David G. Ward

Integrated Adaptive Guidance and Control for Re-Entry Vehicles with Flight Test Results

To enable autonomous operation of future reusable launch vehicles, reconfiguration technologies will be needed to facilitate mission recovery following a major anomalous event. The Air Force’s Integrated Adaptive Guidance and Control program developed such a system for Boeing’s X-40A, and the total in-flight simulator research aircraft was employed to flight test the algorithms during approach and landing. The inner loop employs a modelfollowing/dynamic-inversion approach with optimal control allocation to account for control-surface failures. Further, the reference-model bandwidth is reduced if the control authority in any one axis is depleted as a result of control effector saturation. A backstepping approach is utilized for the guidance law, with proportional feedback gains that adapt to changes in the reference model bandwidth. The trajectory-reshaping algorithm is known as the optimum-path-to-go methodology. Here, a trajectory database is precomputed off line to cover all variations under consideration. An efficient representation of this database is then interrogated in flight to rapidly find the “best” reshaped trajectory, based on the current state of the vehicle’s control capabilities. The main goal of the flight-test program was to demonstrate the benefits of integrating trajectory reshaping with the essential elements of control reconfiguration and guidance adaptation. The results indicate that for more severe, multiple control failures, control reconfiguration, guidance adaptation, and trajectory reshaping are all needed to recover the mission.

Development and Flight Testing of a Parameter Identification Algorithm for Reconfigurable Control

The results are discussed of a series of e ight tests in which a computationally efe cient real-time parameter identie cation and control recone guration algorithm was evaluated. A modie ed sequential least-squares technique was used for identie cation. Signie cant challenges were encountered because of the poor information content of the signals used for identie cation and because of requirements for autonomy, reliability, and fast adaptation. A requirement for real-time operation in e ight-control computers with limited computational throughput imposed additional constraintson the recone gurable controllerand the parameterestimation algorithm. The identie cation algorithm is presented with solutions that were developed to address problems posed by the specie c application. Typical parameter identie cation results from the e ight tests are shown. The e ight tests culminated in a successful landing of an F-16 with a simulated missing elevon.

ADAPTIVE GUIDANCE WITH TRAJECTORY RESHAPING FOR REUSABLE LAUNCH VEHICLES

Reusable launch vehicle (RLV) designs are influenced by weight and other constraints that rarely allow for significant effector redundancy. Therefore, reconfiguration of the inner-loop control system and outer-loop guidance functions can be necessary to compensate for significant control effector failures, aerodynamic uncertainties, and disturbances such as strong winds. Trajectory reshaping is also important in some situations to judiciously manage vehicle energy on-line to meet the flight objectives. This paper presents an Optimum-Path-To-Go (OPTG) algorithm, which is a general framework to perform the on-line trajectory-command generation task. The methodology was applied to Lockheeds X-33 RLV for the approach- and-landing phase of flight, and a Monte-Carlo simulation analysis was used to demonstrate the benefits of the approach. Random increments to the vehicle drag were inserted in the simulation at various downrange locations. Such variations are representative of mismodeled aerodynamics, speedbrake failure, or significant winds. The X-33 implementation of the OPTG was able to reshape the trajectories to result in a safe landing for greater than 93% of the cases. Without the OPTG capability, safe landing was accomplished for only 29% of the drag variations simulated.

On-Line Trajectory Command Reshaping for Reusable Launch Vehicles

To enable autonomous operations in future Reusable Launch Vehicles (RLVs), onboard trajectory command reshaping will be required to facilitate recovery of the mission following a major anomalous event such as an effector failure. The Optimum-Path-To-Go (OPTG) on-line trajectory-reshaping algorithm is presented. In the OPTG methodology, a trajectory database is precomputed off-line covering all variations under consideration. Then, polynomial-based networks are generated which map these variations to basis function coefficients that describe the shape of the trajectories. The networks are then interrogated on-line, and the resulting coefficients are used to generate trajectory commands. Thus, based on the current state of the system, the algorithm will reshape the commanded trajectory to give the best remaining path to the end of the mission segment. For this study, the commanded trajectory was reshaped on-line due to a severe multiple control surface failure. Without reshaping, the vehicle is lost, even with control reconfiguration and guidance adaptation. With trajectory reshaping, the mission is recovered.

Intelligent Control of a Morphing Aircraft

A morphing aircraft is able to drastically alter its planform to optimize performance at very dissimilar flight conditions. Despite significant strides to develop wing structure and actuation systems, much work remains to effectively control both the morphing wing as well as the entire morphing aircraft. The control solution presented in this paper uses modelbased methods that provide precise, closed-loop control of the morphing planform (i.e. wing-shape control) and simultaneously enforce prescribed closed-loop aircraft dynamics (i.e. flight control). The specific planform that is the focus of this research is the N-MAS wing designed by NextGen Aeronautics. At the wing-shape control level, the authors sought to answer two questions: (1) What is the most efficient means of actuating the underlying structure of the N-MAS wing? and (2) Given a fixed set of actuators, how does one precisely manipulate a morphing structure given inherent physical limitations? At the flight-control level, the authors sought to develop a control methodology that can: (1) accommodate different planforms that result in drastically changing plant dynamics, and (2) make the transition between any two configurations while maintaining the stability of the morphing aircraft.

