Search

Mobile ad hoc networks (MANETs) can be useful for providing network coverage in harsh and adversarial environments where nodes may become randomly or systematically disabled in disaster recovery sites or border protection zones. For these networks to be reliable and scalable, robust topology control algorithms must be developed to guide the deployment and geometric configuration of mobile nodes without a priori knowledge of the terrain. This paper studies a fault tolerant differential evolution based topology control mechanism, called TCM-Y, that directs the movements of autonomous vehicles to efficiently and dynamically deploy themselves into a uniformly dispersed configuration. TCM-Y uses a Yao graph inspired fitness function to maintain a minimum desired number of connections for a node with its neighbors while uniformly dispersing autonomous vehicles over an unknown terrain. We evaluate the performance of mobile nodes running TCM-Y in hostile environments where nodes systematically or randomly experience hostility and become disabled. Experimental results obtained from our simulation software show that mobile nodes running TCM-Y perform well in the face of neighbor losses with respect to normalized area coverage, average distance traveled, and average connectivity, while limiting the number of network partitions that may occur as a result of node failures. [ABSTRACT FROM PUBLISHER]

This paper presents heuristic optimization approaches for solving the power dispatch problem of distributed generators in a smart grid for its economic operation. Along with uncertainties in the renewable energy sources namely solar and wind, random behavior of the next generation plug-in vehicles are also considered for solving the problem of power dispatch. Constraints for power balance, power outputs of conventional and distributed generators, grid-vehicle balance and different non-linear characteristics of the generator such as prohibited operating zones and ramp rate limits are included in the problem formulation. The problem is efficiently solved by differential evolution technique to illustrate the effect of the constraints on the operational cost for different test systems and also compared with other evolutionary algorithms. [ABSTRACT FROM PUBLISHER]

In this paper, there are two mutation strategies and two crossover strategies are involved for enhancing solution searching ability of Differential Evolution (DE). These strategies will be activated according to current solution searching status. The elitist mutation will guide particles toward to solution space around the elitist particles, and the random to real-rand mutation can prevent particles form fall into local optimum. Both elitist crossover and one-cut-point crossover can produce potential particles for deeply search the basin of solution space. In the experiments, 25 test functions of CEC 2005 are adopted for testing performance of proposed method and compare it with 4 DE variants. From the results, it can be observed that the proposed method exhibits better than related works for solving most test functions. [ABSTRACT FROM PUBLISHER]

In this article, the particle swarm optimization and differential evolution algorithms inspired by the intrinsic principles of quantum mechanics are presented. These quantum versions of meta-heuristic algorithms, namely quantum inspired particle swarm optimization and quantum inspired differential evolution for multi-level thresholding have been designed to find optimal thresholds of colour images at different levels by exploiting Kapur's entropy as an objective function. The average fitness and the standard deviation of the fitness values are reported. The test results over two test images at different levels certify the efficacy of the proposed methods with reference to precision, computational time, and durability over their classical counterparts. At last, a statistical measure, t-test has been performed among the four methods (two quantum methods and two classical methods) taking two methods in a single grasp to ascertain the supremacy of the results. [ABSTRACT FROM PUBLISHER]

Mention detection is an important component in anaphora resolution. In this paper we present our works on mention detection based on differential evolution (DE). The proposed technique consists of two steps, viz. feature selection and classifier ensemble. In the first step the algorithm performs automatic feature selection for two machine learning algorithms, namely Conditional Random Field (CRF) and Support Vector Machine (SVM). The first step yields a population of solutions, each of which represents a particular feature combination. We generate several models from these feature representations, and combine their decisions by a DE based ensemble technique in the second step of our algorithm. Experiments with a resource poor language show the recall, precision and F-measure values of 67.33%, 88.60% and 76.51%, respectively. [ABSTRACT FROM PUBLISHER]

This paper presents a pattern synthesis method based on two powerful tools of the Evolutionary computing i.e Differential Evolution, Invasive weed optimization and Firefly technique, to provide a significant side lobe reduction. The Antenna elements are embedded on a hemispherical array that leads to one of the geometry of a conformal shape. The array elements are excited with uniform amplitude excitation and progressively phased. The calculation of the element position is presented. The hemispherical array geometry is designed to meet multi objective criteria such as to achieve the desired side lobe reduction, null point detection and main beam enhancement. The exploitative changes of the Invasive Weed Optimization, Differential Evolution and Firefly algorithms have been applied for this array pattern synthesis. Simulation results of the conformal array design have been presented to illustrate the effectiveness and robustness of the Weed technique on the Differential Evolution and Firefly Algorithm. Empirical results clearly demonstrate the ability of IWO outperforming its competitors. By this method, arrays have the ability to produce the desired radiation pattern and could satisfy requirements for many applications. [ABSTRACT FROM PUBLISHER]

