Research Of UAV Attitude Optimization Method Based On Sky Light Polarization Mode

In fact,like humans,most of the creatures in nature need to solve navigation problems.In the long evolutionary process,almost all animals have evolved the necessary navigational capabilities.For example,positioning and direction finding by perceiving polarization distribution mode is one of them.The paper is mainly based on the polarization distribution mode of sky light to realize the navigation and positioning function of the UAV.Meanwhile,combined with the optimization theory and the hybrid filtering idea,the new attitude determination method is designed to provide a new solution for the UAV control system.The paper starts with the analysis of the basic mathematical principles of navigation system attitude and the common coordinate system of navigation system.By comparing and analyzing the characteristics of different attitude description methods,it is determined that the quaternion is used as the mathematical basis for the attitude calculation.Then,the principle of navigation orientation of polarized-light sensor is analyzed in detail.At the same time,a method of fast acquisition of full attitude information based on dual polarization vector is derived theoretically.Considering the experimental test analysis process of polarized light data fusion algorithm,the paper designs a set of UAV attitude measurement experimental platform.For this purpose,the paper analyzes the hardware solutions and software solutions of the experimental platform in detail.Before the flight,the polarization vector sensor extended range method and the flight control system performance test were carried out by using a precise rotary positioning system.Taking the problem of insufficient precision of the existing polarized light navigation and positioning method into account,the paper is based on quaternion method and gradient optimization theory and designs a dynamic step gradient optimization attitude determination method based on polarized-light vector field and gravity vector field,which solves the optimal attitude quaternion using the established target error function model.The test results show that the method not only guarantees the attitude stability in static environment,but also the accuracy of the attitude calculation in dynamic environment is higher than the fixed step gradient optimization method,which shows good dynamic performance.After establishing a gradient-optimized heading solution model based on atmospheric polarization mode,in order to effectively balance the accuracy and real-time of the attitude calculation,the paper designs a hybrid filtering model based on gradient optimization.Can be seen from the test results of static experiments and dynamic experiments,the hybrid filtering model can eliminate the internal noise interference of the sensor to a greater extent,which leads to a further enhancement of the static positioning accuracy.At the same time,under dynamic conditions,the hybrid filtering method overcomes the overshoot problem in the process of gradient optimization estimation,and possesses a better effect of tracking and a more timely response.Subsequently,in order to improve the performance of the carrier in a mobile environment,the initial model structure is improved,and a hybrid filtering adaptive control scheme for suppressing motion acceleration interference is designed.The test results show that the solution can effectively remove harmful acceleration interference in complex motion environments,furthermore,the environmental adaptability and robustness of the polarized light navigation system are improved while the accuracy of the attitude calculation is ensured.