A Retrofit Architecture for Model-Based Adaptive Flight Control

The paper presents an adaptive control framework that is integrated with the production control system of an in-service aircraft. The purpose is to maximize performance and safety for unexpected changes in the dynamics caused by flight control failures, damage, and adverse environmental conditions such as icing. Neural networks are developed from available high fidelity simulation and flight data to encode the dynamics of the nominal closed loop system. A structure learning modeling algorithm is used to address the model selection problem (terms and coefficients) of these neural network function approximators. A constrained parameter identification algorithm provides on-line model corrections that account for uncertainties or changes in the current aircraft dynamics, and the updated estimates are enlisted in a receding horizon optimal controller to provide increments to pilot commands. The increments from the adaptive control law serve to reduce tracking error given the current closed-loop characteristics of the aircraft. A key benefit of the approach is that the adaptation is only significant if the aircraft behavior differs appreciably from the intended closed-loop flying qualities. Furthermore, the control law reconfiguration is included through the control input paths and preserves the structural filters, mode logic, and custom performance and safety software of the original digital flight control system. The retrofit system was integrated with the production F/A-18 control augmentation system, and piloted simulations of inflight refueling, target tracking, and general maneuvering with unforeseen failures to primary aerodynamic control surfaces were performed by US Navy and Boeing pilots. The retrofit software is also implemented in the US Navy fleet support flight control computer, and, in general, real-time hardware in the loop test results support the findings of batch simulation and software-only piloted simulations. The retrofit reconfiguration architecture is summarized; the enabling neural network modeling and system identification methods are discussed, and an overview of the model-based control law is given. F/A-18 piloted simulations and hardware in the loop test results are provided to show the reconfiguration benefits of the method and to substantiate the claim of practical usefulness for fleet aircraft, respectively.

Piloted simulation evaluation of a neural network limit avoidance system for rotorcraft

Piloting a rotorcraft is typically a high gain task, and under adverse conditions the workload may increase to the extent that the pilot is not able to achieve all goals simultaneously. Increased exceedances of operating liits may occur as a result of this high workload. The goal of this research has been to implement in piloted simulations a system that reduces workload and helps pilots to avoid aircraft limits. The system uses neural networks to predict near-future limit exceedances, and alerts pilots to these impending exceedances through tactile cues on the control inceptors and visual cues on the head-up-display. The system was demonstrated in piloted simulations of the UH-6OA and OH-58D and was found to reduce limit exceedances and pilot workload. In these experiments, tactile cues alone generally performed better than visual cues alone, but the combination of visual and tactile cues generally performed best. Pilot comments and handling qualities ratings of the system were highly favorable.

Hybrid Robust Control and Reinforcement Learning for Optimal Upset Recovery

Air safety and flight asset protection benefit greatly from rapid upset recovery. Autonomous recovery is of particular interest due to a recent significant increase in fielded Unmanned Aerial Vehicles (UAVs). Autonomous recovery challenges include complex nonlinear dynamics and large variation in potential upset conditions. A novel UAV upset recovery system is developed that combines the benefits of robust control with the benefits of intelligent learning techniques. Off-line, Reinforcement Learning (RL) techniques are applied to simulation data to discover recovery strategies that improve upon known strategies. When learning is complete, the strategies are provided to an online component. In the event of an upset, the online component is interrogated to determine the best control decision at each control update until the recovery is complete. The online component is designed to easily make use of the best-known recovery strategies, taking advantage of improved strategies as learning matures. The system architecture is partitioned into two components; one which focuses on recovery from high angular rate upsets and another which focuses on recovery from unusual attitude upsets. The input and output sets for both partitions are compact by design to reduce complexity, thereby ensuring the applicability of RL techniques. Two simulation variants of NASA’s Generic Transport Model (GTM) are used, one for training and initial evaluation and another for robustness testing. The results indicate that the learning process frequently finds improvements to best-known strategies, and that learned recovery strategies are robust to uncertainty.

Flight Test Results of an Adaptive Guidance System for Reusable Launch Vehicles

For next generation Reusable Launch Vehicles (RLVs), reconfigurable control, adaptive guidance, and on-line trajectory-command reshaping will often be required to recover the mission in the face of a major anomalous event such as an effector failure. An adaptive guidance system that works in conjunction with a reconfigurable controller and an autonomous trajectory command reshaping algorithm is presented. The guidance law utilizes a backstepping architecture to generate pitch rate commands that drive the inner-loop control system. Under extreme failure conditions the control surfaces can saturate in an attempt to meet commanded moments. In these cases, the guidance feedback gains are reduced to preserve stability margins in the guidance loops. In addition, simulation and flight test results of the complete reconfigurable control/adaptive guidance/trajectory reshaping system are presented for a simulated X-40A RLV. The Total In- Flight Simulator research aircraft was utilized to flight test the X-40A system under a variety of failure conditions. This work was completed in conjunction with the Air Force Research Laboratorys Integrated Adaptive Guidance & Control (IAG&C) program. Both simulation and flight test results indicate the major benefits of the new system. With on-line trajectory reshaping, the vehicle is able to achieve a safe touchdown, whereas the vehicle is lost without trajectory reshaping.

System Identification for Retrofit Reconfigurable Control of an F/A-18 Aircraft

We summarize a retrofit reconfiguration methodology that augments the production flight control system of an F/A-18 to compensate for changes in aircraft stability and control derivatives caused by aerodynamic control surface failures. The retrofit architecture relies on an indirect-adaptive model-following control approach that exploits onboard models of the closed-loop aircraft dynamics in the control computations. An output-error parameter identification process is used off-line to model the nominal behavior of the F/A-18 inner closedloop (airframe and baseline controller). The identified models are employed to describe the desired behavior for the controller and to provide regularization for an online parameter identification process. The online system identification algorithm updates the parameters of the models used for control decision, and a regularized recursive least-squares method is chosen to enable rapid adaptation to unforeseen failures and unmodeled dynamics while being robust to periods of low excitation. The retrofit reconfigurable controller was evaluated in high-fidelity F/A-18 simulations across a wide envelope of flight conditions with single and multiple control surface failures