This paper proposes an evolutionary harmony search (EHS) algorithm for solving economic dispatch (ED) problem with various non-convex objective functions. The proposed EHS is a harmony search algorithm, in which each new created harmony from the harmony memory of harmony search algorithm is based on evolutional idea. The proposed EHS has been tested on 13-unit, 40-unit and 140-unit power systems with different fuel cost functions such as quadratic function, piecewise quadratic function, valve-point effect and prohibited operation zones. The results obtained from the proposed EHS are compared to those from improved coordinated aggregation-based PSO (ICA-PSO), differential evolution (DE), simulated annealing like particle swarm optimization (SA-PSO), improved particle swarm optimization (IPSO), conventional harmony search algorithm, improved harmony search algorithm, and other methods in the literature. The result comparison has indicated that the proposed method can obtain less total costs than the others. Therefore, the proposed EHS can be a favorable method for implementation in the non-convex economic dispatch problems. [ABSTRACT FROM PUBLISHER]

Path planning is an optimal problem how to plan optimal flight path of unmanned aerial vehicle (UAV) in the complex environment of war. A hybrid group search optimizer (GSO) with differential evolution (DE) is proposed to solve UAV path planning problem. Firstly, GSO is applied to update flight path of UAV by the search angle and the distance. DE is used to modify the feasible path of UAV within the search area by means of self-organization and self-regulatory in the evolutionary process. Then, UAV can find the safe path by connecting the chosen points of the coordinates while avoiding the threats area and costing minimum fuel. This approach can accelerate the global convergence speed. Finally, experimental results demonstrate that the proposed GSO-DE algorithm is effective and feasible in UAV path planning. [ABSTRACT FROM PUBLISHER]

The practical economic dispatch (ED) problems have many non-convex characteristics, which makes the searching of the global optimum difficult when using traditional mathematical methods. This paper presents a novel hybrid algorithm (HA) to the ED problems based on the particle swarm optimization (PSO) technique and differential evolution (DE) algorithm. Since the standard PSO has the adversity of premature convergence, the mutation and crossover operators of the DE, as well as the chaotic sequences, are considered to be integrated into the PSO to improve the global searching ability. Moreover, the dynamic penalty function is applied to deal with the constraints. The proposed method is tested on a 15-unit system considering the ramp rate limits, operating zones, and network losses. Also, the results are compared with those of other optimization methods. [ABSTRACT FROM PUBLISHER]

The kinematic design of a machine or mechanism is an important stage in the design methodology. A dexterous workspace for a robot manipulator is a outstanding characteristic that must be considered in the kinematic design, because it permit to reach the points of the workspace with different orientation. With this objective in mind, an optimization problem is formulated. A defined dexterous workspace is taken as an objective function. The design variables are lengths of the links and the angles of the links. The no collision between links are included as constraints. This paper describes the use of metaheuristic optimization techniques such as differential evolution (DE) to solve the above problem. A mechanism is incorporated in the DE algorithm in order to improve its behavior. This mechanism incorporates information of the base and the difference vectors of good trial vector, in an attempt to generate better individuals in the same direction. This mechanism guides the evolution of the population toward a better zone without sacrificing the search capabilities of the DE algorithm. The design of a robot with three revolute joints (3R robot) is considered as a numerical example. [ABSTRACT FROM PUBLISHER]

This paper presents the synthesis of Concentric Circular Antenna Array (CCAA) with hybrid optimization technique combining Differential Evolution with Biogeography Based Optimization (DE/BBO). Biogeography Based Optimization (BBO) is a novel optimization technique based on distribution of organisms over space and time. BBO uses its migration operator to share information between solutions. Differential Evolution (DE) is a robust optimization technique which uses it mutation operator to explore search space for better solutions. DE/BBO combines the mutation operator of DE with migration operator of BBO to improve the quality of solutions and convergence performance. This paper presents the thinning of nine-ring CCAA with centre element using DE/BBO. The main objective of thinning process is to reduce the Side Lobe Level (SLL) while keeping the First Null Beam-Width (FNBW) fixed. Results obtained are compared with Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and BBO techniques. Results show the superior performance of DE/BBO as compared to other techniques, both in terms of SLL and convergence speed. [ABSTRACT FROM PUBLISHER]

