Information Fusion of Conflicting Input Data

Area of Science:

• Industrial Engineering
• Data Science
• Control Systems

Background:

• Modern industrial facilities generate vast amounts of complex data, overwhelming human operators.
• Existing information fusion mechanisms face challenges with large data volumes, epistemic uncertainties, and signal conflicts.
• Current solutions partially address these challenges in data fusion for condition monitoring.

Purpose of the Study:

• To propose a novel multilayered information fusion system, MACRO (multilayer attribute-based conflict-reducing observation).
• To introduce the μBalTLCS (fuzzified balanced two-layer conflict solving) fusion algorithm to mitigate conflicts in data fusion.
• To enhance the reliability of condition monitoring in industrial applications through improved information fusion.

Main Methods:

• Development of the MACRO multilayered information fusion system.
• Implementation of the μBalTLCS fuzzified balanced two-layer conflict solving algorithm.
• Evaluation of the MACRO system in a machine condition monitoring application under laboratory conditions.

Main Results:

• The MACRO system effectively reduces the impact of conflicts on the fusion result.
• Demonstrated superior performance of the MACRO system compared to existing state-of-the-art fusion mechanisms.
• The system provides reliable results consistent with real-world conditions.

Conclusions:

• The proposed MACRO system with the μBalTLCS algorithm offers a significant advancement in industrial data fusion.
• This approach enhances the accuracy and reliability of machine condition monitoring.
• The freely accessible utilized data supports further research and validation.