IMPROVING CONTROL ALLOCATION ACCURACY FOR NONLINEAR AIRCRAFT DYNAMICS

摘要: The forces and moments produced by a vehicle’s aerodynamic control surfaces are often nonlinear functions of surface deflection. This phenomenon limits the accuracy state-of-the-art allocation algorithms since all approaches based on assumption that variable rates linear deflections rate increments not for zero deflections. errors introduced this currently mitigated robustness resulting from feedback laws. A method improving performance control/control system is presented directly attacks inaccuracies these assumptions. approach demonstrated using dynamic inversion-based law developed lifting body with four surfaces. commanded constitute input to inversion forms inner-loop structure. outputs algorithm serve as inputs programming allocator. mixed optimization scheme minimizes norm error deviation preferred value. Whether underlying or in way controls affect state, allocator requires approximation inputs. In work, instead describe due deflection (a subspace, contains element), an affine relationship utilized. allows one modify providing additional intercept term corrects algorithm’s linearity. Since on-line allocators perform computations at flight update rates, decision about how far move next time interval critical. order accurately compute set commands, local slope current operating point used accounted With relatively little computational design overhead, can be improved utilizing dependent accelerations. Simulation experiments show significant improvement tracking when correction used.