This paper considers a scheduling problem for multi-machine two-stage process for continuous production to minimize the total earliness under a constraint of no tardy lots. Both stages consist of plural machines, and the processing machine for each lot is different in each stage. Each unit of a lot is delivered to the second stage as soon as it is processed in the first stage, and it is processed simultaneously in the second stage. Every lot must be completed before or just on the due time of each lot, and the setup time depends on lot sequence. An optimization algorithm for the scheduling problem based on the branch-and-bound method is developed to effectively determine the optimum solution. The branching operation and lower bound value in the proposed algorithm will enhance the computational efficiency. Furthermore, the effectiveness of algorithm is reported by some simulation results, and the possibility of applying the algorithm to the processes with more than 2 stages is also discussed.

View full abstract

This paper studies the learning curves through the group decision support systems, usually referred to as GDSS. GDSS have been used principally for the tasks involving idea generation, planning, and consensus building. GDSS applications cover a wide range. We will emphasize on the use of support systems for wage design, for example. The GDSS in use is an experimental model and constructed to support the activities to obtain the information or intelligence applied to the group of designers. Therefore we will attempt to analyse the relationship between GDSS and learning effects of the group of designers, and concentrate on the analysis of the learning style or learning curve through introducing GDSS. The clarification will be made on the supporting system of wage design which is based as its evaluation standard on the satisfaction degree of learning on the quality of design, by means of dynamic process to attain the generation, moulding and stabilization of knowledge and simulation technique, as well as the learning achievement recognized by the group.

View full abstract

The portfolio selection model proposed by H.Markowitz is a method to manage risk of investment. But there are some criticisms because it is hard to solve the quadratic programming problem when the number of securities to be invested is large. In recent years many fast algorithms have been proposed. In this paper a realistic algorithm to solve Mean-Variance portfolio selection problem is proposed. The idea is originated from the view point of data matrix's rank. This algorithm is an expansion of our previous model to various types of liner constraints. By the algorithm investors can select a portfolio based on their investment policy. In addition to some numerical examples, we discuss the relations of this model to multi-factor model.

View full abstract

It is well-known that asymptotic distribution of the log-likelihood ratio statistics based on the observations from a normal population under the null hypothesis is explained as the chi-square distribution with 2 degrees of freedom in the case that sample size n tends to infinity. However, when sample size n is not sufficiently large, the chi-square distribution with 2 degrees of freedom is not accurate as the approximate distribution of the log-likelihood ratio statistics. Furthermore, the distribution of the log-likelihood ratio statistics under the alternative hypothesis has never been investigated. By the way, when the statistics are distributed in the non-negative space, it is known that the Patnaik's approximation is effective for approximating of the distribution of the statistics. In this paper, we first give the moment generating functions and the cumulant generating functions of the log-likelihood ratio statistics under the null and alternative hypotheses, and investigate the statistical characteristics of the log-likelihood ratio statistics under the null and alternative hypotheses. Thereafter, the Weibull approximation of the distribution of the log-likelihood ratio statistics is proposed in addition to the Patnaik's approximation. The approximate characteristics of the Weibull approximation are also compared with those of the Patnaik's approximation.

View full abstract

Maintenance activities must be considered not only as the activities to keep the equipment condition satisfactory but as ones to support the production. In practice, however, since the relation of maintenance to production is hard to treat quantitatively, most of the maintenance plans focus only on minimizing the maintenance costs incurred. In this paper, considering accomplishment of the scheduled production as realization of the interval availability required to accomplish the production, we present a periodic preventive maintenance policy with block replacement accomplishing the scheduled production. Further we formulate and analyze the policy, then clarify its features.

View full abstract

The purpose of this paper is to describe the development process of a conflict resolution expert system for allocating inheritance to its recipients, to show how effective this system is and to verify the practicability of the system by applying it to the real inheritance problems.

View full abstract

This paper presents a Case-Based Reasoning model for solving production scheduling problems in a corrugated cardboard factory. An ordered tree representation scheme of cases is used to simplify the reasoning process and make it efficient. After providing an algorithm to construct the tree by a bottom-up approach, a scheduling algorithm and three analogical reasoning strategies are developed to make satisfactory schedules based on previous cases. Being applied to the production scheduling problems in the factory, it has been proved that the proposed reasoning model is an effective and available approach.

View full abstract

This paper describes an implementation of genetic algorithm to solve Multiobjective Solid Transportation Problem (MSTP), which is an extension of our previous work. The criteria space approach for multicriteria linear program is embedded in the structure of genetic algorithms (GA) conductive to find out the nondominated points in the criteria space, The proposed algorithm is coded with C language. The computation results show that the proposed algorithms is efficient for solving the multiobjective solid transportation problem.

