Design of an Autonomous Unmanned Aircraft Flight Path ...repository.um.edu.my/31002/1/16 AP2 Inecce.pdf · Design of an Autonomous Unmanned Aircraft Flight Path Controller ... the

Design of an Autonomous Unmanned Aircraft Flight Path Controller

B. M. Albaker, N. A. Rahim UMPEDAC Research Centre, Faculty of Engineering

University of Malaya 50603 Kuala Lumpur [email protected]

Abstract— This paper presents the design of a flight path controller for an unmanned aircraft. Innovative control technique based on nested Proportional–Integral–Derivative (PID) loops is introduced in the design. The flight characteristics of the airframe used in the controller development is analyzed using a three degree of freedom point mass equation of motion. Simulation results of flight controller together with aircraft dynamics are demonstrated in order to show and validate the effectiveness of the proposed controller.

Keywords— Unmanned Aerial Vehicle, flight path controller, PID control and autonomous flight.

I. INTRODUCTION Recently, there has been a rapidly increasing interest in

using UAVs for military and civilian applications including aerial remote sensing, fight forest fires, traffic monitoring, search and rescue…etc. Military UAVs such as Global Hawk and Predator are used for both reconnaissance and combat missions [1]. The importance of UAVs have been illustrated in [2]. Their importance lies in their characteristics and ability for autonomous navigation, making them attractive attribute in the near future.

According to Bekey [3], Autonomy refers to systems capable of operating in the real-world environment without any form of external control for extended periods of time. Long et al., [4] have recently reviewed several software systems and methods that could be used for controlling autonomous vehicles.

Due to high nonlinearities of the aircraft dynamics, a lot of intelligent control techniques have been used for the flight controller to guarantee a smooth desirable trajectory navigation, such as PID control, Neural Network (NN), Fuzzy Logic (FL), Sliding mode control and H-infinity control. A recent review of the literature on some up-to-date controllers can be found in [5, 6].

This paper focuses on the design of autonomous flight path controller, designed at the UMPEDAC Research Centre for the purpose of holding the states of UAVs. This controller is a part in simulating multiple UAVs simulation program to execute higher level functions of automation, guidance, data acquisition and conflict avoidance. Our previous research in these areas can be found in [7-12]. The controller is a behavior

based multi-loop architecture designed to control the flight path for autonomous UAVs.

The remaining of this paper is as follows: Section II presents the platform and dynamics of the UAV that the controller was designed for. Section III describes the design of the flight control system, highlighting the PID loops for state holding. In section IV, the proposed PID controller is applied to UAV dynamics in a closed loop response and the simulation results of both, UAV and controller, are demonstrated. Section V concludes this paper.

II. UAV PLATFORM AND DYNAMICS The first step in designing a controller for any physical

system is to characterize the dynamics of that system. The dynamics of an aircraft used in this paper is point mass three degree of freedom (3DOF) longitudinal and lateral dynamics that include aircraft’s response along the roll and pitch axes. The aircraft’s velocity response to thrust is also modeled. The equations of motion are given by:

(1)

(2)

The simulation utilizes standard 3DOF equations of motion where the Velocity (V), flight path angle (γ), heading (ψ), and North, East and Down Velocities and Positions (VNED, PNED) are the state variables. The heading angle of the UAV is computed with respect to the positive north-axis. The point-mass equations of motion are derived with respect to North, East and Down (NED) local coordinate system from [12]. TVL is the direction transformation matrix that transforms a vector expressed in Local frame into a vector expressed in the aircraft’s velocity frame.

TVL cos γ cos γ sin γsin 0sin γ sin γ cos γ (3)

International Conference on Electrical, Control and Computer Engineering Pahang, Malaysia, June 21-22, 2011

978-1-61284-230-1/11/$26.00 ©2011 IEEE 229

The dynamic characteristics of interest are the aircraft’s response to step inputs. To understand the response of the aircraft, a model is imported into a simulation environment constructed in Matlab/Simulink. The block diagram of the aircraft dynamic model together with the proposed controller is shown in Fig. 1. The inputs to the model are the thrust (Thcmd), Angle of Attack (AoA) and bank angle (Φ). The outputs of the model are the nine aircraft states, illustrated earlier.

