The controllability with input constraints is important in practical problems. Especially, the controllability with positive controls is useful in discussing the vibration control problem of a pendulum system, economic system, etc. And so, this paper presents a necessary and sufficient condition for the controllability to the zero state by positive controls of multiple input discrete-time linear systems. It is derived that the positive controllability of a given system may be reduced to those of its subsystems corresponding to the the Jordan canonical form. Since the sizes of the subsystems are smaller than that of a given system, then the positive controllability test can be simplified very much. Furthermore, the structure of a positive controllable system is revealed. The results in this paper are fundamental and essensial in the control system theory.

抄録全体を表示

This paper proposes a method for transferring the state of a dynamic system to a specified state within a specified time by using a compensation input, and presents a simple formula for obtaining the compensation input under a minimum input-norm condition. Input values at the initial and final times also can be specified, and it is shown that the proposed method includes solutions to fixed-final-state optimal control problems. The paper discusses the robustness of the final-state control performance of the proposed method as compared with feedback control itself, and it is shown that, under a certain condition, the final-state error with the compensation input is less than without it, even if the system parameters are changed. It is also shown that the method can be applied to a closed-loop system with a state observer or a feedback controller, as well as to a state feedback system. Simulation results of a positioning and swing control for a cart and pendulum system are shown as an example.

抄録全体を表示

This note presents a new algorithm for obtaining all polynomial solutions of the Diophantine equation. The proposed algorithm is straightforward. Unlike previous results, it is not necessary to solve any matrix equations or to calculate any inversions of polynomial matrices or to select any matrices satisfying a condition. The computational efforts can be reduced highly.

抄録全体を表示

In this paper we consider a class of dynamical systems with unmatched uncertainties. We propose a new class of state feedback controllers which can guarantee global exponential stability of dynamical systems with unmatched uncertainties as long as a bounding function on perturbations of the uncertainties is available. The state feedback controller developed here is continuous and bounded, and can guarantee faster convergence rate and better transient behavior. A numerical example is given to demonstrate the utilization of our approach.

抄録全体を表示

A new finite dimensional approximation filter is derived for nonlinear systems. For the finite dimensional approximation, a statistical linearization method is employed. The proposed filter is compared with other statistical linearization filter and the extended Kalman filter using an example.

抄録全体を表示

A narrow band eliminating spatial filtering method has been proposed for the on-line defect detection of knitted fabrics. The defects to be detected have several types of defects to be classified. The defects caused by some troubles of the knitting machine give a serious damage on the whole fabrics. Some of the other defects occured by occasional irregularity of the knitting thread give a few temporal damages. For high quality control and high performance manufacturing, it is required not only to detect the defects but also to discriminate the type of defects simultaneously.
This paper describes how to discriminate the type and the size of the defects by using a pair of narrow band eliminating spatial filters. One spatial filter is set perpendicularly, the other is set in an inclined angle. This arrangement of the spatial filters realizes anisotropic signal processing.
The type of defects is discriminated by the anisotropic signal processing, and the size of defects is extracted from the zero-cross width of the output signals of the two spatial filters. The effectiveness of the proposed method is confirmed by computer simulations and experimentally evaluated.

抄録全体を表示

Design methods of the self-tuning control strategy for the system with unknown parameters have been studied up to now. Among them, generalized predictive control and generalized pole-assignment control have been reported as the most succesful for the system with unknown time-delay element. However, these control methods make difficulty to realize for the real plants.
In this paper, we construct a pole-assignment self-tuning controller which is easily to realize for the real plants. This control method enables us to design control system by considering the transient property for the reference signal and the robustness for the disturbance, respectively. And also, in order to treat the system with disturbance as stepwise, we present a modified algorithm for the usual least squares method. Furthermore, we present a global convergence analysis for the proposed pole-assignment self-tuning control algorithm. Finally, we illustrate some numerical simulation results in order to show the effectiveness of the proposed control algorithm.

抄録全体を表示

This paper presents a new structure of nonlinear adaptive control system incorporating a neural network (NN). The control input is given by the sum of the output of a robust adaptive controller and the output of the NN. The role of the NN is to compensate for constructing a linearized model so as to minimize an output error caused by nonlinearities in the controlled system. The role of the robust adaptive controller is to perform the model-matching for the uncertain linearized system to a given linear reference model. One of the distinctive features of the proposed structure is to give an efficient method for calculating the derivative of the system output with respect to the input by using one identified parameter in the linearized model and the internal variables of the NN, which enables to perform the back-propagation algorithm very efficiently. In numerical simulations, we examine the effectiveness of the proposed scheme by applying it to a variety of nonlinear systems, such as a system with backlash.

抄録全体を表示

This paper proposes a design method for an input tracking control system using a neural network. A multi-layered neural network is first trained to model the input-output mapping of a controlled object of unknown, discrete-time nonlinear system. Next, this neuro emulator is used for generating Jacobian information with respect to the controlled object. A simple control law is determined by Lyapunov's method so that the output of the controlled object can follow a reference input. The advantages of this method are: 1) The neural network can emulate the output of the controlled object, and we need no mathematical model of the controlled object. 2) Training of the neural network is carried out off-line, and the control law is determined by Lyapunov's method. Thus, the system output can track a reference input in the sense of stability of linearized system. The simulation results showed the effectiveness of the proposed method. We applied also this method to an input estimating problem, as a dual problem of the tracking problem. The simulation results for this problem are also given.