The potential and effectiveness of a hybrid intelligent PSS controller combining the advantages of both differential evolution (DE) and tabu search (TS) is assessed in this paper. The controller is incorporated in a single machine infinite bus (SMIB) system with a synchronous generator. An analysis is also carried on the quality of results if various parameters of the differential evolution like crossover, mutation and population size of the algorithm is varied. At the end a comparison is made between the controller based totally on differential evolution and the one designed with the hybrid technique. [ABSTRACT FROM PUBLISHER]

The Flexible AC Transmission System (FACTS) devices are low-environmental impact, cost effective and long term proven solutions to transmission line difficulties. Interline Power Flow Controller (IPFC) is a novel FACTS device with an exceptional ability to control power flow among multiline transmission system unlike the other devices in which they are used to manage and compensate single transmission line. This paper attempts to reduce power deviation in a particular transmission line utilizing IPFC by adjusting angles and magnitudes of the series voltage source converters representing IPFC. To the best of our knowledge, there has been no clear study till now proposes employing IPFC to this application. A case-study is presented at the end of this investigation with the aim of simulating the ability of this device to maintain the power flow in a transmission line unchanged. Differential Evolution is employed to optimize the control parameters. The results show the effectiveness of the proposed control scheme in controlling the power flow and diminish its fluctuations. [ABSTRACT FROM PUBLISHER]

Circular Extreme Learning Machine (C-ELM) is an extension of Extreme Learning Machine. Its power is mapping both linear and circular separation boundaries. However, C-ELM uses the random determination of the input weights and hidden biases, which may lead to local optimal. This paper proposes a hybrid learning algorithms based on the C-ELM and the Differential Evolution (DE) to select appropriate weights and hidden biases. It called Evolutionary circular extreme learning machine (EC-ELM). From experimental results show EC-ELM can slightly improve C-ELM and also reduce the number of nodes network. [ABSTRACT FROM PUBLISHER]

This paper presents the results of a comprehensive study on the efficiency of planar parallel mechanisms, considering their kinetostatic performance and also, their workspace. This aim is approached upon proceeding single- and multi-objective optimization procedures. Kinetostatic performance of ten different architectures is analyzed, by resorting to a recent index, kinematic sensitivity. The mathematical framework adopted for the optimization problem is based on genetic algorithm. The results of multi-objective optimization are based on the sets of Pareto points, offering the most reliable decisions to reconciliate between some conflicting objectives. To this end, the ten mechanisms are sorted into two sets based on their type of actuator, namely, five with prismatic actuators and the other ones with revolute actuators. Finally, a comparison between performances of these mechanisms, according to the obtained results, is carried out. [ABSTRACT FROM PUBLISHER]

The conventional K-harmonic means is tend to be trapped by local optima. To resolve this problem, a novel K-harmonic means clustering algorithm using enhanced differential evolution technique is proposed. This algorithm improves the global search ability by applying Laplace mutation operator and logarithmically crossover probability operator. Numerical experiments show that this algorithm overcomes the disadvantages of the K-harmonic means, and improves the global search ability. [ABSTRACT FROM PUBLISHER]

Over the past few decades, differential evolution (DE) has emerged as a versatile algorithm for solving a wide range of global optimization problems arising in various fields. The present study discusses a novel application of DE for thermal optimization of a single inlet T-junction. It is a common but complex problem arising in the field of thermal engineering and can be formulated as a global optimization problem subject to constraints. The current study shows the efficiency of DE in dealing with such problems. [ABSTRACT FROM PUBLISHER]

Parameter estimation is an important problem in engineering and scientific research. To some systems,there is no way can be found to get the real values of their parameters. For some advantages including self-organize, developmental search, self-fitness etc, some intelligent algorithms have been given to solve these problems, such as GA, PSO, ACO etcs. Differntial Evolution is an effective tool for solving global optimization and has been used in some fields. This paper proposes a novel DE algorithm (DPDE) with double populations to solve parameter estimation problem. There are two populations in DPDE. Each population has its own evolutionary model and they finish evolution by different evolutionary model independently. Two Populations implement co-evolution based on local information transfer and share between populations. Result of experiments shows that DPDE is an effective and feasible method in solving parameter estimation problem. [ABSTRACT FROM PUBLISHER]

