• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.1 no.1
• /
• pp.125-130
• /
• 2001

In this paper, we propose a new type of evolution strategies combined with reinforcement learning. We use the variances of fitness occurred by mutation to make the reinforcement signals which estimate and control the step length of mutation. With this proposed method, the convergence rate is improved. Also, we use cauchy distributed mutation to increase global convergence faculty. Cauchy distributed mutation is more likely to escape from a local minimum or move away from a plateau. After an outline of the history of evolution strategies, it is explained how evolution strategies can be combined with the reinforcement learning, named reinforcement evolution strategies. The performance of proposed method will be estimated by comparison with conventional evolution strategies on several test problems.

• Journal of the Korea Society for Simulation
• /
• v.13 no.4
• /
• pp.17-29
• /
• 2004

The focus of this study is to represent a methodology of analysis for integrated logistic system by means of the Evolution Strategies and Data Envelopment Analysis(DEA). The integrated logistic system is composed of AS/RS (Automated Storages and Retrieval System), AGVs(Automated Guided Vehicle System) and Conveyor System. We design the simulation alternatives with choosing the qualitative critical factors for the each subsystem. Evolution Strategies is used to optimize the quantitative critical factors and responses of each alternative. DEA is applied to measure the efficiency of the alternatives in order to select the optimal operation efficiency scheme. The method of analysis which combines Evolution Strategies with DEA can be used to analyze the qualitative and quantitative critical factors in the integrated logistic systems.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.11a
• /
• pp.271-274
• /
• 1997

This paper presents the optimization technique for design of a CMAC network by using an evolution strategies(ES). The proposed technique is designed to find the optimal parameters of a CMAC network, which can minimize the learning error between the desired output and the CMAC network's as well as the number of memory used in the CMAC network. Computer simulations demonstrate the effectiveness of the proposed design method.

• Structural Engineering and Mechanics
• /
• v.34 no.5
• /
• pp.597-609
• /
• 2010

A problem formulation and solution methodology for design optimization of laminated thin-walled composite beams of generic section is presented. Objective functions and constraint equations are given in the form of beam stiffness. For two different problems one for open section and the other for closed section, the objective function considered is bending stiffness about x-axis. Depending upon the case, one can consider bending, torsional and axial stiffnesses. The different search and optimization algorithm, known as Evolution Strategies (ES) has been applied to find the optimal fibre orientation of composite laminates. A multi-level optimization approach is also implemented by narrowing down the size of search space for individual design variables in each successive level of optimization process. The numerical results presented demonstrate the computational advantage of the proposed method "Evolution strategies" which become pronounced to solve optimization of thin-walled composite beams of generic section.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2002.05a
• /
• pp.651-659
• /
• 2002

Component Mode Synthesis (CMS) is a dynamic substructuring technique to get an approximate eigensolutions of large degree-of-freedom structures divisible into several components. But, In practice. most of large structures are modeled by different teams of engineers. and their respective finite element models often require different mesh resolutions. As a result, the finite element substructure models can be non-conforming and/or incompatible. In this work, A hybrid version of component mode synthesis using a localized lagrange multiplier to treat the non-conforming mesh problem was derived. Evolution Strategies (ESs) is a stochastic numerical optimization technique and has shown a robust performance for solving deterministic problems. An ESs conducts its search by processing a population of solutions for an optimization problem based on principles from natural evolution. An optimization example for raising the first natural frequency of a plate structure using beam stiffeners was presented using hybrid component mode synthesis and robust evolution strategies (RES) optimization technique. In the example. the design variables are the positions and lengths of beam stiffeners.

• Structural Engineering and Mechanics
• /
• v.26 no.2
• /
• pp.191-210
• /
• 2007

The three different approaches (reformulations) of evolution strategies (ESs) have been proposed in the literature as extensions of the technique for solving discrete problems. This study implements an extensive research on application, evaluation and comparison of them in discrete optimum design of steel frames. A unified formulation is first developed to explain these approaches, so that differences and similarities between their inherent search mechanisms can clearly be identified. Two examples from practical design of steel frames are studied next to measure their performances in locating the optimum. Extensive numerical experimentations are performed in both examples to facilitate a statistical analysis of their convergence characteristics. The results obtained are presented in the histograms demonstrating the distribution of the best designs located by each approach. In addition, an average improvement of the best design during the course of evolution is plotted in each case to compare their relative convergence rates.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.9 no.6
• /
• pp.650-656
• /
• 1999

In this paper, we study on the convergence of the evolution strategies by introducing the Lamarckian evolution and the Baldwin effect, and propose a random local searching and a reinforcement local searching methods. In the random local searching method some neighbors generated randomly from each individual are med without any other information, but in the reinforcement local searching method the previous results of the local search are reflected on the current local search. From the viewpoint of the purpose of the local search it is suitable that we try all the neighbors of the best individual and then search the neighbors of the best one of them repeatedly. Since the reinforcement local searching method based on the Lamarckian evolution and Baldwin effect does not search neighbors randomly, but searches the neighbors in the direction of the better fitness, it has advantages of fast convergence and an improvement on the global searching capability. In other words the performance of the evolution strategies is improved by introducing the learning, reinforcement local search, into the evolution. We study on the learning effect on evolution strategies by applying the proposed method to various function optimization problems.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.165.1-165
• /
• 2001

The evolution strategy is a good evolutionary algorithm to find the global optimum of a real-valued function. Since many engineering problems can be formulated as real valued function optimization, the evolution strategy is frequently employed in engineering fields. However, in many engineering optimization problems, an optimization parameter is often restricted in the bounded region between two specified values, the minimum and the maximum limit, respectively. Since an offspring individual is generated randomly around a parent individual during mutation process of the evolution strategy, an individual outside the search region can be generated even if the parent is inside the search region. This paper proposes two new boundary-handling techniques for evolution strategies. One is the ...

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.10a
• /
• pp.420-426
• /
• 1998

Evolutionary algorithms are probabilistic optimization algorithms based on the model of natural evolution. Recently the efforts to improve the performance of evolutionary algorithms have been made extensively. In this paper, we introduce the research for improving the convergence rate and search faculty of evolution algorithms by using reinforcement learning. After providing an introduction to evolution algorithms and reinforcement learning, we present adaptive genetic algorithms, reinforcement genetic programming, and reinforcement evolution strategies which are combined with reinforcement learning. Adaptive genetic algorithms generate mutation probabilities of each locus by interacting with the environment according to reinforcement learning. Reinforcement genetic programming executes crossover and mutation operations based on reinforcement and inhibition mechanism of reinforcement learning. Reinforcement evolution strategies use the variances of fitness occurred by mutation to make the reinforcement signals which estimate and control the step length.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.559-564
• /
• 2003

Recently, particle filters have attracted attentions for nonlinear state estimation. They evaluate a posterior probability distribution of the state variable based on observations in simulation using so-called importance sampling. However, degeneracy phenomena in the importance weights deteriorate the filter performance. A new filter, Evolution Strategies Based Particle Filter, is proposed to circumvent this difficulty and to improve the performance. Numerical simulation results illustrate the applicability of the proposed idea.