A Kalman-Filter Based Ballistic Missile Navigation Algorithm And The DSP Reai-Time Simulation

This paper investigates autonomous navigation of a ballistic missile during the free-flight phase. In general, the ballistic missile is navigated using the inertial measurement units (IMUs), whereas a long-time flight will cause a dramatic degradation of the inertial navigation performance for the IMUs biases. Obviously, if the initial state of an inertial navigation system is updated, e.g., by another navigation method, the performance of the inertial navigation system will be improved a lot. In this thesis, an extended Kalman filter (EKF) based navigation method, which uses the orientation information of the Sun, the Earth, and the Moon, has been proposed to complete this mission. In addition, a PC-based mathematic simulation and a DSP-based real-time simulation have been executed to validate the new navigation algorithm.The ballistic determination, using the orientation information of the Sun, the Earth, and the Moon, is based on an obvious fact, that is, once the"Moon-Missile-Earth"angle and the"Sun-Missile-Earth"angle observations are only acquired, also with the known inertial positions of the Sun and the Moon relative to the Earth, the inertial position of the missile will be determined uniquely. It is well known that the geometric determination algorithms are not competing and promising for several apparent drawbacks: relative large errors, completely observable measurements at a single time, etc.An EKF can solve the nonlinear estimation problem with the observations of the"Moon-Missile-Sun"angle and the"Sun-Missile-Earth"angle. The sub-optimal filter, which uses a 6×1 vector (3 position components and 3 velocity components) as well as the corresponding covariance matrix to evaluate the performance of the filter, linearizes the original nonlinear system model and nonlinear observation model at their respective reference sample points. Even if the Moon, or the Sun, or both measurements is possibly not to be acquired in a short time, the EKF can still yield a considerably good navigation performance, which shows the filter's robustness in defect of data.A mathematic simulation has been used to validate the navigation algorithm above and the practicability. First, a"true"ballistic trajectory has been...