Results 1  10
of
29
Designing Efficient And Accurate Parallel Genetic Algorithms
, 1999
"... Parallel implementations of genetic algorithms (GAs) are common, and, in most cases, they succeed to reduce the time required to find acceptable solutions. However, the effect of the parameters of parallel GAs on the quality of their search and on their efficiency are not well understood. This insuf ..."
Abstract

Cited by 220 (5 self)
 Add to MetaCart
Parallel implementations of genetic algorithms (GAs) are common, and, in most cases, they succeed to reduce the time required to find acceptable solutions. However, the effect of the parameters of parallel GAs on the quality of their search and on their efficiency are not well understood. This insufficient knowledge limits our ability to design fast and accurate parallel GAs that reach the desired solutions in the shortest time possible. The goal of this dissertation is to advance the understanding of parallel GAs and to provide rational guidelines for their design. The research reported here considered three major types of parallel GAs: simple masterslave algorithms with one population, more sophisticated algorithms with multiple populations, and a hierarchical combination of the first two types. The investigation formulated simple models that predict accurately the quality of the solutions with different parameter settings. The quality predictors were transformed into populationsizing equations, which in turn were used to estimate the execution time of the algorithms.
A Survey of Parallel Genetic Algorithms
 CALCULATEURS PARALLELES, RESEAUX ET SYSTEMS REPARTIS
, 1998
"... Genetic algorithms (GAs) are powerful search techniques that are used successfully to solve problems in many different disciplines. Parallel GAs are particularly easy to implement and promise substantial gains in performance. As such, there has been extensive research in this field. This survey att ..."
Abstract

Cited by 147 (5 self)
 Add to MetaCart
Genetic algorithms (GAs) are powerful search techniques that are used successfully to solve problems in many different disciplines. Parallel GAs are particularly easy to implement and promise substantial gains in performance. As such, there has been extensive research in this field. This survey attempts to collect, organize, and present in a unified way some of the most representative publications on parallel genetic algorithms. To organize the literature, the paper presents a categorization of the techniques used to parallelize GAs, and shows examples of all of them. However, since the majority of the research in this field has concentrated on parallel GAs with multiple populations, the survey focuses on this type of algorithms. Also, the paper describes some of the most significant problems in modeling and designing multipopulation parallel GAs and presents some recent advancements.
A Parallel Genetic Algorithm for the Set Partitioning Problem
, 1994
"... In this dissertation we report on our efforts to develop a parallel genetic algorithm and apply it to the solution of the set partitioning problema difficult combinatorial optimization problem used by many airlines as a mathematical model for flight crew scheduling. We developed a distributed stea ..."
Abstract

Cited by 66 (1 self)
 Add to MetaCart
In this dissertation we report on our efforts to develop a parallel genetic algorithm and apply it to the solution of the set partitioning problema difficult combinatorial optimization problem used by many airlines as a mathematical model for flight crew scheduling. We developed a distributed steadystate genetic algorithm in conjunction with a specialized local search heuristic for solving the set partitioning problem. The genetic algorithm is based on an island model where multiple independent subpopulations each run a steadystate genetic algorithm on their own subpopulation and occasionally fit strings migrate between the subpopulations. Tests on forty realworld set partitioning problems were carried out on up to 128 nodes of an IBM SP1 parallel computer. We found that performance, as measured by the quality of the solution found and the iteration on which it was found, improved as additional subpopulations were added to the computation. With larger numbers of subpopulations the genetic algorithm was regularly able to find the optimal solution to problems having up to a few thousand integer variables. In two cases, highquality integer feasible solutions were found for problems with 36,699 and 43,749 integer variables, respectively. A notable limitation we found was the difficulty solving problems with many constraints.
Efficient Parallel Genetic Algorithms: Theory and Practice
 Computer Methods in Applied Mechanics and Engineering
, 2000
"... Parallel genetic algorithms (GAs) are complex programs that are controlled by many parameters, which affect their search quality and their efficiency. The goal of this paper is to provide guidelines to choose those parameters rationally. The investigation centers on the sizing of populations, becaus ..."
Abstract

Cited by 36 (1 self)
 Add to MetaCart
Parallel genetic algorithms (GAs) are complex programs that are controlled by many parameters, which affect their search quality and their efficiency. The goal of this paper is to provide guidelines to choose those parameters rationally. The investigation centers on the sizing of populations, because previous studies show that there is a crucial relation between solution quality and population size. As a first step, the paper shows how to size a simple GA to reach a solution of a desired quality. The simple GA is then parallelized, and its execution time is optimized. The rest of the paper deals with parallel GAs with multiple populations. Two bounding cases of the migration rate and topology are analyzed, and the case that yields good speedups is optimized. Later, the models are specialized to consider sparse topologies and migration rates that are more likely to be used by practitioners. The paper also presents the additional advantages of combining multi and singlepopulation parallel GAs. The results of this work are simple models that practitioners may use to design efficient and competent parallel GAs.
Designing Efficient MasterSlave Parallel Genetic Algorithms
, 1997
"... A simple technique to reduce the execution time of genetic algorithms (GAs) is to divide the task of evaluating the population among several processors. This class of algorithms is called "global" parallel GAs because selection and mating consider the entire population. Global parallel GAs are usual ..."
Abstract

