Pazy-winged aircraft aeroelastic control benchmark: An open simulator

A hierarchical nonlinear control benchmark is presented for gust load alleviation (GLA) and maneuver load alleviation (MLA) in a highly flexible winged UAV. The benchmark utilizes a reduced-order nonlinear aeroelastic model, which has been shown to be suitable for real-time control. This work builds on an existing moving-horizon estimation (MHE) and model predictive control (MPC) strategy for flexible aircraft stabilization by building a control framework designed for a wider range of aggressive movements in turbulent environments. The mitigation of induced wing loads is handled by an inner-loop nonlinear MHE (nMHE) and nonlinear MPC (nMPC) that estimates the aeroelastic states and disturbances during maneuvers or gusts. Due to the demanding computational nature of MHE and MPC strategies, the nominal control and motion is handled by an outer-loop LQR regulator. The example Pazy-winged aircraft, which assumes the other components to be relatively rigid compared to the wings, uses shared control surfaces for load alleviation and nominal control. This requires an additional complementary command-fusion scheme which combines the outer-loop (low frequency) and inner-loop (mid/high frequency) control efforts prior to actuator models with saturation and anti-windup. The framework is packaged as a MATLAB simulator ran from a high fidelity open source aeroelastic simulator with example cases, scoring metrics, and post-processing to enable direct comparisons to existing works. Cases involving attitude holds, speed changes, and relatively aggressive turns under turbulence demonstrated satisfactory results in reducing wing root bending moment, wingtip deflection, and rigid-body tracking error (H 2 norm). The goal is a benchmark that can be used for future real-time hardware-in-the-loop control capabilities for a developed flexible aircraft.