Research On BVO/SINS Integrated Navigation System

Strapdown inertial navigation system is a completely autonomous navigation device,which only uses the measurement information of its own inertial unit to acheve the navigation and positioning service,and does not rely on the external equipments,hence it has strong autonomy,anti-interference ability and concealment.However,due to sensor errors and digital errors,the attitude angles and velocity will drift with time,especially the position obtained by quadratic integration has a strong divergence.So a standlone strapdown inertial navigation system cannot obtain and maintain a high positioning accuracy and reliability during the long period.In order to eliminate the divergence of the inertial navigation system,the inertial navigation system is often intrgrated with other navigation information to achieve higher navigation accuracy by using multi-sensor information fusion algorithm.Binocular vision is becoming a hot topic in the field of machine vision autonomous navigation and positioning,which using the image feature information of the consecutive frames to estimate the relative displacement and rotation of the carrier.The estimated results are not bound by time,and can provide a large number of surrounding environment information.In order to overcome the navigation information divergence problem of the strapdown inertial navigation system,this paper introduces the binocular visual odometer into the strapdown inertial navigation system to establish binocular vision odometer/ strapdown inertial integrated navigation system,and make full use of the complementary characteristics of the two systems to improve the positioning accuracy and reliability of the autonomous navigation.The main contents of this paper are described as follows:Firstly,the two subsystems,strapdown inertial navigation system and binocular vision odometer are introduced,which lays the foundation for the fusion of the two systems.For the strapdown inertial navigation system,the basic working principle and error equations are introduced to provide the basis for the establishment of the intergrated navigation state model.For binocular vision odometer,the binocular camera is firstly calibrated.the image correction and Binocular matching of the images collected by the camera are designed.Based on these,the pose estimation algorithm of binocular vision system is designed,and the accuracy of pose estimation is analyzed by simulation.For the integrated system,the principle of the integrated system is introduce before the information fusion,For the integrated system,the principle of the integrated system is introduce before the information fusion,and then Kalman filter,two-stage Kalman filter and adaptive two-stage filtering algorithms are used to fuse the binocular visual odometer and the strapdown inertial navigation system which using the position and attitude information as the system observation.the advantages and disadvantages of the three fusion algorithms in the binocular vision odometer/ strapdown inertial integrated navigation system are analyzed and compared by simulation results.In addition,this paper discusses the temporal and spatial registration problem of the data fusion in binocular vision odometer/ strapdown inertial integrated navigation system,the corresponding solutions for each kind of the problems are proposed,and the validity and feasibility of the scheme are analyzed by simulation experiment.