Maintenance System Using Artificial Intelligence

This book presents a new proposal for load pre-dispatch considering the technical condition of the engines of thermoelectric power plants through the combination of several maintenance and diagnostic techniques based on computational intelligence via Fuzzy logic. The diagnosis of the technical condition of the engines is performed using lubricating oil analysis, vibration analysis, and thermography. With these data and statistical analysis, it is possible to foresee when an engine may fail even in pre-dispatch load. To increase the reliability of the motors and of the electric power supply, the Maintenance Management Program (PGM) was carried out, using management tools, applying only 4 pillars of TPM (total productive maintenance) and combining it with predictive maintenance and diagnostics, thus allowing the management to reduce the corrective shutdowns of the plant's equipment. Some results achieved after implementation are: reduction of annual maintenance cost, reduction of corrective maintenance, increase of MTBF (Mean Time Between Failures).