Application of Expert Systems for Diagnosing Equipment Failures at Central Energy Plants
Abstract
The growing cost of operating and maintaining its central heating plants (CHPs) has forced the Army to seek alternatives to traditional methods of running these facilities. Computer technology offers the potential to automate and assist in many tasks, such as in the diagnosis of equipment malfunctions and failures in Army CHPs. An automated diagnostic tool for heating plant equipment could lower the cost of human labor by freeing personnel for higher priority work. Automatic diagnosis of problems could also reduce downtime for repair, promote thermal efficiency, and improve on-line reliability. Researchers at the U.S. Army Construction Engineering Research Laboratories (USACERL) investigated the application of artificial intelligence (AI) using knowledge-based expert systems to the monitoring and diagnosing of CHP boiler operations. A prototype system (MAD) was developed to Monitor And Diagnose boiler failure or identify inefficient operation, and recommend action to optimize combustion efficiency. The system includes a knowledge base containing rules for diagnosing the condition of major package boiler components. Minimum system requirements for MAD are an IBM-compatible AT-class personal computer (PC) with 640K base memory and 1 megabyte extended memory, 1.5 megabytes of free hard drive space, a color graphics adaptor (CGA), and DOS 3.0 (or higher).
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1993
- Accession Number
- ADA276909
Entities
People
- Christopher Blazek
- Gary Schanche
- Madhavi Kantamneni
- Mark Metea
- Ralph Moshage
Organizations
- Construction Engineering Research Laboratory