Premium
A procedure for designing robust control of systems with bounded uncertain parameters
Author(s) -
Mizukami Koichi,
Wu Hansheng,
Suzumura Fumihiro
Publication year - 1990
Publication title -
electrical engineering in japan
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.136
H-Index - 28
eISSN - 1520-6416
pISSN - 0424-7760
DOI - 10.1002/eej.4391100405
Subject(s) - bounded function , robust control , mathematical optimization , control theory (sociology) , computer science , quadratic equation , control (management) , robustness (evolution) , mathematics , control system , engineering , mathematical analysis , biochemistry , chemistry , geometry , electrical engineering , artificial intelligence , gene
In recent years, the robust control design problem has drawn considerable attention because it allows one to ensure desirable closed‐loop properties in the presence of model uncertainties. As a result, several design methods have been developed. This paper discusses mainly the problem of robust control design of systems with bounded uncertain parameters, and presents a new design method, called “incentive design,” for this design problem. This method is based mainly on such a consideration whereby the designed robust control law can ensure a most favorable value of the cost functional regardless of how the uncertain parameters vary within given bounds. Therefore, in this sense, the existence of the uncertainty in controlled systems has at least no bad effect on the optimal value of the cost functional. This paper gives first the procedure for designing robust control of the systems with bounded uncertain parameters, in general. Then for a class of uncertain linear quadratic systems, a robust control law is designed concretely and a numerical example is presented. It is shown from this derivation and this numerical example that the method proposed in this paper is effective and feasible for some practical control problems with bounded uncertain parameters. The design method developed here may be expected to have some further applications for practical control problems in future.