High-strength bolts of steel bridges gradually loosen in service. In construction and maintenance of steel bridges, it is necessary to develop the method that can infer axial force of high-strength bolts in joints by nondestructive inspection. The loosening of high-strength bolts has to be inspected by experts with hammers in the past. The axial forces of bolts have to be reasoned quantitative for the accurate inspection. A system for inferring axial force of highstrength in steel bridges using waveform data provided by a hitting of automatic looseness detector. The system has been based on a neural network with faculty of pattern recognition. However, that system requires not only such reactionary waveform data but also the grip length of bolts. In the present paper, the system was constructed for inferring axial force and grip length of high-strength bolts without being given such grip length data. Firstly, reactionary waveform data was pretreated through fast Fouriertransform(FFT)for extracting grip length information. The method to construct a system was examined using amplitude spectra and phase-angle spectra after FFT. As a result, the system was obtaind the best reasoning accuracy of the axial force by using both amplitude spectra and phase-angle spectra. This system was demonstrated its usability of reasoning the axial force of high-strength bolts in all existing steel bridges. This system can utilize not only looseness inspection in service but also inadequate tightening torque in construction. By using the system, everybody can infer the axial force in sequence if they hit high-strength bolts without mastery of skills. This study can adjust the lack of expert engineer, and improve an inspection method of structure relied on hunch.

We have studied and developed some systems to generate calligraphic characters by using an electronic pen.These systems have generated Kai-sho style and Hiragana style characters, but they have been unable to generate a variety of styles of character such as Ren-hitsu style. In this parper, we propose one of new systems which can generate both Kai-sho style and Ren-hitsu style characters from the on-line inputted one. The system transforms Kai-sho style into Ren-hitsu style.

Recently, a lot of researches for knowledge acquisition using Genetic Algorithm(GA)have been reported. Two approaches, Michigan approach and Pittsburgh approach, have been mainly used for these researches. Michigan approach, in which one rule corresponds to one chromosome, has the advantage to acquire effective each rule. However, the evaluation function for each movement of the robot has to be established and it is difficult to evaluate each rule. Pittsburgh approach evaluates a set of rules according to the movement of the robot. It has the advantage to establish the evaluation function easily. However, some of the rules can be meaningless, ineffective, or anomalous one another while the robot acquired the objective movements. The acquired rules can have low stability for change of environment. On the other hand, the mechanisms of the biological developmental process from only one egg cell have been investigated in biology. The existence of some of the controlling genes called "Homeobox Genes" in DNA has been discovered in many animals including human beings. These genes are controlling activation/inactivation of other structural genes as transcription factor in the developmental process. The biological DNA acquires the indirect design system by these homeobox genes. DNA coding method, which has been proposed by authors, is suitable for knowledge representation. This method can be thought that it is able to realize the biological structure like homeobox genes and the mechanism of the developmental process effectively. This paper introduces a part of the mechanism of the homeobox genes, inactivation of other structural genes, into the DNA coding method. In this method, each rule can inactivate the other rules according to the situation. It is expected that this method makes effective, understandable for human, and high stability rules. It is also expected that this method can make the set of rules separated into some effective roles. This paper applies this method to the acquisition of control rules for a mobile robot, and studies the effectiveness of this method showing the acquired rules by the simulation. The robot moves to goal avoiding the obstacles using inputs from the sensors. The result shows that only a few but effective and meaningful rules for the movement of the robot can be acquired by this method. The acquired rules also show they are separated into some important roles automatically. This paper also compares this method with the normal GA using a penalty into the evaluation function to reduce the number of rules. It shows the effectiveness of this method.

Recently, automobile manufacturers continue large effort for developing the low emission and fuelefficient vehicle, and various researches have been made. One of the most important researches is to develop the engine that can practically work in the ideal combustion condition with fuel as little as possible. However, the tolerance is narrow for normally driven with a small air fuel ratio. The combustion condition becomes unstable, when the air fuel ratio is below the tolerance. The unstable combustion condition makes a fluctuation of torque and an increase of NOx in exhaust gas. Conventional method for solving the problem stated above is to control the air fuel raito by sensing an unstable combustion condition. But the control method predicting the combustion condition is not proposed yet. Generally the predictive control has many advantages, but is required high prediction accuracy. It is considered that to predict the combustion condition of an automobile engine is quite difficult. By the way, the chaos is noticed as the theory that can treat the nonlinear dynamical system, and the research has energetically been advanced. For predicting the combustion condition of an automobile engine, we applied the chaos theory. That is : the peak pressure of succeeding combustion is forecasted, using only combustion pressure time series of cylinders inside at a constant load. In this paper, the purpose, method and result of our research are reported.

A technique of quantifying three-way data obtained from subjective evaluation is proposed. Fuzzy subsets are defined in the model parameter space by mapping all personal data into the parameter space after the average model from the average data is constructed. This is a new approach that differs from the traditional fuzzy statistical analysis including the fuzzy quantification techniques. In this paper, this technique is applied to the quantification method of type 1, and a new type of quantification model is proposed. The method of inferring the combination of design elements from evaluations is also considered based on the obtained fuzzy quantification model.

This paper is dealing with the fuzzy clustering method which combines the deterministic annealing approach and the fuzzy entropy. Regularizing fuzzy c-means with the fuzzy entropy, we obtain the Fermidirac type distribution function as the membership function. So, we derive formulas of the partition function and the Helmholtz free energy that give the Fermi-Dirac function, and re-formulate the method as the fuzzy clustering by deterministic annealing. Then we examine the meanings of Fermi-Dirac function and the fuzzy entropy from the statistical mechanical point of view, and show that the mechanism of this fuzzy clustering is not other than the Fermi-Dirac statistics. This paper also gives the sample algorithm which performs deterministic annealing of the membership function. Numerical experiment demonstrates that the algorithm can cluster a given set of data points properly.