Toshiyuki Kitamori

A new algorithm for finite spectrum assignment of single-input systems with time delay

Finite spectrum assignment for time-delay systems is the elimination of delay operators from the characteristic function of the closed-loop system and the arbitrary assignment of poles. The control consists of polynomials in the delay operator and finite Laplace transforms. An algorithm for computing the control matrix is presented. In particular, the control matrix over rational functions of a delay operator is computed and expanded to partial fractions. Partial fractions are systematically transformed to finite Laplace transforms. >

Design of a PI-type state feedback optimal servo system

A method of designing a type-one optimal servo system is described. The design uses an exact model matching method with pole-zero cancellations. The initial values of the integrators are set according to the plant initial conditions at the beginning of control. This servo system is optimal in the sense that the performance index, comprising the quadratic forms of both the error vector and the deviation control vector from its steady-state, is minimized. The sensitivity to plant parameter variations can be reduced without loss of optimality for the nominal plant. Moreover, this servo system has the same robustness characteristic as typical optimal servo systems designed using different performance indices. Two numerical examples illustrate the efficiency of this servo system.

Spectrum assignment and parametrization of all stabilizing compensators for time-delay systems

The parameterization of the set of all stabilizing compensators for linear systems with commensurate time-delays is discussed. Two definitions of stability are adopted for closed-loop stability: asymptotic stability and finite-spectral stability (which is asymptotic stability with the closed-loop system having only finitely many spectra). Parametrizations of all compensators attaining the closed-loop stability are given. It is noted that, using the parametrizations obtained, one can reduce robust control problems to corresponding interpolation problems.<<ETX>>

A Design Method for Nonlinear Control Systems Based Upon Partial Knowledge About Controlled Objects

Abstract A unified method is developed to design control systems in which the controlled object and the controller and/or compensator are not necessarily linear. The study proceeds in three steps. Firstly, some fundamental properties of calculation are listed which enable us to compute control system input-output relation and to design the control systems in the model matching sense. Secondly, an actual expression of the input-output relation is given in such a form as a natural extension from linear to nonlinear, and calculation basis is established. Thirdly, a desired system model in the proposed expression is given and design in the sense of partial model matching is carried out. Some sample response curves are shown to demonstrate effectiveness of the method

Fault tolerance of decentralized adaptive control

A decentralized adaptive control system, in which the local feedback gains are tuned by the sigma -modified adaptive law so as to achieve small output errors with small feedback gains, is proposed. Stability analysis and fault tolerance analysis are performed, and it is shown that if the stability of one of the stable subsystems breaks down and if the control input becomes zero, the local feedback gains of other subsystems are tuned automatically so as to keep the stability of the overall system. The interconnection may involve distributed systems such as a time delay system.<<ETX>>

A design method for repetitive control systems with multiple periodic signal generators based upon the Bezout equation

The authors consider repetitive control systems with multiple internal models for periodic signals where output tracking and disturbance rejection can simultaneously be achieved. To obtain such systems, a repetitive compensator with multiple periodic signal generators connected in series is used. Such systems are designed for multivariable controlled objects using the Bezout equation.<<ETX>>

Reliable Stabilization based on a Multi-compensator Configuration

Abstract This paper proposes a new multi-compensator configuration for reliable stabilization problem. In the proposed configuration, each compensator observes the outputs of the other compensators. This configuration provides reliable stabilization against actuator failures. A sufficient condition for the stability with this configuration is given in this paper. The transfer function of the closed-loop system is given explicitly.

Design of Control Systems on the Basis of Partial Knowledge About Nonlinear Objects Represented by Volterra Series and its Application to Artificial Control of Respiration

Abstract A method is developed to design nonlinear control systems in which controlled objects, and controllers and/or compensators are described by Volterra series. On the basis of some fundamental properties of operations characterized by Volterra functional operators, two kinds of control systems are synthesized providing their input-output relation with that of a reference model which satisfies appropriate technical requirement. The controlled objects are assumed to be not perfectly described, i.e. only important parameters are assumed to be known and others to be unknown. Based on such a limited knowledge of controlled objects, control systems are designed by computer-aided partial model matching technique, which assures the robust control for parameter changes. This method is applied to the design of control systems of artificial respiration, whose dynamics satisfies the requirement of medical treatment.

Continuous-time model reference adaptive control for an unknown nonlinear system

Abstract A model reference adaptive control is presented for an unknown continuous-time non-linear plant. The controller consists of an estimator, a generator, and a table. The estimator puts the plant non-linear characteristics into the table, from which the generator synthesizes the control input to realize the model reference. An error convergence proof under a boundedness assumption and a simple numerical simulation are presented for verification of the viability of the proposed control system.

Partial Model Matching Method Conformable to Physical and Engineering Actualities

Abstract In this paper, a unified design method is presented which covers from linear continuous-time to sampled-data, from linear to smooth nonlinear, and from SISO to MIMO decoupling control. The method has been developed paying special attention to conformity with physical and engineering actualities. The conformity brings about a very simple design theory which calls for simple manipulation on formal polynomials with ascending power of s , the differential operator, even for sampled-data control system design. Sample designs are given to show the validity of the method.