Research On The Acquisition Method Of Course And Attitude Information Using The Bionic Polarized Skylight Sensor Array

The skylight polarization pattern, as one natural property of the earth, is the specific polarization state distribution due to the atmospheric scattering of polarized light. Bionic polarized navigation is suitable, stable and has an important significance in unfamiliar environment, which acquires the course information by means of the skylight polarization pattern. Taking the celestial polarized skylight sensor array as the implementation means, and through the study of bionic structure, information acquisition analysis, characteristics extraction and attitude generation method, we explore the mapping between the spatial characteristics of the skylight polarization pattern and the spatial location of the sensor array.The main contents are included as follows:(1) According to the mechanism of desert ant’s compound eye to the polarized light, we design the spherical array of polarized skylight sensor, and analyze the information processing method. Additionally, we propose a course and attitude acquisition method by extracting the spatial characteristics of the skylight polarization pattern.(2) Referring to the angle ambiguity of every sensor unit in the array, we realize an independent4-channel measurement model. Based on this model, we solve the angle ambiguity problem and improve the effective angle range dramatically. Besides, we design and test the polarized light navigation sensor.(3) On the basis of the E-vector distribution of the skylight polarization pattern, we propose two algorithms to estimate the Sun’s position. The corresponding simulation experiments are given to test the algorithms under different kinds of places and time. The results show that acquisition of the Sun’s position is very precisely.(4) We design and construct a celestial sky imaging polarimetric system to analyze the accuracy of the Sun-finding algorithms under different conditions in real skies. The results demonstrate that the algorithms are effective when part of the sky is covered by the cloud and canopy. Additionally, we propose an acquisition method of3D attitude information by choosing the Sun as one characteristic of the skylight polarization pattern when the imaging polarimetric system is tilted from the horizon.