Eiichi Muramatsu

Feedback error learning control without recourse to positive realness

Feedback error learning (FEL) method has been proposed by Kawato et al. as a scheme of brain motor control in neurophysiology. Recently, Miyamura and Kimura have established a control theoretical validity of the FEL method in the framework of adaptive control and proved its stability based on the strictly positive realness. In this note, we propose another scheme for the feedback error learning method which does not require any positive real conditions.

Feedback-error Learning for Explicit Force Control of a Robot Manipulator Interacting with Unknown Dynamic Environment

Force control of a robot manipulator is important for the robot to perform physical interaction with its manipulated objects as well as its environment. Usually, the environmental dynamics is unknown and during interactions the environmental dynamics will influence the robots control loop. In this research, based on the fact that the transfer functions from the robots control torque to the environmental reaction force is biproper, a novel 2 degree of freedom adaptive control approach is presented and is applied for the explicit force control of the robot manipulator. In this approach, both force feedback and feedforward controllers are involved in the robots control system, the feedback control is set as constant while the feedforward controller is adjusted adaptively online to approach the inverse of the force control transfer function. Using this approach, exact force response without any loop delay can be realized. Computer simulations show the effectiveness of this control approach

Parameter and state estimation for uncertain linear systems by output estimation error

Parameter and state estimation is considered for uncertain linear time-invariant systems. Estimates of the state and parameters are computed by differential equations with estimation error between the plant output and its estimate by the observer. Design method for the estimator is proposed via conditions on matrix gains multiplied by the output estimation error. Multiple observers are utilized to derive the conditions and design method.

Prediction of discrete-time signals via adaptive estimation

This paper considers prediction problems for discrete-time signals. We introduce a virtual system that generates the signals to be predicted. The prediction problem is reduced to an estimation problem for the dynamics of the virtual system. In order to estimate the dynamics, an adaptive scheme is introduced. The scheme estimates the state and some parameters of the virtual system from the discrete-time signals.

Stability analysis of a class of second-order switched systems via eigenvalues of matrices

Stability condition is considered for a class of second-order piecewise linear systems. The state space of the system is divided by partitions into some subsystems. The state trajectory evolves along with the dynamics of the subsystem in which the state exists. A necessary and sufficient condition for the system to be stable is derived. The condition is expressed via the eigenvalues of the matrices which specifies the dynamics of the subsystems.

Adaptive Internal Model Control of MIMO Systems via State Equations

This paper considers an internal model control scheme for multi-input multi-output systems which contain unknown parameters in the coefficient matrices of the state equations. The internal model is adaptively constructed from input and output signals of the plant. We first propose a linear parametric representation for the unknown multi-input multi-output systems. Then, we present an adaptation method such that the output of the internal model is asymptotically identical to that of the plant, guaranteeing stability of the closed-loop system.

Two-Degree-of-Freedom IMC Paramertization of Multivariable Systems

Abstract This paper presents the systematic design method of two-degree-of-freedom parametrization of multivariable systems. It is shown that the four elements are required. The first one is to yield the inner-outer factorization, where the outer is low-pass filter to adjust the output response to the reference. The second is to stabilize the unstable plant and to adjust the response to the disturbance. The third is for internal stability and the last one is to yield the steady state characteristics. The design methods of them are presented. The total controllers are realized as observer-based state feedback controller.