Simulation Of A Blackboard System With Multiple Reasoning For System Identification In Adaptive Control Applications  

Oleh:

Readle, John C. ; Alder, Christopher

Jenis: Article from Journal - ilmiah internasional
Dalam koleksi: Simulation vol. 69 no. 4 (Apr. 1993), page 219-230.
Topik: Adaptive control; blackboard system; control simulation; fuzzy logic; genetic algorithms; intelligent systems Smalltalk-80; system identification
Fulltext: 219.pdf (1.06MB)

Isi artikel

Adaptive control systems have been developed and used for a number of decades. There are still a number of problems with their operation, which limits their wider industrial adoption. This paper addresses one aspect of adaptive control, namely the on-line system identification of the time-varying dynamic process being controlled. The approach adopted is to use a blackboard system with algorithmic, fuzzy logic, and evolutionary reasoning to identify the underlying time-varying system model. A framework to implement this approach is proposed. The framework is simulated, and the paper describes the choice of language (Smalltalk-80) for, and the development of, this simulator. A number of simulation runs are described and discussed. Based on the simulations it is concluded that the framework is very effective, particularly with respect to its robustness under the sort of poor conditions experienced in real adaptive control applications. The application of this approach to on-line system identification of a real plant is the principal recommendation.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali