Parallel Implementation Of Genetic Algorithm Research Articles

Multi-GPU Island-Based Genetic Algorithm

Genetic algorithms are effective in solving many optimization tasks. However, the long execution time associated with it prevents its use in many domains. In this paper, we propose a new approach for parallel implementation of genetic algorithm on graphics processing units (GPUs) using CUDA programming model. This paper introduces a novel implementation of the genetic algorithm exploiting a multi-GPU cluster. The proposed implementation employs an island-based genetic algorithm where every GPU evolves a single island. Recently there has been a growing interest in developing parallel algorithms using graphic processing units (GPU) also referred as GPU computation. Advances in the video gaming industry have led to the production of low-cost, high performance graphics processing units that possess more memory bandwidth and computational capability than central processing units (CPUs). As GPUs are available in personal computers, and they are easy to use and manage through several GPU programming languages, graphics engines are being adopted widely in scientific computing.

Stopping rules for a parallel genetic algorithm

A novel method for the implementation of parallel genetic algorithms is introduced to locate the global minimum of a multidimensional function inside a rectangular hyperbox. The algorithm relies on a client - server model and incorporates an enhanced stopping rule. A number of experiments were conducted in order to measure the effects in termination by using the termination rule either on server machine or on clients. The method is tested on a series of well - known test functions as well as neural network training and the results was compared against another parallel genetic algorithm method. The results from the experiments are reported in terms of test error and amount of generations.

Parallel implementation of genetic algorithm on FPGA using Vivado high level synthesis

Genetic algorithm (GA) is one of most popular evolutionary search algorithms that simulates natural selection of genetic evolution for searching solution to arbitrary engineering problems. However,...

Stopping rules for a parallel genetic algorithm

A novel method for the implementation of parallel genetic algorithms is introduced to locate the global minimum of a multidimensional function inside a rectangular hyperbox. The algorithm relies on...

Parallel implementation of genetic algorithm on FPGA using Vivado high level synthesis

Genetic algorithm (GA) is one of most popular evolutionary search algorithms that simulates natural selection of genetic evolution for searching solution to arbitrary engineering problems. However, it is computationally intensive and will become a limiting factor for evolving solution to most of the real life problems as it involves large number of parameters that needs to be determined. Fortunately, there are some parallel platforms such as field programmable gate array (FPGA) that can be adopted to overcome this constrains by improving its latency. So, efficient parallel implementation of GA was proposed where each step of GA was exploited to improve its computational task. Moreover, many optimization and parallelisation techniques were adopted and applied to achieve high speed up. The results show that 43 speed up is achieved compared with the typical one. Moreover, higher speed up can be achieved with larger input size.

High-Performance Parallel Implementation of Genetic Algorithm on FPGA

Genetic Algorithms (GAs) are used to solve search and optimization problems in which an optimal solution can be found using an iterative process with probabilistic and non-deterministic transitions. However, depending on the problem's nature, the time required to find a solution can be high in sequential machines due to the computational complexity of genetic algorithms. This work proposes a parallel implementation of a genetic algorithm on field-programmable gate array (FPGA). Optimization of the system's processing time is the main goal of this project. Results associated with the processing time and area occupancy (on FPGA) for various population sizes are analyzed. Studies concerning the accuracy of the GA response for the optimization of two variables functions were also evaluated for the hardware implementation. However, the high-performance implementation proposes in this paper is able to work with more variable from some adjustments on hardware architecture.

Error-Safe, Portable, and Efficient Evolutionary Algorithms Implementation with High Scalability.

We present an efficient massively parallel implementation of genetic algorithms for chemical and materials science problems, solely based on Java virtual machine (JVM) technologies and standard networking protocols. The lack of complicated dependencies allows for a highly portable solution exploiting strongly heterogeneous components within a single computational context. At runtime, our implementation is almost completely immune to hardware failure, and additional computational resources can be added or subtracted dynamically, if needed. With extensive testing, we show that despite all these benefits, parallel scalability is excellent.

PDoublePop: An implementation of parallel genetic algorithm for function optimization

A software for the implementation of parallel genetic algorithms is presented in this article. The underlying genetic algorithm is aimed to locate the global minimum of a multidimensional function inside a rectangular hyperbox. The proposed software named PDoublePop implements a client–server model for parallel genetic algorithms with advanced features for the local genetic algorithms such as: an enhanced stopping rule, an advanced mutation scheme and periodical application of a local search procedure. The user may code the objective function either in C++ or in Fortran77. The method is tested on a series of well-known test functions and the results are reported. Program summaryProgram title: PDoublePopCatalogue identifier: AFBJ_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AFBJ_v1_0.htmlProgram obtainable from: CPC Program Library, Queen’s University, Belfast, N. IrelandLicensing provisions: GNU GPL v3No. of lines in distributed program, including test data, etc.: 26048No. of bytes in distributed program, including test data, etc.: 161286Distribution format: tar.gzProgramming language: GNU-C++, GNU-C, GNU Fortran - 77, MPI.Computer: The tool has been tested on Linux and FreeBSD. The tool is designed to be portable in all systems running the GNU C++ compiler, with Open MPI or LAM MPI.Operating system: Any running the GNU C++ compiler.Has the code been vectorised or parallelized?: YesRAM: 200KBClassification: 4.9, 4.12, 6.5.Nature of problem:A series of problems in science and engineering usually can be formulated as a problem of minimizing a function of many variables. The so called local optimization techniques are frequently trapped in local minima, that it sub optimal solutions. For that reason researchers should use more advanced methods that aim to estimate the global minimum of the function.Solution method:A stopping rule and a periodical application of local search are utilized in conjunction with parallel genetic algorithms in order to solve optimization problems. The parallelization is performed using the MPI programming interface.Running time:Depending on the objective function.

DWFS: a wrapper feature selection tool based on a parallel genetic algorithm.

Many scientific problems can be formulated as classification tasks. Data that harbor relevant information are usually described by a large number of features. Frequently, many of these features are irrelevant for the class prediction. The efficient implementation of classification models requires identification of suitable combinations of features. The smaller number of features reduces the problem’s dimensionality and may result in higher classification performance. We developed DWFS, a web-based tool that allows for efficient selection of features for a variety of problems. DWFS follows the wrapper paradigm and applies a search strategy based on Genetic Algorithms (GAs). A parallel GA implementation examines and evaluates simultaneously large number of candidate collections of features. DWFS also integrates various filtering methods that may be applied as a pre-processing step in the feature selection process. Furthermore, weights and parameters in the fitness function of GA can be adjusted according to the application requirements. Experiments using heterogeneous datasets from different biomedical applications demonstrate that DWFS is fast and leads to a significant reduction of the number of features without sacrificing performance as compared to several widely used existing methods. DWFS can be accessed online at www.cbrc.kaust.edu.sa/dwfs.

An efficient crossover architecture for hardware parallel implementation of genetic algorithm

In this article a new architecture for hardware implementation of genetic algorithm in reconfigurable embedded systems is presented. The main idea is based on the efficient use of a genetic algorithm's crossover operator to enhance the speed of algorithm to reach an optimal solution. In this article a new crossover called DSO and also two new architectures for implementation of crossover operators are introduced to provide suitable solutions for solving the problems related to fitness function of the genetic algorithm. At first, some optimum operators are selected and then utilized in a new parallel architecture to increase the speed and accuracy of algorithm convergence. Finally, based on reusability of existing resources, the main idea of the article is introduced to improve the performance of the algorithm and finding the optimal solution. The properties of FPGAs such as flexibility and parallelism help this purpose.

Genetic Algorithm for the History Matching Problem

In this work we present a study of genetic algorithms for the automatic history matching problem of reservoir simulation. The history matching process is an inverse problem that searches a set of parameters that minimizes the difference between the model performance and the historical performance of the field. This model validation process is essential and gives credibility to the predictions of the reservoir model.We studied a Parallel Genetic Algorithm implementation, several tests were performed and the preliminary results are presented and discussed in this work.

Parallelization of genetic algorithms using Hadoop Map/Reduce

In this paper we present parallel implementation of genetic algorithm using map/reduce programming paradigm. Hadoop implementation of map/reduce library is used for this purpose. We compare our implementation with implementation presented in [1]. These two implementations are compared in solving One Max (Bit counting) problem. The comparison criteria between implementations are fitness convergence, quality of final solution, algorithm scalability, and cloud resource utilization. Our model for parallelization of genetic algorithm shows better performances and fitness convergence than model presented in [1], but our model has lower quality of solution because of species problem.

Parallel Genetic Algorithm for Intelligent Model Parameter Extraction of Metal-Oxide-Semiconductor Field Effect Transistors

Equivalent circuit model of semiconductor devices associated with a set of optimized parameters currently plays a central role in the circuit design and semiconductor manufacturing communities. An intelligent model parameter extraction system that simultaneously integrates evolutionary and numerical optimization techniques for optimal characterization of sub-100 nm metal-oxide-semiconductor field effect transistors (MOSFETs) has recently been advanced [1]. In this article, to accelerate the extraction process, parallelization of the genetic algorithm (GA) for the intelligent model parameter extraction system of MOSFETs is developed. The GA implemented in the extraction system is mainly parallelized with a diffusion scheme on a PC-based Linux cluster with message passing interface libraries. Parallelization of GA is governed by various factors, which affect the quality of extracted parameters and its computational efficiency. The result obtained in this study shows that the diffusion GA is superior to an isolated GA, and the superiority of the diffusion GA becomes significant when the number of MOSFETs to be optimized is increased. Theoretical estimation and preliminary numerical implementation of parallel GA show that there exist an optimal number of processors with respect to the number of devices to be extracted. Benchmark results, such as speedup and efficiency including accuracy of extraction are presented and discussed for different sets of realistic multiple sub-100 nm devices to show the robustness and efficiency of the method. Practical implementation of the parallel GA approach benefits the engineering of device model parameter extraction in nowadays semiconductor manufacturing industry.

