ADAPTIVE-CONTROL AND INTELLIGENT AUTOMATION

Adaptive filters and controllers have the capability to automatically adjust their tuning parameters to maintain desired performance in a changing environment. The basic design concepts behind those algorithms will be presented and some of their capabilities described. These algorithms can be used in process control to provide information about the process that complements that provided by sensors and analytical instruments. For instance, nonlinear adaptive control could provide on-line estimates of physical and chemical parameters of a system. In addition, due to the recent availability of affordable digital signal processing chips, these algorithms can easily be imbedded in sensors and analytical instruments to provide them with new capabilities. For instance, analytical instruments could be able to detect and compensate interferences, using adaptive noise cancellation; to adapt their parameters to changing conditions and to diagnose performance degradation, using adaptive control; and to automatically find the optimal analysis conditions, using extremal control. Finally, progress in neural networks may allow those instruments to acquire learning capabilities and to automatically interpret results using pattern recognition techniques. © 1990.