• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.40 no.1
• /
• pp.95-104
• /
• 2017

This paper deals with solution methods for discrete and multi-valued optimization problems. The objective function of the problem incorporates noise effects generated in case that fitness evaluation is accomplished by computer based experiments such as Monte Carlo simulation or discrete event simulation. Meta heuristics including Genetic Algorithm (GA) and Discrete Particle Swarm Optimization (DPSO) can be used to solve these simulation based multi-valued optimization problems. In applying these population based meta heuristics to simulation based optimization problem, samples size to estimate the expected fitness value of a solution and population (particle) size in a generation (step) should be carefully determined to obtain reliable solutions. Under realistic environment with restriction on available computation time, there exists trade-off between these values. In this paper, the effects of sample and population sizes are analyzed under well-known multi-modal and multi-dimensional test functions with randomly generated noise effects. From the experimental results, it is shown that the performance of DPSO is superior to that of GA. While appropriate determination of population sizes is more important than sample size in GA, appropriate determination of sample size is more important than particle size in DPSO. Especially in DPSO, the solution quality under increasing sample sizes with steps is inferior to constant or decreasing sample sizes with steps. Furthermore, the performance of DPSO is improved when OCBA (Optimal Computing Budget Allocation) is incorporated in selecting the best particle in each step. In applying OCBA in DPSO, smaller value of incremental sample size is preferred to obtain better solutions.

• Journal of Electrical Engineering and Technology
• /
• v.8 no.1
• /
• pp.53-63
• /
• 2013

To guide the decision making of the expert engineer specialized in power system operation and control; the practical OPF solution should take in consideration the critical situation due to severe loading conditions and fault in power system. Differential Evolution (DE) is one of the best Evolutionary Algorithms (EA) to solve real valued optimization problems. This paper presents simple Differential Evolution (DE) Optimization algorithm to solving multi objective optimal power flow (OPF) in the power system with shunt FACTS devices considering voltage deviation, power losses, and power flow branch. The proposed approach is examined and tested on the standard IEEE-30Bus power system test with different objective functions at critical situations. In addition, the non smooth cost function due to the effect of valve point has been considered within the second practical network test (13 generating units). The simulation results are compared with those by the other recent techniques. From the different case studies, it is observed that the results demonstrate the potential of the proposed approach and show clearly its effectiveness to solve practical OPF under contingent operation states.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.599-602
• /
• 2003

A new technique is presented to construct predictive models of plasma etch processes. This was accomplished by combining a backpropagation neural network (BPNN) and a random generator (RC). The RG played a critical role to control neuron gradients in the hidden layer, The predictive model constructed in this way is referred to as a randomized BPNN (RG-BPNN). The proposed scheme was evaluated with a set of experimental plasma etch process data. The etch process was characterized by a 2$^3$ full factorial experiment. The etch responses modeled are 4, including aluminum (Al) etch rate, profile angle, Al selectivity, and do bias. Additional test data were prepared to evaluate model appropriateness. The performance of RC-BPNN was evaluated as a function of the number of hidden neurons and the range of gradient. for given range and hidden neurons, 100 sets of random neuron gradients were generated and among them one best set was selected for evaluation. Compared to the conventional BPNN, the proposed RC-BPNN demonstrated about 50％ improvements in all comparisons. This illustrates that the RG-BPNN of multi-valued gradients is an effective way to considerably improve the predictive ability of current BPNN of single-valued gradient.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.449-449
• /
• 2000

Many researchers have studied architectures for non-Neumann's computers because of escaping its bottleneck. To avoid the bottleneck, a neuron-based computer has been developed. The computer has only neurons and their connections, which are constructed of the learning. But still it has information processing facilities, and at the same time, it is like as a simplified brain to make inference; it is called "neuron-computer". No instructions are considered in any neural network usually; however, to complete complex processing on restricted computing resources, the processing must be reduced to primitive actions. Therefore, we introduce the instructions to the neuron-computer, in which the most important function is implications. There is an implication represented by binary-operators, but general implications for multi-value or fuzzy logics can't be done. Therefore, we need to use Lukasiewicz' operator at least. We investigated a neuron-computer having instructions for general implications. If we use the computer, the effective inferences base on multi-value logic is executed rapidly in a small logical unit.

• Journal of IKEEE
• /
• v.18 no.4
• /
• pp.593-602
• /
• 2014

One of the purposes of an artificial neural netowrk(ANNet) is to implement the largest number of functions as possible with the smallest number of nodes and layers. This paper presents a CoreNet which has a multi-leveled input value and a multi-leveled output value with a 2-layered ANNet, which is the basic structure of an ANNet. I have suggested an equation for calculating the capacity of the CoreNet, which has a p-leveled input and a q-leveled output, as $a_{p,q}={\frac{1}{2}}p(p-1)q^2-{\frac{1}{2}}(p-2)(3p-1)q+(p-1)(p-2)$. I've applied this CoreNet into the simulation model 1(5)-1(6), which has 5 levels of an input and 6 levels of an output with no hidden layers. The simulation result of this model gives, the maximum 219 convergences for the number of implementable functions using the cot(${\sqrt{x}}$) input leveling method. I have also shown that, the 27 functions are implementable by the calculation of weight values(w, ${\theta}$) with the multi-threshold lines in the weight space, which are diverged in the simulation results. Therefore the 246 functions are implementable in the 1(5)-1(6) model, and this coincides with the value from the above eqution $a_{5,6}(=246)$. I also show the implementable function numbering method in the weight space.