Multi-objective synthesis for microwave components (e.g. integrated transformer, antenna) is in high demand. Since the embedded electromagnetic (EM) simulations make these tasks very computationally expensive when using traditional multi-objective synthesis methods, efficiency improvement is very important. However, this research is almost blank. In this paper, a new method, called Gaussian Process assisted multi-objective optimization with generation control (GPMOOG), is proposed. GPMOOG uses MOEA/D-DE as the multi-objective optimizer, and a Gaussian Process surrogate model is constructed ON-LINE to predict the results of expensive EM simulations. To avoid false optima for the on-line surrogate model assisted evolutionary computation, a generation control method is used. GPMOOG is demonstrated by a 60GHz integrated transformer, a 1.6GHz antenna and mathematical benchmark problems. Experiments show that compared to directly using a multi-objective evolutionary algorithm in combination with an EM simulator, which is the best known method in terms of solution quality, comparable results can be obtained by GPMOOG, but at about 1/3-1/4 of the computational effort. [ABSTRACT FROM AUTHOR]

Timetabling problem is a kind of combinatorial optimizations. However, it is very difficult to be solved the timetabling problem from the enormous combination total number and the complexity of the limitation condition. In this study, we develop the timetabling algorithm for Osaka International University (OIU). However, it is difficult to design and solving the problem with many constraints. So, in this paper, we design the reduced model and propose a solution method by Differential Evolution (DE). This study is a new trial such as to apply the DE to timetabling problem. Furthermore, we propose one of generating method of initial population. [ABSTRACT FROM PUBLISHER]

Dynamic economic emission dispatch (DEED) problem is used to determine the optimal generation schedule of online generating units by minimizing both fuel cost and emission simultaneously under load demand constraint, ramp rate constraint and other constraints. DEED with the consideration of valve-point effects is a complicated non-linear constrained multi-objective optimization problem with non-smooth and non-convex characteristics. The first purpose of this paper is to solve the DEED problem using a hybrid method which combines differential evolution (DE) and sequential quadratic programming (SQP). DE is used as a global optimizer and SQP is used as a fine tuning to determine the optimal solution at the final. A five-unit system with non-smooth fuel cost function has been taken to illustrate the effectiveness of the proposed method compared with other methods. The second purpose of this paper is to modify the DEED problem such that its optimal solution can be periodically implemented. [ABSTRACT FROM PUBLISHER]

Wireless sensor networks have many applicable areas including environmental monitoring, medical care, and military surveillance. Energy awareness is an essential issue of data dissemination while designing routing protocols on these networks. Implementations of routing protocols based on optimization methods present energy efficient solutions. In this paper, we utilized the feature of simple and fast search mechanism of Differential Evolution (DE) algorithm to optimize energy consumption of clustering based routing protocols. A WSN routing implementation of DE is proposed and its performance is compared with the performance of other DE implementations. [ABSTRACT FROM PUBLISHER]

In this study two nature inspired optimization methods, namely particle swarm optimization and differential evolution, are proposed for the distributed detection problem. Performances of the methods are investigated through simulations where the detection capability of a parallel distributed topology consisting of N local sensors and a central processor is optimized. Simulation results have revealed that the performance of the differential evolution, in terms of convergence, is considerably better than that of the particle swarm optimization. [ABSTRACT FROM PUBLISHER]

Landing footprint of an entry vehicle provides critical information for mission planning. Conventional methods calculate it through solving a family of multi-constraints optimal control problems. It is difficult to solve. To generate the boundary of it, we proposed a new scheme based on differential evolution(DE). Utilizing DE's parallelism, each run generates one side instead of one point of the boundary. To get the full boundary, just need run the algorithm two times. The algorithm's merits include accurate, fast, not relying on simplification of the system, all control variables are under consideration and facilitating parallel programming. [ABSTRACT FROM PUBLISHER]

This work addresses the reduction of power consumption of the AODV routing protocol in vehicular networks as an optimization problem. Nowadays, network designers focus on energy-aware communication protocols, specially to deploy wireless networks. Here, we introduce an automatic method to search for energy-efficient AODV configurations by using an evolutionary algorithm and parallel Monte-Carlo simulations to improve the accuracy of the evaluation of tentative solutions. The experimental results demonstrate that significant power consumption improvements over the standard configuration can be attained, with no noteworthy loss in the quality of service. [ABSTRACT FROM PUBLISHER]

In this paper, we proposed an evolutionary algorithm for joint estimation of amplitude and DOA of signals impinging from far field on a uniform linear array. We used Differential Evolution algorithm with Mean Square Error as a fitness evaluation function. This fitness function defines an error between actual and estimated signal and is derived from Maximum Likelihood Principle. It does not require any permutations to link it with the angles estimated in the previous snapshot. It requires only single snapshot to converge and produce fairly good results even in the presence of low Signal to Noise Ratio. The usefulness and competence of proposed algorithm is tested on the basis of large number of Monte-Carlo simulations and its statistical analysis. [ABSTRACT FROM PUBLISHER]

