Adaptive Learning in the Design of Pipe Network Layout Geometry

		/vol02/j_davids/	© Copyright 1995
		Volume 02	Received: Accepted:	---- ----

Adaptive Learning in the Design of Pipe Network Layout Geometry Jim W. Davidson and Ian C. Goulter
Abstract
	A rule-based learning system for optimising the layout geometry of pipe networks is described. The system adaptively learns production rules through autonomous experimentation with sample problems. The learning system consists of three components: an agitator, a rule formulator, and a production system. The agitator and rule formulator allow new rules to be derived and added to the rule base to continually extend the ability of the production system to improve solutions. The rule base is self-organising through periodic reassessment of the effectiveness of the rules on the basis of their recorded performance, with less effective rules being eliminated. Two rule bases are compared on the basis of performance on a sample problem and although the rule bases are of similar size, the rule base that has evolved over a longer period significantly outperforms the other.

Full Text

Multimedia Links
(none)

Reference Links
(none)

Citation Reference

Get viewers
for PS & PDF

[CI Editor] [Site Manager]