Efficient Nonlinear Algorithm for Drag Tracking in Entry Guidance

AbstractIn drag acceleration guidance for anentry vehicle, a drag-vs-velocity or drag-vs-energyprofile needs to be tracked. If there are modeling errors, tracking laws derived from feedback linearizationmay have poor performances. Tracking laws using nonlinear model predictive control (NMPC) have satisfactory tracking accuracy, but they are computational intensive. A new tracking law concerning both tracking accuracy and computational efficiency is proposed based on NMPC. First, receding-horizon optimization in NMPC is replaced by random searching. Second, at each guidance cycle, corresponding trajectories in the predictive horizon are obtained using random generated constant bank angles, and the commanded bank angle is the value leading to the minimum performance index. Third, bank rate constraint is used to shrink the searching space before the startup of random searching. Fourth, a second-order filter is used to estimate the true drag accelerations, and an additional logic modifying the tracking law is adopted to take into account the downrange error. Finally, the new tracking law is tested by the simulation of 500 entry cases with modeling errors. Simulation results indicate that the new tracking law has both high tracking accuracy and high computational efficiency and therefore is more suitable for online missions