This paper presents an improved differential evolution based on Gaussian disturbance for multi-objective optimization. Differential evolution algorithm is often trapped in local optima and converges slowly. In this paper, Gaussian disturbance is employed to increase the variety of the individual to improve its performance. External archive is employed to reserve the non-dominated solutions and crowding-distance calculation is introduced in order that the solutions in the neighborhood of the solutions with smallest and largest function values or locating in a lesser crowded region will have higher probability to be preserved. The improved differential evolution algorithm is tested on several classical multi-objective optimization benchmark functions. The simulation results show that the improved algorithm can obtain the better solutions and they are widely spread on the true Pareto optimal front. [ABSTRACT FROM PUBLISHER]

Design of non-uniform circular antenna arrays is one of the important optimization problems in electromagnetic domain. While designing a non-uniform circular array the goal of the designer is to achieve minimum side lobe levels with maximum directivity. In contrast to the single-objective methods that attempt to minimize a weighted sum of the four objectives considered here, in this article we consider these as four distinct objectives that are to be optimized simultaneously in a multi-objective (MO) framework using one of the best known Multi-Objective Evolutionary Algorithms (MOEAs) called NSGA-II. This MO approach provides greater flexibility in design by producing a set of final solutions with different trade-offs among the four objective from which the designer can choose one as per requirements. To the best of our knowledge, other than the single objective approaches, no MOEA has been applied to design a non-uniform circular array before. Simulations have been conducted to show that the best compromise solution obtained by NSGA-II is far better than the best results achieved by the single objective approaches by using the differential evolution (DE) algorithm and the Particle Swarm Optimization (PSO) algorithm. [ABSTRACT FROM PUBLISHER]

A hybridized optimization technique called particle swarm optimization with differentially perturbed velocity, is employed to two different types of thinned antenna array design problems and the performance is compared with that of particle swarm optimization and differential evolution. Two objectives to be reduced are the side lobe level in principal planes and the number of active elements. In the first case, thinning of a linear antenna array of 100 isotropic elements is considered. In the second case, thinning of a planar antenna array of 10×10 elements is considered. [ABSTRACT FROM PUBLISHER]

Differential Evolution (DE) is a very simple but powerful stochastic global optimizer. Since its inception, it has proved to be a very efficient and robust technique for function optimization and has been used to solve problems in many scientific and engineering fields. The robustness of DE has drawn the attention of many researchers all over the world resulting in a lot of variants of the basic algorithm with improved performance. This paper presents the performance evaluation of four DE variants on non-convex Economic Load Dispatch (ELD) problem. It discusses the various adaptations made on a classical DE algorithm to improve exploration and exploitation of a search space. ELD is a powerful tool in power system optimization problems. The problem becomes complex due to the presence of throttle losses in a multi valve steam generating unit. Throttle losses cause valve point loading effect which makes non convex cost characteristics. Three different test cases are studied and implemented using different DE variants and their performance with varying dimensionality is studied. [ABSTRACT FROM PUBLISHER]

The problems related to the classical approaches used in electro magnetics determined the search for alternative methods that can be faster, more powerful and easy to use. Differential evolution (DE) is such a tool, its efficiency and robustness laying in its simple structure which can be easily modified. In this paper, the characteristics of DE in the context of evolutionary algorithms, along with some hybridization approaches are given. Also, the latest applications related to the use of this algorithm for various types of optimization from the electromagnetic field are presented and discussed. [ABSTRACT FROM PUBLISHER]

We present a differential evolution based topology control mechanism, called TCM-DE, for the decision making process of evolutionary and autonomous systems that adaptively reconfigures spatial configuration in MANETs. We introduce quantitative metrics to evaluate performance of our TCM-DE with respect to uniform distribution, total terrain covered by communication areas of all nodes, and distance traveled by each node until a desired network topology is reached. Voronoi tessellation for configurations of mobile nodes is used to create two uniformity metrics. Physical relocation of mobile nodes is a power consuming task. Therefore, minimizing the average distance each node travels (ADT) until the network reaches a desired distribution is an important indicator for the performance of MANET nodes. Another important performance metric is the network area coverage (NAC) achieved by all nodes. NAC is used to measure the speed of network convergence and the efficiency of node deployment. Experimental results from our simulation software shows that TCM-DE performs well with respect to NAC, ADT, and Voronoi-based uniformity evaluation techniques. [ABSTRACT FROM PUBLISHER]

