Design And Analysis Of Velocity Self-Estimation System Based On Sequential Images

Velocity is one of the most important parameters for UAV navigation, especially when velocity was used to locate its position. Closely to the need of military equipments development, velocity measurement related ingredients was studied and a radial velocity measurement method based real-time sequential images was proposed and used to estimate the UAV radial velocity and the landscape orientation velocity when the moon lander was landing to the moon surface in this dissertation. Meanwhile a velocity vector measurement method based on matching of real-time images and reference image was introduced, and the topography problem related in this method was thoroughly studied.One image matching method based on biological vision was use in the first and the second velocity measurement method, while three image matching method including the method based on biological vision, normalized covariance inter-correlation method and least squares image matching method were sued in the third velocity measurement method for accuracy comparison. In order to verify the accuracy of the image matching method based on biological vision, a large number of simulation experiments were carried out in this dissertation.After matching and tracking ground features among sequential images given by the visible or infrared camera, the first velocity measurement method used the camera frame rate and the flying parameters given by the inertial sensors and laser rangefinders to get the instantaneous velocity. While UAV was flying in uniform velocity, the average velocity was received by data process.The first and the second velocity measurement methods had a lot in common, the significant difference was that during the measurement, the flying height gradually fell down. So how to eliminate its influence was the key point. The reference map coordinates in the third velocity measurement method was Gauss plane coordinates, the matching points and the velocity directly measured were both in it. But the geodetic survey coordinates which is needed in UAV navigation is ellipsoid coordinates. In order to get the estern and the western velocity, the difference was analysed and a solution was proposed in this dissertation. Meanwhile a method to get the velocity modification of the inertial navigation system was proposed here. The radial velocity measurement method and the velocity vector measurement method were tested in flying experiments, and the error factors were analysed. The possibility of correct matching acquisition was studied and a method to estimate the possibility of correct velocity acquisition was proposed.