Title :
Control optimization for high performance system
Author :
Cotsaftis, Michel
Author_Institution :
LTME/ECE, Paris, France
Abstract :
A regular approach for realizing high performance industrial or technical systems rests upon the development of better performing components with a controller acting with its specific features. As this implies the mastery of a larger frequency domain than the present coverage of actuators, robustification of controllers is required, as for modern ones with improvement over classical PID to better split the frequency band into a lower range for power input and a high gain for performance, and a higher range with small gain for robustness. However, as stability is only simple, the limit is often reached for moderate performance even with high quality components for partly known and/or intrinsically too noisy systems, where performance level and power requirement cannot be simultaneously met. Reassessing the problem, another controller giving asymptotic stability is proposed, and further improvement by distributing the controller´s location in an optimum design view of the system is analyzed on a general nonlinear split n-dimensional dynamical system
Keywords :
asymptotic stability; multidimensional systems; nonlinear dynamical systems; optimal control; robust control; PID; actuators; asymptotic stability; control optimization; high performance system; nonlinear multidimensional dynamical system; optimum design; power requirement; robust control; stability; Actuators; Control systems; Electrical equipment industry; Frequency domain analysis; Industrial control; Nonlinear control systems; Performance gain; Robust control; Stability; Three-term control;
Conference_Titel :
Systems, Man, and Cybernetics, 2000 IEEE International Conference on
Conference_Location :
Nashville, TN
Print_ISBN :
0-7803-6583-6
DOI :
10.1109/ICSMC.2000.886548