Title :
AMOS-a new hybrid evolutionary algorithm for continuous time systems
Author :
Santosuosso, Giovanni L.
Author_Institution :
Dipt. di Ingegneria Elettronica, Rome Univ., Italy
Abstract :
A novel evolutionary algorithm called atomic metaphor optimization strategy (AMOS) is proposed, which is designed for real-time analog optimization problems. This new evolutionary algorithm is integrated with the continuous time adaptive observer algorithm based on the Lyapunov stability theory, developed for classes of approximating functions with linear parametrization. The combined hybrid algorithm is applied to the online modeling of continuous-time nonlinear systems, via a nonlinearly parametrized neural approximation of the system dynamics
Keywords :
Lyapunov methods; continuous time systems; evolutionary computation; nonlinear systems; observers; simulated annealing; AMOS; Lyapunov stability theory; atomic metaphor optimization strategy; continuous time adaptive observer algorithm; continuous time systems; hybrid evolutionary algorithm; nonlinear systems; nonlinearly parametrized neural approximation; real-time analog optimization problems; simulated annealing; Approximation algorithms; Continuous time systems; Data structures; Evolutionary computation; Genetic algorithms; Linear approximation; Neural networks; Real time systems; Simulated annealing; Stability;
Conference_Titel :
Decision and Control, 2001. Proceedings of the 40th IEEE Conference on
Conference_Location :
Orlando, FL
Print_ISBN :
0-7803-7061-9
DOI :
10.1109/.2001.980988
