an optimization model of sintering primary and secondary burdening based on linear programming (LP) was established. Not only was the mathematical model devoted to reducing the total cost of sintering, but it also tried to reduce the emission of sulfur-containing substances. At the same time, it was restricted by the chemical composition index. After analyzing the batching process, this paper proposes a combination of genetic algorithm and the chicken swarm algorithm, analyzes its working process, and proves that the improved genetic chickens hybrid algorithm has stronger optimal through the test function. The genetic chickens hybrid algorithm based on linear programming is used in the first and second ingredients optimization of sintering process, and then it is compared with other ways in terms of reducing the cost of sintering and reducing emissions of sulphuric materials. The results show that it is feasible in theory and reliable in practice to apply the intelligent hybrid optimization algorithm to sintering proportioning optimization.

With the development of the Internet, malicious URL attacks have shown a complex and changeable situation, and have always been one of the main factors that threaten network security. In order to detect malicious URLs more effectively, a random walk strategy is proposed to improve the Sparrow search algorithm, which is compared and analyzed with the Sparrow search algorithm, the gray wolf optimization algorithm and the whale optimization algorithm. After a benchmark test function experiment, it is verified that the improved sparrow search algorithm has the characteristics of fast convergence speed and high optimization precision. In order to further verify the practicability of the improved algorithm, the optimal solution of the random walk strategy optimization sparrow search algorithm is used as the initial neuron connection weight and threshold information of the BP neural network and applied to the classification of malicious URLs. The results show that the random walk strategy optimization sparrow search algorithm has a certain ability to optimize the BP network, which can improve the classification accuracy of malicious URLs.(Abstract)

Aiming at the problem of slow convergence speed and ease of falling into local optimum when solving high dimensional problems, this paper proposes an improved chicken swarm optimization algorithm. The improved chicken swarm optimization includes four aspects, namely, cock position update mode, hen position update mode, chick position update mode, and population update strategy, so it is abbreviated as ICSO-RHC. On the basis of algorithm improvement, the influence of the number of retained elite individuals and control parameters on the convergence speed of the algorithm is discussed. The calculation results of the test function show that when the number of elite individuals in the population is 1, and the control parameters is a random number uniformly distributed between [0, 1], the algorithm has a faster convergence speed. In addition, in order to verify the performance of ICSO-RHC, 30 test functions and CEC 2005 benchmark functions were selected. The calculation results of these test functions show that the success rate of ICSO-RHC is significantly higher than other algorithms, both for low-dimensional and high-dimensional optimization problems. The average iteration number and average running time are significantly lower than other algorithms. Finally, ICSO-RHC and other improved algorithms in the literature are used to optimize the parameters of four practical engineering problems. The optimization results show that the statistical results obtained by ICSO-RHC are significantly better than other algorithms. The calculation results of the test functions and the actual engineering problems show that the performance of ICSO-RHC proposed in this paper is significantly better than other algorithms.

With the successive launch of multiple intelligent high-speed railways, the functions of the CTC have continued to expand. Intelligent adjustment, big data processing, and multifunctional interaction have become the new outstanding features of the CTC. In order to ensure the stability and reliability of the new generation of intelligent CTC, the testing methods of system functions need to be continuously improved and innovated. CTC functional testing can be divided into two aspects: test case design and test environment design. Comprehensively considering the characteristics of CTC, this paper introduced the design method of CTC functional test cases in detail, from the aspects of functional logic forward verification, functional logic boundary protection, performance testing, compatibility testing, business scenario applications, etc. At the same time, an environmental structure that is convenient for simulating test input conditions was designed according to the test function category, including mirror structure test environment, parallel structure test environment, information recharge test environment, dynamic simulation test environment, etc., and the characteristics and principles of the structure were analyzed one by one.

The traditional numerical calculation methods of electrostatic field include finite element method, finite difference method and so on. Due to the time-consuming meshing, the meshless method is gradually used by scholars for physical field analysis. At present, the meshless local Petrov-Galerkin method and the meshfree weak-strong method have been reported in various fields with their better accuracy and wide applicability. This paper presents an improved meshless strong-weak coupling method based on test function. The scattered nodes are used to represent the problem domain and boundary, and the shape function is constructed by the moving least square method. For the nodes where the local integration domain does not intersect the global derivative boundary, the meshless strong form method is used to construct the basic equation. For other nodes, the meshless local weak formula method is used. The exponential function is used as a test function for weak formula expressions. By changing the coefficient of the exponential function, the shape of the test function is controlled, and the influence of the change in the size of the influence domain on the calculation accuracy is reduced. The case analysis shows that this method has higher accuracy and better convergence effect than the traditional meshless method. The proposed method has high precision and computational efficiency in processing electrical numerical calculations, and has strong astringency. This method has better application prospects than traditional methods.