Figure 1. UAV simulation block diagram using three degree of freedom

equations of motion

A. Simulation Setup At the beginning of the simulation, the trim values are

estimated at a given operating condition for level, un-accelerated flight of the aircraft model. An aircraft is setup at an initial condition of airspeed (V) of 25 m/sec and altitude (Alt) of 300m. For that condition, the values for AoA and throttle position, that make the derivation of the states zero, are computed. The control inputs are centered around these trim values.

B. Dynamic Simulation The aircraft responses to control inputs are analyzed by

applying step commands and monitoring the output states. A pulse input command with amplitude of 3º and pulse width of 100sec is applied to the AOA after the first 1000sec from time of running the simulation. This first simulation is a test of the altitude and flight pan angle changes in response to a step in AoA command. Another pulse command is applied to the bank angle command with amplitude of 40º and width of 100sec after the second 1000sec for the purpose of changing the direction of a UAV. Finally a pulse command is injected at the third 1000sec to the throttle input command. The interval from 3000sec to 4000sec is to test the velocity change in response to throttle command. Fig.2 shows the applied control set commands as a scenario for the purpose of showing system response.

Fig.3 demonstrates the resultant aircraft states (V, flight plan angle and heading as well as the down position) in response to control inputs. The simulation shows the poor response of the UAV to the applied signals. First, an increase in AOA leads to an increase in flight path angle with a decrease in velocity. Both states are dampened out over the 50 sec after applying the AoA pulse. The heading oscillates over 100sec then dampens out at a value of -2º in the second time interval. In addition to that, the altitude decreases rapidly to 150m in same time pulse. Finally, as the third throttle pulse applied, the V oscillates then stabilized at its original value of 25 m/sec.

The bank angle stabilizes at this point and the UAV’s altitude increases rapidly to around 400m. This behavior is consistent with an unstable short period mode. It is clear that the aircraft will require activate stabilization and control in the throttle and roll and pitch axes.

Figure 2. Applied control input commands in the simulation scenario.

Figure 3. Responses of the UAV dynamics illustrating the four output states: V, heading, flight path angle and finally down position.

III. AIRCRAFT FLIGHT PATH CONTROL DESIGN In order to achieve stable autonomous flight, the

instabilities must be addressed. This section discusses the PID

230

control structure developed to stabilize and control the aircraft. To fly the aircraft autonomously, the controller must be capable of controlling the heading, altitude and airspeed of the aircraft. For manual control, it is desirable that the controller also accepts AOA, bank angle and throttle commands. To accomplish this goal, a controller constructed of nested PID loops has been developed. The throttle command is controlled via inner PID loop that stabilize the throttle, whereas the AOA and Bank angle are controlled with outer loops. These loops are explained in the following subsections.

A. Bank Angle Controller This loop is responsible for controlling the heading of the

aircraft. It generates a bank angle from the heading error. This bank angle serves as the commanded roll angle for the UAV dynamics. This loop is shown in Fig.4.

Figure 4. Block diagram of the heading angle controller

B. Throttle Controller The control variable for speed is throttle. The purpose of

this loop is to control the aircraft’s airspeed by adjusting the throttle. This loop derives the throttle servo. The control law is to calculate throttle command from velocity and x-axis acceleration feedback as illustrated in Fig.5.

Figure 5. Block diagram of the throttle controller

C. Pitch Controller The primary control variable for altitude is angle of attack.

A pitch can be generated directly from altitude feedback. For acceleration based, the control law determines a normal velocity command which is converted to an angle of attack using additional pitch dynamics models. Therefore, this loop generates a commanded pitch angle from the altitude error and normal acceleration value. The output of this loop is connected

to the AOA command. Fig. 6 illustrates the dependency of the pitch controller to the altitude error and normal velocity.

Figure 6. Block diagram of the pitch controller

IV. CLOSED-LOOP SIMULATION The control structure is implemented in Matlab/Simulink.

The controller is connected to the UAV dynamic model in a closed loop as shown in Fig.1 above. The input to the UAV model is the desired velocity, altitude and heading. The performances of the controllers are simulated by changing the states, altitude, heading and velocity commands, in three stages. The simulation is started by executing the simulator with an initial operating condition of 0º heading, 25 m/sec velocity and altitude of 300m set at its trimmed values with level un-accelerating flight. The altitude command is then issued to change the state of UAV’s Altitude to 350 after the first 1000sec. Heading loop is simulated by giving a step input in the desired heading of 90º after passing 2000sec of simulation time. Finally, at 3000sec of simulation time, the response of the velocity loop is simulated in the same manner by applying a step input with a commanded velocity of 20m/sec. The results of the simulations are demonstrated in Fig.7.

