Optimal guidance and nonlinear estimation for interception of decelerating targets

Optimal guidance and nonlinear estimation algorithms are formulated for interception of an accelerating target vehicle during boost. For an interceptor with two-axis control of translational acceleration, time to go may be selected to null the component of commanded acceleration along the uncontrolled axis. A nine-state, extended Kalman filter is formulated, in a Cartesian-inertial frame. The filter dynamics model includes a vector-differential equation for the thrust acceleration vector of the target during a gravity-turn maneuver. With angle measurements from a strapdown seeker, very small miss distances can be achieved, despite large estimation errors in range, because of the time-to-go algorithm. Monte Carlo simulations are used to generate theoretical collision probabilities as functions of sensor measurement accuracy and filter update rate.