According to the characteristics of classical PSO algorithm(s), this paper uses the spatial equal-scale segmentation method (PESS) to effectively reduce the data dimension because of the low efficiency of PSO algorithm caused by big data and large population size. Diversity selection and global optimization, according to the law of evolution selection, the local optimization is focused on the two extremes of minimum value and maximal value, thus reducing the number of calculations of the fitness function of the algorithm. On this basis, we introduce a time-hop and elite Gaussian mixture strategy that can overcome the particle plunging into local optimum, while improving the efficiency of the algorithm, the generalization ability of the algorithm is taken into account. In order to prove the performance of FSPSO, we chose PSO and the self-organizing migration algorithm (SOMA) evolved from PSO that were tested in 30, 60 and 100 comparison tests on 5 different test function set with different degrees of complexity. The experimental results show that we proposed FSPSO algorithm has good effectiveness, robustness, complexity and universality in solving global optimization problems.

This paper proposed a method to determine the optimal location and sizing of Distributed Generation (DG) for power loss minimization using Ant Lion optimizer (ALO). The analysis also covers the effect of DG installation to voltage improvement and maximum system loadability index. ALO is an optimization algorithm that based on the nature interaction between ants and antlions. The nature interaction consists of five steps of hunting prey such as random walk of ants, building traps, entrapment of ants in traps, catching preys, and rebuilding traps. The ALO algorithm is tested on IEEE 69-bus distribution test system. The system will find the optimal location and sizing with corresponding load increase until it reaches the maximum system loadability (MSL) of the network. DG are usually attached to the end terminal at the load side of the system that refers to a technology that generate electricity at or near where it will be used such as solar panels and combined heat and power. The result of test function shows that the proposed algorithm can provide accurate yet competitive result in terms of power loss minimization, maximum system loadability enhancement and consistency.

Aiming at the problem of incomplete evaluation index of relay guidance task assignment, an evaluation function of the situational advantage and performance advantage of the guidance machine and target was designed. A relay assignment model was established. By adding greedy strategy, Tent chaotic initialization strategy and adaptive inertia weight strategy, Multi-Genetic Algorithm Tent- Dragonfly Algorithm (MGAT-DA) is proposed, the solution of the scheme is completed, and simulation is performed. First, the related theory of the single-target intelligent algorithm is introduced; second, the intelligent algorithm, the dragonfly algorithm, is introduced, and some improvements are made; again, the test function is used to test its optimized performance, and the experiment is compared with the traditional algorithm. ; Finally, the improved algorithm is used to solve the relay guidance task assignment scheme.Secondly, based on the linear programming to solve the optimal task allocation scheme, the guidance advantage under this scheme is 1.3667370376.

Quantum electron transport in modern semiconductor devices can be described by a Wigner equation which is formally similar to the classical Liouville equation. The stationary Wigner equation has a singularity at zero momentum (k=0). In order to get a non-singular solution it is necessary to impose a constraint for the solution at k=0 which gives the constrained Wigner equation. We introduce a Petrov-Galerkin method for the solution of the corresponding constrained sigma equation. The constraint in the Wigner equation is interpreted as an extra test function and is naturally incorporated in the method.

The discontinuous Galerkin time-domain (DGTD) method used for the computation of electromagnetic fields in 3-D structures obtains the weak solution form of the Maxwell’s equations through spatial semidiscrete approximation and integration of test function. Numerical flux needs to be added between unstructured tetrahedral elements to compensate for boundary conditions. The establishment of corresponding surface mapping between mesh elements is a prerequisite for the numerical flux transfer between meshes. Based on fast sorting algorithm, this paper designs a grid correspondence judgment algorithm called ‘Sorting and Querying’, which reduces computational complexity from O(n^2) to O(nlogn). The numerical results show that the algorithm is faster than the similar algorithms and easier to program even in the case of a large number of grids.

Aiming at the defects of chicken swarm optimization algorithm, such as easy to fall into local optimal, premature convergence and slow convergence, a chicken swarm optimization algorithm based on quantum behavior is proposed in this paper. A quantized potential well model is established based on the individual information of chicken swarm. According to the existing individual extremum and global extremum obtained by the original updating formula, Monte Carlo random sampling is adopted to complete the updating of individual extremum, and the search is conducted at a parallel Angle near individual extremum and global extremum, which improves the local search performance of the algorithm. At the same time, the convergence of quantum-behavior chicken swarm optimization algorithm is discussed in this paper, and QCSO is proved to be a globally convergent optimization algorithm. The optimization capability of QCSO is tested by using basic test function, and the results show that the optimization performance of this algorithm is greatly improved compared with the original algorithm.

