Adaptive pattern recognition approach for dynamic system control using neural networks

Now showing item 14 of 1858

Title:	Adaptive pattern recognition approach for dynamic system control using neural networks
Author:	Lee, Dennis Tak-Fat
Description:	An adaptive pattern recognition approach implemented using neural networks for control is proposed, and its performance is compared with other conventional and modern approaches. The new design utilizes self-organization and predictive estimation capabilities of neural-net computing. Real-time adaptation is facilitated by the error-based, on-line learning scheme implemented on a cluster-wise segmented associative memory system. The goals of a neural network control system are to be able to manipulate a large number of input and output variables, follow input commands, stabilize the system and satisfy multiple control objectives. Experimental investigation using computer simulation for case studies of the single area megawatt-frequency and megavar-voltage control problem is presented. It is demonstrated that the neural network system is capable of modeling highly nonlinear systems, detecting changes in the dynamic process conditions and stabilizing the system.
Permanent Link:	http://rave.ohiolink.edu/etdc/view?acc_num=case1059488640 http://hdl.handle.net/2374.OX/17498
Date:	1991

Files in this item

Files	Size	Format	View
There are no files associated with this item.

Now showing item 14 of 1858