Research On Autonomous Navigation For Planets And Moons Landing Exploration

Navigation method is an important research contents in deep space exploration. Autonomous optical navigation is a new navigation method which has been widely used in recent years. However the traditional image matching method can not apply to aircraft’s navigation for planets and moons landing exploration very well with large calculation quantity and low efficiency. The main studying contents of this thesis are developing a new method of optical navigation which applies to planets and moons landing exploration aircraft, based on graph theory and image matching method.The principal research of optical navigation method based the characterized graphics matching includes two aspects: the star map recognition method and the object’s surface features matching method. Determine the aircraft attitude through the star map matching, measure location and speed of aircraft through the object’s surface features matching. Characterized graphics matching method includes the establishment of the guide star catalogue, image preprocessing, feature points of the object’s surface extracting, the characterized graphics constructing and matching.First, process the star map, it includes the establishment of the guide star catalogue through an equal partition method of celestial sphere, acquires the location coordinates of the star points in the camera coordinate system after image gray processing, anti-white and stretch processing; And then handle the image of object’s surface, collect the center of the celestial craters on the object’s surface by the edge detection, edge token and edge matching. Giving the construction and the matching method of the characterized graphics, using the property of relative distance being unchanged between two characterized points, establishing the connected rules of two characterized points, the method is adopted to overcome the problem of graphics matching with translation, rotation and scale changes.The algorithm is tested by multiple real lunar pictures and star images, the emulational experiment results show that the effectiveness of the method.Navigation method based on the characterized graphics matching is validated by designing a numerical simulation. The simulations reveal that the method that can gain the position, speed and attitude of aircraft with high precision, in near real-time is an effective solution to planets and moons precise landing.