In this paper, we propose a new environment that supports cooperative development of Multi-agent systems (MAS). For this purpose, we suggest a system called GuardNet. In GuardNet, which is also realized as MAS, there exists an agent called Guardant for every agent in target MAS. Guardants can store requirement specifications of their for target and relevant agents, and exchange them with each other on behalf of agent programmers. Another feature of GuardNet is agent-imitating facilities for debugging and testing. In agent-imitating, a Guardant behaves as a proxy either for the target agent or relevant agents. Using agent-imitating facilities, we can test the agent system which includes incomplete agents.
In the Osaka bay areas, where many regional development projects have been planned and carried out, the infrastructures such as electricity and gas are being consolidated. It is necessary to make optimum design considering future demands for those utilities. Qualitative response model based on utility value is effective to estimate demands for the project under indefinite conditions, especially the Logit model is popular tool to estimate demands. However, this model does not be clarified the choice behavior of a decision-maker in its inducing process, and is called in question in the supposition. This report showed that the Logit model can be derived from information entropy without positing an extreme value distribution. Besides, this paper showed that new aggregate conception was suggested for statistical calculation and that it was verified by applying to the amusement institutions.
When realizing a practical face identification system, higher robustness against fluctuation of the lighting condition is required. In this paper, a new face template-matching method dedicating to facial features (eye, nose and mouth) is proposed. It uses a special template for each facial feature, which consists of sub-divided region with compensated intensity. First, the size and incline of the input face image is normalized. Secondly, the facial features are extracted and separated into some small regions. Next, the intensity values in each region are compensated according to their average and variance. Finally, the correlation coefficients between the compensated features and templates are evaluated as an identification score. Experiments with face images taken in various lighting conditions prove that the proposed method is more robust than conventional methods based on the normalized correlation coefficient.
In this paper, we deal with fixed-order H∞ controller design problems, based on bilinear matrix inequalities. BMI problems are nonlinear, non-differentiable and non-convex, and therefore very difficult to obtain global solutions numerically. First, we convert BMI problems into differentiable problems introducing so-called shift-parameters. Next, combining the shift-parameter scheme with quasi-Newton methods, we propose a new computational technique as BMI solvers. Lastly, we demonstrate the effectiveness of our approach by simulations, with applications to the design of fixed-order H∞ controllers.
We first discuss the requirements for community information sharing. We then propose an information representation called weak information structures to integrate heterogeneous information such as static information (e.g. local sites information) and dynamic information created in word-of-mouth communication. The weak information structures connects various information media without defining the semantics rigorously. By leaving the interpretation of the semantics to tacit human background knowledge, it becomes compact and robust. We have developed an information sharing system for community called InfoCommon which provides people with intelligent assistance for exchanging and sharing knowledge and ideas. We have evaluated InfoCommon at the ICMAS'96 Mobile Assistant Project.
This paper presents a design method of optimal servo systems with frequency weights from the view point of the inverse LQ problem. This design method allows us to assign desirable closed-loop transfer functions directly and in addition, to obtain the desired optimal control law in a closed form in terms of design parameters and the system parameters. A numerical example is also shown to illustrate the design method proposed here as well as to show its usefulness.
A procedure is developed to design a D/A converter via sampled-data control theory. A multirate converter with oversampling is designed to reconstruct a delayed original analog signal via H∞ and H2/H∞ design method. In spite of the multirate and delay elements, these problems are shown to be reducible to a finite dimensional discrete-time problem and furthermore to linear matrix inequality (LMI) conditions. Some numerical examples are presented to illustrate the results.