A Redundancy Metric Set within Possibility Theory for Multi-Sensor Systems

Area of Science:

• Intelligent Systems Engineering
• Information Theory
• Sensor Fusion

Background:

• Intelligent technical multi-sensor systems often exhibit inherent or designed information redundancy.
• Sensor redundancy is crucial for improving robustness against failures, facilitating failure detection, and reducing computational load.
• Existing methods struggle with imprecise, incomplete, or noisy sensor data, and detecting spurious redundancy.

Purpose of the Study:

• To propose a novel metric for quantifying the degree of redundancy between sensors in intelligent technical systems.
• To leverage possibility theory for modeling imprecise and incomplete sensor information effectively.
• To ensure the metric avoids detecting spurious redundancy.

Main Methods:

• Developed a metric within the framework of possibility theory to quantify sensor redundancy.
• Defined redundancy specifically for possibilistic information.
• Specified requirements for an effective redundancy metric.
• Detailed the information processing steps for the proposed metric.

Main Results:

• The proposed metric quantifies sensor redundancy using possibility theory.
• The approach effectively models imprecise and incomplete sensor data.
• Qualitative evaluation on three technical datasets demonstrated the metric's utility.
• The metric is designed to avoid identifying spurious redundancy.

Conclusions:

• The developed metric offers a robust way to measure sensor redundancy in technical and cyber-physical systems.
• Possibility theory provides a suitable framework for handling uncertain sensor information.
• The metric contributes to more reliable and efficient multi-sensor system design by addressing data uncertainty and redundancy.