Sensor Data Fusion with Z-Numbers and Its Application in Fault Diagnosis

Area of Science:

• Engineering
• Computer Science
• Information Science

Background:

• Sensor data fusion is crucial for fault diagnosis.
• Existing methods often overlook sensor reliability, focusing mainly on data fuzziness.
• Uncertainty in sensor information, including randomness and fuzziness, requires robust handling.

Purpose of the Study:

• To propose a novel method for sensor data fusion that accounts for both fuzziness and reliability.
• To enhance the robustness and reliability of fault detection systems.
• To address the limitations of existing methods in handling sensor reliability.

Main Methods:

• Utilizing Z-numbers to simultaneously model the fuzziness and reliability of sensor data.
• Applying Dempster-Shafer (D-S) evidence theory to fuse uncertain information represented by Z-numbers.
• Developing a combined Z-number and D-S evidence theory framework for sensor data fusion.

Main Results:

• The proposed method provides a more robust measure of reliability for sensor data.
• Fusion of complementary information from multiple sensors reduces fault recognition uncertainty.
• Enhanced reliability in fault detection is achieved through improved data processing.

Conclusions:

• The integration of Z-numbers and D-S evidence theory offers a powerful approach for sensor data fusion.
• This method effectively handles the dual nature of uncertainty (fuzziness and reliability) in sensor data.
• The approach significantly improves the accuracy and dependability of fault diagnosis systems.