In this paper, we describe FFEAT - a library for GPU-based implementation of evolutionary algorithms based on Torch. We discuss limitations of GPU computing and how they affect implementations of evolutionary algorith...
详细信息
ISBN:
(纸本)9781450392686
In this paper, we describe FFEAT - a library for GPU-based implementation of evolutionary algorithms based on Torch. We discuss limitations of GPU computing and how they affect implementations of evolutionary algorithms and other population-based heuristics. Using FFEAT, we implement a number of different types of nature inspired algorithms, including evolutionary algorithms, evolution strategies, and particle swarm optimization. We investigate the performance of such algorithms in a number of benchmarks and with varying algorithm settings. We showthat in some cases, we can obtain an order of magnitude speed-up by running the algorithm on a GPU compared to running it on a CPU.
This paper assesses the potential for mechanised assistance in the formulation of schedulability tests. The novel idea is to use evolutionary algorithms to semi-automate the process of deriving response time analysis ...
详细信息
ISBN:
(纸本)9781728144030
This paper assesses the potential for mechanised assistance in the formulation of schedulability tests. The novel idea is to use evolutionary algorithms to semi-automate the process of deriving response time analysis equations. The proof of concept presented in this paper focuses on the synthesis of mathematical expressions for the schedulability analysis of messages on Controller Area Network (CAN). This problem is of particular interest, since the original analysis developed in the early 1990s was later found to be flawed. Further, as well as known exact tests that have been formally proven, there are a number of useful sufficient tests of pseudo-polynomial complexity and closed-form polynomial-time upper bounds on response times that provide useful comparisons.
Genetic algorithms (GAs) and other evolutionary algorithms (EAs), as powerful and broadly applicable stochastic search and optimization techniques have been successfully applied in the area of management science, oper...
详细信息
Genetic algorithms (GAs) and other evolutionary algorithms (EAs), as powerful and broadly applicable stochastic search and optimization techniques have been successfully applied in the area of management science, operations research and industrial engineering. In the past few years, researchers gave lots of great idea for improvement of evolutionary algorithms, which include population initialization, individual selection, evolution, parameter setting, hybrid approach with conventional heuristics etc. However, though lots of different versions of evolutionary computations have been created, all of them have turned most of its attention to the development of search abilities of approaches. In this paper, for improving the search ability, we focus on how to take a balance between exploration and exploitation of the search space. It is also very difficult to solve problem, because the balance between exploration and exploitation is depending on the characteristic of different problems. The balance also should be changed dynamically depend on the status of evolution process. Purpose of this paper is the design of an effective approach which it can correspond to most optimization problems. In this paper, we propose an auto-tuning strategy by using fuzzy logic control. The main idea is adaptively regulation for taking the balance among the stochastic search and local search probabilities based on the change of the average fitness of parents and offspring which is occurred at each generation. In addition, numerical analyses of different type optimization problems show that the proposed approach has higher search capability that improve quality of solution and enhanced rate of convergence.
Medium-voltage distribution network expansion planning involves finding the most economical adjustments of both the capacity and the topology of the network such that no operational constraints are violated and the ex...
详细信息
ISBN:
(纸本)9783319116839;9783319116822
Medium-voltage distribution network expansion planning involves finding the most economical adjustments of both the capacity and the topology of the network such that no operational constraints are violated and the expected loads, that the expansion is planned for, can be supplied. This paper tackles this important real-world problem using realistic yet computationally feasible models and, for the first time, using two instances of the recently proposed class of Gene-pool Optimal Mixing evolutionary algorithms (GOMEAs) that have previously been shown to be a highly efficient integration of local search and genetic recombination, but only on standard benchmark problems. One GOMEA instance that we use employs linkage learning and one instance assumes no dependencies among problem variables. We also conduct experiments with a widely used traditional Genetic Algorithm (GA). Our results show that the favorable performance of GOMEA instances over traditional GAs extends to the real-world problem at hand. Moreover, the use of linkage learning is shown to further increase the algorithm's effectiveness in converging toward optimal solutions.
A method based on evolutionary algorithms for the estimation of parameters in excitation systems is shown. Usually, for these parameters, typical values or values provided by the manufacturer are used. These values do...
详细信息
ISBN:
(纸本)9781424402878
A method based on evolutionary algorithms for the estimation of parameters in excitation systems is shown. Usually, for these parameters, typical values or values provided by the manufacturer are used. These values do not reflect the recent conditions of operation of the system. When estimation of parameters is confronted, this is done based on methods supported on the domain of frequency, which demand an arduous work of adjustment, also can be based on methods that force to linearize the model around an operation point and omitting certain types of non-linearities like limiters. The evolutionary method presented estimates parameters with signals in the time domain. It can use real data originated in the operation of the and directly executes the adjustment of the results. This evolutionary method may be applied for specific functions of the system or in the whole excitation system. It can include the non-linear properties of the system, such as the saturation in the case of an operation that reaches the limits. The evolutionary method was applied to the model of an excitation system in the hydroelectric plant of Salvajina in Colombia.
evolutionary computation can increase the speed and accuracy of pattern recognition in multispectral images, for example, in automatic target tracking. We have developed two classes of evolutionary algorithms for expl...