Tuning Electric Power System (EPS) controllers for small-signal enhancement is a very hard task. Nowadays, the most used techniques for this task involves a trial-and-error process. These techniques consider a set of operating conditions of the system and multiple modes of oscillation, making it very difficult for a human designer to keep track of so many variables and parameters. This results in a sequential approach, where one controller is tuned, placed in the system, and then the designer proceeds to tune the next controller. This paper proposes a method for simultaneously tuning controllers considering several operation conditions at once. The tuning is done in a coordinated way, avoiding interactions among the controllers. A differential evolution technique is used to perform the automatic tuning method proposed in this work. One important point to remark is that this algorithm can tune different types of controllers at once (PSSs and PODs in this case). Results show satisfactory performance of the tuned controllers, showing the effectiveness of the proposed technique. Furthermore, a significant productivity gain can be achieved if the engineer in charge of this design only supervises the automatic process, instead of performing all the calculations himself/herself. This is one of the main contributions of the present paper. [ABSTRACT FROM PUBLISHER]

Differential Evolution (DE) is a powerful stochastic algorithm for real-coded optimization. However, DE has a problem as well as other traditional stochastic optimization algorithms: difficult to optimize search spaces that are little globally convex. Thus, it is difficult for DE and traditional algorithms to optimize some practical problems where globally convex cannot be assumed. As one of the solution for this problem, we propose Differential Evolution on Scattered Parents (DE-SP) that re-selects the individuals on each dimension when the mutant individual is calculated and some children individuals' candidates unconditionally become the children individuals. We have implemented 2 types of optimization experiment to verify the performance of DE-SP: Noisy Function 1 (NF1), that is a benchmark problem with little globally convex, Noisy Function 2 (NF2), that is the one with globally convex. And we compared the performance of DE-SP with those of 15 types of algorithms. Thereby, it is confirmed that DE-SP was the most stable algorithm to optimize little globally convex spaces among 15 comparative algorithms from the experiment of optimizing NF1. And it is comfirmed that DE-SP can optimize globally convex spaces as well as DE from the experiment of optimizing NF2. [ABSTRACT FROM PUBLISHER]

Due to real-world optimization problems become increasingly complex. Algorithms are with higher efficiency and higher solution searching ability for finding global optimal solution in reasonable computing time is always needed. Thus, in this paper, an improved DE is proposed for solving large scale global optimization. The proposed method is incorporated with the population manager to eliminate redundant particles or to hire new ones or to maintain population size according to the solution searching status to make the process more efficient. Besides, a local search strategy is also involved to enhance population's solution search ability. Experiments were conducted on ten CEC 2012 test functions to present performance of the proposed method. The proposed method exhibits better performance than other three related works in solving most test functions. [ABSTRACT FROM PUBLISHER]

In this paper, a new optimization method based on a combination of Differential Evolution (DE) and Evolution Strategy (ES), which belong to both Meta-Heuristics and Evolutionary Computation, are developed as a fast approximation optimization method. A weak point of DE is weak local search ability and considerable computation time for obtaining a good approximate solution. The proposed method aims at compensation for the weak points of DE by connecting ES with strong local search capacity. The effectiveness of the proposed method is confirmed through the numerical experiments. [ABSTRACT FROM PUBLISHER]

Training artificial neural networks (ANNs) is a complex task of great importance in problems of supervised learning. Evolutionary algorithms (EAs) are widely used as global searching techniques for optimization in scientific and engineering problems, and these approaches have been introduced to ANNs to perform various tasks, such as connection weight training and architecture design. Recently, a novel optimization algorithm called Group Search Optimizer (GSO) was introduced, which is inspired by animal searching behaviour and group living theory. In this paper, we present two new hybrid GSO approaches, one based on opposite populations and the other based on opposite populations and a modified Differential Evolution (DE) strategy. We also applied the Weight Decay (WD) heuristic to enhance the generalization power of networks. Experimental results show that the proposed GSO approaches are able to achieve better generalization performance than Levenberg-Marquardt (LM), Opposite Differential Evolution (ODE) and traditional GSO in real benchmark datasets. [ABSTRACT FROM PUBLISHER]

