Simplified interval observer scheme: a new approach for fault diagnosis in instruments

Area of Science:

• Control Systems Engineering
• Fault Detection and Isolation (FDI)
• Process Monitoring

Background:

• Existing fault diagnosis schemes like Simplified Observer Scheme (SOS), Dedicated Observer Scheme (DOS), and Generalized Observer Scheme (GOS) rely on multiple observers and input measurements.
• These conventional methods can be complex and challenging to implement in processes where input measurement is difficult, such as biologic reactors.
• The need for simplified and efficient FDI methods is critical for ensuring the reliability of dynamic systems.

Purpose of the Study:

• To propose a novel fault diagnosis scheme, the Simplified Interval Observer Scheme for Fault Detection and Isolation (SIOS-FDI).
• To develop an FDI method that requires minimal measurements, specifically only one output and no input.
• To enhance the ease of fault diagnosis in processes with challenging input measurement conditions.

Main Methods:

• Development of the Simplified Interval Observer Scheme (SIOS-FDI).
• Utilizing a single observer that processes only one output signal.
• Eliminating the need for any input measurements in the fault detection process.

Main Results:

• The proposed SIOS-FDI scheme successfully detects unique faults in sensors.
• The method simplifies fault diagnosis by not requiring input measurements.
• Demonstrated applicability in complex environments like biologic reactors where input measurement is difficult.

Conclusions:

• SIOS-FDI offers a significant simplification for fault diagnosis in dynamic processes.
• The scheme is particularly advantageous for systems where measuring all inputs is impractical.
• This approach enhances the robustness and implementability of fault detection systems.