This paper presents solutions for single objective optimization problems with developed algorithm from simulated annealing based on a simple (μ + λ) -ES, It is divided into two algorithms, separated mutation (SM1) and survival mutation (SM2). After that, compared with randomized local search and simulated annealing. The test function is part of the IEEE WCCI 2020 on the topic of CEC-C2 single objective bound constrained optimization. This research has chosen the basic functions in the test such as Bent cigar function, rastrigin function, high conditioned elliptic function, HGBat function, rosenbrock’s function, griewank’s function, discus function, expanded schaffer’s function, weierstrass function, sphere function, natyas function, levi function N.13, himmelblau’s function, and three-hump camel function. These functions are attract attention and competition. A results of SM1 and SM2 can solve single objective optimization problems better than RLS and SA. In high conditioned elliptic, The fitness value of RLS is equal to 3.96E-11, The fitness value of SA is equal to 8.12E-10, The fitness value of SM1 is equal to 5.39E-14 and The fitness value of SM2 is equal to 9.70E-15, It let us show the efficiency of SM2 that can get better results than SM1.

In order to cooperate with the promotion of high-performance processor BWDSP1042, the cooperation of hardware and software is required, and the library function is one of the most basic and core software in high-performance computing. Its performance will directly affect the efficiency of various upper-level scientific and engineering calculations. Therefore, the implementation of library functions is closely related to the development of hardware platforms. To this end, this study designed an optimized algorithm function library with common algorithm standards based on its unique instruction characteristics and resource structure. First, based on the signal processing requirements of the basic function library, based on the Matlab function library, and The experience in signal processing selects the development functions required for each module, and based on the selected functions, the C language library is developed. Then, combined with the C language algorithm framework, the assembly language is written. The assembly language function library should be consistent with the algorithm flow of the C language function library, and the calculation error with the Matlab language library function is within a reasonable range. Finally, the running time is compared with the actual running period and the theoretical running period by using the clock cycle as a unit to check whether the operation efficiency of the assembly function meets the efficiency requirements of the library function. By making full use of bwdsp1042's computing resources in the study of this subject, the effects of reducing cycle and improving efficiency have been achieved. The relative error of the results obtained by the test function is in the order of 10-7, which is within the tolerance range. In summary, the library function development work based on the bwdsp1042 processor in this research can actively cooperate with the launch of the processor, laying a foundation for the localization of core devices.

With the advent of big data era, complex optimization problems with many objectives and large numbers of decision variables are constantly emerging. Traditional research about multi-objective particle swarm optimization (PSO) focuses on multi-objective optimization problems (MOPs) with small numbers of variables and less than four objectives. At present, MOPs with large numbers of variables and many objectives (greater than or equal to four) are constantly emerging. When tackling this type of MOPs, the traditional multi-objective PSO algorithms have low efficiency. Aiming at these multi-objective large-scale optimization problems (MOLSOPs) and many-objective large-scale optimization problems (MaOLSOPs), we need to explore thoroughly parallel attributes of the particle swarm, and design the novel PSO algorithms according to the characteristics of distributed parallel computation. We survey the related research on PSO: multi-objective large-scale optimization, many-objective optimization, and distributed parallelism. Based on the aforementioned three aspects, the multi-objective large-scale distributed parallel PSO and many-objective large-scale distributed parallel PSO methodologies are proposed and discussed, and the other future research trends are also illuminated.

In current epoch, the economic operation of micro-grid under soaring renewable energy integration has become a major concern in the smart grid environment. There are several meta-heuristic optimization techniques available under different categories in literature. One of the most difficult tasks in cost minimization of micro-grid is to select the best suitable optimization technique. To resolve the problem of selecting a suitable optimization technique, a rigorous review of six meta-heuristic algorithms (Whale Optimization, Fire Fly, Particle Swarm Optimization, Differential Evaluation, Genetic Algorithm, and Teaching Learning-based Optimization) selected from three categories (Swarm Intelligence, Evolutionary Algorithms, and Teaching Learning) is conducted. It presents, a comparative analysis using different performance indicators for standard benchmark functions and proposed a smart micro-grid (SMG) operation cost minimization problem. A proposed SMG is modeled which incorporates utility connected power resources, e.g., wind turbine, photovoltaic, fuel cell, micro-turbine, battery storage, electric vehicle technology, and diesel power generator. The proposed work will help researchers and engineers to select an appropriate optimization method to solve micro-grid optimization problems with constraints. This paper concludes with a detailed review of micro-grid operation cost minimization techniques based on an exhaustive survey and implementation.