详细信息
ISBN:
(纸本)9780819468482
evolutionary computation can increase the speed and accuracy of pattern recognition in multispectral images, for example, in automatic target tracking. We have developed two classes of evolutionary algorithms for exploiting multispectral imagery. The first method treats the clustering process. It determines a cluster of pixels around specified reference pixels so that the entire cluster is increasingly representative of the search object. An initial population (of clusters) evolves into populations of new clusters, with each cluster having an assigned fitness score. This population undergoes iterative mutation and selection. Mutation operators alter both the pixel cluster set cardinality and composition. Several stopping criteria can be applied to terminate the evolution. An advantage of this evolutionary cluster formulation is that the resulting cluster may have an arbitrary shape so that it most nearly fits the search pattern. The second algorithm class automates the selection of features (the center-wavelength and the bandwidth) for each population member. For each pixel in the image and for each population member, the Mahalanobis distance to the reference set is calculated and a decision is made whether or not this pixel belongs to a target. The sum of correct and false decisions defines a Receiver Operating Curve, which is used to measure the fitness of a population member. Based on this fitness, the algorithm decides which population members to use as parents for the next iteration.
Tie-in spools must be designed to resist a large number of onerous load combinations. These loads include gravitational, temperature, pressure and environmental loads along with various imposed displacements. Addition...
详细信息
ISBN:
(纸本)9780791855379
Tie-in spools must be designed to resist a large number of onerous load combinations. These loads include gravitational, temperature, pressure and environmental loads along with various imposed displacements. Additionally, there are several design constraints that must be satisfied. Due to the three-dimensional geometric freedom of the spool there are many possible design scenarios that could be evaluated in the search for the optimum solution. It is the responsibility of the pipeline design engineer to use their own judgment and experience to find the best possible solution within the design period. Traditionally a trial and error design approach is used in an iterative manner. This method is typically slow and labor intensive and can be too focused on one design concept at the expense of others that are potentially superior. On similar engineering problems with many design parameters automated non-linear optimization routines have been shown to be very effective. Specifically, applying evolutionary algorithms is a robust, time-effective and adaptable approach. Such a tool assists the engineer in finding superior design solutions and assists in searching the entire design space. To test this design method, a multi-objective evolutionary algorithm has been applied to two semi-constrained spool design problems. The spool design has been modeled using finite element analysis. First, the algorithm was applied to the optimization of spool geometry for multiple design objectives. Within 24-hours of runtime the algorithm was able to find superior solutions to those found using a traditional iterative approach. Also, the trade-off between conflicting design objectives could be quantified and visualized to enable the designer to select the most appropriate candidate. The second problem evaluated was the placement of supports to mitigate the onset of vortex induced vibration (VIV). The algorithm was again able to quickly find a better solution and quantify the tradeoff betwe
This work addresses the problem of building representative subsets of benchmarks from an original large set of benchmarks, using statistical analysis techniques. The subsets should be developed in this way to include ...
详细信息
ISBN:
(纸本)9781424422326
This work addresses the problem of building representative subsets of benchmarks from an original large set of benchmarks, using statistical analysis techniques. The subsets should be developed in this way to include only the necessary information for evaluating the performance of a computer system or application. The development of representative workloads is not a trivial procedure, since incorrectly selecting benchmarks the representative subset can produce erroneous results. A number of statistical analysis techniques have been developed for identifying representative workloads. The goal of these approaches is to reduce the dimensionality of the original set of benchmarks prior to identifying similar benchmarks. In this work we propose a combination of Independent Component Analysis (ICA) and evolutionary Algorithm (EA) as a more efficient way for reducing the computational complexity of the problem and the redundant information of the original set of benchmarks. Experimental results validate that the proposed technique generates more representative workloads than prior techniques.
In this paper we analyze a new method for an adaptive variation of evolutionary algorithms (EAs) population size: the Self-Regulated Population size EA (SRP-EA). An empirical evaluation of the method is provided by co...
详细信息
ISBN:
(纸本)3540389903
In this paper we analyze a new method for an adaptive variation of evolutionary algorithms (EAs) population size: the Self-Regulated Population size EA (SRP-EA). An empirical evaluation of the method is provided by comparing the new proposal with the CHC algorithm and other well known EAs with varying population. A fitness landscape generator was chosen to test and compare the algorithms: the Spear's multimodal function generator. The performance of the algorithms was measured in terms of success rate, quality of the solutions and evaluations needed to attain them over a wide range of problem instances. We will show that SRP-EA performs well on these tests and appears to overcome some recurrent drawbacks of traditional EAs which lead them to local optima premature convergence. Also, unlike other methods, SRP-EA seems to self-regulate its population size according to the state of the search.
The mutation is one of the operators that is used by many evolutionary algorithms (EA) to diversify the population (solutions). It can enhance the algorithm exploration of the problem search space and improve the evol...
详细信息
ISBN:
(纸本)9781665474849;9781665474832
The mutation is one of the operators that is used by many evolutionary algorithms (EA) to diversify the population (solutions). It can enhance the algorithm exploration of the problem search space and improve the evolution process. This paper introduces a novel mutation technique that is based on a recently investigated mutation bias pattern in the Arabidopsis thaliana plant [1]. The proposed mutation technique is called an essential mutation. The proposed method uses the. parameter to control the amount of distance we can be from the parent's fitness. Three different configurations are studied and the best results are obtained when epsilon=0. It is compared against five well-known mutation techniques which are Boundary, Non-uniform, MPT, and Polynomial on standard benchmark functions. The obtained results show the superiority of the proposed essential mutation in terms of best solution and convergence speed in most of the test functions.
暂无评论