China has been increased the significance of its presence in the global capitalistic economy. This is brought to us by not only dedicated and low labor cost of direct workers, but also the high productivity and potentiality that entrepreneurs, administrations, and engineers in China has ever cultivated. Japanese manufacturers are now getting too late for producing commodity products competing with China. They have to strengthen global competitiveness by producing sophisticated hardware based upon science and technology, software, contents, and service that meet one's sensitivity. Also, creation of new values integrated by the above products is absolutely imperative for enhancing their international competitiveness.

View full abstract

This paper presents recent trends in some behavioral (or descriptive) models of individual decision making and group decision making. For individual decision making some models are shown to explain the violations of expected utility models. For group decision making a model to describe the ethical consensus formation process among multi-agent conflicting decision maker is shown. Some realistic applications are mentioned.

View full abstract

Bio-mimetic research of motor control is an important approach not only for basic scientific understanding of biological system but also for developing next generation of human friendly robot systems. This review mainly focuses on arm and body manipulation ability. The recent topics of our diffusion-based learning and spatial generalization of optimal control for redundant arms, the sufficient condition for the arm's environmental adaptability and passivity, the whole body cognitive manipulation are introduced.

View full abstract

The purpose of our recent study is to explore the circumstances in which cortical neurons generate spikes in vivo. First, we found that, on the basis of a few non-dimensional statistical coefficients, standard neurospiking models are inconsistent with biological spiking data. Second, we found that the distribution of some non-dimensional statistical coefficients of actual spiking data is largely dependent on the recording site. Third, we found that, on the basis of a new measure of local variation of inter-spike intervals, neurons in each area can be classified into different groups possessing unique spiking statistics and corresponding functional properties.

View full abstract

In this paper, we look back upon the circumstances from the Hanshin Awaji great earthquake to the project called Dai-Dai-Toku sponsored by Ministry of Education, Culture, Sports, Science and Technology, and the following points are considered. (i)Why is it that there was no motion before the earthquake disaster? (ii)Why we took 7 years to realize the project. (iii)What were the main reasons for the realization?

View full abstract

Rough set approaches have been developed and extended to treat dominance relatons on attribute values. However, the monotonicity between conditional attributes and a decision attribute is implicitly assumed. Therefore non-monotonic preference information tables cannot be analyzed by the previous rough set approaches. In this paper, we propose a new rough set approach to analyze non-monotonic preference information tables. We show that all decision rules with minimal conditions can be extracted by extending the conventional decision matrix.

View full abstract

In this paper, we study cooperative fuzzy games with some restrictions on coalitions.We define restricted games by feasible coalition sets which consist of permissible fuzzy coalitions. A special class of feasible coalition sets called partition sets is discussed. Moreover, we propose the core and the Myerson value as solution concepts for cooperative fuzzy games with restrictions on coalitions.

View full abstract

In this paper, we formulate quadratic programming problems whose decision variables correspond to behavioral strategies of players in a special class of games defined in extensive form, i.e., two-person non-zero-sum games with incomplete information. By showing that optimal solutions to the formulated problems are equilibrium solutions of the games, we give computational methods for obtaining the equilibrium solutions of the games.

View full abstract

This paper studies the problem of competitive and cooperative behavior on multi-agentsystem. Each agent which consist in multi-agent system have additional decision-making process besides the motive of the game theory. We proposed the method of communication between agents, and confirmed that the method is useful for the system to carry out more advanced and intelligent behavior.

View full abstract

This paper presents a new computer program named MIDASS, which is concerned with multiobjective extension of interactive decision analysis support systems in uncertain decision environments, and in particular discusses its characteristics as interactive decision support systems. The MIDASS combines the existing computer programs: MAP, an assessment and calculation program for multiattribute utility analysis, and IDASS, an interactive decision analysis support program under uncertainty. Our new program is particularly concerned with supporting decision maker's judgmental evaluation processes interactively and mitigating the difficulty in keeping consistent assessment in subjective decision making under uncertainty.