Aimed at the problem of the grey wolf optimizer (GWO) algorithm of low convergence speed and easily falling into local optimum, this paper proposes a nonlinear control parameter strategy based on sinusoidal function (GWO-SIN) and a nonlinear control parameter combination (GWO-COM) strategy. In order to verify the validity and advantages and disadvantages of the proposed nonlinear control parameter strategies, selected the adjustment parameter values of GWO-COM from five different adjustment parameter values by test results of simulation of four benchmark functions. Then, this paper compares against the original GWO and four GWO based on different nonlinear control parameter strategies on a set of well-known benchmark functions. Experimental results show that the nonlinear control parameter combination (GWO-COM) strategy is superior to the others and the IGWO (an improved grey wolf algorithm for global optimization), and the optimization result of the GWO-SIN algorithm is similar to GWO-COM in the multi-peak test function.

A vast amount of different electronic oscillators with chaotic behavior have been proposed in literature, but for this study a simple two transistor oscillator has been chosen. The low cost of such a circuit makes it attractive for secure communication between wireless sensor nodes. However, for practical use it is necessary to know, how the deviation of nominals of circuit elements and a supply voltage influences the behavior of the chaotic oscillator and synchronization between distinct circuits. Theoretical analysis was performed using simulations of chaotic oscillators in LTspice program, and MATLAB software has been selected to analyze the obtained data. To evaluate the chaotic behavior of the system, the MATLAB 0–1 test function was employed, and the autocorrelation function was used to evaluate the synchronization of the system. The most sensitive elements were identified, and the regions of stable operation revealed.

Today the problem of lacking effective calibration means for two key test function Indicators(delay characteristics and precision) of substation merging unit (MU) performance test equipment still exists. Under the background, this paper presents a calibration method for MU performance test equipment, develops a relevant calibration system and introduces the structure and technical indicators of the calibration system. The calibration system is based on the hardware system of high performance ARM processor, field programmable gate array (FPGA) and high precision AD chip. Then, a pre-circuit for quick sampling of voltage and current is designed to reduce sampling delay error, a high precision amplitude and phase algorithm of fixed frequency sampling is applied to realize high precision measurement of sampling signals, and a precise and flexible test strategy for delay characteristics is adopted. Finally, the comparative test of MU test equipment is carried out by using this calibration system, which basically solves the practical problem that the core performance of substation MU test equipment cannot be detected.

This paper presents a hybrid algorithm based on the invasive weed optimization (IWO) and particle swarm optimization (PSO), named IW-PSO. By incorporating the reproduction and spatial dispersal of IWO into the traditional PSO, exploration and exploitation of the PSO can be enhanced and well balanced to achieve better performance. In a set of 15 test function problem, computational results, preceded by analysis and selection of IW-PSO parameters, show that IW-PSO can improve the search performance. In the other comparative experiment with fixed iteration, the IW-PSO algorithm is compared with various more up-to-date improved PSO procedures appearing in the literature. Comparative results demonstrate that IW-PSO can generate quite competitive quality solution in stability, accuracy and efficiency. As evidenced by the overall assessment based on two kinds of computational experience, IW-PSO can effectively obtain higher quality solutions so as to avoid being trapped in local optimum.

The paper presents an algorithm and computer software for optimization of the permanent magnet synchronous motor. The grey wolf optimization algorithm has been applied. In-house software has been developed in the Borland Delphi environment. First, the correctness of the algorithm has been tested using analytical test function. The comparison of two non-deterministic algorithms: (a) gray wolf optimizer (GWO) and (b) genetic algorithm (GA) have been performed. Next, the optimization of the permanent magnet synchronous motor has been carried out. The multiplicative objective function has been composed from three PM motors parameters.

In many applications, the feature selection plays an important role, as best feature can bring out the accurate results. The features selected must represent the entire dataset. Here we have considered “Sequential Forward Selection” for feature extraction and used refined dragonfly algorithm to approach and to migrate from the best and worst features respectively. We improvised the conventional dragonfly algorithm by adding the convergence and fitness functions. To access the accuracy of the algorithm we introduced the fitness function. This paper has discussed about the general hunting behaviour of the dragonfly and dragonfly algorithm (DA) with convergence and fitness functions. A comparative study was shown for the best search agent position between modified DA and traditional DA, at the same time test function values of refined dragonfly algorithm (RDA) is compared with whale optimization algorithm (WOA) and Tornadogenesis Optimization algorithm (TOA). We have evaluated refined DA on the 23 benchmark function corresponding values are shown in experiment.

