This paper aims to discuss a method for assigning numbers to subjective degrees that are measured with verbal categories such as "very tall". We call this method "method of fuzzy categories" and derive it from the model for subjective rating process published by Yoshikawa et al. (1991). This method makes the psychological scale value for each category from the membership function of linguistic truth value that has the same verbal hedge as the category has. This method is appropriate to handle vagueness of categories, difference of meaning between persons and effects of context. We carry out psychological experiment to verify this method. Nineteen random dots patterns are used as stimuli. Eight male subjects rate subjective numerousness of those patterns with method of fuzzy categories and fuzzy graphic rating scale simultaneously. The results obtained are : First, the data obtained by two methods are equal in degree of matching between two fuzzy set and degree of linearity. Next, the data by method of fuzzy categories satisfies two axioms of interval scale. Therefore those results show that this method works as phychological scaling.

View full abstract

Generally, there are two types of psychological impressions caused by noise stimulus. One is an instantaneous, temporal psychological impression following the randomly fluctuating noise. The other is a total psychological impression which comes some time after the instantaneous one. We have already shown through psychological experiment that the former impression can be regarded as a fuzzy event. In this study, we focus on estimating and/or predicting the latter impression by fuzzy reasonig based on one of the evaluation indices of the former one excited by the random noise of arbitrary probability distribution type. This trial is important in the actual engineering field of noise counter-plan. Concretely, we first establish a psychological impression concerning noisiness (the former one) and annoyance (the latter one) divided into 7 categories, and conduct an indoor psychological experiment. Next, we show the importance of not only the average impression of the former one but also of its variance, and existance of the non-linear relationship between above two pychological impressions based on the experimentally obtained date. From these findings, we propose an estimation and/or prediction procedure of the latter impression by use of the well-known if-then rule. And we confirm its validity by applying it to the actually observed data. Moreover, we discuss the usefulness and advantages of our proposed method by comparing them with the linear regression method conventionally used in psychology.

View full abstract

History of the development of emotion psychology, especially in which the two theories concerning mixed emotions, adopted in this research is briefly explained at first. And a production method of the image codes corresponding to the mixed emotional words and a way of how to realize a mechanism to evoke some emotional change affected by a sequence of words are explained. In the next, the related part of the theory for subjective observation model [10] to the application of emotion processing system is explained. And a construction methodology of the system which has two functions is dascribed, i.e.; the one is that the evoked emotion is transitioned depending on the observation, and the other is that the aggregated emotion by several subjective observations settles to the objective emotion. Lastly, some attractive results by the simulated emotion processing system are shown concerning on what kinds of emotions are evoked, and on how the emotional transition is done by subjective observations, when some poems are used for the input signals.

View full abstract

The application of fuzzy theory was limited to the engineering field in the beginning, but now it began to apply fuzzy theory to non-engineering fields. Many researchers find that fuzzy theory is also useful for the evaluation problems. This paper deals with the application of fuzzy reasoning to the total evaluation problem. It is shown that there are several differences between the structures of fuzzy conctrol and fuzzy evaluation, which make the difference in designing the rules of fuzzy reasoning. As an example, the data retrieval system considering customer's satisfaction is presented.

View full abstract

This paper describes a new technique of the decision making by using hierarchical fuzzy integrals, which is called HFI. HFI is a technique that is regarded as the hierarchy process of fuzzy integrals. It is shown that using the fuzzy integrals, we can evaluate an object of evaluation by plural evaluation-items. In order to apply the fuzzy integrals to the decision making, it is established so as to integrate the evaluation objects by the same time. Although the fuzzy integral is an useful way for the subjective evaluation, it has a thorny problem that how to determine fuzzy measures if the references are increased. So it is shown that to improve the above problem, the concept of HFI which is the main idea of this paper will be contributed to new areas of the subjective evaluations.

View full abstract

Using fuzzy theory, this study sets up a general equilibrium model, where is economic model. In this paper, decisions of economic agents consisting of firms and households are represented by a fuzzy set. Since prices of goods are ambiguous and are affected by the nature of the goods themselves as well as the location of agents, therefore a good's price is represented by a fuzzy set varying with the agent. Using the Extension Principle and fuzzy mapping, the model's economic behavior is defined. In contrast to traditional theory, economic equilibrium is defined as the state in which fuzzy prices are not adjusted. We show the existence of that state. Moreover, we extend the model to the case where decisions and prices have non-comparable ambiguities, and where the membership functions reach an N-dimensional point. In such a case, we show the existence of the equilibrium state in the same manner.

View full abstract

A physical test (a physical checkup and a sports test) is done every year in every school. The data-base system supporting the evaluation of physical test data has been developed by the statistical analysis. On the other hand, some kinds of fuzzy regression analysis have been proposed by introducing the fuzzy concept into the statistical analysis. However, few applications of the fuzzy regression analysis using fuzzy input data and/fuzzy output data have been reported. The purpose of this study is the application of some kinds of fuzzy regression analysis to the evaluation of physical tests. Through our study, two major conclusions are derived. First, we had better use the statistical regression analysis, when data have outliers and we study the individual balance. Second, we had better use the fuzzy regression analysis, when data do not have outliers and we study the whole tendency.