View full abstract

The American policy for the Kyoto Protocol has a large influence on the price of international CO₂ emissions trading. Since Russia is shown to be unique supplier of CO₂ emissions trading, its price is determined as a monopolistic price. In this paper, we consider the carbon tax and the emissions trading as the counter policy against the global warming problem, and evaluate the monopolistic price for the cases that the United States join or not, using the mathematical model.

View full abstract

According to deregulation on power system, it becomes important to control line flow. For power flow control serial controllers attract researchers. It is considered that optimal allocation problems of serial controllers and optimal operating problems become more important. These problems are quit nonlinear and different from conventional optimal power flow problems. In this paper, a new method to solve these problems with interior point technique is proposed and it is shown by numerical experiments that the proposed method is effective.

View full abstract

In IT society in recent years, use of information communication apparatus is indispensable, in order to fully utilize information communication technology. However, a setup and operation are difficult for the conventional infor- mation apparatus. Moreover, there is a problem to which use becomes impossible by the case. In this research, constru- ction of a life support system is tried using the concept of invisible computing. In this system, by a user's needs, inform- ation is acquired automatically and advice and assistance are performed to a user.

View full abstract

In this paper, we obtain the dynamic traffic information, predict the future road situation from obtained traffic information, and create the prediction map database reflecting the road situation. By comparing the dial-a-ride system considering the road situation with that without considering the road situation, the validity of considering a road situation was showed.

View full abstract

A Search engine for supporting environmental risk analysis was developed. A robot-type search system with a ranking program that evaluates quality of information related to environmental risks included in WWW pages was designed. To determine whether a page contains important information or not, the system utilizes the words listed in the index of "Handbook of Risk Research". If words inputted by a user is found in a chapter in the Handbook, words listed in the index and found in the same chapter are considered as the "important words", which are highly weighted and used for ranking the search result.

View full abstract

This paper presents the application of 2-level problem in network industry market. The relationship of players in the market can be regarded as the gaming equilibrium of 2-level problem between the upper class as suppliers and lower class as the customers or passengers. In this research, the characteristic of the market is discussed, and specifically the practical model to evaluate policy scenarios in domestic air passengers flow market is formulated. As the formulation is expressed as optimization problems, strict convergence solutions and genetic algorithms are adopted. The former is the program design following formula faithfully. The latter is to rapidly obtain the approximate solution as the heuristic solution.

View full abstract

It has been disputed whether "Ogawabon-Shinsousenjimon" was an originally written work or a work made by unknown copying technique, "transparency copy". One of factors for the judgement is the strength of stroke and press. In this paper, through the examination of the gray value distribution with the image processing of character's images, we compare characters of "Ogawabon-Shinsousenjimon" with those of two old works; one is known as an originally written work and the other is known that it was written by "transparency copy".

View full abstract

In this paper, we consider a DC programming problem constrained by a convex set and a reverse convex set which is defined by the complement of the interior of a compact convex set X. When X is not necessarily a polytope, we propose an inner approximation method. The algorithm utilizes inner approximation of X by a sequence of polytopes to generate relaxed problems. Then, every accumulation point of the sequence of optimal solutions of relaxed problems is an optimal solution of the original problem.

View full abstract

Huber M-estimator cost function is the most popular cost function used in the robust linear regression to deemphasize outliers. Recently, Mangasarian and Musicant have proposed a quadratic programming model for Huber M-estimator and have shown that it is computationally efficient. In this paper, we apply a generalized Newton method to their quadratic programming model and examine its efficiency.