Aiming at solving the shortcomings of traditional wolf swarm algorithm, such as low search efficiency due to fixed search direction and radius, and easy to fall into local optimum when updating rules, an improved wolf swarm algorithm is proposed in this paper. Firstly, the exploding rules of fireworks algorithm are introduced to reduce the exploring steps of wolves near prey and increase the searching direction in order to improve the ability of local exploration, while the wolves far from prey the searching radius are increased and the searching direction are decreased in order to improve the global search ability. Secondly, the moving step size of running behavior is improved automatically to adjust the step size according to the location of each wolf, which can improve the efficiency of running behavior. Finally, the updating rules of wolf swarm algorithm is improved for the individual wolf swarm is selected according to probability to enhance the global optimization ability of the algorithm. The performance test in test function and path planning simulation experiment results show that the improved wolf swarm algorithm has faster convergence speed and higher convergence accuracy.

The complex characteristics of random variant underwater acoustic channel challenge the underwater communication theory and application, i.e. channel modelling and symbol estimation, which has become a focusing problem of research circle. In order to detect the time-varying underwater channels, a real time channel tracking method based on parameter fault detection algorithm is proposed to track the random variation of the underwater channel, which can help the transmitter to determine at what time it is necessary to train the newly changed channel. The basic idea is that the variant channel parameters can be reflected by the swift variation of the designed fault detection index. Simulation results show that the parameter fault detection algorithm can track the stochastic time varying underwater channel dynamically since the test function responds swiftly when the channel changes.

This paper introduces a new strategy by means of the social wound treatment for improving optimization in population-based metaheuristic methods. This strategy is inspired on the behavior observed in social animals when dealing with injuries at both social and individual levels. An implementation of this strategy is done in the so-called wound treatment optimization (WTO) algorithm. A benchmark on several test functions for optimization was conducted to compare the response between the simple particle swarm optimization (PSO) and the WTO. Also, an application of WTO in decentralized system of robots was developed aiming to learn suitable parameters in the navigation controller of robots. This is the first time that social wound treatment is considered in these optimization methods.

The Central Force Optimization (CFO) algorithm is a new multi-dimensional search-determined heuristic optimization algorithm. The results obtained by using CFO algorithm are unstable and easy to fall into local optimum. To solve this shortcoming, we propose a new algorithm for central gravity optimization using Niche and Opposition-Based Learning. Based on the extension theory, the algorithm model is constructed. The niche is divided into groups, we settled shared area for these groups. The adaptive sharing of the particles in the shared area can effectively prevent the algorithm from prematurely converging and enhance the global search ability. The optimal particle is introduced into the elite reverse learning strategy to enhance the development of the solution space and improve the accuracy of the algorithm. The performance of the algorithm is evaluated by the test function, and the results show that the optimization performance of the algorithm is significantly improved.

International audience; Camera relocalization is needed in several applications such as augmented reality or robot navigation. However, it is still challenging to have a both real-time and accurate method. In this paper, we present our hybrid method combing machine learning approach and geometric approach for real-time camera relocalization from a single RGB image. We introduce our sparse feature regression forest to improve the machine learning part. In our regression forest, we propose a novel split function, that uses a whole feature vector instead of classical binary test function to improve the accuracy of 2D-3D point correspondences. Moreover, we use sparse feature extraction (SURF features) to reduce time processing. The results indicate that our method is the only real-time hybrid method (50ms per frame). We also achieve results as accurate as the best state-of-the-art methods (hybrid methods) and outperform machine learning based and sparse feature based methods.; La relocalisation de caméra est nécessaire dans plusieurs applications telles que la réalité augmentée ou la navigation de robots. Cependant, il est encore difficile d'avoir une méthode à la fois précise et temps réel. Dans cet article, nous présentons notre méthode hybride combinant une approche d'apprentissage automatique et une approche géométrique pour la relocalisation de caméra en temps réel à partir d'une seule image RGB. Nous introduisons notre forêt de régression s'appuyant sur ces caractéristiques éparses pour améliorer la partie apprentissage automatique. Dans notre forêt de régression, nous proposons une nouvelle fonction de séparation qui utilise un vecteur caractéristique entier au lieu d'une fonction de test binaire classique pour améliorer la précision des correspondances 2D-3D des points. De plus, nous utilisons l'extraction de caractéristiques éparses (caractéristiques SURF) pour réduire le temps de traitement. Les résultats indiquent que notre méthode est la seule méthode hybride temps réel (50ms par image). Nous obtenons également des résultats aussi précis que les meilleures méthodes de l'état de l'art(méthodes hybrides) et surpassons les méthodes basées sur l'apprentissage automatique et les méthodes basées sur des caractéristiques éparses.