Traditional differential evolution (DE) mutation operators explore the search space with no considering the information about the search directions, which results in a purely stochastic behavior. This paper presents a DE variant with self-navigation ability for multi-objective optimization (MODE/SN). It maintains a pool of well designed DE mutation operators with distinct search behaviors and applies them in an adaptive way according to the feedback information from the optimization process. Moreover, we deploy the neural network, which is trained by the extreme learning machine, for mapping an artificially generated solution in the objective space back into the decision space. Empirical results demonstrate that MODE/SN outperforms several state-of-the-art algorithms on a set of benchmark problems with variable linkages. [ABSTRACT FROM PUBLISHER]

This paper presents a differential evolution (DE) algorithm to solve the optimum switching angles problem in the harmonic elimination pulse width modulation (HE-PWM) inverter. Reformulated equations for two-level schemes are proposed to satisfy the DE algorithm. Optimum solutions are found for odd value of switching angles (odd N) case within all range of modulation index M. The DE algorithm needs a relative small amount of generations to reach accurate solutions with appropriate control parameters. The comparison of some common variants in the DE shows that all variants may generate the same precision results by using their best specific crossover rate and mutation factor values. MATLAB/Simulink simulation results confirm that the solutions obtained by the proposed technique can be applied to eliminate the entire undesired specific harmonics as expected by the proposed formulation. [ABSTRACT FROM PUBLISHER]

In this paper, an on-line training PIDNN controller using an improved DEPSO algorithm for trajectory tracking of the ball and plate system is proposed. Since the ball and plate system is a typical under-actuated system with inherent nonlinearity and coupling between its parameters, the accurate mathematical model is difficult to be derived, so that a lot of nonlinear control and intelligent control methods are used for the ball and plate system control. The control method using a PID neural network is one of the intelligent methods. In this paper, an improved particle swarm optimization method based on differential evolution algorithm (DEPSO) is used to train the weighting factors of multilayered forward neural network. This PIDNN control method based on DEPSO algorithm can overcome the shortcoming of the BP algorithm which is easy to get into local minimum. At the same time, the simulation results of tracking control for ball and plate system show that the proposed PIDNN controller has simple structure, nice static and dynamic characteristics. [ABSTRACT FROM PUBLISHER]

An essential component of an autonomous robotic swarm system is the ability to form an accurate map of its environment. In this paper, we consider the fuzzy approach to map generation using a swarm of robots fitted with multi-sonar sensors. While adopting an approach to map making similar to that in [1] we have reduced and simplified the number of parameters employed without significant loss of accuracy. These parameters are first optimized using a synthetic image of the target environment using differential evolution. A parallel implementation of the differential evolution optimizer is developed in MPI on a 200 core cluster. The algorithm has been tested on real-time data through experiments with three Parallax Stingray robots, each equipped with three sonars, with good results. [ABSTRACT FROM PUBLISHER]

Economic power dispatch (EPD) is an important tool for optimal operation and planning of modern power systems. To solve effectively EPD problems, most of the conventional calculus methods rely on the assumption that the fuel cost characteristic of a generating unit is a continuous and convex function, resulting in inaccurate dispatch. This paper presents the design and application of an enhanced differential evolution (EDE) algorithm for the solution of the economic power dispatch problem, where the nonsmooth and nonconvex characteristics of the generators, such as valve-point effects and multi-fuel options of the practical generator operation are considered. The 10-generator benchmark test system with valve-point loading effects and multi-fuel options was used for testing and validation purposes. The results obtained demonstrate the effectiveness of the proposed method for solving the nonsmooth economic dispatch problem. [ABSTRACT FROM PUBLISHER]

This paper is presented with a hybrid technique named as Efficient Hybrid Simulated Annealing (EHSA) algorithm. The new algorithm is implemented to the non-convex economic load dispatch (ELD) problem so as to minimize the total generation cost when considering the linear and non linear constraints. In EHSA algorithm, the velocity of the particle in Particle Swarm Optimization (PSO) is improved by the mutation operator of Differential Evolution (DE) thus a global solution is obtained. This solution is set as an initial solution to Simulated Annealing (SA) to reach an optimal solution. The proposed methodology is implemented for economic dispatch problem by considering the constraints such as generator constraints, transmission loss, ramp rate limits and prohibited operating zones. This method is experimented on a 3 units test system to get an optimized result. By using this optimization algorithm we obtain a minimum operating cost with minimum transmission loss and better convergence speed. [ABSTRACT FROM PUBLISHER]

