August 1997, Volume 3, No. 8
|
|
August 1997, Volume 3, No. 8 Lafarge Optimizes Cement Production with Intelligent Software |
|
|
Thermocouples at the intake and back-end discharge are used, along with non-contact infrared sensors and pyrometers to measure material inside the kiln, but even these are limited to surface measurements, and dust in the kiln can compromise the accuracy of these. In addition, temperature and feed rates are related to the level of dust in the process. Among the complexities of both product and process is the natural variation of limestone coming from the quarries. Since all of the physical changes occurring in the kiln are highly complex and cannot be measured by conventional means, an accurate mathematical model could not be developed. Lafarge decided to try to mimic the decision making of the operator, who does not know all the equations underlying the process, but who knows what to do when the kiln is too hot. The result was a system that modeled the kiln with mathematical equations in conjunction with the knowledge and decision making of the operator. Having a graphic interface and object-oriented technology at the fundamental level was a major criteria for the system. In an object-oriented system, processes and equipment are defined as graphic icons that can be manipulated like blocks in a flow chart. Engineers can directly configure large chunks of the control systems by moving the icons, rather than by writing complex computer code. When a change is made, the system automatically understands and makes the appropriate changes to the process logic. Lafarge uses the G2 intelligent system software from Gensym (Cambridge, Mass.; http://www.gensym.com). G2 uses a rule-based inference engine that can reason about the collected data much the way a human does. The rules themselves closely resemble natural language, and may say, "If the kiln is a little bit too hot, then cut the fuel by 3%." Lafarge uses seven levels of linguistic hedges like "a little, a lot, or very" to express the degree of membership in various sets of process conditions on the "if" side of a rule set. In turn, "then" statements characterize various control actions to be taken. Under the previous control system, some of the operator decisions were still manual, which left a margin of error. Now, with the expert system tied in a closed and automated control loop, standard deviation from real-time temperature models has been cut in half. The benefit relates to the intelligent software's ability to direct control actions through the supervisory control application and remote I/O, that was not possible with the prior control system. G2 monitors and can make appropriate adjustments on a minute-by-minute basis. The widespread application of intelligent software has provided a common technical culture among the worldwide plants of this company. In the past, personnel at various plants believed that their problems were unique. Lafarge has established that a common operating philosophy can be successful at any of their facilities. Monetarily, the system has shown savings in burner fuel alone estimated at $250,000 per plant, per year. Additional benefits are realized in reduced emissions and in an increase in the replacement interval of refractory bricks by as much as 50%. |
|
|
Intelligent Manufacturing Resources: Contents of Current Issue | Intelligent Manufacturing Home Page Subscriber Access — Index of Past Issues | Free Access — Index of Past Issues Search | Subscriptions | Reprints | Links E-mail to the Editorial Department of Intelligent Manufacturing: Lionheart Site Resources: Lionheart's Home Page | What's New | Site Index | Newsroom | Contact Software Surveys | Online Sponsors | Visitor Survey Other Publications: APICS — The Performance Advantage BDO Seidman, LLP Comprehensive Guide to Manufacturing Software Operations Research/Management Science Today | Intelligent Systems Report Subject Areas: Manufacturing & Resource Planning Operations Research and Management Science | Intelligent Systems | Lionheart's Publishing Services Copyright © 2020 by Lionheart Publishing, Inc. All rights reserved. Lionheart Publishing, Inc. 2555 Cumberland Parkway, Suite 299, Atlanta, GA 30339 USA Phone: +44 23 8110 3411 | e-mail: | URL: http://lionhrtpub.com |