In recent years, with the rise of artificial intelligence and deep learning, as an evolutionary algorithm based on probability model, estimation of distribution algorithm has been widely research and development. The estimation of distribution algorithm without the traditional genetic operation such as crossover and mutation, is a new kind of evolution model. As an algorithm based on probabilistic mode, the estimation of distribution algorithm establishes a probabilistic model describing the solution space of optimization problems. With the emergence for big data, the convergence of the algorithm and the requirements for solving precision are also increasing. This paper attempts to improve the distribution estimation algorithm. The optimal population of each iteration is found through the location update of each iteration of the Particle Swarm Optimization (PSO) algorithm. The simulation test was carried out with ten benchmark test function. The proposed algorithm was compared with the GA_EDA9improved genetic algorithm) and the basic distribution estimation (EDA) algorithm. Experimental results show that the new algorithm is superior to GA_EDA and basic EDA in terms of convergence and accuracy.

Long term optimal reservoir operation is a nonlinear optimization problem. This paper adopts Teaching Learning Based Optimization (TLBO) to solve the Long term scheduling for reservoirs. Firstly, the basic test function is used to verify that the TLBO algorithm has high accuracy. Then the TLBO algorithm is applied to the Long term optimal reservoir operation model and the competitive selection and semi-feasible domain are introduced to deal with the constraint conditions. Comparing the TLBO algorithm with the genetic algorithm and the dynamic programming algorithm, solving the available TLBO algorithm has the advantages of high calculation accuracy, short computation time.

Aimed at the shortcomings of traditional PID control method, such as large overshoot, long dynamic response time and low control accuracy in wire rod looper control, an improved wolf pack algorithm optimizing approximation of fractional order control method is proposed. The method uses fractional order controller to improve the control precision of the looper control system, to realize the rational approximation of fractional calculus operator online optimization using the improved wolf pack optimization algorithm. The gravitational search mechanism introduced into the wolf pack algorithm effectively improves the global search ability and search efficiency of the wolf pack algorithm, and adds dynamic disturbance strategy to the optimal region of wolf pack algorithm to improve the search accuracy and avoid falling into the local optimum value. The improved wolf pack algorithm is applied to test the multimodal test function, the test results show that the wolf pack optimization algorithm based on gravitational search mechanism has better global search ability and higher search accuracy. The application of which into the fractional order wire rod loop control system improves the control accuracy, speed and robustness of loop system, and its comprehensive performance is better than other comparison algorithms.

With the development of power market, the current plant-level load distribution not only conforms with the requirements of load economy in scheduling, but also accords with the demands of rapid load adjustment time. Based on the existing economic model and rapidity model of optimal load distribution, a mathematical model of multi- objective fusion load distribution is established. Global memory and convergence accuracy are added to the standard Gravitational Search Algorithm (GSA). The simulation experimental results of GSA and Improvement Gravitational Search Algorithm (IGSA) is verified by the peak multidimensional test function. At the same time, a multi-objective continuous optimization gravitational search algorithm is proposed, which contains non-dominance sorting, crowding distance calculation and elite strategy in Nondominated Sorting Genetic Algorithm II (NSGA-II). IGSA was applied to multi-objective load distribution problems and compared with the NSGA-II algorithm. Pareto set was obtained through simulation experiments. A multi-attribute decision method based on pole and entropy is applied to select the optimal solution of load distribution. From the experimental results, IGSA can achieve better performance in multi-objective load distribution of power plant.

We present our data modeling and implementation for testing a device of elementary digital as well as analog functions provided with real-world failure models as a jumper construction. A modeled open circuit failure is done by unplugging the (normally plugged) respective jumper, and analogously a modeled short circuit failure is done by plugging the (normally unplugged) respective jumper. The DUT is supplied by a current limited voltage power supply. The programmable current limitation activates the overcurrent relay to protect the device in case of short circuit failure. The ATE resources for stimuli, measure and test evaluation are provided by an adequate uC-Board with programmable analog-to-digital io-channels. The test automation is implemented by the firmware of the uC, which corresponds to the test program of a conventional ATE including the pin-mapping, the test function library and the test flow (schedule). With this in mind, we are speaking about the Micro-ATR (uATR). Within the use case above, we discuss the challenge for localization of the so called dominated (dominent) faults, which are not detectable uniquely due to the dominating faults within the scope of the stuck-at failure modeling. At this case the testing accuracy has to be enhanced. We solve this problem definition by deploying external current sources to provide additional stimuli at the given test points. Our approach results in the overall data modeling, which covers the test tree, the test flow, and the diagnose flow for fault localization.

