Research On The Algorithm Of Inertial/visual Integrated Navigation System

With the advancement of science and technology,navigation and positioning technology has become an indispensable part in daily life.With its superior autonomy and rich navigation information,inertial navigation technology is always an important technology in navigation.Among them,strapdown inertial navigation system is the mainstream research direction.Due to the improvement of the manufacture level of inertial devices,inertial navigation technology is widely used in military,aviation and aerospace fields.In addition,the visual navigation system is also one of the widely used methods in navigation.In recent years,due to the rapid development of technologies such as robots,artificial intelligence and AR,visual navigation technology has gradually entered our lives.According to visual navigation,more comprehensive information can be obtained,and many additional functions can be realized based on this.However,both inertial navigation technology and visual navigation technology have their own advantages and limitations,and the fusion of the two has gradually become a new hot research topic.In this research background,this paper carried out the algorithm research on inertial / visual integrated navigation system,the main work of this paper is as follows:(1)Briefly introduce the research status of inertial navigation,visual navigation and integrated navigation system of the two,and expound the research background and significance of this article.(2)Summarize the algorithm flow of strapdown inertial navigation system.The attitude,velocity and position updating algorithm are deduced in detail,and the error equations of strapdown inertial navigation system is given,which provide a theory basis for the establishment of the state equation of the integrated system.(3)Introduce the camera imaging principle,give out the algorithm flow of the binocular visual navigation system,and explain the details in the algorithm flow.The algorithm extracts ORB feature points for matching and tracking to perform pose estimation.Optimize the visual system by a nonlinear algorithm(BA).In the maintenance of key frames,a sliding window method is used to maintain the scale of the local map.We use the KITTI public data set to testify the effect of the algorithm.The experimental results show that the visual navigation algorithm has good accuracy and reliability.(4)In order to improve the stability of the system,based on the idea of indirect Kalman filter algorithm,the strapdown inertial navigation system and the stereo visual navigation system are loosely integrated.The algorithm flow of the integrated system is expounded,the algorithm principle of initial alignment of inertial navigation is given,and the state equation and observation equation of the integrated system are derived.Similarly,the algorithm was experimentally tested using the KITTI public data set.Experimental results show that the results of the integrated navigation system are superior to the single system,and the accuracy has been improved to a certain extent.