Figure 7. Close-loop response with the proposed controller

231

The simulation shows that adequate performance can be achieved with PID control. The simulation results show that the proposed PID structure in this paper can adequately control the altitude, velocity and heading angles in response to step inputs.

V. CONCLUSION This paper has outlined the design of flight path algorithm

used to control an autonomous fixed wing UAV. It began by analyzing the dynamics of the flying aircraft by applying step inputs to a 3DOF aircraft model. This model had lateral and longitudinal instabilities that have corrected with the flight controller constructed with nested PID loops. The paper concluded by simulating the closed loop performance of the controller and aircraft dynamics by applying step inputs. Through simulations, it was found that the controller could adequately control the velocity, altitude and heading of the aircraft.

REFERENCES [1] G. L. Sinsley, L. N. Long, A. F. Niessner and J. F. Horn, “Intelligent

Systems Software for Unmanned Air Vehicles,” In the proceedings of the 46th AIAA Aerospace Sciences, 2008-0871, 2008.

[2] Department of Defense (DoD), “U.S. Army Unmanned Aircraft Systems Roadmap 2010-2035,” Office of the Secretary of Defense, US Fort Rucker, Alabama, 2010.

[3] G. A. Bekey, Autonomous Robots: From Biological Inspiration to Implementation and Control, The MIT Press, 2005.

[4] L. N. Long, S. D. Hanford, O. Janrathitikarn, G. L. Sinsley and J. A. Miller, “A Review of Intelligent Systems Software for Autonomous

Vehicles,” IEEE Symposium on Computational Intelligence for Security and Defense Applications, Hawaii, pp. 69 – 76, April 2007.

[5] H. Chao, Y. Cao and Y. Chen, “Autopilots for small Fixed-Wing Unmanned Air Vehicles: A Survey,” In the Proceedings of the IEEE International Conference on Mechatronics and Automation, China, pp. 3144-3149, Aug. 2007.

[6] R. S. Christiansen, “Design of An Autopilot For Small Unmanned Aerial Vehicles,” M.Sc. Thesis, Electrical and Computer Engineering, Brigham Young University, Utah, USA, Aug. 2004.

[7] B. M. Albaker, N. A. Rahim, "Signal Acquisition and Parameters Estimation of RF Pulse Radars using Novel Method," IETE Journal of Research, Vol. 55, Issue 3, pp. 128-134, May-June 2009.

[8] B. M. Albaker, N. A. Rahim, "Detection and Parameters Interception of a Radar Pulse Signal Based on Interrupt Driven Algorithm," Journal of Scientific Research and Essays, Vol. 6, Issue 6, March 2011.

[9] B. M. Albaker, N. A. Rahim, “Intelligent Conflict Detection and Awareness for UAVs,” In proceedings of the 3rd IEEE Conference on Innovative Technologies in Intelligent Systems and Industrial Applications (CITISIA 2009) , Monash University, Malaysia, pp. 261-264, 25-26 July 2009.

[10] B. M. Albaker, N. A. Rahim, “Straight Projection Conflict Detection and Cooperative Avoidance for Autonomous Unmanned Aircraft Systems,” In proceedings of the 4th IEEE conference on Industrial Electronics and Applications, (ICIEA 2009), Xian P. R. China, pp.1956-1960, Vols.1-6, 25-27 May 2009.

[11] B. M. Albaker, N. A. Rahim, “Unmanned Aircraft Collision Avoidance System Using Cooperative Agent-Based Negotiation Approach,” International Journal of Simulation, System, Science and Technology, IJSSST, Vol. 11, Issue 4, pp. 1-8, Sept. 2010.

[12] B. M. Albaker, N. A. Rahim, “Autonomous Unmanned Aircraft Collision Avoidance System Based on Geometric Intersection,” International Journal of Physical Sciences (IJPS), Vol. 6, Issue 3, pp. 391-401, Feb. 2011.

[13] P. H. Zipfel, Modeling and Simulation of Aerospace Vehicle Dynamics, 2nd Edition American Institute of Aeronautics & Astronautics, (AIAA), 2007.

232