This paper presents a novel improved grey wolf optimization (IGWO) to avoid local minimization and premature convergence. In the proposed IGWO algorithm, a nonlinear control parameter based on cosine function is presented to ensure a faster convergence rate of late iteration. Genetic algorithm is introduced into the IGWO to avoid premature convergence and trapping in local minima, in which the probability of cross-over and mutation is dynamically adjusted according to the swarm's fitness value. The new approach is compared against the original GWO and GA on a set of well-known benchmark test function. Experimental results show that the presented algorithm is superior to the other two comparative approaches.

Considering the problem that chicks only follow the hens to search for food and the roosters has weak global search capability in original chicken swarm optimization (CSO) algorithm, Mutation Chicken Swarm Optimization (MCSO) was proposed on the basis of Nonlinear Inertia Weight. The proposed algorithm first introduces the variation factor in the chicken particle position updating formula and then introduces the non-linear decreasing weight based on a parabola opening upwards in the rooster location update formula to improve the chicken location update formula and the rooster location update formula, thus the mutation chicken swarm optimization based on nonlinear inertia weight is easier to obtain the global optimal solution than original chicken swarm optimization algorithm. The optimization performance of the proposed algorithm was verified by six test function, the experimental results are compared with the chicken swarm optimization and differential evolution (DE) algorithm. Then, the mutation chicken swarm optimization based on nonlinear inertia weigh (NW-MCSO) was applied to the Anti-saccharification activity prediction. The experimental results show that the proposed algorithm has strong global search ability, high precision and better optimization performance compared with chicken swarm algorithm and differential evolution algorithm.

For effective image classification and classification of image database expansion in the correct rate is not high, this paper proposes an image classification method based on LSSVM and Powell-HS algorithm, this method used the test function to test the Powell-HS algorithm and the search efficiency is high; then using KPCA to extract the image features of the samples as LSSVM model the input, and compared to the other 3 methods, the results show that the new method is applied to image classification, and the classification accuracy has been greatly improved.

The artificial bee colony (ABC) algorithm is a well-known algorithm for solving numerical optimal problems. In this paper, a biologically adaptive artificial bee colony (BAABC) algorithm is proposed. In BAABC, the initialization of population is performed by the Halton Sequence so the population is generated in potential solution space. Besides, the size of population is varying according to status of solution searching. If BAABC finds a good solution, the size of population is increased for enhancing solution searching and vice versa. To examine the proposed method, fifteen evaluated functions are adopted in experiments. The result shows that BAABC performs better than related ABC methods on most of the test function.

Due to the lack of diversity of the initial population, the multi-objective particle swarm optimization algorithm easily falls into the local optimal value during the iterative process. The method of piecewise logistic chaotic map is introduced to increase the randomness of initial population. A disturbance variable is used to weaken the dependency on global optimal value. A segmented maintenance of the external file is used to select the particle which is more representative for the population. A monitoring selection mechanism is used to improve the population jump out of local optimum. The strategy for eliminating the final particle one by one is used to clip the external file. The validity of the proposed algorithm is proved by comparing with the other algorithms on the test function. And the proposed algorithm has been used to solve the vehicle routing problem.

In this work, a hybrid method that includes metaheuristic algorithms has been proposed for solving optimization problems. The proposed method was implemented as employing three metaheuristic algorithms which are Artificial Bee Colony, Differential Evolution and Particle Swarm Optimization in an order. The success of the developed method is presented by testing on 12 continuous optimization test functions which are widely used in the literature. The experimental results show that the proposed hybrid method gives better results than the individual algorithms that make up it.

This paper proposes a multi-objective optimization algorithm based on the membrane computing. Inspired by the theory of membrane optimization, the membrane structure, multiple sets and reaction rules is employed to tackle multi-objective optimization issues. Aiming at adaptability of algorithm, the cross-over and mutation mechanism of the genetic algorithm are introduced to combine with membrane framework. Moreover, for the sake of improving the diversity of global search solution, the non-dominated sorting and crowding distance are used to update external archive. The experimental results demonstrate that the proposed algorithm is not only practicable and efficient but also capable of obtaining the approximate Pareto front in KUR and ZDT test function.

The multi-objective optimization algorithms are used as the best optimizer in many design issues. One of the main challenge for these algorithms is that increasing the number of objective functions leads poor performing of the algorithm. Reducing the selection pressure is the main reason of this phenomenon. In order to overcome this problem, the population diversity should be controlled. In this regard, this study developed a new evolutionary algorithm through resolving this dilemma. In the proposed method, a measure is considered for estimating the diversity of individuals in the population to adaptively control the rate of population diversity. A new fitness evaluation is also provided in this paper for assessing the fitness of chromosomes. So in this schema the selection of chromosomes is based on their contribution to population diversity in addition to being based on their fitness. The obtained results proved that the performance of the proposed algorithm has been improved through various tests. It is worth noting that the potential of the proposed method is examined on the most recent test function that presented in this field.

