How can AI procedures become more effective for manufacturing? (original) (raw)
The problcmls of AI effectivent~s in manufacturing for Design, Scheduling, Control and Proce~-.; Diagnosis are considered. We have developed an effective dialog procedure for a designer. This procedure I~ei~ him to identify file no..xh.'d parameters of a d~igned product, i.e., to distinct acceptable paranleters, utm acceptable parameters and paran~eters that require additional design sttgly. In ~lx'x'luling we developed a new intelligent procedure to formulate and find an effective schedule. Often st,ch approach can avoid CtmlplicattvJ time-ctmsunling ctmlpt,tations. In Ctmtrol we developed simple and robust control lm~cedures, which join the advantages of pur~ctmveutitmal interpolation and fuzzy ctmtrol methtxts for design of quick and cost-effective controllers. In Prc~e~s Diagnosis we overct~ne some difficulties of such known methods as neural networks, linear discriminant analysis and the method of nearest ncighlx~rs. The main difficulties which we overcome are related to the speed of a dynamic learning protein and reliability of diagnosis. We also make the extracted diagnostic regularities easily understandable by a manufacturing expert. This arcroach was succe~fidly tlsed for ~vt~'al tasks related to cngineering and medical probhmls.
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.