View full abstract

The Second-Order Cone Complementarity Problem (SOCCP) is a wide class of problems containing the Nonlinear Complementarity Problem (NCP) and the Second-Order Cone Programming Problem (SOCP). Recently, Fukushima, Luo and Tseng extended some merit functions and their smoothing functions for NCP to SOCCP. Moreover, they derived computable formulas for the Jacobians of the smoothing functions and gave conditions for the Jacobians to be invertible. In this paper, based on smoothing and regularization methods, we propose a concrete algorithm for solving SOCCP. Moreover, we show that our algorithm is globally and quadratically convergent under the monotonicity assumption.

View full abstract

In this paper, we propose a new approximate method using the affine scaling(AS) algorithm for the quadratic assignment problem(QAP). By relaxing the 0-1 valued constraint of the problem, we apply the AS algorithm to it. This method can use a large step-size and find the next solution in the AS direction within the feasible region. Through some experiments, we verified that the proposed method was more effective in solving the QAP than conventional methods.

View full abstract

In this paper, we propose a framework for obtaining the optimal car service to elevator operation problems by applying a branch-and-bound method based on the deterministic optimization model. In designing the model, it is assumed that the movement of cars and passengers synchronized with discrete time, and all passengers arriving to the hall is known beforehand. The assignment of passengers to cars and the processing order of the services to passengers are taken as decision variables in designing a branch-and-bound solution. In the computational experiments, the Call-Dispatching and Selective-Collective rule is taken as a typical rule and the results obtained by using this rule and those by the branch-and-bound method are compared and examined.

View full abstract

In recent years, in the field of bond investment, default has come to be recognized. In this paper, we treat multi-period bond portfolio selection problem with default risk. Uncertain elements associated with bond investment, such as future annual interest, default and so on, are treated as possibilistic variables and fuzzy random variables. The problem is formulated as a fuzzy random linear programming problem. Using fuzzy expected values and the necessity measure, we reduce the problem to the linear programming problem. Since the problem has a special structure called staircase structure, we apply an algorithm based on the nested decomposition method.

View full abstract

We study deadlock avoidance in a flexible manufacturing system (FMS) that consists of two machines and a robot. Yalcin and Boucher have presented an automata based approach for synthesizing a deadlock avoidance supervisor through an example. We follow their approach and present a systematic procedure to synthesize a deadlock avoidance supervisor by using the discrete event system theory.

View full abstract

Time delayed feedback control is one of practical methods for stabilizing unstable periodic orbits embedded in chaotic attractors. This control method has an open problem of estimating domain of attraction towards stabilized orbits in function space. In this paper, domain of attraction in two types of Duffing system under the control method is numerically discussed. The disturbance to initial conditions towards stabilized orbits shows the domain of attraction possibly has self-similar structures in their boundaries.

View full abstract

In this paper, a data-driven design method of PID controllers is proposed based on unfalsified control concept. This method does not require constrcting a mathematical model for controller design. Therefore, this method can be applied to complex plants including nonlinear elements. A sharp estimate of the falsified region on the parameter plane can be obtained by the input-output responses of the plant excited by sinusoidal reference inputs for many frequencies.

View full abstract

In this paper, we deal with a control problem of a 4-wheeled mobile vehicle that is a typical example of nonholonomic systems. We introduce a coordinates transformation that is valid for almost all state to have a chained form like system.We transform it again to a hybrid time-state form. We propose a simple control scheme described by hybrid automata to control the system to the origin and extend the control scheme to restrict the behavior of a state variable by a supervisory controller.

View full abstract

In this report, system identification problem in which input and output data are measured by sensors with saturation characteristics is considered. A novel system identification method which estimates both system parameters and saturation error is proposed. Effectiveness of the proposed method is examined through a numerical example.

View full abstract

This paper proposes analysis method of state space region with guaranteed L₂/L_∞ performance for control systems having direct-feedthrough matrix from both disturbance input and saturating output to saturating input.We define the region and characterize it based on the circle criterion.It is the key idea of our analysis method using the circle criterion to derive a new constraint of magnitude bound on the saturating input by considering the fact that both the disturbance input and the saturating output are L_∞ bounded signals.Our result can reduce conservatism of the existing circle approach which deals with the direct-feedthrough matrix using a global sector bound on the saturation nonlinearity.