This paper presents a deep analysis of literature on the problems of optimization of parameters and structure of the neural networks and the basic disadvantages that are present in the observed algorithms and methods. As a result, there is suggested a new algorithm for neural network structure optimization, which is free of the major shortcomings of other algorithms. The paper describes a detailed description of the algorithm, its implementation and application for recognition problems.

Nowadays, the attraction of the optimization techniques based on artificial intelligence has increased the due to obtaining the its successful results on the difficult optimization problems in various areas. The artificial bee colony (ABC) algorithm based on swarm intelligence and the differential evolution (DE) algorithm improved by inspiring the natural biological evolution mechanism is the most popular of the artificial intelligence optimization methods. The exploration ability of these optimization techniques is generally very good. However, there is a disadvantage of these algorithms for the difficult optimization problem. The exploitation capability of the algorithms is usually not sufficient. In this study, the investigation results of the performance for an improved hybrid optimization method (HOM) to overcome their disadvantages occurred in optimization process of the difficult problems by combining ABC with DE algorithms are presented. While the improved algorithm keeps the exploration ability by retaining their standard updating strategy in the employed bees step, it enhances the exploitation skill by using powerful mutation and crossover strategies of the DE algorithm into onlooker bees step. The HOM having very high convergence speed and performance is a novel and robust optimization method based on meta-heuristic approach. In order to appraise the performance of the HOM in the testing step, a sequence of the classical benchmark function has been utilized and the performance results of the proposed method are compared with the performance of the standard DE, ABC algorithms. The reached numerical results in this study illustrated that the search performance of the presented HOM is better than the standard ABC, DE algorithms.

Various methods of identifying cellular automaton models and their state transition functions, in particular, have been proposed. Among these, a method that uses l 1 -norm optimization can be applied to multi-valued cellular automaton models. When solving this optimization problem, it is important to reduce the computational cost as the size of the problem increases. In this paper, we present a method for reducing the number of constraints in the optimization problem using the divide-and-conquer method.

In real-world applications, a majority of optimization problems belong to dynamic optimization, which involves optimization over time and whose underlying space is infinite-dimensional. In this paper, a novel dynamic optimization technique based on state transition algorithm (STA) is investigated to solve dynamic optimization problems (DOPs). Firstly, an infinite dynamic optimization problem is converted into a finite nonlinear programming problem (NLP) based on the control vector parameterizations (CVP), in which, instead of discretizing both state and control variables, only the control profiles are approximated by piecewise constant functions. Then, a stochastic global method is proposed to solve the resulting nonlinear programming problem based on state transition algorithm. By applying it in two well-known dynamic optimization instances, simulation results show that the proposed approach is able to obtain similar or even better performance indicators compared with other methods in terms of computational cost, solution accuracy and stability.

The Ethylene cracking is highly nonlinear, complex process with many constraints because of the complex running state of cracking furnace groups. For the solutions to multi-objective problems of yield, cost and benefits, this paper puts forward a biogeography-based multi-objective optimization algorithm with hybrid migration (BBMOHM). The hybrid migration strategy combines the self-adaptive migration with the simplex operator and the replication migration, which strengthens the convergence and the search ability, to avoid falling into local optimum. For the goal of utilizing the dominance information between individuals, this algorithm adopts the migration model of dominance degree to assist migration strategies improving the performance of the algorithm. Finally, experiments about the ZDT test function and the multi-objective optimization problem of the ethylene cracking process is verified the BBMOHM achieves good performance on convergence and distribution.

In this paper a new evolutionary algorithm, Leadership Based Optimization (LBO) will be presented. It will be tested on some standard benchmark functions to assess his optimization capability. To understand the potentiality of this algorithm, it has been compared with other Evolutionary Algorithms, such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). Finally the algorithm will be tested on the optimization of the array factor of a planar array antenna.

In optimization problems, nature inspired algorithms are able to generate near optimal solutions faster than other optimization algorithms. Based on nature intelligence, these algorithms are preferable especially when the function to be optimized is computationally intensive. In this paper it is proposed an image registration procedure based on the Artificial Bee Colony algorithm. First, its performances are compared to those of Particle Swarming and Cuckoo Search algorithms by optimizing some benchmark functions and then by determining the parameters of geometric transforms in image registration procedures.