Font Size: a A A

Study Of Restraining False Alarm Based On Kalman Filtering In Collision Warning Of Two Spacecrafts

Posted on:2017-03-02Degree:MasterType:Thesis
Country:ChinaCandidate:G ZhangFull Text:PDF
GTID:2282330509957363Subject:Aerospace engineering
Abstract/Summary:
In the recent years, as space debris gradually increase, the research on restraining false alarm in collision warning of spacecrafts becomes more and more important in order to ensure normal operation of the in-orbit spacecraft. In the paper, the way of restraining false alarm in collision warning between spacecraft and space target is studied based on the relative motion dynamic model and Kalman filtering method.First, C-W equation is introduced in the paper. The relative dynamic equation diverges as the time increases. We studies two spacecrafts with almost same orbital altitude in the paper. Therefore, C-W equation can be simplified deeply. The term that causes divergence has not been eliminated in the simplified relative equation, it makes the result converge by decreasing magnitude of coefficient of time variable. Second, it chooses a combination of the ground-based method and the space-based method when measuring the relative state. In view of different nonlinear systems, EKF and UKF were used to estimate the relative state that between the two space targets, respectively. By simulation, it can be known that UKF has higher accuracy and faster convergence speed than EKF. Therefore, it is suitable to use the simplified equation and the UKF algorithm of Doppler radar for the measurement equipment. Finally, the false alarm problem in the collision warning process is studied. We predict the error of the relative state by using different relative dynamic equations, introduces a collision probability model, and calculate the collision probability. The way of restraining false alarm in collision warning is analyzed from three aspects: the relative dynamic equation, the early warning threshold, and the measurement error. The influence of the initial state measurement error on the collision probability is also analyzed.
Keywords/Search Tags:relative dynamic model, kalman filtering, collision probability, restraining false alarm
Related items