COMPARISON AND EXPERIMENTAL EVALUATION OF MULTIVARIABLE FREQUENCY-DOMAIN DESIGN TECHNIQUES

Summary The multivariable Nyquist array, and characteristic locus design techniques are evaluated by experimental application to the double effect, computer controlled, pilot plant evaporator at the University of Alberta. These frequency domain design procedures and the resulting controllers are also compared with multivariable “optimal-quadratic” regulators, and with a conventional multiloop control scheme. All of the multivariable control approaches were found to produce better control of the evaporator than the multiloop scheme, and the best of the controllers designed using frequency domain techniques ( output feedback ) compared favourably with the best “optimal” state feedback controllers. The direct Nyquist array technique is preferred by the authors but all frequency domain techniques give good insight into system performance and, from the point of view of both the designer and the plant operator, are convenient, practical and robust.

[1]  P. McMorran,et al.  Extension of the inverse Nyquist method , 1970 .

[2]  E. V. Bohn,et al.  Use of matrix transformation and system eigenvalues in the design of linear multivariable control systems , 1963 .

[3]  H. Rosenbrock.  Design of multivariable control systems using the inverse Nyquist array , 1969 .

[4]  H. H. Rosenbrock,et al.  Stability and the eigenvalues of G(s) , 1975 .

[5]  A. MacFarlane.  Return-difference and return-ratio matrices and their use in analysis and design of multivariable feedback control systems , 1970 .

[6]  I. Horowitz,et al.  Superiority of transfer function over state-variable methods in linear time-invariant feedback system design , 1975 .

[7]  David Q. Mayne.  The design of linear multivariable systems , 1973 .

[8]  J. J. Belletrutti,et al.  The characteristic locus design method , 1973 .

[9]  G. C. Barney,et al.  Application of multivariable control theory to hydraulic models , 1975 .

[10]  H. H. Rosenbrock,et al.  Bounds for transfer functions in multivariable systems , 1973 .

[11]  F. M. Hughes,et al.  Application of Frequency Response Methods to Nuclear Power Station Once-Through Boiler Control , 1974 .

[12]  H. Nicholson,et al.  Inverse-Nyquist-array design with minimum sensitivity , 1976 .

[13]  Mee.  Specifications and criteria for multivariable control system design , 1976 .

[14]  A. MacFarlane,et al.  Multivariable Nyquist-Bode and multivariable Root-Locus techniques , 1976 .

[15]  G. B. Price.  Bounds for determinants with dominant principal diagonal , 1951 .

[16]  A. MacFarlane,et al.  Survey paper: A survey of some recent results in linear multivariable feedback theory , 1972 .

[17]  Ching-Hwang Hsu,et al.  A proof of the stability of multivariable feedback systems , 1968 .

[18]  Dale E. Seborg,et al.  Disturbance localization in linear systems by eigenvector , 1977 .

[19]  A. Ostrowski.  Note on bounds for determinants with dominant principal diagonal , 1952 .

[20]  N. Munro,et al.  Applications of the inverse Nyquist array design method , 1976, 1976 IEEE Conference on Decision and Control including the 15th Symposium on Adaptive Processes.

[21]  H. H. Rosenbrock,et al.  Progress in the Design of Multivariable Control Systems , 1971 .

[22]  E. Bohn,et al.  Stabilization of linear multivariable feedback control systems , 1960 .

[23]  J. Clay,et al.  Multivariable Control of Bath Recovery Plant For Aluminum Electrolysis , 1974 .