• Proceedings of the KIEE Conference
• /
• 1994.11a
• /
• pp.408-410
• /
• 1994

This Paper Presents a learning strategy for neural networks based on genetic algorithms and evolution strategies. Genetic algorithms and evolution strategies are used to train weights of feedforward neural network to solve problems faster than neural network, especially backpropagation. Simulations are performed exclusive-OR problem, full-adder problem, sine function generator to demonstrate the effectiveness of neural-GA-ES.

• Proceedings of the KIEE Conference
• /
• 1994.11a
• /
• pp.359-361
• /
• 1994

This paper propose a Neural - GA-ES DC motor speed controller. The purpose is to achieve accurate trajectory control of the motor speed. A feedforward neural network structure is used for the controller. Genetic algorithm and evolution strategy is used for learning controller. Simulations are performed to demonstrate the effectiveness of proposed genetic algorithm and evolution strategy with neural structure.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.41.4-41
• /
• 2001

In this paper, we propose a LQR(Linear Quadratic Regulator) controller design for the active suspension using two-degree-of-freedom quarter-car model. We can improve the inherent suspension problem, the tradeoff between ride quality and suspension travel by selecting appropriate weights in the LQR-objective function. Because any definite rules for selecting weights do not exist, we replace the designer´s trial and error with the optimization-algorithm, ES(Evolution Strategy). Using the ES, we can find the proper control gains for selected frequencies, which have major effects on the vibrations of the vehicle´s state variables.

• Proceedings of the KSR Conference
• /
• 2009.05b
• /
• pp.165-170
• /
• 2009

The Korean high-speed rail network has been made a remarkable growth recording the world's 4th rank passenger transportation scale during last 5 years after opening the revenue service of Kyoung-bu high-speed line in 2004. However, in spite of it's outstanding growth, Korean Rail technology should meet a demand of intensive technology development in order to prepare a severe competition with an advance parties of worldwide high-speed rail technology. In this paper, the characteristics and new trend of world's rail industry was reviewed based on the most recent statistics and papers of worldwide rail network from UIC and WCR32008. In conclusion, new R&D strategy with choice an concentration for the Korean high-speed rail industry was suggested based on an analysis on the trend of evolution of the state of the art technologies in worldwide high-speed rail system.

• Journal of the Korean Society for Railway
• /
• v.10 no.3 s.40
• /
• pp.306-312
• /
• 2007

This paper presents the optimal design method of the train repair facility based on the simulation analysis. The train is divided into the power car, motorized car and passenger car for the simulation process analysis and train repair facility is composed of each subsystems such as a blast, dry and wash workshop. In simulation analysis, we consider the critical (dependent) factors and design (independent) factors for the optimal design. Therefore, a simulation optimization uses Evolution Strategy (ES) in order to find the optimal design factors. Experimental results indicate that simulation design factors are sufficient to satisfy the conditions of dependent variables. The proposed analysis method demonstrates that simulation design factors determined by the simulation optimization are appropriate for real design factors in a real situation and the accuracy and confidence for the simulation results are increased.

• Journal of Magnetics
• /
• v.18 no.1
• /
• pp.65-73
• /
• 2013

In this paper, design and performance analysis of robust and inexpensive permanent magnet-assisted synchronous reluctance generators (PMa-SynRG) for tactical and commercial generator sets is studied. More specifically, the optimal design approach is investigated for minimizing volume and maximizing performance for the portable generator. In order to find optimized PMa-SynRG, stator winding configurations and rotor structures are analyzed using the lumped parameter model (LPM). After comparisons of stator windings and rotor structure by LPM, the selected stator winding and rotor structure are optimized using a differential evolution strategy (DES). Finally, output performances are verified by finite element analysis (FEA) and experimental tests. This design process is developed for the optimized design of PMa-SynRG to achieve minimum magnet and machine volume as well as maximum efficiency simultaneously.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.12
• /
• pp.268-277
• /
• 2020

This study emphasizes that the survival strategy of universities should be a co-evolution strategy based on ecological thinking. Therefore, the purpose of the research is to present a theoretical framework for dividing the university innovation ecosystem into four stages and building a co-evolution strategy for each step, as universities play a prominent role in regional innovation ecosystems. Thus, our research method focused on literature research, and the theoretical framework for the university innovation ecosystem used Moore's Enterprise Ecosystem Model (1996). The university's ecological innovation strategy is divided into four stages of development, and a step-by-step co-evolution strategy is presented. Findings are summarized as follows. The pioneering stage involves the creation of values of the university-led innovation ecosystem. The expansion stage focuses on the establishment of critical mass. The authority stage covers maintaining authority and bargaining power. The renewal stage features continuous performance improvement. In particular, this theoretical model of the university-regional innovation ecosystem is meaningful in that it provides a theoretical basis for enhancing the effectiveness of government financial support projects, and for individual universities, it provides a framework for strategies suitable for their ecosystem building process.

• Fashion Information and Technology
• /
• v.10
• /
• pp.48-55
• /
• 2013

• Progress in Superconductivity and Cryogenics
• /
• v.17 no.2
• /
• pp.36-40
• /
• 2015

Electron cyclotron resonance (ECR) ion source is an essential component of heavy-ion accelerator. For a given design, the intensities of the highly charged ion beams extracted from the source can be increased by enlarging the physical volume of ECR zone [1]. Several models for ECR ion source were and will be constructed depending on their operating conditions [2-4]. In this paper three simulation models with 3, 4 and 6 solenoid system were built, but it's not considered anything else except the number of coils. Two groups of optimization analysis are presented, and the evolution strategy (ES) is adopted as an optimization tool which is a technique based on the ideas of mutation, adaptation and annealing [5]. In this research, the volume of ECR zone was calculated approximately, and optimized designs for ECR solenoid magnet system were presented. Firstly it is better to make the volume of ECR zone large to increase the intensity of ion beam under the specific confinement field conditions. At the same time the total volume of superconducting solenoids must be decreased to save material. By considering the volume of ECR zone and the total length of solenoids in each model with different number of coils, the 6 solenoid system represented the highest coil performance. By the way, a certain case, ECR zone volume itself can be essential than the cost. So the maximum ECR zone volume for each solenoid magnet system was calculated respectively with the same size of the plasma chamber and the total magnet space. By comparing the volume of ECR zone, the 6 solenoid system can be also made with the maximum ECR zone volume.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.39 no.1
• /
• pp.23-34
• /
• 2002

This paper proposes a new neuro-fuzzy system that is fast trained by evolution strategy and describes application results of the proposed system to nonlinear system identification to show its usefulness. As training methods of neuro-fuzzy systems, modified error back-propagation algorithms and genetic algorithms have been used so far. However, the former has some drawbacks such as long training time, falling to local optimum, and experimental selecting of learning rates and the latter has difficulty in precise searching solutions because genetic algorithms represents solutions as genotype individuals. The evolution strategy we used can do precise search because its individuals are represented as phenotype real values, it seldom falls into a local optimum, and its training speed is faster than error back-propagation algorithms. We apply our neuro-fuzzy systems to nonlinear system identification. It was found from experiments that training speed is fast and the training results were considerably good.