An integrated framework for intelligent monitoring, diagnosis and supervisory control of processes

In process industries, monitoring, fault diagnosis and supervisory control are the key tasks. In the past two decades, it has become increasingly obvious that the performance of process control systems depends not only on the control algorithms but also on the integration of these algorithms into a concise operational framework in terms of safety, environmental protection, maintenance strategy, and general monitoring to identify poor performance and detect faults. Hence, there is a need to create a virtual environment which enables comprehensive assessment of process operations and can identify the factors which causes undesirable outcomes. This paper presents a conceptual framework for integrating emerging data pre-processing technology with knowledge discovery and pattern recognition from process data techniques, and combines various solution approaches for designing state space based process monitoring, diagnosis and control systems. The resulting system would help to solve the problems of process information management and serve as an intelligent aide for the operator