Accession Number : ADA302927

Title :   Smart Composite Damage Assessment System Based on the Neural Network,

Corporate Author : NATIONAL AIR INTELLIGENCE CENTER WRIGHT-PATTERSON AFB OH

Personal Author(s) : Yungang, Tao ; Baoqi, Tao

PDF Url : ADA302927

Report Date : 04 DEC 1995

Pagination or Media Count : 13

Abstract : This article introduces a new type of composite material damage assessment system. The system in question is composed of embedded optical fiber sensor arrays, shape memory alloy wires, and Kohenen self-organizing neural network processors. Material damage detection is realized by embedded optical fiber sensor arrays. High speed neural network parallel distribution processors composed of TM5320C25 high speed parallel processors and IBM PC/386's carry out simulations, realizing sensor output signal real time processing. In conjunction with that, corresponding control signals are produced to actuate shape memory alloy (SMA) wires in order to change material stress configurations and delay material destruction. (MM)

Descriptors :   *NEURAL NETS, *COMPOSITE MATERIALS, FIBER OPTICS, DAMAGE ASSESSMENT, PARALLEL PROCESSORS, MEMORY DEVICES, ARTIFICIAL INTELLIGENCE, TRANSLATIONS, OPTICAL DETECTORS, CHINA, CHINESE LANGUAGE.

Subject Categories : Laminates and Composite Materials
      Cybernetics

Distribution Statement : APPROVED FOR PUBLIC RELEASE