Background information

Unmanned Aerial Vehicles (UAV) and Unmanned Aircraft Systems (UAS) are becoming increasingly popular and advantageous in both the defense industry, and civilian or commercial activities. According to Consumer Technology Association, 2.4 million personal drones were sold in the United State alone in 2016, more than double the 1.1 million sold in 2015.

An unmanned aircraft systems (UAS) refers not only to the aircraft, but to all supporting equipment used in the system including (but not limited to) sensors, microcontrollers, software, ground-station computers, and communication hardware. The focus of this project is on small fixed-wing aircraft and its guidance, navigation and control subsystems. Note that the classification of “small” aircraft according to the DoD refers to UAVs with a maximum gross takeoff weight between 0-20 pounds, normal operating altitude less than 1,200 feet above ground level, and airspeed less than 100 knots. [1]

UAVs are most often divided into two categories: fixed-wing and rotary wing. Fixed-wing UAVs consist of a rigid wing with predetermined airfoil which make flight possible by generating lift caused by the UAV’s forward airspeed – these aircraft will be the focus of this project. Rotary wing UAVs work similar to fixed-wing, however constant aircraft forward movement is not necessary to produce airflow over the blades – instead the blades (rotors) themselves are in constant motion to create airflow which generates lift. The primary advantage of a fixed-wing UAV is that is consists of a much simpler structure in comparison with a rotary-wing aircraft. The fixed-wing structure has more efficient aerodynamics which provide advantages such as longer flight durations and higher operating speeds. Additionally, fixed-wing structure is more straightforward to model in simulation, and thus will be used throughout this project.

 

Problem Statement and Objectives

This project attempts to design lateral and longitudinal autopilots using two different types of controllers to achieve adequate course-heading tracking, and altitude and airspeed hold, respectively, both in simulation and on a physical fixed-wing system. The first control technique for lateral and longitudinal autopilots to be simulated and implemented to the physical system is known as successive-loop closure (SLC) with proportional-integral-derivative (PID) gain-tuning. The second technique for simulation and physical implementation utilizes the linear quadratic regulator (LQR) method of control. It should be noted that, due to time restrictions, only the SLC control method was implemented to the physical system.

The primary objective of this design project will be to create control algorithms to command roll, pitch, airspeed, altitude and course heading of a small fixed-wing unmanned aerial vehicle (UAV). A flight simulator will be developed from the ground up in MATLAB/Simulink following procedures and methods described in detail in [2]. The simulation will model a Zagi HP Electric Flying Wing RC Aircraft using the lateral and longitudinal coefficients provided by [2] and found in Appendix A of this report. The physical system will be a Zagi HP Electric Flying Wing with a Pixhawk 2.1 autopilot, Here GNSS GPS Module, Holybro airspeed sensor, and ArduPilot ground control station (GCS) and firmware.

The goal is to create and implement multiple control design models for the fixed-wing autopilot and to observe relative advantages and disadvantages of each. The initial control design will be successive loop closure to command inertial position and attitude of the UAV via lateral and longitudinal autopilots. The second control method used will be a more robust, albeit more complicated design, utilizing a linear quadratic regulator (LQR) control algorithm to achieve similar results in both lateral and longitudinal autopilots.

The objective of the project overall will be to compare these control techniques both in simulation and the physical domain. It is desired to find which controller provides better performance in simulation as well as that which is more feasible in reality.