Showing papers on "Crossover published in 2014"

Differential evolution based on covariance matrix learning and bimodal distribution parameter setting

Abstract: Point out the drawbacks of the crossover operator and the parameter settings of differential evolution (DE).Propose a novel DE variant based on covariance matrix learning and bimodal distribution parameter setting, named CoBiDE.Verify the effectiveness of CoBiDE by many experiments. Differential evolution (DE) is an efficient and robust evolutionary algorithm, which has been widely applied to solve global optimization problems. As we know, crossover operator plays a very important role on the performance of DE. However, the commonly used crossover operators of DE are dependent mainly on the coordinate system and are not rotation-invariant processes. In this paper, covariance matrix learning is presented to establish an appropriate coordinate system for the crossover operator. By doing this, the dependence of DE on the coordinate system has been relieved to a certain extent, and the capability of DE to solve problems with high variable correlation has been enhanced. Moreover, bimodal distribution parameter setting is proposed for the control parameters of the mutation and crossover operators in this paper, with the aim of balancing the exploration and exploitation abilities of DE. By incorporating the covariance matrix learning and the bimodal distribution parameter setting into DE, this paper presents a novel DE variant, called CoBiDE. CoBiDE has been tested on 25 benchmark test functions, as well as a variety of real-world optimization problems taken from diverse fields including radar system, power systems, hydrothermal scheduling, spacecraft trajectory optimization, etc. The experimental results demonstrate the effectiveness of CoBiDE for global numerical and engineering optimization. Compared with other DE variants and other state-of-the-art evolutionary algorithms, CoBiDE shows overall better performance.

Swarm intelligence based algorithms: a critical analysis

Abstract: Many optimization algorithms have been developed by drawing inspiration from swarm intelligence (SI). These SI-based algorithms can have some advantages over traditional algorithms. In this paper, we carry out a critical analysis of these SI-based algorithms by analyzing their ways to mimic evolutionary operators. We also analyze the ways of achieving exploration and exploitation in algorithms by using mutation, crossover and selection. In addition, we also look at algorithms using dynamic systems, self-organization and Markov chain framework. Finally, we provide some discussions and topics for further research.

A novel fruit fly optimization algorithm for the semiconductor final testing scheduling problem

Abstract: In this paper, a novel fruit fly optimization algorithm (nFOA) is proposed to solve the semiconductor final testing scheduling problem (SFTSP). First, a new encoding scheme is presented to represent solutions reasonably, and a new decoding scheme is presented to map solutions to feasible schedules. Second, it uses multiple fruit fly groups during the evolution process to enhance the parallel search ability of the FOA. According to the characteristics of the SFTSP, a smell-based search operator and a vision-based search operator are well designed for the groups to stress exploitation. Third, to simulate the information communication behavior among fruit flies, a cooperative search process is developed to stress exploration. The cooperative search process includes a modified improved precedence operation crossover (IPOX) and a modified multipoint preservative crossover (MPX) based on two popular structures of the flexible job shop scheduling. Moreover, the influence of the parameter setting is investigated by using Taguchi method of design-of-experiment (DOE), and suitable values are determined for key parameters. Finally, computational tests results with some benchmark instances and the comparisons to some existing algorithms are provided, which demonstrate the effectiveness and the efficiency of the nFOA in solving the SFTSP.

An effective hybrid immune algorithm for solving the distributed permutation flow-shop scheduling problem

Abstract: In this article, an effective hybrid immune algorithm (HIA) is presented to solve the distributed permutation flow-shop scheduling problem (DPFSP). First, a decoding method is proposed to transfer a job permutation sequence to a feasible schedule considering both factory dispatching and job sequencing. Secondly, a local search with four search operators is presented based on the characteristics of the problem. Thirdly, a special crossover operator is designed for the DPFSP, and mutation and vaccination operators are also applied within the framework of the HIA to perform an immune search. The influence of parameter setting on the HIA is investigated based on the Taguchi method of design of experiment. Extensive numerical testing results based on 420 small-sized instances and 720 large-sized instances are provided. The effectiveness of the HIA is demonstrated by comparison with some existing heuristic algorithms and the variable neighbourhood descent methods. New best known solutions are obtained by the HI...

