AN ARTIFICIAL NEURAL NETWORK BASED-SYSTEM FOR AUTOMATIC GAS RECOGNITION

Abstract

The economic realities of productivity, quality, and reliability for the industrial societies of the 21st century are placing major demands on existing manufacturing technologies To meet both present and anticipated requirements, new and improved methods are needed. It is now recogmzed that these methods must be based on the puweitiJI techniqueemploying computer­ assisted systems full computer compatib!lirv of all components and system must be clllned till. With the widespread usc ,1!' the artiticml neural n<.'twurks a

complex problems -;uch as Image, vn1cc. pattel!I, < nd signal processmg are ideallv done. App!Jcation of the ai1Ificial neural network on gas rce<Jhrnition is mvestigatcd m this thesis_ trying to (>Jfer more p(merlul, and accurate momtoring and recognition s'y stc:m t<n impwving work safctv and reducing the du\vntime A complete autPmatic gJs recognition svstcm was desif,.'lled Tht: syst<:m include dectruchemical gas -;ensor _ A/ [) card, and the digita! computer_ bc:siJe the neecssarv circmts for installation and 1nterfacmg S,Jftwan: programs have been designed and Implemented to pertollli the tasks uf data >:<JIIecting, and the neural network simulatidn, trainmg_ anJ testing. Also manv -;o1hvarc packages are used.