View full abstract

This paper presents an off-line design method of a reference governor for constrained systems subject to a class of bounded disturbance.The present method assures the infinite-time constraint fulfillment by settling the state variable into the maximal constrained positively invariant (CPI) set at the final instance of reference management.In order to improve the transient response, we introduce a performance index taking account of the tracking error.Then the design of the reference governor is formulated as the minimization of worst-case performance index under the constraint that the state should be settled into the maximal CPI set for any bounded disturbance in the prescribed class.We show that this minimization problem is reduced to a convex optimization problem.The effectiveness of the present method is demonstrated by a numerical example.

View full abstract

Simulations are performed for deployment and retrieval of a tethered satellite applying nonlinear receding horizon control. Since the tether cannot be pushed, the control tension is constrained to be nonnegative, and therefore, it is not necessarily guaranteed that the tethered satellite does not collide with the mother ship. To operate the satellite safely avoiding such a situation, switching conditions are given for the control input in this paper. As a result, it is guaranteed that the state of satellite can be restricted within a specified region even if it falls uncontrollable.

View full abstract

LQ actuator selection problem for multi-output system is discussed, in which actuator sequence and input sequence are decided by minimizing quadratic control performance. This problem has combinatorial property, so it is difficult to get optimal solution. If comparability in given actuators holds combinatorial property can be dissolved. In this paper we yield sufficient condition for comparability of actuators in the multi-output system.

View full abstract

In this paper, we propose a new switching vibration control method; switching between passive and active. The vibration control device is pendulum type mass damper, which works as passive and active mass damper. Experimental results show that the proposed switching control has better performance than full time passive mass damper and active mass damper.

View full abstract

State-Dependent Riccati Equation (SDRE) is a recently developed technique for control of nonlinear systems. SDRE allows us to apply linear control thoery to the design for nonlinear controllers. However, the reports on the experimental results of SDRE method are not so many, so far. It is due to the difficulty in solving Algebraic Riccati Equation in real time. This paper proposes a new method for solving SDRE in real time, which involves the use of Quasi-Newton method, and shows simulation and experimental results.

View full abstract

This paper proposes a design method of a controller to swing up the parallel double inverted pendulums simultaneously. The method is based on the energy control by using a Lyapunov function. Close to the upright point, the unstable mode in the homoclinic orbit is avoided by using switching control. Meanwhile, we propose an additional controller to restrict the cart travel while controlling the two pendulums. The present method is examined by an experiment using a parallel double inverted pendulums system.

View full abstract

A new design scheme of PID controllers is proposed . This scheme is designed based on the IMC which is a kind of the model driven controllers.

View full abstract

This paper is concerned with sliding mode control for a linear system with time delay in the state. We consider here a design method of a hyperplane as a switching surface and a simple control which brings the trajectory on the surface. And, by using the zero dynamics technique, it is shown that the sliding mode system on the surface is stable. The stabilizability condition is given by a matrix inequality which becomes an LMI (linear matrix inequality) in a special case.

View full abstract

A simple control method is proposed for a linear discrete time system, which is tentatively called a simple predictive control. The method is extended for a system with time lag. A method so as to introduce an integrator is also discussed.

View full abstract

Power electronic systems become nonlinear hybrid systems due to the marvelous switching functions of power semiconductor devices, which makes it difficult to analyze and design the systems in systematic and unified ways. This paper surveys the problems in analysis and design of power electronic systems, focusing on how to treat the hybrid features. The computer-based analysis and design methods are also discussed.

