A novel angle computation and calibration algorithm of bio-inspired sky-light polarization navigation sensor

Area of Science:

• Robotics and Autonomous Systems
• Bio-inspired Navigation
• Sensor Technology

Background:

• Traditional navigation systems like Inertial Navigation Systems (INS) and Global Navigation Satellite Systems (GNSS) have limitations, including cumulative errors and susceptibility to signal denial.
• Animal navigation strategies, particularly those utilizing sky polarization patterns, offer a promising alternative for robust positioning.
• Existing polarization-based navigation methods may not fully leverage sensor data or incorporate comprehensive calibration.

Purpose of the Study:

• To develop and evaluate a bio-inspired POLarization Navigation Sensor (POLNS) for enhanced navigation capabilities.
• To introduce an improved method for calculating the input polarization angle using all sensor outputs via Least Squares estimation.
• To present a novel sensor calibration algorithm addressing installation angle errors and sensor biases.

Main Methods:

• Construction of a bio-inspired POLarization Navigation Sensor (POLNS) to detect skylight polarization patterns.
• Application of Least Squares estimation to optimize input polarization angle computation from all POLNS outputs.
• Development and detailed discussion of a new sensor calibration algorithm accounting for installation errors and biases.
• Performance evaluation through simulations and real-world data testing against existing algorithms.

Main Results:

• The Least Squares approach provides optimal input polarization angle estimation.
• The new calibration algorithm effectively addresses installation angle errors and sensor biases.
• Simulations and real-world tests demonstrate the superiority of the proposed algorithms.
• The POLNS system shows enhanced feasibility and effectiveness in practical navigation scenarios.

Conclusions:

• The developed POLNS, coupled with advanced angle estimation and calibration algorithms, offers a robust and effective navigation solution.
• This bio-inspired approach overcomes limitations of traditional INS and GNSS, particularly in challenging environments.
• The findings suggest significant advancements in autonomous navigation technology through the utilization of celestial polarization patterns.