Research On Integrated Navigation Systems Of Vision Inertia And Satellite

Low cost micro-mechanical inertial navigation system is used more and more widely during recent years in the field of consumer electronics, specially in multi-rotor aircraft, camera stabilization platform and virtual reality or augmented reality devices. In order to meet the market demands and achieve the mass production of micro-mechanical gyroscope, silicon is chosen as the main material of micro-mechanical inertial device at this stage. However, for silica-based micro-mechanical gyroscope, the output noise is large and the bias is easily influenced by the environment temperature, which cause that the navigation system can provide attitude information only rather than position or velocity information. The satellite navigation system and monocular vision navigation system can fulfill the task of positioning and velocity measurement in indoor and outdoor respectively, but restricted by the navigation system size and other factors, the attitude measurement is failed. In this paper, based on the low cost micro-mechanical inertial navigation system, satellite navigation system and vision navigation system, an integrated navigation algorithm has been designed, and an integrated navigation system which can be used in indoor and outdoor environment has been implemented. The main research contents and the results of this paper are summarized as follows:1. The coordinate systems of attitude estimation system has been establish, and the transformation relation of coordinate systems have been deducted. An output model of micro-mechanical inertial device has been deducted, and based on the model, a calibration method has been provided. In the method, three-axis turntable has been used to calibrate the scale factor, bias and installation error matrix. Based on the principle of discrete Extended Kalman Filter, an attitude estimation algorithm has been implemented.In the algorithm, the attitude quaternion is chosen as state variable, and the attitude angle is chosen as observations. The hardware and software of micro-mechanical attitude estimation & satellite positioning system has been implemented, and the results of system testing are provided.2. The coordinate systems of monocular vision positioning system has been establish, and the transformation relation of coordinate systems have been deducted. A camera model has been deducted, and the calibration method has been provided. The identification and verification methods of planar-coded marker has been given. In the methods, Open CV and Ar Uco are used. The position and attitude estimation algorithms of camera has been designed, and the vision positioning system has been implemented by actual hardware, and the results of system testing are provided.3. According to the characteristics of low cost navigation system, a filter model ofunified integrated navigation has been presented. Based on the model, the micro-mechanical inertia & satellite integrated navigation system, the micro-mechanical inertia & vision integrated navigation system as well as the micro-mechanical inertia &satellite & vision integrated navigation system have been designed and implemented,and tested in static or dynamic.4. According to the output model of micro-mechanical inertial sensor, a hardware simulation system has been implemented with appropriate attitude estimation algorithm.Under the guidance of the experimental method, some original data has been sampled in micro-mechanical attitude estimation & satellite positioning system and vision positioning system, and the data has been processed under designed integrated navigation algorithm with computer to analyses the performance of algorithm.