View full abstract

In Hard disk drives (HDD), a mode-switching control (MSC) method has been widely applied as a motion controller. This MSC is defined that there are several servo modes and each servo mode is switched from one to another during a servo action. The design points are how to set initial values of controller at switching and how to define the switching condition. The MSC has been studied since 1970s in HDD industry, and an initial value compensation (IVC) method has been developed to determine appropriate initializing of controller and switching conditions in 1990s. The MSC with IVC has been extended to consider the suppression of mechanical resonance excitation and the reduction of acoustic noise during seek-settling. In order to enhance both fast movement and precise positioning accuracy capabilities, this type of MSC should be discussed and re-designed in the frame of hybrid system theory.

View full abstract

Safety is the first requirement of any mechanical systems. In order to assure the safety, several research and development have been carried out and some of them are well-organized and become international standards. By the way, due to the rapid development of computer technology, many mechanical systems are controlled by computers. This makes the mechanical system complex in safety issues. To tackle such complexity, the hybrid dynamical system theory could be one key issue. In this paper, the possibility of applications of hybrid dynamics system theory to the safety of machinery is discussed based on the international standards and seismic analysis.

View full abstract

This paper studies a discretization method of continuous-time controllers taking account of the intersample behavior of the resulting sampled-data systems. Based on the frequency response of sampled-data systems, we propose two fundamental schemes to determine desirable frequency response characteristics of the discretized compensator such that deterioration of closed-loop characteristics is expected to be suppressed; one is based on open-loop treatment, while the other is based on closed-loop treatment. Once desirable frequency response characteristics are determined, a discretized compensator is found that approximately attains such characteristics. We also discuss how to choose reasonable weights in these methods.

View full abstract

This paper studies a robustness analysis method based on positive-realness of sampled data systems. It is demonstrated with a numerical example that robust stability margin against uncertainty of physical parameters in the plant can be analyzed easily especially because we can deal with sampled-data systems with direct feedthrough terms.

View full abstract

This paper is concerned with I/O based optimal control of Hamiltonian systems, which is applicable to electro-mechanical systems. We propose a new learning control scheme based on the self-adjoint structure of the variational of those systems. This method dose not require precise information about target systems such as the physical parameters of them. Furthermore experiments demonstrate the effectiveness of the proposed method.

View full abstract

A two-degree-of-freedom design method of LQI servo systems has been proposed based on optimal control theory. The feature of this method is that tracking characteristics are determined optimally with respect to a quadratic-integral performance index while the integral compensator (i.e., the integral gain) can be designed independently without affecting tracking characteristics. This paper considers to apply the H∞ techniques to the design of the integral compensator. We apply such a design scheme to positioning control of a pneumatic servo system and show its effectiveness through experiments.

View full abstract

Heavy vegetable harvesting and organic compost fertilizing requires farmers to do hard operations. In this paper as the 3rd step study, to achieve robust performance in the presence of uncertainty, a new control system using robust gain-scheduling and μ-synthesis is designed and its validity is confirmed by watermelon harvesting experiment.

View full abstract

Directional stability radius is a tool to analyze stability for parametric uncertain polynomials. The computation method for the directional stability radius is formulated as a least squares problem with linear constraints. In this report, we reformulate the method as a nonnegative least squares problem, and apply a known efficient computation algorithm to the problem.

View full abstract

It is known that for uncertain systems with delta-operator diamond characteristic polynomials, extreme point results do not hold. Stability of the diamond polynomials is, however, equivalent to that of some segment polynomials contained due to Edge theorem. In this paper, we propose two stability analysis methods for delta-operator segment polynomials.

View full abstract

Control systems with input saturations exhibit nonlinear behavior when operating in saturation, and saturation becomes essential to determine the domain of attraction of the system. We show piecewise linear lyapunov functions proposed in this paper can be used to estimate the domain of attraction for given system. An example is used to illustrate the obtained results.

View full abstract

In this report, we present an application of QE to common Lyapunov function probrems. First, we make a short introduction of QE and present two examples of seeking for existence conditions of a common quadratic Lyapunov function by QE. Next, we consider the relation between the existence condition of a common quadratic Lyapunov function and that of a common infinity-norm Lyapunov function.

View full abstract