Self-adaptive robust algorithm applied on compact real-time polarization sensor for navigation

Abstract

Celestial navigation has remarkable performance in autonomous navigation without error accumulation. For supplying orientation information to a portable vehicle in real time, a compact polarization sensor with high precision and strong robustness was constructed in this study based on the mechanism of insect polarization-opponent neurons structure. The sensor, composed of polarization-dependent module and microcontroller unit, could supply heading angle for real-time navigation. To enhance the robustness of sensor in a complex environment, a self-adaptive algorithm based on information entropy theory was proposed. The algorithm could eliminate the effect of random obstacles in the sensor field under constantly changing light conditions. The laboratory and outdoor tests indicated that the sensor, with ±0.1 deg measured error and ±0.4 deg fluctuation under natural obstacle, had potential application prospects in polarization navigation.