Cellular genetic algorithm technique for the multicriterion design optimization

There have been increased activities in the study of genetic algorithms (GA) for problems of design optimization. The present paper describes a fine-grained model of parallel GA implementation that derives from a cellular-automata-like computation. The central idea behind the Cellular Genetic Algorithm approach is to treat the GA population as being distributed over a 2-D grid of cells, with each member of the population occupying a particular cell and defining the state of that cell. Evolution of the cell state is tantamount to updating the design information contained in a cell site, and as in cellular automata computations, takes place on the basis of local interaction with neighboring cells. A focus of the paper is in the adaptation of the cellular genetic algorithm approach in the solution of multicriteria design optimization problems. The proposed paper describes the implementation of this approach and examines its efficiency in the context of representative design optimization problems.

A SAND approach based on cellular computation models for analysis and optimization

Genetic algorithms (GAs) have received considerable recent attention in problems of design optimization. The mechanics of population-based search in GAs are highly amenable to implementation on parallel computers. The present article describes a fine-grained model of parallel GA implementation that derives from a cellular-automata-like computation. The central idea behind the cellular genetic algorithm (CGA) approach is to treat the GA population as being distributed over a 2-D grid of cells, with each member of the population occupying a particular cell and defining the state of that cell. Evolution of the cell state is tantamount to updating the design information contained in a cell site and, as in cellular automata computations, takes place on the basis of local interaction with neighbouring cells. A special focus of the article is in the use of cellular automata (CA)-based models for structural analysis in conjunction with the CGA approach to optimization. In such an approach, the analysis and optimization are evolved simultaneously in a unified cellular computational framework. The article describes the implementation of this approach and examines its efficiency in the context of representative structural optimization problems.

A hybrid parallel genetic algorithm approach for graph coloring

Graph-Coloring problem (GCP) deals with assigning labels (colors) to the vertices of a graph such that adjacent vertices do not get the same color. Coloring a graph with minimum number of colors is a well-known NP-hard problem. In this paper a new permutation based representation of graph coloring problem is solved using a parallel genetic algorithm (PGA). Migration model of parallelism is used with Message passing interface (MPI) for implementation of parallel genetic algorithm. Three-crossover operators namely greedy partition crossover (GPX), Uniform independent set crossover (UISX), and Permutation-based crossover (PX) are used. The performance of the three crossover operators is investigated in terms of convergence and execution time for standard benchmark graphs. The results show that GPX performs well in terms of convergence and PX in terms of execution time. The three crossover operators in parallel genetic algorithm outperform the serial genetic algorithm approximately by a factor of three. The paper is also validated with the static wavelength assignment problem in optical networks.

Scalable Parallel Genetic Algorithms

Genetic algorithms, search algorithms based on the genetic processes observed in natural evolution, have been used to solve difficult problems in many different disciplines. When applied to very large-scale problems, genetic algorithms exhibit high computational cost and degradation of the quality of the solutions because of the increased complexity. One of the most relevant research trends in genetic algorithms is the implementation of parallel genetic algorithms with the goal of obtaining quality of solutions efficiently. This paper first reviews the state-of-the-art in parallel genetic algorithms. Parallelization strategies and emerging implementations are reviewed and relevant results are discussed. Second, this paper discusses important issues regarding scalability of parallel genetic algorithms.

An Investigation of PWB Layout by Genetic Algorithms to Maximize Fatigue Life

Thermal considerations in printed wiring board (PWB) assemblies are becoming increasingly important as packaging constraints shrink and power use escalates. In this paper, we provide a study on the potential for a genetic algorithm-driven PWB layout design tool to improve the thermal performance of such assemblies. As a case study, the thermomechanical fatigue of surface mounted leadless chip carriers on an FR4 epoxy board is used. We have found that by utilizing appropriate formula-based engineering approximations, the efficiency of parallel implementations of genetic algorithms in finding near-optimal and results makes this approach effective as an explorative “scouting” approach to identify promising board configurations for more computationally expensive evaluations such as finite element method.

Stack Filter Design Using a Distributed Parallel Implementation of Genetic Algorithms

Stack Filter Design Using a Distributed Parallel Implementation of Genetic Algorithms

Parallel genetic algorithm implementation in multidisciplinary rotor blade design

Abstract : The present paper describes an adaptation of genetic algorithms in the design of large-scale multidisciplinary optimization problems. A hingeless composite rotor blade is used as the test problem, where the formulation of the objective and constraint functions requires the consideration of disciplines of aerodynamics, performance, dynamics, and structures. A rational decomposition approach is proposed for partitioning the large-scale multidisciplinary design problem into smaller, more tractable subproblems. A design method based on a parallel implementation of genetic algorithms is shown to be an effective strategy, providing increased computational efficiency, and a natural approach to account for the coupling between temporarily decoupled subproblems. A central element of the proposed approach is the use of artificial neural networks for identifying a topology for problem decomposition and for generating global function approximations for use in optimization. (AN)