Neural learning control of missile interception against dynamics uncertainty and target evasive maneuver

The integrated guidance and control (IGC) design for the advanced interceptors against targets in progress of implementing evasive maneuvers is investigated. During interception, the target's unknown maneuver accelerations, and the extra angle of attack (AOA) caused by wind effects, are considered as disturbances. For the sake of enough hit-to-kill accuracy under multisource disturbances, the neural networks (NNs) and disturbance observer (DOB) are applied to solve the model inaccuracy and external disturbance. As a result, a novel DOB-based neural controller can accomplish high tracking accuracy and hit-to-kill interception performance can be accomplished. Simulations studies demonstrate the validity and advantage bring by the DOB-based neural control.