Methodology for Modelling of Physical Processes by Acquisition ofSilent Knowledge from Operators

M. Kristiansen (Denmark)

Keywords

Modelling, reasoning, knowledge discovery, data fusion.

Abstract

Knowledge acquisition of operator knowledge is mainly a silent knowledge source and it is rarely used for producing models of industrial processes, named process models. An example of a model made from operator knowledge is fuzzy logic. In the presented work methodologies to formalise and make use of acquired operator knowledge are developed and presented. The principle is to formalise operator knowledge through interviews to basic descriptions, named single parts, which are clear to the operator. The single parts can then afterwards be aggregated to data, named aggregated parts, usable for modelling. The presented methodologies for acquiring operator knowledge can be used for cheaper and faster production of process models. Especially when comparing with more traditional ways by making expensive empirical experiments and by using rarely available analytical knowledge. Sawing costs for making process models is interesting for the industry because models are required for automation of processes as e.g. welding, painting and milling. The methodology is demonstrated on a welding example and it shows reliable results.

Important Links:



Go Back


IASTED
Rotating Call For Paper Image