Intelligold - An expert system for gold plant process design

Citation
Vm. Torres et al., Intelligold - An expert system for gold plant process design, CYBERN SYST, 31(5), 2000, pp. 591-610
Citations number
13
Categorie Soggetti
AI Robotics and Automatic Control
Journal title
CYBERNETICS AND SYSTEMS
ISSN journal
01969722 → ACNP
Volume
31
Issue
5
Year of publication
2000
Pages
591 - 610
Database
ISI
SICI code
0196-9722(200007/08)31:5<591:I-AESF>2.0.ZU;2-U
Abstract
Gold mining projects are a rare opportunity in the minerals industry. They require relatively small capital and give high profitability and fast retur n on investment compared with other mineral projects. To expand or maintain gold production, continuous development of new deposits and fast implement ation of new mining sites are needed. Process design is one of the major is sues. As simple and easily extractable ores are almost all exhausted, there is a need for a consistent approach to deal with increasing complexity and decreasing or stagnant gold prices. Process design must consider ore genes is, mineralogical characteristics, ore behavior in available metallurgical processes, linkage with the mining method, environmental impact, and econom ic issues. The type of work and environment involved makes this application ideal for using AI tools such as expert systems, fuzzy logic, and neural n etworks. This paper presents Intelligold, an expert system for project deve lopment teams to use at the preliminary evaluation and conceptual project s tages. Information and knowledge from geology/mineralogy, processing, and e conomics are organized, and recommendations on process options and estimate d costs and revenue are given. The "knowledge-building" method is described , together with implementation and verification. Success in building this s ystem suggests application to other ores such as copper and complex base me tals.