Metamodels using response surface methodology (RSM) are used for the optimality analysis of linear programming (LP). They have the form of a simple polynomial, and predict the optimal objective function value of an LP for various levels of the constraints. The metamodelling techniques for optimality analysis of LP can be applied to large-scale LP models. What is needed is some large-scale application of the techniques to verify how accurate they are. In this paper, we plan to use the large scale LP model, strategic transport optimal routing model (STORM). The developed metamodels of the large scale LP can provide some useful information.

A number of combat simulation models are scattered and the analytic solution approaches have experienced very difficult computational efforts. Today´s computer communication technology let people to do many unrealistic things possible and the use of those technologies is becoming increasingly prevalent throughout the military operation. Both DIS and ADS are welled defined computer aided military simulations. This study discusses a simulation of stochastic combat network modeling through Internet space. We have developed two separate simulation models, one for clients and another for server, and validated for conducting studies with these models. The object-oriented design was necessary to define the system entities and their relationship, to partition functionality into system entities, and to transform functional metrics into realizations derived from system component behaviors. Heterogeneous forces for each side are assumed at any battle node. The time trajectories for mean number of survivors and combat history at each node, some important combat measures, and relative difference computations between models were made. We observe and may conclude that the differences exit and some of these are significant based on a limited number of experiments.

This paper deals with the problem of Customs trying to stop a Smuggler attempting to ship a cargo of perishable contraband(or information) across a strait within the limited period. The papers concerning these fields referred by Thomas, Baston et al, and Garnaev, have treated games dealing with one, two, and three patrol boats against one smuggler ship respectively. In this paper the author proposed a game model dealing with N patrol ships against one smuggler ship with the N boats being identical. The strategies for the gamers where games are conducted between them are shown in tables 3.3-1 and the game values thereof are in table 3.2-3. The author has exhaustively filled up the assumptions and conditions in the paper to justify the above-mentioned results so that they may not loose the generality.

Cost estimation has posed a significant challenge to estimators, planners, and managers in both government and military. Considerable historical evidence shows that accurate cost estimation has been difficult to achieve across a wide range of projects, including weapon systems. This paper introduces new cost estimating concept, CAIV(Cost As an Independent Variable) and a cost estimating case study using PRICE model, computer aided parametric estimating models(CAPE) for K1 tank cost estimate. CAIV concept is to set realistic but aggressive cost objectives easily in each acquisition program and to achieve cost, schedule, and performance objectives considering various managing risks with a project manager and industry teams. The Price model is one of computer aided cost estimating models and widely used in U.S. defense system analysis as a tool for CAIV. We analyze theories, inputs, outputs of the PRICE model and present a case study for K1 tank to estimate costs in requirement and concept phase, program and budgeting phase, and life cycle phase. Finally we obtain results that the Price model can be used in various phases of PPBEES depending upon available data and time.

Taguchi´s parameter design is to determine the optimal settings of design parameters of a product or a process such that the characteristics of a product exhibit small variabilities around their targer values. His analysis of the problem has focused only on a single characteristic or response. However the quality of most products is seldom defined by a characteristic, and is rather the composite of a great number of characteristics which are often interrelated and nearly always measured in a variety of units. The critical problem in dealing with multiple characteristics is how to compromise the conflict among the selected levels of the design parameters for each individual characteristic. In this paper, Methodology using SN ratio optimized by unvariate technique is proposed and a parameter design procedure to achieve the optimal compromise among several different response variables is developed. One existing case study is solved by the proposed method and the results are compared with ones by the sum of SN ratios, the expected weighted loss, the desirability functions, and EXTOPSIS model.

In this paper, we consider the shortest path network problem with fixed charges. A turn penalty algorithm for the shortest path problem with fixed charges or turn penalties is presented, which is using the next node comparison method. The algorithm described here is designed to determine the shortest route in the shortest path network problem including turn penalties. Additionally, the way to simplify the computation for the shortest path problem with turn penalties was pursued.

