• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1990.04a
• /
• pp.59-64
• /
• 1990

The cable structures undergo large deformation because of its highly flexibility. Therefore, we must take account of its geometric nonlinearity before analysis and find the equiribrated shape of cable structures. To solve these problems, a numerical procedures included nonlinear near theory which is applicable to general cable net, flexible transmission lines and suspended cable roofs, are presented in this paper. Now, this procedures are devided two parts : the one is to obtain the equibrated shape and stress of the cable structures applied uniform load by flexibility iteration method, the other is to analysis the equibrated structures subjected to nodal external forces by nonlinear finite element mothed. Its accuracy and efficiency are found to be comparable to some of other method and, in some aspect, it is mere applicable to cable structures.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.20 no.43
• /
• pp.299-319
• /
• 1997

Since a convergence strategy in interactive procedure is the origin of a lot of the restrictions and inconvenience of many methods, a new flexible convergence strategy is necessary for developing that is applicable for interactive MOLP as well as interactive multiple objective integer programming(MOIP). A method for computing the volume of the interval defined subsets of weighting vector space is developed using Sobol's LP$_{i}$, Generator. Since the interactive procedure requires a number of iterations, with a fixed number of solutions presented per iteration, a convergence strategy is needed to control the rate at which the interactive procedure converges to a final solution. Thus, in this paper, a new convergency strategy is developed.

• Journal of Electrical Engineering and Technology
• /
• v.3 no.2
• /
• pp.207-212
• /
• 2008

This paper presents a robust and computationally efficient optimal design algorithm for electromagnetic devices by combining an adaptive response surface approximation of the objective function and($1+{\lambda}$) evolution strategy. In the adaptive response surface approximation, the design space is successively reduced with the iteration, and Pareto-optimal sampling points are generated by using Latin hypercube design with the Max Distance and Min Distance criteria. The proposed algorithm is applied to an analytic example and TEAM problem 22, and its robustness and computational efficiency are investigated.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.50 no.10
• /
• pp.464-472
• /
• 2001

This study uses distributed parameter systems as the spatial discretization technique, modelling in lumped parameter systems, and applies fast WALSH transform and the Picard's iteration method to high order partial differential equations and matrix partial differential equations. This thesis presents a new algorithm which usefully exercises the optimal control in the distributed parameter systems. In exercising optimal control of distributed parameter systems, excellent consequences are found without using the existing decentralized control or hierarchical control method. This study will help apply to linear time-varying systems and non-linear systems. Further research on algorithm will be required to solve the problems of convergence in case of numerous applicable intervals.

• Proceedings of the KIEE Conference
• /
• 1999.07g
• /
• pp.3199-3201
• /
• 1999

Convolutional coding with Viterbi decoding is known as a powerful method for forward error correction among many kinds of channel coding methods. This paper presents a soft decision Viterbi decoder which has systolic array trace-back architecture[1]. Soft decision is known as more effective method than hard decision and most of digital communication systems use soft decision. The advantage of using a systolic array decoder is that the trace-back operation can be accomplished continuously in an array of registers in a pipe-line fashion, instead of waiting for the entire trace-back procedure to be completed at each iteration. Therefore it may be suitable for faster communication system. We described operations of each module of the decoder and showed results of the logic synthesis and functional simulation.

• Journal of the korean Society of Automotive Engineers
• /
• v.8 no.1
• /
• pp.52-61
• /
• 1986

An iterative viscous-potential flow procedure has been developed and used to predict aerodynamic characteristics of automobiles in ground proximity. The method is capable of predicting the effects of separated flows. The viscous-potential flow iteration procedure provides the connection between potential flow, boundary layer and wake modules. The separated wake is modeled in the potential flow analysis by thin sheets across which exists a jump in velocity potential. The ground effect is properly accounted for by placing a body image in the potential flow calculation. The agreement between theory and experiment is good and, thus, demonstrates that the method can be used in the preliminary design stage.

• Proceedings of the KIEE Conference
• /
• 2004.11b
• /
• pp.24-27
• /
• 2004

This paper proposes a new algorithm of underground cable fault location based on the analysis of distributed parameter circuit. The proposed method firstly makes voltage and current equations for each of cores and sheathes respectively, and then establishes an equation of the fault distance according to the analysis of the fault conditions. Finally the solution of this equation is calculated by Newton-Raphson iteration method. The effectiveness of this proposed algorithm has been proven through PSCAD/EMTDC simulations.

• Proceedings of the KIEE Conference
• /
• 2001.07d
• /
• pp.2084-2086
• /
• 2001

This study uses distributed parameter systems as the spatial discretization technique, modelling in lumped parameter systems, and applies orthogonal transform and the Picard's iteration method to high order partial differential equations and matrix partial differential equations. In exercising optimal control of distributed parameter systems, excellent consequences are found without using the existing decentralized control or hierarchical control method.

• Communications for Statistical Applications and Methods
• /
• v.12 no.2
• /
• pp.285-294
• /
• 2005

This paper deals with the comparison of parameter estimation methods in a 3-parameter Kappa distribution which is sometimes used in flood frequency analysis. Method of moment estimation(MME), L-moment estimation(L-ME), and maximum likelihood estimation(MLE) are applied to estimate three parameters. The performance of these methods are compared by Monte-carlo simulations. Especially for computing MME and L-ME, three dimensional nonlinear equations are simplified to one dimensional equation which is calculated by the Newton-Raphson iteration under constraint. Based on the criterion of the mean squared error, L-ME (or MME) is recommended to use for small sample size( n$\le$100) while MLE is good for large sample size.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.22.5-22
• /
• 2001

Robust control of a GFR composite beam with a distributed PVDF sensor and piezo-ceramic actuator is presented En this paper. Modal analysis method and modal coordinates are introduced to obtain the state educations of the structural system. 1st and 2nd natural frequencies are considered In the modeling, because robust control theory which is robustness to structured uncertainty is adopted to suppress the vibration. If the controllers designed by H$\^$$\infty$/ theory do not satisfy control performance, it is improved by ${\mu}$-synthesis method with D-K Iteration so that the ${\mu}$-controller based on the structured singular value satisfies the nominal performance and robust performance.