BASIC NEURAL COMPUTING CONCEPTS ARE DESCRIBED AND THEN APPLIED TO DEVELOPING AN ADAPTIVE PID (APID) CONTROLLER. AUTOMATICALLY RETUNING THE PARAMETERS IN THE PID CONTROLLER RESULTS IN BETTER CONTROL OF THE PLANT, IMPROVED QUALITY, AND REDUCED MAINTENANCE COSTS. SIMULATION RESULTS SHOWED THAT THE APID CONTROLLER REDUCED THE ERROR VARIANCE BY 50% OVER A RANGE OF OPERATING CONDITIONS. A PID CONTROLLER TUNED TO THE APPLICATION IS GRAPHICALLY COMPARED TO THE APID.
| Number of Pages : | 21 |
| Published : | 06/01/1991 |
