Integrated Targeting, Guidance, Navigation, and Control for Unmanned Aerial Vehicles

Abstract

The goal of this dissertation research is to demonstrate the integration of targeting, guidance, navigation, and control (TGNC) functions for real-time implementation onboard unmanned aerial vehicles (UAVs) for a wide range of applications. This allows us to create a robust and accurate integrated TGNC software platform for UAVs, which enables them with real-time capabilities and leverages the flight autonomy. Target-relative guidance, real-time targeting and re-targeting capabilities are of great interest in today's UAV technology. This research proposes new guidance and estimation methods as well as new extremal control laws for UAV applications. In particular, this research focuses on quadcopter applications. The proposed guidance methods represent an extension of the existing explicit translational guidance (E-guidance) to include rotational guidance and exponential braking guidance to reach target points. The proposed estimation method is a hierarchical mixture of experts (HME) framework with extended Kalman filters (EKFs) and a modified softmax function to provide state and parameter estimations for navigation solutions. The proposed research utilizes the Hamiltonian formalism with the indirect method to solve the optimal control problem, which replaces existing PID control laws with extremal control laws based on first-order optimality conditions. Three illustrative examples demonstrate integration of targeting, guidance, and control functions for takeoff, waypoint, and roll maneuvers of quadcopters. It is shown that the proposed HME framework with acoustic parameters demonstrates a viable navigation solution. Implementation of the TGNC functions through the proposed HME, target-relative guidance, and extremal control with simulated acoustic parameter measurements demonstrates a completely integrated TGNC software system for quadcopters. Novelties of the proposed research include extension of E guidance, simulating an exponential braking guidance law to reach a target point, determination of the switching function for max-intermediate thrust arcs, and design and validation of a HME framework to provide navigation solutions. The proposed research results can be used to address environmental and agricultural problems that utilize UAVs. This research has been funded, in part, by the NASA EPSCoR ACTUAS (Autonomous Control Theory - Unmanned Aerial Systems) project. The core research contributions are deriving E Guidance for rotational maneuvers extending E Guidance to higher order integration methods, integrating E Guidance with extremal control satisfies the boundary conditions to yield an extremal for the guided trajectory, and incorporating acoustics with EKFs in HME shows the impact of considering several different models and parameters to have accurate state estimation. Future work may encompass determining accurate dynamic thrust and acoustic models and considering second-order conditions to determine optimal control with a corresponding trajectory.