| Bionic polarized light navigation is an autonomous navigation method that uses the polarized skylight pattern as the compass.It has a large coverage and is insensitive to interference caused by human factors.It has important military significance and research value for autonomous navigation in a satellite-refused environment.In order to solve the key problem of obtaining the position of the sun in the research of bionic polarized light navigation,this paper proposes two characterization methods of polarized skylight pattern ? shaped characteristics,and combined with the harmony search algorithm,the solar position solving model is established,and solve the position of the sun for bionic polarized light navigation.The main research work is as follows:(1)By studying the distribution characteristics of the polarized skylight pattern under different weather conditions,it is found that the polarized skylight pattern ? shaped characteristic is a significant characteristic of the polarized skylight pattern with the characteristics of "stability,uniqueness,ease of measurement,and resolvability".How to extract the ? shaped characteristics from the E-vector of the polarized skylight pattern,and model the representation of the ? shaped characteristics of the polarized skylight pattern from different aspects.(2)The FSIM index of feature similarity is used as the evaluation standard for finding the characteristics image that best matches the desired ? shaped characteristics from the database of polarized skylight pattern ? shaped characteristics model,thus transforming the problem of finding the correct solar position into a global optimization problem An improved harmony search algorithm is proposed,a solar position solving model is established,and the solar position information is solved from the measured polarized skylight pattern image.(3)Conducted experimental research,collected data on polarized skylight pattern under different weather conditions,analyzed the influence of clouds on the measured ? shaped characteristics,and verified the model and the paper proposed in this paper through the measured experiments under different weather conditions The feasibility and effectiveness of the algorithm.By comparing the experimental results under the same experimental conditions,it is found that the ? shaped characteristics model based on edge extraction is better than the ? shaped characterization model based on equivalent sampling in solving the solar position. |
PDF Full Text Request