Neusa Maria Franco de Oliveira

UAV autopilot controllers test platform using Matlab/Simulink and X-Plane

Presently there is vast interest in UAV (Unmanned Aerial Vehicle) development given its civilian and military applications. One of the main UAV components is the autopilot system. Its development invariably demands several lab simulations and field tests. Generally after an UAV crash few parts remain usable. Thus, before embedding an autopilot system, it has to be exhaustively lab tested. With educational and research purposes in autopilot control systems development area, a test platform is herein proposed. It employs Matlab/Simulink to run the autopilot controller under test, the flight simulator X-Plane with the aircraft to be commanded, a microcontroller to command model aircraft flight control surfaces and a servo to drive these control surfaces. These resources are interconnected through data communication buses. So that, the autopilot controller designed on Matlab/Simulink is tested by controlling an aircraft on X-Plane. The inputs given to the aircraft flight control surfaces in the X-Plane are simultaneously sent to the microcontroller which translates these commands into effective servo movement control. Through this platform, designed autopilot systems can be applied into models similar to real aircraft minimizing risks and increasing flexibility for design changes. As study case, tests results from a roll attitude autopilot system are presented.

Hardware-In-the-Loop Simulation with X-Plane of Attitude Control of a SUAV Exploring Atmospheric Conditions

The attitude control law for fixed-wing small unmanned aircraft proposed in this paper is constructed based on two phases of a flight: stable flight and maneuvering flight. In the maneuvering flight, the aircraft deflects the main control surfaces (ailerons and elevator), whereas on the stable flight only the trim tabs are deflected. The switch between the two flights is done when the aircraft enters a zone in which the difference between the aircraft’s attitude and the reference value that the airplane needs to reach is greater than a predetermined value. The control laws are implemented on an on-board computer and are validated though Hardware-In-the-Loop (HIL) simulations, between the hardware and the flight simulator X-Plane, which simulates the unmanned aircraft dynamics, sensors, and actuators. The paper proves that this implementation can reduce the rise time and the overshoot, compared with traditional PID implementations. In order to analyze the behavior of the SUAV in these situations, it was performed simulations with Wind Gust and levels of Turbulence, using the X-Plane features.

Test platform to pitch angle using hardware in loop

The low cost electronic component and sensors makes the development of UAVs (Unmanned Aerial Vehicles) more accessible. Consequently, more students are interested in the skills necessary to integrate an UAV developing group. A very important component in a UAV is the automatic pilot. Aiming Education and Research in Digital Automatic Pilot development a flexible system to test flight controllers is proposed. The system uses the structure known as ¿hardware in the loop¿ in which the aircraft model is simulated in a PC (Personal Computer) and the designed controller is implemented in a microcontroller. The proposed system is appropriate to be used in laboratory classes in which the students can test the controllers previously designed using control theory and can also practice the use of interfacing devices, necessary to convert the analog signals provided by the sensors to the digital form used by the processors, and communication devices, necessary to the information exchange between the PC and the microcontroller. The laboratory practice is such that the student can compare the results obtained using the conventional simulation approach, in which a control loop consisted of the designed controller and the aircraft model transfer function is simulated in a computer, and the results obtained using the proposed platform. Implementing the designed controller in a microcontroller that can be embedded in an UAV exposes the student to a more real situation, considering also the possible delays between a command given by the controller and the change of a parameter in the aircraft. The results presented here were obtained implementing a PID (Proportional, Integral and Derivative) controller in a microcontroller and doing all the connections necessary to a PC implementing a longitudinal motion model of an aircraft.

A describing function approach to limit cycle controller design

The design of robust limit cycle controllers introduced here can be used for autonomous systems with separable single-input-single-output nonlinearities. Considering a system with unavoidable limit cycle and an uncertain linear subsystem, the objective is to design a controller such that the variation in limit cycle amplitude and frequency due the uncertainty is as small as possible in the controlled system for the worst uncertainty considered. The method consists of quasi-linearization of the nonlinear element via a describing function (DF) approach and then shaping the loop to reach desired limit cycle characteristics. As the DF method is used, loop shaping takes place in the Nyquist plot

Central Processing Unit for an Autopilot: Description and Hardware-In-the-Loop Simulation