View full abstract

Social phenomena such as economical variation contain many kinds of fuzziness. This paper explains the results of applying Possibilistic Regression Models to the Forcast of Advertising Expenditures in Japan. And also we makes clear whether these models are useful or not. We analyze using following three types of models, 1st is that the regression coefficient is fuzzy, 2nd is that the input data is fuzzy itself, 3rd is calculated by using the orthogonal array in factorial experiment.

View full abstract

A problem on estimation of a membership function is formulated as to estimate the expected value of a fuzzy random variable. This formulation makes it possible to discuss the estimation problem objectively. First it is shown that fuzziness can be estimated when information of individual membership functions is available. Secondly a statistical method is proposed for estimation and applied to a practical example. Moreover it is pointed out that a method based on frequency obtained from repeated measurement, one of the former methods, cannot be demonstrated its reasonability.

View full abstract

This paper represents the method of detecting different driving environments such as highway roads, mountainous roads, traffic congestion, and so on by analyzing the driver's behavior. In order to assess this situation, a lot of attention was paid to the relationship between the driving environment and the driver's behavior, for example, accelerator pedal operation and/or braking operation. This relationship is expressed by the fuzzy relation matrix whose elements are of an interval value between 0.0 and 1.0,that is defined as "Interval Fuzzy Relation Equation". This matrix is generated by analyzing experimental data statistically. In this relationship, the characteristics of the driving environments are the predictor variables, and the characteristics of the driver's behavior are the criterion variables. Therefore, the prediction that driving environments from the driver's behavior is needed to solve the inverse problem in this relationship. In our approach, we considered fuzziness of the driver's behavior. Additionally, the matrixes of both the above-mentioned predictor and criterion variables contained expressed interval values as well. We have ascertained the method of detecting driving environments from the driver's behavior and from the study of this inverse fuzzy relation problem. We applied this new technology to an actual functioning automotive system and estimated the performance of this method.

View full abstract

Predicted volume of traffic is the most important factor for the feed forward ventilation control of a tunnel on the expressway. Several prediction methods were applied to the traffic prediction. However, it is difficult to obtain high accuracy of traffic prediction by conventional methods. Such as the self regression analysis, kalman filter, and other methods were developed to solve the problem. We developed the prediction method for traffic based on fuzzy reasoning. This prediction method used the traffic models and the traffic measured past one hour as the inference parameters, and it infers the traffic in next one hour period. We made a practical application of this prediction method based on fuzzy reasoning and obtained high precise performance. This paper shows the traffic prediction technique based on fuzzy reasoning, and the results which applied on the practical system.

View full abstract

This paper proposes a path planning and traveling speed control method applying fuzzy logic for autonomous vehicle, in order to realize the more efficient transportation system in the automated factory. This method provides the satisfactory traveling route and traveling speed beforean autonomous vehicle begins to travel.

View full abstract

This paper deals with an electronic equipment reliability MTBF(mean time between failures). When electronic equipment and/or precision equipment are used in field, there are usually large differences between a design MTBF and a real MTBF. This problem is mainly dependent on an initial failure of equipment which is caused by a poor workmanship of worker in the maker, but this problem may continue long period. In this paper, some MTBFs of an airborne equipment at each period in field are predicted with a fuzzy inference method, and each predicted MTBF is compared with the real MTBF in field. It is found that those MTBFs are similar about ± 10 %. After that, this paper insists that the fuzzy inference method is useful to solve the reliability problems and to make a planning for supplementation of the spare parts/equipment in field.

View full abstract

This paper deals with problems of conventional demand supervisory/control methods used for equipments such as electric melting furnance with a large load variations. It also explains the fuzzy demand supervision/control mathod and shows the result of application to a practical system.

View full abstract

In this note, we define an implication operation in fuzzy algebras. That is, we consider the operation as an operation satisfying a certain axiom system. We give three types of axiom systems of the implication operation. It follows that in each kind of axiom systems the interpretation of the operation can be uniquely defined in the simplest fuzzy algebra 3. We also give some examples of the operation.

View full abstract

Acquisition of control rules have been actively studied for improving performances of control systems. This paper deals with a collision avoidance problem for avoiding a moving object. The controlled object is a large ship with a large inertia. It has been difficult to obtain the collision avoidance rules which coincide with operators' knowledge. This paper shows that the operators' avoiding rules can be acquired directly from data, which operators probably observe, using a Fuzzy Neural Network (FNN). This paper also shows that the FNN can obtain portions of the fuzzy rules for the inferences of the static and dynamic degrees of danger and the decision table based on the degrees of danger.

View full abstract

The present study aims to develop a support system for high school students' choosing their career. First, We investigated the structure of their career motives by means of the principal component analysis. Second, we proposed a career decision making model on the basis of fuzzy reasoning. The validity of the model was confirmed and the potential applications were outlined.

View full abstract