Single Fault Diagnosis Method of Sensors in Cascade System Based on Data-Driven

Area of Science:

• Control Engineering
• System Safety
• Data-Driven Analytics

Background:

• Cascade systems are widely used, increasing the need for enhanced reliability and safety.
• Fault detection and diagnosis are critical for maintaining operational integrity.
• The complexity of double closed-loop systems presents challenges for accurate fault diagnosis.

Purpose of the Study:

• To propose a real-time fault diagnosis method for single sensor faults in second-order valued systems within cascade systems.
• To improve the reliability and safety of widely applied cascade systems through advanced fault detection and diagnosis.

Main Methods:

• Development of static fault detection, location, and estimation models using off-line data.
• Real-time calibration of static models with on-line data to create dynamic fault diagnosis models.
• Implementation of real-time calibration procedures, workflow, and anti-interference measures.

Main Results:

• Experimental validation demonstrating the effectiveness and accuracy of the proposed data-driven fault diagnosis method.
• Successful detection, location, and estimation of single sensor faults in second-order valued systems.
• The method proved suitable for general cascade systems.

Conclusions:

• The proposed real-time fault diagnosis method is valid and accurate for cascade systems.
• Data-driven approaches offer a robust solution for complex fault diagnosis in industrial applications.
• This method contributes to enhancing the overall safety and reliability of cascade systems.