DESIGN, ANALYSIS AND REAT-TIME REALIZATION OF ARTIFICIAL NEURAL NETWORK FOR CONTROL AND CLASSIFICATION

دانشگاه آزاد اسلامی اصفهان(خوراسگان)

منو

رکورد قبلی

رکورد بعدی

" DESIGN, ANALYSIS AND REAT-TIME REALIZATION OF ARTIFICIAL NEURAL NETWORK FOR CONTROL AND CLASSIFICATION "

Document Type	:	Latin Dissertation
Language of Document	:	English
Record Number	:	148540
Doc. No	:	ET20332
Main Entry	:	PUXUAN DONG
Title Proper	:	DESIGN, ANALYSIS AND REAT-TIME REALIZATION OF ARTIFICIAL NEURAL NETWORK FOR CONTROL AND CLASSIFICATION
Note	:	This document is digital این مدرک بصورت الکترونیکی می باشد
Abstract	:	DONG, PUXUAN. Design, Analysis and Real-Time Realization of Artificial NeuralNetwork for Control and Classification. (Under the direction of Dr. Griff L. Bilbro.)Artificial neural networks (ANNs) are parallel architectures for processing informationeven though they are usually realized on general-purpose digital computers. This researchhas been focused on the design, analysis and real-time realization of artificial neuralnetworks using programmable analog hardware for control and classification.We have investigated field programmable analog arrays (FPAAs) for realizing artificialneural networks (ANN). Our research results and products include a general theoreticallimit on the number of neurons required by an ANN to classify a given number of datapoints, a design methodology for the efficient use of specific FPAA resources in ANNapplications, several multi-chip FPAA implementations of ANNs for classificationexperiments, several single-chip FPAA implementations of analog PID controllers for anunmanned ground vehicle (UGV), experimental evaluation of FPAA PID controllers witha conventional digital PID controller on a UGV, and finally a single-chip FPAAimplementation of a (non-linear) ANN controller.
Subject	:	Electericl tess
	:	برق
electronic file name	:	TL43459.pdf
Title and statement of responsibility and	:	DESIGN, ANALYSIS AND REAT-TIME REALIZATION OF ARTIFICIAL NEURAL NETWORK FOR CONTROL AND CLASSIFICATION [Thesis]