• Journal of Industrial Technology
• /
• v.24 no.A
• /
• pp.99-107
• /
• 2004

This paper considers reinforcement learning control with the self-organizing map. Reinforcement learning uses the observable states of objective system and signals from interaction of the system and the environments as input data. For fast learning in neural network training, it is necessary to reduce learning data. In this paper, we use the self-organizing map to parition the observable states. Partitioning states reduces the number of learning data which is used for training neural networks. And neural dynamic programming design method is used for the controller. For evaluating the designed reinforcement learning controller, an inverted pendulum of the cart system is simulated. The designed controller is composed of serial connection of self-organizing map and two Multi-layer Feed-Forward Neural Networks.

• 제어로봇시스템학회:학술대회논문집
• /
• 1989.10a
• /
• pp.635-642
• /
• 1989

We formulated the multi-robot trajectory problem into a series of NLP problem, each of which is that of finding the optimal tip positions of the robots for the next time step. The NLP problem is composed of an objective function and three constraints, namely: a) Joint position limits, b) Joint velocity limits, and c) Collision-avoidance constraints. By solving a series of NLP problem, optimally coordinated trajectories can be determined without requiring any prior path information. This is a novel departure from the previous approach in which either all paths or at least one path is assumed to be given. Practical application of the developed method is for optimal synthesis of multiple robot trajectories in off-line. To test the validity and effectiveness of the method, numerical examples are illustrated.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.26 no.9
• /
• pp.1779-1786
• /
• 2002

When the topology optimization is applied to the design of compliant mechanism, unexpected displacements of input and output port are generated since the displacement control is not included in the formulation. To devise a more precise mechanism, displacement constraint is formulated using the mutual potential energy concept and added to multi-objective function defined with flexibility and stiffness of a structure. The optimization problem is resolved by using Finite Element Method(FEM) and Sequential Linear Programming(SLP). Design examples of compliant mechanism with displacement constraint are presented to validate the proposed design method.

• Proceedings of the KSME Conference
• /
• 2003.04a
• /
• pp.1441-1446
• /
• 2003

The design optimization of the plate heat exchanger with staggered pin arrays for a fixed volume is performed numerically. The flow and thermal fields are assumed to be a streamwise-periodic flow and heat transfer with constant wall temperature and they are solved by using the finite volume method. The optimization is carried out by using the sequential linear programming (SLP) method and the weighting method is used for solving the multi-objective problem. The results show that the optimal design variables for the weighting coefficient of 0.5 are as follows; S=6.497mm, P=5.496mm, $D_1=0.689mm$, and $D_2=2.396mm$. The Pareto optimal solutions are also presented.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2003.05a
• /
• pp.744-748
• /
• 2003

Developing an interactive satellite multi­media system, such as a digital video broadcasting (DVB) return channel via satellite (RCS) system, is gaining popularity over the world To accommodate the increasing traffic demand we are motivated to investigate an alternative for improving return channel utilization We develop an efficient method for optimal packet scheduling in an interactive satellite multimedia system using hybrid CDMA/TDMA channels. We formulate the timeslot-code assignment problem as a binary integer programming (BIP) problem, where the throughput maximization is the objective, and decompose this BIP problem into two sub-problems for the purpose of solution efficiency. With this decomposition, we promote the computational efficiency in finding the optimal solution of the original BIP problem Since 2001, ETRI has been involved in a development project where we have successfully completed an initial system integration test on broadband mobile Internet access via Ku-band channels using the proposed resource allocation algorithm.

• Journal of Mechanical Science and Technology
• /
• v.17 no.12
• /
• pp.1938-1948
• /
• 2003

This paper represents that an enhanced genetic algorithm (EGA) is applied to optimal design of a squeeze film damper (SFD) to minimize the maximum transmitted load between the bearing and foundation in the operational speed range. A general genetic algorithm (GA) is well known as a useful global optimization technique for complex and nonlinear optimization problems. The EGA consists of the GA to optimize multi-modal functions and the simplex method to search intensively the candidate solutions by the GA for optimal solutions. The performance of the EGA with a benchmark function is compared to them by the IGA (Immune-Genetic Algorithm) and SQP (Sequential Quadratic Programming). The radius, length and radial clearance of the SFD are defined as the design parameters. The objective function is the minimization of a maximum transmitted load of a flexible rotor system with the nonlinear SFDs in the operating speed range. The effectiveness of the EGA for the optimal design of the SFD is discussed from a numerical example.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.23 no.3
• /
• pp.27-36
• /
• 1998

This work has improved the method of Kim and Gal's (6) in that of requiring less response of the DM(decision maker) and ease of reply. The underlying notion is the MCDC(maximally changeable dominance cone) for describing all efficient solutions under the particular preference structure. According to the DM's partial preference expression, enlarging the MCDC is achieved, which results in reducing the solutions needed to take into consideration. The cone generators corresponding to the DM's response are added to the MCDC, which results the MCDC is enlarged. Adopting the scheme of pairwise comparison as a means of acquiring preference attitude, an improved interactive method is proposed. And also, a scheme of choosing a reference point is suggested to achieve the computational efficiency.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.2
• /
• pp.355-361
• /
• 2010

This paper addresses a mathematical approach to decision making in a real-world material distribution situation. The problem is characterized by a low-volume and highly-varied mix of products, therefore there is a lot of material movement between the facilities. This study focuses especially on the transportation scheduler with a tool that can be used to quantitatively analyze the volume of material moved, the type of truck to be used, production schedules, and due dates. In this research, we have developed a mixed integer programming problem using the minimum cost, multiperiod, multi-commodity network flow approach that minimizes the overall material movement costs. The results suggest that the optimization approach provides a set of feasible solution routes with the objective of reducing the overall fleet cost.

• Journal of the military operations research society of Korea
• /
• v.16 no.1
• /
• pp.130-141
• /
• 1990

This paper presents an allocation model of test times of software modules. An evaluation measure of modules which is based on the data flow path between modules is used at test phase. An evaluation measure is expressed by the possible path number of data flow for each module, which can be interpreted as an importance of individual module. Three criteria : module test time, module reliability, and system reliability is considered in this model. Multi-objective programming, hence, is used to solve this model.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.5B
• /
• pp.295-303
• /
• 2007

An algorithm is proposed to seek a local optimal solution of the network utility maximization problem in a wireless mesh network, where the architecture being considered is an infrastructure/backbone wireless mesh network. The objective is to achieve proportional fairness amongst the end-to-end flows in wireless mesh networks. In order to establish the communication constraints of the flow rates in the network utility maximization problem, we have presented necessary and sufficient conditions for the achievability of the flow rates. Since wireless mesh networks are generally considered as a type of ad hoc networks, similarly as in wireless multi-hop network, the network utility maximization problem in wireless mesh network is a nonlinear nonconvex programming problem. Besides, the gateway/bridge functionalities in mesh routers enable the integration of wireless mesh networks with various existing wireless networks. Thus, the rate optimization problem in wireless mesh networks is more complex than in wireless multi-hop networks.