• The Mathematical Education
• /
• v.26 no.1
• /
• pp.41-45
• /
• 1987

In this paper, we give several fixed point theorems in a complete metric space for two multi-valued mappings commuting with two single-valued mappings. In fact, our main theorems show the existence of solutions of functional equations f($\chi$)=g($\chi$)$\in$S$\chi$∩T$\chi$ and $\chi$=f($\chi$)=g($\chi$)$\in$S$\chi$∩T$\chi$ under certain conditions. We also answer an open question proposed by Rhoades-Singh-Kulsherestha. Throughout this paper, let (X, d) be a complete metric space. We shall follow the following notations : CL(X) = {A; A is a nonempty closed subset of X}, CB(X)={A; A is a nonempty closed and founded subset of X}, C(X)={A; A is a nonempty compact subset of X}, For each A, B$\in$CL(X) and $\varepsilon$>0, N($\varepsilon$, A) = {$\chi$$\in$X; d($\chi$, ${\alpha}$) < $\varepsilon$ for some ${\alpha}$$\in$A}, E$\sub$A, B/={$\varepsilon$ > 0; A⊂N($\varepsilon$ B) and B⊂N($\varepsilon$, A)}, and (equation omitted). Then H is called the generalized Hausdorff distance function fot CL(X) induced by a metric d and H defined CB(X) is said to be the Hausdorff metric induced by d. D($\chi$, A) will denote the ordinary distance between $\chi$$\in$X and a nonempty subset A of X. Let R$\^$+/ and II$\^$+/ denote the sets of nonnegative real numbers and positive integers, respectively, and G the family of functions ${\Phi}$ from (R$\^$+/)$\^$s/ into R$\^$+/ satisfying the following conditions: (1) ${\Phi}$ is nondecreasing and upper semicontinuous in each coordinate variable, and (2) for each t>0, $\psi$(t)=max{$\psi$(t, 0, 0, t, t), ${\Phi}$(t, t, t, 2t, 0), ${\Phi}$(0, t, 0, 0, t)} $\psi$: R$\^$+/ \longrightarrow R$\^$+/ is a nondecreasing upper semicontinuous function from the right. Before sating and proving our main theorems, we give the following lemmas:

• Journal of Information Processing Systems
• /
• v.9 no.1
• /
• pp.89-102
• /
• 2013

The paper presents a fuzzy based impulse noise filter for both gray scale and color images. The proposed approach is based on the technique of boundary discriminative noise detection. The algorithm is a multi-step process comprising detection, filtering and color correction stages. The detection procedure classifies the pixels as corrupted and uncorrupted by computing decision boundaries, which are fuzzified to improve the outputs obtained. In the case of color images, a correction term is added by examining the interactions between the color components for further improvement. Quantitative and qualitative analysis, performed on standard gray scale and color image, shows improved performance of the proposed technique over existing state-of-the-art algorithms in terms of Peak Signal to Noise Ratio (PSNR) and color difference metrics. The analysis proves the applicability of the proposed algorithm to random valued impulse noise.

• Proceedings of the KIEE Conference
• /
• 1987.07b
• /
• pp.1195-1198
• /
• 1987

This paper develops algorithms of element generation, addition, multiplication and inversion based on GF($2^m$). Since these algorithms are implemented by general purpose computer, these are more efficient than the conventional algorithms(Table Lookup, Euclid's Algorithm) in each operation. It is also implied that they can be applied to not only the normally defined elements but the arbitrarily defined ones for constructing multi-valued logic function.

• Journal of IKEEE
• /
• v.15 no.4
• /
• pp.354-362
• /
• 2011

One of the unsolved problems in Artificial Neural Networks is related to the capacity of a neural network. This paper presents a CoreNet which has a multi-leveled input and a multi-leveled output as a 2-layered artificial neural network. I have suggested an equation for calculating the capacity of the CoreNet, which has a p-leveled input and a q-leveled output, as $a_{p,q}=\frac{1}{2}p(p-1)q^2-\frac{1}{2}(p-2)(3p-1)q+(p-1)(p-2)$. With an odd value of p and an even value of q, (p-1)(p-2)(q-2)/2 needs to be subtracted further from the above equation. The simulation model 1(3)-1(6) has 3 levels of an input and 6 levels of an output with no hidden layer. The simulation result of this model gives, out of 216 possible functions, 80 convergences for the number of implementable function using the cot(x) input leveling method. I have also shown that, from the simulation result, the two diverged functions become implementable by precalculating the weight values. The simulation result and the precalculation of the weight values give the same result as the above equation in the total number of implementable functions.

• 제어로봇시스템학회:학술대회논문집
• /
• 1999.10a
• /
• pp.141-144
• /
• 1999

The authors have recently developed a new method for identification of Volterra kernels of nonlinear systems by use of pseudorandom M-sequence and correlation technique. And it is shown that nonlinear systems which can be expressed by Volterra series expansion are well identified by use of this method. However, there exist many nonlinear systems which can not be expressed by Volterra series mathematically. A nonlinear system having backlash type nonliear element is one of those systems, since backlash type nonlinear element has multi-valued function between its input and output. Since Volterra kernel expression of nonlinear system is one of the most useful representations of non-linear dynamical systems, it is of interest how the method of Volterra kernel identification can be ar plied to such backlash type nonlinear system. The authors have investigated the effect of application of Volterra kernel identification to those non-linear systems which, accurately speaking, is difficult to express by use of Volterra kernel expression. A pseudorandom M-sequence is applied to a nonlinear backlash-type system, and the crosscorrelation function is measured and Volterra kernels are obtained. The comparison of actual output and the estimated output by use of measured Volterra kernels show that we can still use Volterra kernel representation for those backlash-type nonlinear systems.