DARPA, managed by the U.S. Air Force Research Laboratory:
An Integrated, Multi-Layer Approach to Software-Enabled Control:
Mission Planning to Vehicle Control

Principal Investigator: Professor Gary Balas

Project Summary: 2000


The objective of this project is development of a unified design framework to synthesize and simulate individual vehicle management systems. This program uses innovations in nonlinear control, fault detection, reconfiguration and tactical trajectory generation to advance technologies associated with command and control of a Uninhabited Air Vehicle (UAV). The goal is to embed in the on-board vehicle management system significant high level functionality to extend the autonomy and lifetime of UAVs.


Theory, algorithms and software modules are being developed under this program for on-line control customization of Uninhabited Air Vehicles (UAVs). On-line control customization (OCC) will enable a dramatic increase in military effectiveness by increasing the level of autonomy in UAVs and the probability of mission success and survivability, and by expanding the range of UAV missions while reducing air vehicle fatigue and life cycle costs. The benefits to the military of OCC advances include use of extremely aggressive maneuvering of UAVs to achieve mission directives, accommodation of goal changes in real-time, life-extending control, and a reduced need for hardware redundancy while allowing more complex control strategies without increased software production and verification costs. A key component of our research is the integration of our OCC algorithms into the Open Control Platform (OCP) software infrastructure.

Our approach to individual vehicle control is to blend on-line optimization with off-line robust, nonlinear controllers. Both approaches to control require an accurate model of the nonlinear dynamics of the UAV. The off-line controllers will be designed using linear, parameter varying (LPV) control techniques. The benefit of LPV synthesis techniques is that very aggressive, multivariable inner-loop flight controllers can be synthesized off-line that guarantee stability of a large subset of the nonlinear flight envelope. In addition to the standard tracking and disturbance rejection flight control performance objectives, LPV controllers can be synthesized that schedule in real-time as a function of overall mission objectives, i.e. threats, failure, environment, longevity. The LPV framework has also been extended to allow for variable sample-time implementation of the controller. Real-time implementation of LPV controllers is similar to existing gain-scheduled controllers and therefore can be directly integrated within the OCP software infrastructure.

The on-line control technique combines receding horizon control (RHC), which solves an optimal open-loop receding-horizon problem, with the closed-loop Lyapunov functions (CLF) generated by the LPV design. The LPV CLF is used as a final cost penalty in the RHC optimization. The combination of RHC with a stabilizing CLF final cost penalty provides a stability guarantee for the RHC controller. The RHC optimization is based on full state feedback with set point regulation and doesn't include plant and model mismatch, external disturbances, command tracking and state operability constraints. RHC uses the full nonlinear model of the UAV to solve for the "optimal" control inputs on-line. Hence, RHC provides real-time adaptation to mission goals, changes in the system dynamic, constraints, and environmental factors and disturbances provided appropriate CLFs can be defined for each scenario. RHC will add significant autonomy to the UAV allowing event-based RHC for mission reconfiguration, smart allocation of resources, event-driven reconfiguration due to either mission requirements or system/component failures as well as performance customization.

To validate the proposed control design approaches, a public domain model of an F-16 aircraft, based on NASA Technical Paper 1538, was constructed to serve as our UAV. The baseline F-16 Block A controller was included in the nonlinear simulation for comparison. A quasi-LPV model of the nonlinear aircraft dynamics was synthesized and used to design an LPV inner-loop flight controller for the up and away flight envelope. Off-line LPV controllers have been synthesized that schedule on altitude, Mach number, angle-of-attack and mission level goals. RHC/LPV full state feedback controllers have been synthesized that show a performance improvement based on the ideal, nonlinear airframe model.