The location where temperature and wind speed sensors are installed is usually restricted when air-conditioning a large multi-purpose indoor space. In some areas of the space, however, it is often impossible to obtain sensor readings that serve as representative values for control. Estimating the temperature and wind speed of any particular spot is therefore essential in terms of effective control. This paper presents a method for estimating these distributions based on fuzzy-logic models. The main advantage of this method is that both fuzzy rules and their membership functions are directly generated from physical I/O data, thereby building theses models. First, a series of steps of the method is shown, and then a two-variable nonlinear function is identified as an example and its fundamental characteristics are analyzed. Next, modeling is performed using temperature and wind-speed distribution data on large-space air-conditioning based on many input variables and a high degree of nonlinearity. Through this approach, this paper will demonstrate that the models thus built are effective in estimating such distributions.

View full abstract

There have been many studies about the cost of so-called joint products for the long time, and stillit is very hard to determine the appropriate cost allotment to joint products. In the present study we deal with the problem from the viewpoint of macro economic consistency. Several new ideas such as qs-system, set cost, standard units will be introduced, and the Choquet integral with respect to fuzzy measures will be used to describe the present model, and to make an approach to the study about how to determine an appropriate cost assignments to joint products with set costs of composite commodities.

View full abstract

A computer system asks elderly people questions about their health and quality of life and usually allows them the alternatives of yes or no as their answers. A human doctor, on the other hand, has some conversations with their patients and often receives more ambiguous answers than a straightforward yes/no. This paper proposes a method to extract affirmative/negative degrees of the answers to yes-no questions in order to judge speakers' intentions from their utterances. We paid special attention to adverbs, interjections, direct/indirect representation predicates and defined their affirmative values based on the result of question-naire. We also defined a formula to calculate affirmative value change caused by different verb aspects. As a result, our proposed method can determine an overall affirmative/negative intention to a given question by analyzing users' utterances. Finally, in order to verify the validity of our method, we present the result of its application to our web-based analytical system interface for elderly people.

View full abstract

For decision making problems involving uncertainty, there exist two typical approaches: a stochastic programming based on the probability theory and a fuzzy programming based on the fuzzy set theory, and they have been developed in various ways. In this paper, we focus on multiobjective linear programming problems invoving random variable coefficients. After reformulating them based on the variance minimization model which aims to minimize variances of objective functions in view of risk avoidance under chance constrained conditions, we attempt to apply the interactive fuzzy satisficing method to the formulated problems. To be more specific, introducing fuzzy goals to consider the fuzziness of the decision maker's judgement for objective functions and applying the interactive fuzzy satisficing method to derive a satisficing solution for the decision maker through interaction, we propose a new decision making method as a hybrid of the stochastic programming and the fuzzy programming.

View full abstract

Although fuzzy reasoning has been used for decision making models for career choices, it has been pointed out that constructing of the fuzzy rules for fuzzy reasoning models are relatively difficult. Qualitative Comparative Analysis, which was developed by Ragin [5] and has been used for the qualitative case studies in the field of social sciences, is suitable for identifying the configurations of causal conditions. Therefore, we propose to use Qualitative Comparative Analysis as a complementary tool for fuzzy reasoning, that is, to use Qualitative Comparative Analysis for our constructing the fuzzy rules for a fuzzy reasoning model. The effectiveness of our method is discussed through an example of high school students' career decision making.

View full abstract

It is necessary to estimate time series inflow traffic to do appropriate traffic control on urban expressway. However, it is difficult to estimate time series inflow traffic by the exponential smoothing method because it creates nonlinear and irregular fluctuation. This paper proposes to estimate time series inflow traffic on urban expressway based on chaos theory and neural network. These methods can be applied to describe nonlinear and irregular fluctuation of time series data. The forecasted inflow traffics are used for input data of traffic simulation model. Applicability of the proposed methods is examined by the estimation accuracy of traffic congestion on the expressway. As a result, it is known that the estimation accuracy of the proposed methods is better than the exponential smoothing method.

View full abstract