Crisscross optimization algorithm and its application

Abstract: How to improve the global search ability without significantly impairing the convergence speed is still a big challenge for most of the meta-heuristic optimization algorithms. In this paper, a concept for the optimization of continuous nonlinear functions applying crisscross optimization algorithm is introduced. The crisscross optimization algorithm is a new search algorithm inspired by Confucian doctrine of gold mean and the crossover operation in genetic algorithm, which has distinct advantages in solution accuracy as well as convergence rate compared to other complex optimization algorithms. The procedures and related concepts of the proposed algorithm are presented. On this basis, we discuss the behavior of the main search operators such as horizontal crossover and vertical crossover. It is just because of the perfect combination of both, leading to a magical effect on improving the convergence speed and solution accuracy when addressing complex optimization problems. Twelve benchmark functions, including unimodal, multimodal, shifted and rotated functions, are used to test the feasibility and efficiency of the proposed algorithm. The experimental results show that the crisscross optimization algorithm has an excellent performance on most of the test functions, compared to other heuristic algorithms. At the end, the crisscross optimization algorithm is successfully applied to the optimization of a large-scale economic dispatch problem in electric power system. It is concluded that the crisscross optimization algorithm is not only robust in solving continuous nonlinear functions, but also suitable for addressing the complex real-world engineering optimization problems.

A memetic algorithm for enhancing the robustness of scale-free networks against malicious attacks

Abstract: The robustness of the infrastructure of various real-life systems, which can be represented by networks and manifests the scale-free property, is of great importance. Thus, in this paper, a new memetic algorithm (MA), which is a type of effective optimization method combining both global and local searches, is proposed to enhance the robustness of scale-free (RSF) networks against malicious attacks (MA) without changing the degree distribution. The proposed algorithm is abbreviated as MA–RSF MA . Especially, with the intrinsic properties of the problem of optimizing network structure in mind, a crossover operator which can perform global search and a local search operator are designed. In the experiments, both synthetic scale-free networks and real-world networks, like the EU power grid network and the real Internet at the level of autonomous system (AS), are used. MA–RSF MA shows a strong ability in searching for the most robust network structure, and clearly outperforms existing local search methods.

Repairing the crossover rate in adaptive differential evolution

Abstract: Differential evolution (DE) is a simple yet powerful evolutionary algorithm (EA) for global numerical optimization. However, its performance is significantly influenced by its parameters. Parameter adaptation has been proven to be an efficient way for the enhancement of the performance of the DE algorithm. Based on the analysis of the behavior of the crossover in DE, we find that the trial vector is directly related to its binary string, but not directly related to the crossover rate. Based on this inspiration, in this paper, we propose a crossover rate repair technique for the adaptive DE algorithms that are based on successful parameters. The crossover rate in DE is repaired by its corresponding binary string, i.e. by using the average number of components taken from the mutant. The average value of the binary string is used to replace the original crossover rate. To verify the effectiveness of the proposed technique, it is combined with an adaptive DE variant, JADE, which is a highly competitive DE variant. Experiments have been conducted on 25 functions presented in CEC-2005 competition. The results indicate that our proposed crossover rate technique is able to enhance the performance of JADE. In addition, compared with other DE variants and state-of-the-art EAs, the improved JADE method obtains better, or at least comparable, results in terms of the quality of final solutions and the convergence rate.

Multiobjective bilevel optimization for production-distribution planning problems using hybrid genetic algorithm

Abstract: Currently, the production-distribution planning problems are usually modeled as single-objective bilevel programming problems. However, many real world production-distribution planning problems involve several objectives simultaneously for decision makers at two different levels when the production and the distribution processes are considered. In this paper, a multiobjective bilevel production-distribution planning model with equilibrium between supply and demand is set up, in which the distribution company is the leader who controls the distributing process with the aims to minimize its overall cost, and the manufacturer is the follower who controls the production process with the aims to minimize its overall cost and storage cost. So in the proposed model, the leader has one objective and the follower has two objectives. To solve the model efficaciously, the lower level problem follower's problem is transformed into an equivalent single-objective programming problem by a weighted aggregation method. As a result, the multiobjective bilevel optimization problem is transformed into a single-objective bilevel optimization problem. To solve the transformed problem efficiently, a uniform design scheme is applied to generate some representative weight vectors and initial population. Thereafter, a uniform design based crossover and exponential mutation are designed, and a local search scheme is applied. Based on all these, a hybrid genetic algorithm is proposed. Finally, two real word problems are solved successfully by the proposed algorithm, and the effectiveness and efficiency of the proposed algorithm are also tested by other test problems.

Selecting the most efficient genetic algorithm sets in solving unconstrained building optimization problem

Abstract: Effective optimization of unconstrained building optimization problem involves coupling a building energy simulation program with an optimization evolutionary algorithm such as the genetic algorithm (GA). The aim of this paper is to find the most appropriate GA set that obtains the optimum, or near optimum, solutions in a reasonable computational time (less numbers of simulations). Twelve control parameter sets of binary encoded GA are tested to solve unconstrained building optimization problems that are coupled with EnergyPlus simulation program. The results show that population size is the most significant control parameter and that the crossover probability and mutation rate have insignificant effects on the GA performance. In general, a binary encoded GA with small population sizes can be used to solve unconstrained building optimization problems by around 250 building simulation calls. In particular, the smaller population size of about 5 individuals helps reach the optimum solution faster than larger population sizes.

