The Study Of UAV GPS/SINS Integration Navigation Based On MEMS

With the rapid development of unmanned aerial vehicle (UAV), the requirement for thenavigation system is also increasing day by day, and integrated navigation mode combinedwith global positioning system (GPS) and strap-down inertial navigation system (SINS)based on MEMS technology is both the mainstream of the current and development trendof the future, so it’s necessary to improve its reliability and accuracy. But in practicalapplications, integrated navigation will not be able to continue to work because of signalloss and interference. In order to solve this problem, this article discusses the navigationwith the aid of BP neural network, and the GPS/SINS system are studied and analyzed indepth.First of all, the composition and characteristics of GPS and SINS are analyzed, and itsworking principle including speed measurement, Orientation, force equation of SINS, mechanical layout and error equation and so on is also introduced and derived.Secondly, gesture decoding algorithms are analyzed in detail in SINS attitudealgorithm. In view of the rotation vector method, the precision of the algorithm is improvedrespectively from three directions which are increasing the sample, using angular rate andusing fast loop to increase the updating frequency of angle incremental considering the badenvironment of the coning motion. Then the corresponding improved algorithm is derived,and the experimental contrast show that optimized algorithm has superiority of higherprecision and smaller drift error without increasing computational complexity, which alsodiscussed and used almost all ideas for increasing accuracy of rotation vector algorithm.The GPS/SINS integrated navigation system is designed by using loose couplingmethod and open-loop indirect method. The principle and contents of the Kalman filteringalgorithm is discussed, and three kinds of data fusion including Kalman filter (KF), theextended Kalman filter (EKF) and the Unscented Kalman Filter (UKF) are deduced. Thenthe experiment shows that integrated navigation of GPS and SINS is obviously better than the single navigation in precision, and it is better when using UKF.Considering the situation of blockage of satellite’s signals, the method of BP neuralnetwork auxiliary is put forward and the navigation work mode is designed. Theexperiment shows that the BP can aid navigation system and ensure the navigationprecision when lack of GPS signal, so as to improve the adaptability and reliability of thenavigation system.