Research On Observation Gaps And Sensor Management Of Low-earth Orbit Constellation

Low-earth orbit satellite equiped with long band infrared stare sensor is able to detect and track the dim object in the space such as missile in mid-course, satellite and other spacecraft. By means of the proper design of satellite constellation, sensor resources form a dynamic network which is capable of surveille the spacecrafts around world. It overcomes the shortage of ground based and sky based surveillance systems whose mobility and detect area is limited. Thus it is valuable for space surveillance and hurtle early warning in application. The research on observation gaps during the target's mid-course phase and sensor management algorithm based on the LEO satellite constellation apply to constellation performance evalueation and sensor resources utilization respectively. These researches have significance on satellite sensor resources utilization and optimization.The research work in this dissertation focuses on observation gaps during the target's mid-course phase and sensor management algorithm based on the LEO satellite constellation. The contributions and innovation points are as follows:1,The cause of the observation gaps is studied; the computational model of the observation gap is formed. We also work out the indexes of the observation gaps by which we can evalue the performance of the constellation. Three types of walker-δconstellations are evaluated by the indexes: the number of gaps, the movement area of the gap, the size of the gap in the simulation. Some discoveries are also found from the simulation: different gaps have different movement areas, the movement areas of gaps'are symmetrical by the equator. The dissertation helps to design a proper observation constellation.2,The framework of LEO satellites scheduling for tracking mid-course targets is established. Two kinds of scheduling algorithms are designed——Convariance control method which minimizes the switch number of the sensors and simplified scheduling method which balances the sensor resources and maximizes the total time window. These two methods are compared with the method based on Posterior Cramer-Rao Lower Bound (PCRLB). Simulation shows that each method has its own advantage. PCRLB method ensures the high precision of target tracking during the scheduling periods. Convariance control method minimizes the switch times of sensors and also gets the satisfactory precision. And simplified scheduling method consumes the least time in computation and gets a high precision estimations, too.