Title :
On-line system identification using additive dynamic neural networks. An invariant imbedding approach
Author_Institution :
Inst. de Cibernetica, Univ. Politecnica de Catalunya, Barcelona, Spain
Abstract :
In this work additive dynamic neural models are used for the identification of nonlinear plants in online operation. In order to accomplish this task an invariant imbedding method and matrix calculus has been applied to the variational solution of the parameter identification problem to obtain its online version. The work also includes a complexity study of the developed solution
Keywords :
identification; matrix algebra; neural nets; variational techniques; additive dynamic neural networks; complexity study; invariant imbedding approach; invariant imbedding method; matrix calculus; nonlinear plants; online operation; online system identification; parameter identification problem; variational solution; Artificial neural networks; Biological system modeling; Delay lines; Feedforward neural networks; Multi-layer neural network; Neural networks; Neurons; Nonlinear dynamical systems; State-space methods; System identification;
Conference_Titel :
Neural Networks for Identification, Control, Robotics, and Signal/Image Processing, 1996. Proceedings., International Workshop on
Conference_Location :
Venice
Print_ISBN :
0-8186-7456-3
DOI :
10.1109/NICRSP.1996.542745
