J. Kobalicek et al., EXAMPLES OF THE APPLICATION OF FUZZY EXPERT SYSTEMS TO RUBBER COMPOUNDING, Plastics, rubber and composites processing and applications, 19(3), 1993, pp. 159-173
A classical quantitative (analytical and/or statistical) analysis is n
ot appropriate for some ill-defined and/or very complex polymer/rubber
engineering problems. Therefore, a new form of analysis using fuzzy m
athematics has been applied. The fuzzy model can utilize semiqualitati
ve and semisubjective data which are to a certain level inconsistent.
To minimise an information loss of valuable knowledge a conventional p
re-processing (usually statistical analysis) of this knowledge is elim
inated. A primary knowledge (e.g., experimental results) is used direc
tly. Conventional experimental and others' records (usually tables) us
ed in plastics/rubber engineering are not suitable for an efficient un
certainty reasoning because of bad knowledge acquisition methods used
to develop them. A revitalization of these valuable records is a retro
spective application of knowledge engineering algorithms. The final go
al of the revitalization is to insert additional information items (e.
g., expert guess) to increase the reasoning power. Two realistic case
studies (rubber compounds conventional knowledge revitalization) are g
iven.