View full abstract

Information about the process capability is important in process control and quality control. Especially the process capability index C_p is well known. Recently C_<pk> considering the bias of the process mean and C_<pm>, C_<pmk> influenced by the concept of the Taguchi Methods are also proposed. These indices are based on the concept that the process capability should be measured by the ratio of products within the specification limits. In this study, we suggest a new index from a different standpoint of the traditional indices. Consequently, the process capability is evaluated based on the division between the quality in the present process and the target quality, as the limit quality, defined by process mean and standard deviation. Additionally we mention sample size to assure the confidence about the proposed index.

View full abstract

In cement industry, which consumes a great deal of energy, fuel planning is very important to minimize purchase and delivery costs of fuel. The optimal fuels chosen from the plural sources must be distributed to the optimal plants which are subject to many kinds of constraints. The authors not only introduce a formulation of this fuel planning with MIP but propose "negotiation prices" which are calculated with the sensitivity analysis data obtained in the process of optimization and are very useful for decision makers who must negotiate with fuel suppliers to distribute fuel to more than one plant for more than one time period.

View full abstract

This paper studies the effects of order delay in Kanban systems. In the Kanban systems, the orders are released not immediately after a demand arrives, but after the production or transportation for the demand is available. Therefore, the release of order is delayed during the two points of time. In this paper, the order delay in the Kanban systems is discussed, and a mathematical model of the Kanban system and that of the system without the order delay are developed. With simulation experiments for the developed models, the effects of order delay upon performance measures are investigated.

View full abstract

Transaction costs can be classified into several factors, such as liquidity and timing. Commission fee is of particular importance, because transaction cost affects investor's return directly. But there are few papers dealing with transaction costs, because of its difficulty in solving associated optimization problems. In this paper we propose a portfolio optimization model to manage transaction costs using integer programming. Our model can deal with concave commission fee in Japan. Applying our model to Tokyo Stock Exchange 1st Section, we obtained some knowledge about their effects on the resulting portfolios.

View full abstract

The purpose of this study is to clarify the effects of personal contacts in the advanced information system environment by analyzing the empirical data, and consider about the information system environment and the personal contacts there. Findings are that if they belong to an advanced information-based organization, then they have much computer work and personal contacts, and if they have much computer work and personal contacts, then their work motivations go up, and if they have only a little computer work and many personal contacts, then their stress becomes weak, and so this suggests that the personal contact is an important concept when we study the information systems. Finally the policies for increasing the personal contact opportunities include offering the work which contains more personal contacts, offering the space for personal contacts, holding jobsite meetings frequently and using the information and communication technology.

View full abstract

The present study proposes a VRP model coping with distribution problems which focus on the total efficiency of a business cycle rather than simply on the effective transportation. The objective of the model is to maximize the frequency of delivery for a given demand in each location under the leveling operation of dispatching. It contributes to the reduction of inventory without increasing working hours and shortage. An algorithm is also presented for solving a problem formulated by this model. We apply the model and the algorithm to an actual VRP situation of automated selling machines for canned beverages to discuss its advantages and usefulness.

View full abstract

While learning a new skill, the evaluation of learning speed change is important for understanding the relationship between learning ability and the effect of training method. The traditionally used learning constant is a quantitative index for the learning speed of an entire learning period, and it is not applicable for evaluating any learning speed change during that period. The present study, therefore, proposes to adopt a learning constant method for dynamic evaluation of learning speed change. A guideline is suggested to select the optimum period of length for moving learning constants through a series of simulation study. The method is applied for a practical computer operation task to validate its effectiveness.

View full abstract

This paper analyzes a facility layout problem in advanced manufacturing systems with alternative processing routes. In a traditional quantitative facility layout problem, minimizing the total value of "the distance value multiplied by the work flow value" between any two facilities, the work flow between facilities was assumed to be constant. However, there are several alternative processing routes for any job in recent machining systems, so that the work flow becomes changeable. This paper formulates a facility layout problem with alternative processing routes, using a closed queueing network model. The original layout problem is decomposed into the work flow decision problem and the layout position decision problem. An efficient method exploiting an approximate CQN model and a heuristic method using a simulated annealing are proposed to solve these subproblems, respectively. Some numerical experiments are shown to demonstrate that this method efficiently yields a near optimal layout with high accuracy.

View full abstract

In this paper, we expand the network simplex algorithm for solving network structured linear programming problems. This algorithm performs the simplex operations directly on the graph or network itself. We expand this algorithm by means of a fuzzilize and multiplize of objective functions. And we apply this expanded algorithm to minimum cost network flow problems with multiple objective functions such as cost, time and so on. The numerical example is also given, and it shows that the algorithm is very effective with respect to the real-world network problems under the computer configurations.

View full abstract