• Journal of the Korean Institute of Intelligent Systems
• /
• v.8 no.6
• /
• pp.58-69
• /
• 1998

This paper introduces a novel neuro-fuzzy system based on the polynomial fuzzy neural network(PFNN) architecture. The PFNN consists of a set of if-then rules with appropriate membership functions whose parameters are optimized via a hybrid genetic algorithm. A polynomial neural network is employed in the defuzzification scheme to improve output performance and to select appropriate rules. A performance criterion for model selection, based on the Group Method of DAta Handling is defined to overcome the overfitting problem in the modeling procedure. The hybrid genetic optimization method, which combines a genetic algorithm and the Simplex method, is developed to increase performance even if the length of a chromosome is reduced. A novel coding scheme is presented to describe fuzzy systems for a dynamic search rang in th GA. For a performance assessment of the PFNN inference system, three well-known problems are used for comparison with other methods. The results of these comparisons show that the PFNN inference system outperforms the other methods while it exhibits exceptional robustness characteristics.

• Journal of the Korean Society of Systems Engineering
• /
• v.12 no.2
• /
• pp.115-125
• /
• 2016

An improved RRS (Reactor Regulating System) logic is implemented in this work using systems engineering approach along with GA (Genetic Algorithm) deemed as providing an optimal solution to a given system. The current system works desirably and has been contributed to the safe and stable NPP operation. However, during the ascent and decent section of the reactor power, the RRS output reveals a relatively high steady state error and the output also carries a considerable level of overshoot. In an attempt to consolidate conservatism and minimize the error, this research proposes applying genetic algorithm to RRS and suggests reconfiguring the system. Prior to the use of GA, reverse-engineering is implemented to build a Simulink-based RRS model and re-engineering is followed to apply the GA and to produce a newly-configured RRS generating an output that has a reduced steady state error and diminished overshoot level.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.8 no.5
• /
• pp.61-71
• /
• 1998

Genetic Algorithm(GA) is a method of approaching optimization problems by modeling and simulating the biological evolution. GA needs large time-consuming, so ti had better do on a parallel computer architecture. Our proposed system has a VLSI-oriented interconnection network, which is constructed from a viewpoint of fractal geometry, so that self-similarity is considered in its configuration. The approach to Parallel Genetic Algorithm(PGA) on our proposed system is explained, and then, we construct the classifier system such that the set of samples is classified into weveral classes based on the features of each sample. In the process of designing the classifier system, We have applied PGA to the Traveling Salesman Problem and classified the sample set in the Euclidean space into several categories with a measure of the distance.

• Proceedings of the KAIS Fall Conference
• /
• 2006.05a
• /
• pp.316-319
• /
• 2006

Connectionist networks, also called neural networks, have been broadly applied to solve many different problems since McCulloch and Pitts had shown mathematically their information processing ability in 1943. In this thesis, we present a genetic neuro-control scheme for nonlinear systems. Our method is different from those using supervised learning algorithms, such as the backpropagation (BP) algorithm, that needs training information in each step. The contributions of this thesis are the new approach to constructing neural network architecture and its training.

• Journal of the Institute of Convergence Signal Processing
• /
• v.11 no.3
• /
• pp.250-254
• /
• 2010

As the number of input and fuzzy set of a fuzzy system increase, the size of the rule base increases exponentially and becomes unmanageable (curse of dimensionality). In this paper, tree architectures of fuzzy controller (TAFC) is proposed to overcome the curse of dimensionality problem occurring in the design of fuzzy controller. TAFC is constructed with the aid of AND and OR fuzzy neurons. TAFC can guarantee reduced size of rule base with reasonable performance. For the development of TAFC, genetic algorithm constructs the binary tree structure by optimally selecting the nodes and leaves, and then random signal-based learning further refines the binary connections (two-step optimization). An inverted pendulum system is considered to verify the effectiveness of the proposed method by simulation.

• Proceedings of the KIEE Conference
• /
• 2002.07d
• /
• pp.2486-2489
• /
• 2002

Pool of the central nervous system. In this thesis, we present a genetic DNU-control scheme for unknown nonlinear systems. Our method is different from those using supervised learning algorithms, such as the backpropagation (BP) algorithm, that needs training information in each step. The contributions of this thesis are the new approach to constructing neural network architecture and its training.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.11 no.2
• /
• pp.809-818
• /
• 2019

The current trend in shipyard industry is to reduce the weight of ships to support the reduction of CO₂ emissions. In this study, the stiffened plate was optimized that is used for building most of the ship-structure. Further, this study proposed the hybrid Genetic Algorithm (GA) technique, which combines a genetic algorithm and subsequent optimization methods. The design variables included the number and type of stiffeners, stiffener spacing, and plate thickness. The number and type of stiffeners are discrete design variables that were optimized using the genetic algorithm. The stiffener spacing and plate thickness are continuous design variables that were determined by subsequent optimization. The plate deformation was classified into global and local displacement, resulting in accurate estimations of the maximum displacement. The optimization result showed that the proposed hybrid GA is effective for obtaining optimal solutions, for all the design variables.

• Journal of the Society of Naval Architects of Korea
• /
• v.43 no.2 s.146
• /
• pp.156-163
• /
• 2006

The accuracy of estimates of hydrodynamic coefficients can be increased by using sensitivity-maximizing inputs. In this study, sensitivity-maximizing actuator commands of a submersible, which are sequences of bang-bang type commands, were obtained using Genetic Algorithm (GA) optimization technique. By comparing the total sensitivity values, deduced actuator inputs were found to be superior to the other sea trials. Based on the sensitivity distribution of conventional sea trials and sensitivity distribution results through deduced input scenario a review of submersibles' manoeuvring equations of motion was conducted .

• Proceedings of the KIEE Conference
• /
• 1996.07b
• /
• pp.1356-1358
• /
• 1996

Hopfield's suggestion of a neural network model for associative memory aroused the interest of many scientists and led to efforts of mathematical analyses. But the Hopfield Network has several disadvantages such as spurious states and capacity limitation. In that sense many scientists and engineers are trying to use a new optimization algorithm called genetic algorithm. But it is hard to use this algorithm in Hopfileld Network because of the fixed architecture. In this paper we introduce another method to determine the weight of Hopfield type network using Genetic Algorithm.

• Ocean Systems Engineering
• /
• v.2 no.4
• /
• pp.271-287
• /
• 2012

Due to its large size, a ship is first divided into scores of blocks and then each block is constructed through various shops, such as the assembly shop, the painting shop, and the outfitting shop. However, each block may not be directly moved to the next shop and may be temporarily laid at a block stockyard because the working time in each shop is different from each other. If blocks are laid at the block stockyard without any planning, the rearrangement of the blocks by a transporter is required because the blocks have the different in and out time. In this study, a block layout method based on the genetic algorithm was proposed in order to minimize the rearrangement of the blocks in the block stockyard. To evaluate the applicability of the proposed method, it was applied to simple layout problems of the block stockyard. The result shows that the proposed method can yield a block layout that minimizes the total relocation cost of moving obstacle blocks in the block stockyard.