A Study of Nuclear Plant Heat Rate Optimization Using Nonlinear Artificial Intelligence and Linear Statistical Analysis Models

منو

" A Study of Nuclear Plant Heat Rate Optimization Using Nonlinear Artificial Intelligence and Linear Statistical Analysis Models "

Document Type	:	Latin Dissertation
Language of Document	:	English
Record Number	:	151614
Doc. No	:	ET23406
Main Entry	:	C l a d e Irvine
Title Proper	:	A Study of Nuclear Plant Heat Rate Optimization Using Nonlinear Artificial Intelligence and Linear Statistical Analysis Models
Note	:	This document is digital این مدرک بصورت الکترونیکی می باشد
Abstract	:	Tlic ernplusis of this dissertation is on developing methods by \vhich a combination of rnulti\-ariatc analysistcchiques (MAT) and artificial intelligence (AI) procedures can be adapted to on-line. real timemonitoring systems for impro\ing nuclear plant thermal efficienq. Present-da\- first principle nlodclsin\.otvc performing a heat balance of piant systems and the reactor coolant s\.stcni. Typical variablesin\.oIved in the plant data acquisition system usually nuniber one-to-two thousand. The goal of the currentwork is tu.ofold. First. simulate the heat n t e wid1 hlAT and A1 computcr models. The sccond objective isto selectively reduce the number of predictors to o n i . the most imponant variables. induce snlaIIperturbations around normal operating levels. and ex-aluate changes in the nia-~tudc of plant eff~cienq. ItIS mticipared ~1131 making snlali cilangcs will improi-e the thennal efficient\- of the plant and lead tosupplcnicntap cost s;~vings.Conclusions of this report are several. A s ~ n ~ i t i \ ~ i h analysis shotved tile reduction of input \.ambles bydin~cnsionaIith. reduction. i.e.. principal component analysis or factor ana(.sis. rc~no\~cs iduableinformation. Predictors can simply be eliminated from the input space. but dimensionality reduction of theinput niatris is nor an alternative option. Houxxxr, perturbation modeling does require data to bestandardized and collincar variables rcmovcd. Filtering of input data is not reconirnendcd cscepl.....-....,....-...,..tested for theQ1 PC1 bus cardBoth these projects mere sofixare des elopment efforts tonards contributing to dlfferentaspects of Roboucs and lZ1echatronics projects m the Controls and Roboucs Group..
Subject	:	Electericl tess
	:	برق
electronic file name	:	TL46645.pdf
Title and statement of responsibility and	:	A Study of Nuclear Plant Heat Rate Optimization Using Nonlinear Artificial Intelligence and Linear Statistical Analysis Models [Thesis]