A real-time knowledge-based system for intelligent monitoring in complex, sensor-rich environments

Abstract

Abstract Real-time and on-line optimisation methods for control of large-scale systems use real-time sensor data to optimise the system's performance. The data need to be pre-processed for faults to safeguard against inappropriate control actions. Conventional real-time signal processing techniques are likely to fail because of the system's complexity and the processing speed required. A generic, real-time knowledge-based system—FLASH (FLexible AI for Signal Handling in real time)—has been developed to detect, diagnose and replace faulty sensor readings. Its effectiveness in processing signals for control in a complex, sensor-rich environment has been demonstrated in real-time environmental control of greenhouse microclimate.