Original Article/Research Selecting the most efficient genetic algorithm sets in solving unconstrained building optimization problem

Abstract: Effective optimization of unconstrained building optimization problem involves coupling a building energy simulation program withan optimization evolutionary algorithm such as the genetic algorithm (GA). The aim of this paper is to find the most appropriate GA setthat obtains the optimum, or near optimum, solutions in a reasonable computational time (less numbers of simulations). Twelve controlparameter sets of binary encoded GA are tested to solve unconstrained building optimization problems that are coupled with EnergyPlussimulation program.The results show that population size is the most significant control parameter and that the crossover probability and mutation ratehave insignificant effects on the GA performance. In general, a binary encoded GA with small population sizes can be used to solveunconstrained building optimization problems by around 250 building simulation calls. In particular, the smaller population size ofabout 5 individuals helps reach the optimum solution faster than larger population sizes. 2014 The Gulf Organisation for Research and Development. Production and hosting by Elsevier B.V. All rights reserved.

The Crossover Region Between Long-Range and Short-Range Interactions for the Critical Exponents

Abstract: It is well know that systems with an interaction decaying as a power of the distance may have critical exponents that are different from those of short-range systems. The boundary between long-range and short-range is known, however the behavior in the crossover region is not well understood. In this paper we propose a general form for the crossover function and we compute it in a particular limit. We compare our predictions with the results of numerical simulations for two-dimensional long-range percolation.

An improved NSGA-II algorithm for multi-objective lot-streaming flow shop scheduling problem

Abstract: Crossover and mutation operators in NSGA-II are random and aimless, and encounter difficulties in generating offspring with high quality. Aiming to overcoming these drawbacks, we proposed an improved NSGA-II algorithm (INSGA-II) and applied it to solve the lot-streaming flow shop scheduling problem with four criteria. We first presented four variants of NEH heuristic to generate the initial population, and then incorporated the estimation of distribution algorithm and a mutation operator based on insertion and swap into NSGA-II to replace traditional crossover and mutation operators. Last but not least, we performed a simple and efficient restarting strategy on the population when the diversity of the population is smaller than a given threshold. We conducted a serial of experiments, and the experimental results demonstrate that the proposed algorithm outperforms the comparative algorithms.

Crossover Heterogeneity in the Absence of Hotspots in Caenorhabditis elegans

Abstract: Crossovers play mechanical roles in meiotic chromosome segregation, generate genetic diversity by producing new allelic combinations, and facilitate evolution by decoupling linked alleles. In almost every species studied to date, crossover distributions are dramatically nonuniform, differing among sexes and across genomes, with spatial variation in crossover rates on scales from whole chromosomes to subkilobase hotspots. To understand the regulatory forces dictating these heterogeneous distributions a crucial first step is the fine-scale characterization of crossover distributions. Here we define the wild-type distribution of crossovers along a region of the C. elegans chromosome II at unprecedented resolution, using recombinant chromosomes of 243 hermaphrodites and 226 males. We find that well-characterized large-scale domains, with little fine-scale rate heterogeneity, dominate this region’s crossover landscape. Using the Gini coefficient as a summary statistic, we find that this region of the C. elegans genome has the least heterogeneous fine-scale crossover distribution yet observed among model organisms, and we show by simulation that the data are incompatible with a mammalian-type hotspot-rich landscape. The large-scale structural domains—the low-recombination center and the high-recombination arm—have a discrete boundary that we localize to a small region. This boundary coincides with the arm-center boundary defined both by nuclear-envelope attachment of DNA in somatic cells and GC content, consistent with proposals that these features of chromosome organization may be mechanical causes and evolutionary consequences of crossover recombination.

Multi-objective optimization in rule-based design space exploration

Abstract: Design space exploration (DSE) aims to find optimal design candidates of a domain with respect to different objectives where design candidates are constrained by complex structural and numerical restrictions. Rule-based DSE aims to find such candidates that are reachable from an initial model by applying a sequence of exploration rules. Solving a rule-based DSE problem is a difficult challenge due to the inherently dynamic nature of the problem.In the current paper, we propose to integrate multi-objective optimization techniques by using Non-dominated Sorting Genetic Algorithms (NSGA) to drive rule-based design space exploration. For this purpose, finite populations of the most promising design candidates are maintained wrt. different optimization criteria. In our context, individuals of a generation are defined as a sequence of rule applications leading from an initial model to a candidate model. Populations evolve by mutation and crossover operations which manipulate (change, extend or combine) rule execution sequences to yield new individuals.Our multi-objective optimization approach for rule-based DSE is domain independent and it is automated by tooling built on the Eclipse framework. The main added value is to seamlessly lift multi-objective optimization techniques to the exploration process preserving both domain independence and a high-level of abstraction. Design candidates will still be represented as models and the evolution of these models as rule execution sequences. Constraints are captured by model queries while objectives can be derived both from models or rule applications.