In this paper, we propose an automatic image enhancement tool for smart phone by using interactive differential evolution (IDE). From a remarkable progress of the camera sensor in mobile devices, people take pictures with their mobile phone instead of a digital camera. However, as they are not satisfied with their images in spite of the progress, they still want to edit their images by using mobile applications, which are usually complex and cause user fatigue, especially for beginners. To reduce it and make a simple interface, we exploit IDE, which is a kind of interactive evolutionary computation. Let the user IDE to evaluate the individuals. Because of the small parameters of the differential evolution (DE), we could make the tool simply and overcome the user fatigue. DE is also an efficient and fast evolutionary algorithm which uses the difference of the vectors. Subjective test shows the usefulness of the tool. [ABSTRACT FROM PUBLISHER]

Many real world optimization problems are dynamic optimization problems (DOPs) whose optima change over time. In this paper, we propose new variants of differential evolution (DE) to solve DOPs. A hybrid method that combines population core based multi-population strategy and prediction strategy and new local search scheme is introduced into DE to enhance its performance for solving DOPs. The population core based multi-population strategy is useful to maintain the diversity of population by using the multi-population and population core concept. The prediction strategy is useful to rapidly adapt to the dynamic environment by using the prediction area. The local search scheme is useful to improve the searching accuracy by suing the new chaotic local search method. Experimental results on the moving peaks benchmark show that the proposed schemes enhance the performance of DE in the dynamic environments. [ABSTRACT FROM PUBLISHER]

Differential Evolution (DE) is an evolutionary algorithm. DE has been successfully applied to optimization problems including non-linear, non-differentiable, non-convex and multimodal functions. There are several mutation strategies such as the best and the rand strategy in DE. It is known that the best strategy is suitable for unimodal problems and the rand strategy is suitable for multimodal problems. However, the landscape of a problem to be optimized is often unknown and the landscape is changing dynamically while the search process proceeds. In this study, we propose a new and simple method that detects the modality of landscape being searched: unimodal or not unimodal. In the method, some points on the line connecting the centroid of search points and the best search point are sampled. When the objective values of the sampled points are changed decreasingly and then increasingly, it is thought that one valley exists. If there exists only one valley, the landscape is unimodal and a greedy strategy like the best strategy is adopted. Otherwise, the rand strategy is adopted. Also, the sampled points realize global search in the region spanned by all search points and realize local search near the best search point. The effect of the proposed method is shown by solving some benchmark problems. [ABSTRACT FROM PUBLISHER]

An Equity Market Neutral Portfolio (EMNP) is an assortment of long and short positions that ensures a riskless portfolio in terms of its exposure to the relevant market benchmark. While a naïve formulation of the EMNP optimization problem can be easily solved using linear programming techniques, the inclusion of the Risk Budget constraint on the high risk assets, together with the other EMNP specific constraints of zero net market exposure, close-to-zero portfolio beta and zero financial leveraging, besides the bounding constraints imposed on the long-short positions and high risk assets, can turn the problem difficult for direct solving using traditional methods. This work aims to solve such a complex constrained EMNP optimization problem using a meta-heuristic method viz., Differential Evolution (rand/1/bin) with Hall of Fame (DE HOF). The DE HOF exploits a penalty function strategy and employs weight standardization procedures to ensure faster convergence and an efficient tackling of complex constraints to yield optimal portfolios within realistic time. Experimental studies which include a rigorous out of sample performance analysis have been undertaken on the Bombay Stock Exchange data set (BSE 200: March 1999–March 2009) which included both upturns and downturns in the global markets. [ABSTRACT FROM PUBLISHER]

In recent years there has been a growing trend in the application of Memetic Algorithms for solving numerical optimization problems. They are population based search heuristics that integrate the benefits of natural and cultural evolution. In this paper, we propose an Adaptive Memetic Algorithm, named LA-DE which employs a competitive variant of Differential Evolution for global search and Learning Automata as the local search technique. During evolution Stochastic Automata Learning helps to balance the exploration and exploitation capabilities of DE resulting in local refinement. The proposed algorithm has been evaluated on a test-suite of 25 benchmark functions provided by CEC 2005 special session on real parameter optimization. Experimental results indicate that LA-DE outperforms several existing DE variants in terms of solution quality. [ABSTRACT FROM PUBLISHER]

An Elite Multi-Group Differential Evolution algorithm for unconstrained single objective optimization is proposed. In the novel algorithm, the population is divided into sub-groups with different parameters setting to balance the global and local search ability. The good information collected in the search process is exchanged among groups. Experiments are conducted on seven commonly used benchmark functions and two new constructed harder test functions which are useful to test the local search ability of the algorithms and the proposed algorithm shows its effectiveness and efficiency. [ABSTRACT FROM PUBLISHER]