The purpose of this paper is to find the most vital arc in the minimum cost flow problem. The most vital arc is the arc whose removal results in the greatest influence in the costs or the amount of demands in a given minimum cost flow network. This problem can be well applied to the conflict situations such as military logistics network or communications network. In this situation, network user wants to know which arcs are the most vital to him so that he can reinforce these arcs against attack, while interdictor wants to destroy these arcs which increase the distance of the shortest path most through the network. When one of arcs is removed from the network of the minimum cost flow problem, two kinds of situations can be occurred ; breaking feasibility and increasing cost. In case of breaking feasibility, the rank of arcs are determined using the amount of modified flow in a related network which is made of modifying the optimal alternative of the minimum cost flow problem. The rank of arcs with the increased costs are determined by using a method which finds the directed cycle with the minimum cost in a related network.

When infeasible interior-point methods are applied to large-scale linear programming problems, they become unstable and cannot solve the problems if convergence rates of primal feasibility, dual feasibility and duality gap are not well-balanced. We can balance convergence rates of primal feasibility, dual feasibility and duality gap by introducing control parameters. As a result, the stability and the efficiency of infeasible interior-point methods can be improved.

The purpose of this study is to present a method for determining the k most vital arcs in the maximum flow problem using genetic algorithms. Generally, the problem which determine the k most vital arcs in maximum flow problem has known as NP-hard. Therefore, in this study we propose a method for determining all the k most vital arcs in maximum flow problem using genetic algorithms. First, we propose a genetic algorithm to find the k most vital arcs removed at the same time and then present the expression and determination method of individuals compatible with the characteristics of the problem, and specify the genetic parameter values of constitution, population size, crossover rate, mutation rate and etc. of the initial population which makes detecting efficiency better. Finally, using the proposed algorithms, we analyzed the performance of searching optimal solution through computer experiment. The proposed algorithms found all alternatives within shorter time than other heuristic methods. The method presented in this study can determine all the alternatives when there exists other alternative solutions.

We consider a Location-Allocation Problem with the Cost of Land(LAPCL). LAPCL has extremely huge size of problem and complex characteristic of location and allocation problem. Heuristics and decomposition approaches on simple Location-Allocation Problem were well developed in last three decades. Currently, genetic algorithm(GA) is used widely at combinatorics and NLP fields. A lot of research show that GA has efficiency for finding good solution. Our main motive of this research is developing of a GA in LAPCL. We found that LAPCL could be reduced to trivial problem, if locations were given. In this case, we can calculate fitness function by simple technique. We propose fourth alternative genetic algorithm. Computational experiments are carried out to find a best algorithm.

Lesson learned from the Gulf War, Information systems should be interoperable and integrated for taking information superiority. But most of Information systems that have been developed and operated by military departments are stove-pipe systems that are built over and over again incompatible ways even when requirements are the same, or slightly difference between systems. As a results of this development way, those systems have poor interoperability and portability. But, If some common functions could be extracted, developed as a set of extensible building blocks, the software for common functions could be reusable. Therefore interoperability would be improved because common software is used across systems for common functions. This concept led the development of COE(Common Operating Environment). US and UK COE have been developed and being implemented for the purpose of achieving above goals. Recently Korean Military try to develop the architecture of COE, but it is not easy because of the lack of COE concept, COE development methodology as well as expert In this paper we suggest Korean Common Operating Environment Implementing Strategy based on analysis of US and UK COE that have being implemented to their military systems.

In this paper, we consider a new preventive replacement policy for the system which deteriorates while it is in operation with an increasing failure rate. The system is subject to two types of failure. A type 1 failure is repairable while a type 2 failure is not repairable. In the new policy, a system is replaced at the age of $t_p$ or at the instant the$\textsc{k}^{th}$ type 1 failure occurs, whichever comes first. However, if a type 2 failure occurs before a preventive replacement is performed, a failure replacement should be made. We assume that a type 1 failure can be rectified with a minimal repair. We also assume that a replacement takes a non-negligible amount of time while a minimal repair takes a negligible amount of time. Under a cost structure which includes a preventive replacement cost, a failure replacement cost and a minimal repair cost, we develop a model to find the optimal ($\textsc{k},t_p$) policy which minimizes the expected cost per unit time in the long run while satisfying a system availability constraint.

This paper is concerned with the application of generalized linear interactive modelling(GLIM) to the binary categorical data. To analyze the categorical data given by a contingency table, finding a good-fitting loglinear model is commonly adopted. In the case of a contingency table with a response variable, we can fit a logit model to find a good-fitting loglinear model. For a given $2^4$ contingency table with a binary response variable, we show the process of fitting a loglinear model by fitting a logit model using GLIM and SAS and then we estimate parameters to interpret the nature of associations implied by the model.