抄録全体を表示

In this paper, we consider position and attitude control of flexible structures using the Robust Model Matching method.
The Robust Compensator designed in the Robust Model Matching is attached to the existing controller, and improves the sensitivity for parametric variations independently of the reference input responses. The purpose of the position and attitude control of the flexible structures is to get fast and accurate responses with its elastic vibrations suppressing, therefore the Robust Model Matching is effective on the control system design of the flexible structures.
Here, we have considered two problems on applying the Robust Model Matching to a control system of flexible structures. One is a treatment for the truncated error, the other is the design method of the Robust Compensator from the point of view of stabilities against the model uncertainties. In the former, we defined the equivalent disturbances including the truncation error. In the latter, we proposed the method which decides the parameters contained in the Robust Compensator in consideration of stability against additive variations. By the proposed method, some attitude control systems of a flexible artificial satellite apparatus are designed, and their performances are shown by simulations.

抄録全体を表示

The present paper proposes a method to estimate hand impedance characteristics including inertia and viscosity as well as stiffness in multi-joint human arm movement. While subjects maintain a given hand position, small external disturbance to the hand is applied by a manipulandum. Time changes of the hand displacements and forces caused by the disturbance are measured, and the hand impedance are estimated using a second-order linear impedance model.
Firstly, accuracy of the estimated impedance is analyzed applying the method to the estimation of mass of the objects and mechanical spring coefficients. Then, the hand impedances of human arm during maintenance of posture are estimated, and the experimental results show the following: 1) the estimated inertia matrices of human hand agree with computed values using a two-joint arm model, 2) the orientations of stiffness fields are almost the same as the experimental results of Mussa-Ivaldi et. al (1985), 3) the amplitude of the estimated stiffnesss is considerably lower than the one of the results of Mussa-Ivaldi et. al (1985), 4) the orientation of viscosity ellipses almost agree with the one of the stiffness ellipses.

抄録全体を表示

In this paper we deal with an efficient method for obtaining a suboptimal solution of combinatorial optimization problems. We propose a hybrid approach of genetic algorithm (GA) and Lagrange relaxation method (LR) where the solution space of the given problem is partitioned into several small subspaces and only promising subspaces are searched for obtaining a suboptimal solution. The most promising subspace is found by GA where a lower bound of each subspace obtained by LR is used for evaluating the performance (degree of promise) of each subspace. The final suboptimal solution is obtained by searching an optimal or suboptimal solution in the most promising subspace. A numerical example of jobshop scheduling is included, and the computational burden and the accuracy of the solution are compared among the solutions obtained by our hybrid approach, GA, and LR.

抄録全体を表示

This paper proposes a new strategy for motion planning in robotics. When robots performs some tasks, they work along with motion plans. The plans should be effective for the robots. The proposed strategy applies a Genetic Algorithm (GA) to optimize the motion planning. To evaluate the planned motion, the strategy also applies fuzzy logic to a fitness function. The fitness function is referred to as Fuzzy Critic. The Fuzzy Critic evaluates populations in the GA with respect to multiple factors, while traditional fitness functions do with respect to only one factor. Depending on the goals of the tasks, human operators can easily determine inference rules in the Fuzzy Critic because of the fuzzy logic. In this paper, the strategy determines a path for a mobile robot which moves from a starting point to a goal point while avoiding obstacles in a work space and picking up loads on the way. Simulations illustrate the effectiveness of the proposed strategy.

抄録全体を表示

Tactile information processing was studied based on psychophysical experiments in order to design a tactile display for “Hand with Sensory Feedback” of tele-existence robots or vitrual reality devices. Firstly, the identification characteristics of tactile sense concerned with the quality of materials were investigated when human fingertips touched the surface of various materials in a vertical direction. In addition to obtaining the tactile characteristics, the temperature change of the skin surface of a fingertip was measured when the fingertip contacted the surface of materials. From these experimental results, the temperature difference from the normal skin temperature was found to be very important to recognize what kind of materials contacted the skin surface. Next, the tactile display for presenting the quality of materials was designed on the basis of the above experimental results. The tactile display system was composed of a Peltier element and a PID controller. It was ascertained that the tactile display has the ability to easily present a similar quality of materials to reals objects.

抄録全体を表示

A neural network is often expected to have the noise-resisting ability, which is the property that the outputs keep the desired signals against the inputs perturbed with some noise to some extent. In this paper, the learning problem which realizes the noise-resisting ability is formulated as an optimization problem minimizing the maximum valued output error function, and “Restriction Learning Method” is proposed to solve it.
This learning starts by solving a restricted learning problem using standard input patterns without noises as the training signals. In progress of learning, other input patterns with noises which give the maximum errors are generated iteratively under the synaptic weights obtained by a restricted problem, and the corresponding error functions are supplemented into the set of error functions in the restricted problem. Lastly, when the necessary and sufficient number of inputs with noises as the training signals are generated, the learning terminates in acquisition of the resistibility against arbitrary perturbation within a certain range. The learning results for the simple recognition problem show that the noise-resisting ability is obtained by the restriction learning method.

抄録全体を表示

In this paper, the effect of float shape on the pulsation error of area-flowmeter is examined. For some shapes of float pulsation errors are calculated on the basis of equation of motion taking into the float shape factor and are compared with measuruments and show good agreement.

抄録全体を表示

Temperature pattern instrumentation is useful to the diagnostics of thermal plants and equipments. This paper deals with the contour acquisition from 2-dimensional density distribution data by chain code technique and comprehensive understandings of the contour structure. The computer simulation revealed that this approach would be appropriate to the diagnosis of a thermal process in manufacturing industries.

抄録全体を表示