This paper presents a new model reference approach to the design of composite feedback-feedforward controllers using I/O data of the plant. The method is direct, as no plant-identification step is required to design the controller components. The basic idea is that of interpreting the open- loop I/O measurements on the plant as closed-loop data produced by a "virtual" reference signal that can be computed by back-propagat ing the measured output of the plant through the reference model; thus, the controller design reduces to a standard identification problem, in which the output signal to be matched is the measured input of the plant. Both a deterministic (noise-free) setting and a stochastic setting, for the case of noisy data due to unknown disturbances acting on the plant, are considered. In both cases, the controller designed by the new method is shown to coincide under suitable conditions with the one given by a standard (model-based) indirect method applied to the (unknown) data-generating model.
[1]
Karl Johan Åström,et al.
Computer-Controlled Systems: Theory and Design
,
1984
.
[2]
Graham C. Goodwin,et al.
Adaptive filtering prediction and control
,
1984
.
[3]
Lennart Ljung,et al.
System Identification: Theory for the User
,
1987
.
[4]
K. Åström.
Introduction to Stochastic Control Theory
,
1970
.
[5]
Kumpati S. Narendra,et al.
Identification and control of dynamical systems using neural networks
,
1990,
IEEE Trans. Neural Networks.
[6]
Stephen P. Boyd,et al.
Linear controller design: limits of performance
,
1991
.