Abstract This article discusses the ccxnplete controller design strategy required for implementing Multiple Model Control (MMC), applied to nonlinear systems. It is shown that the multiple model design can be recast into a supervisory arrangement, where a global supervisor is utilised to select the appropriate controller from a fixed family set. Unlike current techniques where Fuzzy Validity Functions or Bayesian Estimates are utilised in the selection mechanism, the apprоаϿһ of a Multiple Model Observer (MMO) is employed for the selection architecture within the supervisor. This notion, is a natural extension of the MMC design. Switching between the individual controllers is accomplished bumplessly by using a Multiple Model Bumpless Transfer Mechanism, thus producing a smooth and continuous control signal as the plant alters continuously. The design of the complete controller is segmented into two sections; Supervisory and Global Regulator Controllers, with bumpless transfer implemented within the regulatory controller design. All sections ensure stability and robustness in face of plant uncertainties.
[1]
G. Goodwin,et al.
Hysteresis switching adaptive control of linear multivariable systems
,
1994,
IEEE Trans. Autom. Control..
[2]
A. B. Poore,et al.
On the dynamic behavior of continuous stirred tank reactors
,
1974
.
[3]
John Doyle,et al.
Structured uncertainty in control system design
,
1985,
1985 24th IEEE Conference on Decision and Control.
[4]
Pierre Apkarian,et al.
Self-scheduled H∞ control of linear parameter-varying systems: a design example
,
1995,
Autom..
[5]
P. Ramadge,et al.
Model and controller selection policies based on output prediction errors
,
1996,
IEEE Trans. Autom. Control..
[6]
A. Stephen Morse,et al.
Control Using Logic-Based Switching
,
1997
.
[7]
A. Morse.
Supervisory control of families of linear set-point controllers Part I. Exact matching
,
1996,
IEEE Trans. Autom. Control..
[8]
T. Johansen,et al.
Constructing NARMAX models using ARMAX models
,
1993
.