This paper describes the architecture of the Central Processing Unit (CPU) of Pegasus AutoPilot, which is an academic autopilot, in the developmental stage, for small Unmanned Aerial Vehicles (UAVs). The data manager process and control laws, implemented on dedicated hardware, are described. In order to verify, validate, and optimize the system a Hardware-In-the-Loop (HIL) simulation, between the CPU and the flight simulator X-Plane is performed. X-Plane simulates the other systems of the autopilot, such as the sensors and actuators. The system is designed to facilitate the disconnection of the flight simulator and the connection of the real navigation hardware and control surface manager drive, as a plug and play device. The described control loops, consisting of inner and outer loops, controls the aircraft’s attitude and maintains a constant altitude, direction, and speed. The work presented can also be used as a guide for those who wants to begin to use Hardware-In-the-Loop Simulations using X-Plane.

An algebraic approach to the design of robust limit cycle controllers

The design of robust limit cycle controllers introduced here can be used for autonomous systems with separable single-input-single-output nonlinearities and unavoidable limit cycles. The objective is to design a controller to secure specified oscillation amplitude and frequency. The method consists of quasi-linearization of the nonlinear element via a describing function (DF) approach and then shaping the loop to reach desired limit cycle characteristics. As the DF method is used, loop shaping takes place in the Nyquist plot.

Work in Progress: Virtual Instrumentation in Electronic Warfare Education

Aiming education and research in electronic warfare (EW), the Brazilian Air Force acquired an electronic warfare training system. Its operation integrated with tools of Virtual Instrumentation (VI) has didactic potential to be explored either in local or remote laboratory courses. The Work in Progress presented in this paper has as objectives to study the concepts and techniques involved with VIs and to apply them in the context of practical instruction of EW in academy. The first phase of study consisted in the development of an interface (hardware) for computer control of radar jamming pod and the study of the data format used by it. The data format study was necessary in order to make possible the communication between the VI and the radar jamming pod. LabVIEW graphical language was chosen to develop the VI. Other tools to allow control the entire electronic warfare training system will be developed in next phase and discussed here

Initial experimental procedures for modeling and identification of a fixed wing aerial vehicle

Modeling and identification of an aerial vehicle will allow engineers to design and implement guidance and control systems using the mathematical model acquired on flight. Using the mathematical model, several improvements and benefits shall be obtained, such as designing control subsystem using Modern Control Theory; the expected results at the workbench test will correspond more accurately to the real flight data. The model described here is a nonlinear model for fixed wing. This model uses several geometric, inertial and aerodynamic parameters. The procedures used to determine these parameters will be presented along the discourse of this article. To determine these aerodynamic parameters, it is frequently necessary to acquire flight data using embedded electronic systems on the aerial vehicle. Therefore, accuracy is key when acquiring reliable data from these sensors used where several lab and ground tests were performed.

Proposal of hardware-in-the-loop control platform for small fixed-wing UAVs

Unmanned Aerial Vehicles (UAVs) have received considerable attention from the academic community and technology solutions companies, given their civilian and military applications. The autopilot system shall be thoroughly tested in lab because an accident may cause irreversible damage to the UAV. This article presents a proposal of a Hardware-In-the-Loop platform for testing and validation of a small fixed-wing aircraft. Besides a communication system, it was developed a deflection measurement platform of the aircraft control surfaces. This measurement platform was validated, enabling future work on implementing embedded control algorithms.

Hardware-in-the-loop simulation of an attitude control with switching actuators for SUAV

The attitude control law for fixed-wing small unmanned aircraft proposed in this paper is constructed based on two phases of a flight: stable flight and maneuvering flight. In the maneuvering flight, the aircraft deflects the main control surfaces (ailerons and elevator), whereas on the stable flight only the trim tabs are deflected. The switch between the two flights is done when the aircraft enters a zone in which the difference between the aircrafts attitude and the reference value that the airplane needs to reach is less than a predetermined value. The control laws are implemented on an on-board computer and are validated though Hardware-In-the-Loop (HIL) simulations, between the hardware and the flight simulator X-Plane, which simulates the unmanned aircraft dynamics, sensors, and actuators. The results presented prove that this implementation can reduce the rise time and the overshoot, compared with traditional implementations.