Accession Number : ADA292576

Title :   Process Optimization Through Discovery and Integration of Materials and Processing Properties. Phase 1.

Descriptive Note : Final rept. 1 Jul-31 Dec 94,

Corporate Author : AI WARE INC CLEVELAND OH

Personal Author(s) : Yip, Percy

PDF Url : ADA292576

Report Date : JAN 1995

Pagination or Media Count : 22

Abstract : The research objective is to examine process discovery methods for increasing the productivity of materials research with use of computational intelligence paradigms Phase 1 work focused on the study of procedures and enabling technologies to achieve the ultimate goal. The overall work consists of four important tasks. Task 1 was related to process modeling and monitoring. Once this task has been achieved, appropriate control actions can be generated based on an estimation of the current state of a process. In task 2, we focused our study on a feedforward/feedback optimal control methodology with use of neural networks and evolutionary programming. Task 3 was an illustration how a pattern of film thickness estimated from task 1could be used to formulate an objective function. With such an objective function, appropriate control actions can be generated via an optimal control software. Task 4 was concerned with development of an intelligent visualization display enabling users to understand and discover the process behavior. (AN)

Descriptors :   *NEURAL NETS, *OPTIMIZATION, *COMPUTER PROGRAMMING, MATHEMATICAL MODELS, SOFTWARE ENGINEERING, AUTOMATION, THIN FILMS, LEARNING MACHINES, CHEMICAL VAPOR DEPOSITION, EPITAXIAL GROWTH, FEEDBACK, ARTIFICIAL INTELLIGENCE, PRODUCTIVITY, COMPUTER AIDED MANUFACTURING, CONTROL THEORY, ELLIPSOMETERS.

Subject Categories : Cybernetics
      Mfg & Industrial Eng & Control of Product Sys

Distribution Statement : APPROVED FOR PUBLIC RELEASE