Cited by 27 (4 self)
 Add to MetaCart
A simple technique to reduce the execution time of genetic algorithms (GAs) is to divide the task of evaluating the population among several processors. This class of algorithms is called "global" parallel GAs because selection and mating consider the entire population. Global parallel GAs are usually implemented as masterslave programs and require constant interprocessor communication. This will affect their performance, but most investigations of these algorithms ignore the penalty caused by communications. This paper presents an analysis of the execution time of global parallel GAs that includes a simple model of the time used in communications and shows that there is an optimal number of processors that minimizes the execution time. To further reduce the execution time we recommend the use of hybrids that combine global and coarsegrained parallel GAs. 1 Introduction A simple technique to parallelize genetic algorithms is to divide the task of evaluating the population among seve...
Experiences with FineGrained Parallel Genetic Algorithms
 Annals of Operations Research
, 1996
"... this paper we present some results of our systematic studies of finegrained parallel versions of the island model of genetic algorithms and of variants of the neighborhood model (also called diffusion model) on the massively parallel computer MasPar MP1 with 16k processing elements. These parallel ..."
Abstract

Cited by 21 (2 self)
 Add to MetaCart
this paper we present some results of our systematic studies of finegrained parallel versions of the island model of genetic algorithms and of variants of the neighborhood model (also called diffusion model) on the massively parallel computer MasPar MP1 with 16k processing elements. These parallel genetic algorithms have been applied to a range of different problems (e.g. traveling salesperson, capacitated lot sizing, ressource constrained project scheduling, flow shop, and warehouse location problems) in order to obtain an empirical basis for statements on their optimization quality.
Parallel Strategies for Metaheuristics
"... We present a stateoftheart survey of parallel metaheuristic developments and results, discuss general design and implementation principles that apply to most metaheuristic classes, instantiate these principles for the three metaheuristic classes currently most extensively used  genetic metho ..."
Abstract

Cited by 14 (4 self)
 Add to MetaCart
We present a stateoftheart survey of parallel metaheuristic developments and results, discuss general design and implementation principles that apply to most metaheuristic classes, instantiate these principles for the three metaheuristic classes currently most extensively used  genetic methods, simulated annealing, and tabu search, and identify a number of trends and promising research directions.
A Hardware Architecture For A Parallel Genetic Algorithm For Image Registration
 in Proceedings of IEE Colloquium on Genetic Algorithms in Image Processing and Vision
, 1994
"... this paper the nature of Parallel Genetic Algorithms is described followed by the application of genetic algorithms to vision systems. A description of a hardware architecture for vision systems is detailed along with various modifications to improve the implementation. A simulation is used to produ ..."
Abstract

Cited by 11 (1 self)
 Add to MetaCart
this paper the nature of Parallel Genetic Algorithms is described followed by the application of genetic algorithms to vision systems. A description of a hardware architecture for vision systems is detailed along with various modifications to improve the implementation. A simulation is used to produce results that verify the effectiveness of the hardware architecture and finally conclusions and future work are discussed. Parallel Genetic Algorithms Parallel genetic algorithms can be categorised into three types, 'standard', 'coarse grained' and 'fine grained'. Standard GAs can be implemented on a parallel architecture by distributing the evaluation process over a number of processors (Fogarty 1990). Coarse grained genetic algorithms run several populations of genes in parallel. After a number of generations (G) the separate populations export a set of individuals (n) to other neighbouring populations. G generations is termed an 'epoch' or migration period. Normally a single processor will monitor each population and the processor swaps individuals every epoch thus parallelising the problem over the number of populations (processors) within the multiprocessor system (Figure 1). A variety of topologies have been used to define 'neighbours' typically a simple grid or hypercube (Figure 2) is used with each node corresponding to a processor (Tanese 1989, Cohoon 1990, Muhlenbein 1991, Xu 1992). Theoretical studies of a coarse grain GA have been done by Petty & Leuze 1989. Figure 1: Grid Topology Figure 2: Hypercube Topology (4 Dimensional) Finegrained parallel genetic algorithms act on each member of the population in parallel. Consequently each member of the population performs crossover with its immediate neighbours, where the neighbourhood is defined by the topology and s...
On Efficient Communication in Distributed Genetic Algorithms
, 1994
"... yuvaKDschema.co.il To what extent is distribution beneficial to the search quality and computational resources used by a genetic algorithm execution? Most distributed genetic algorithms rely on communicating genetic information, in the form of individual solutions, between concurrently evolving popu ..."
Abstract

Cited by 10 (0 self)
 Add to MetaCart
yuvaKDschema.co.il To what extent is distribution beneficial to the search quality and computational resources used by a genetic algorithm execution? Most distributed genetic algorithms rely on communicating genetic information, in the form of individual solutions, between concurrently evolving populations. Another way of effectively using the additional information generated by the parallel executions is the profiling approach to communication, where populations decide whether their own performance is satisfactory, relative to the global average improvement curve. Thus, communication between populations takes the form of improvement histories. This is shown to improve on the traditional communication approach, in terms of both solution quality and execution performance. 1
A Hardware Engine for Genetic Algorithms
, 1997
"... A genetic algorithm (GA) is an optimization method based on natural selection. Genetic algorithms have been applied to many hard optimization problems including VLSI layout optimization, boolean satisfiability, power system control, fault detection, control systems, and signal processing. GAs have b ..."
Abstract

Cited by 9 (1 self)
 Add to MetaCart
A genetic algorithm (GA) is an optimization method based on natural selection. Genetic algorithms have been applied to many hard optimization problems including VLSI layout optimization, boolean satisfiability, power system control, fault detection, control systems, and signal processing. GAs have been recognized as a robust generalpurpose optimization technique. But application of GAs to increasingly complex problems can overwhelm software implementations of GAs, causing unacceptable delays in the optimization process. This is true of any nontrivial application of GAs if the search space is large or if realtime performance is necessary. It follows that a hardware implementation of a GA is desirable for application to problems too complex for softwarebased GAs. Hardware's speed advantage and its ability to parallelize offer great rewards to genetic algorithms. Speedups of 12 orders of magnitude have been observed when frequently used software routines were implemented in hardware...