Unmanned aerial vehicles (UAVs) are increasingly gaining popularity due to the considerable technological advances in the fields of microelectronics and MEMS (Micro Electro mechanical systems ).
This provides a huge range of high-performance microcontrollers and sensors of several types, elements which are essential for an UAV. The key advantages of using UAVs against conventional manned aircraft are: (i) smaller costs of production and maintenance, (ii) greater flexibility allowing difficult maneuvers or movement in locations hard to access with conventional aircraft, (iii) ability to sustain long hours of flight, and (iv) elimination of the risk that a manned aircraft exposes to its crew.
The main objective of the research described in this paperwas to develop an UAV capable to capture high definition aerial images and videos in an efficient way with a relatively low cost. To fulfill this it was decided to develop an UAV of small proportions, of the type MAV (Micro Air Vehicle).
The aircraft system we used in this project was a multirotor design, specifically the quad-rotor. With this vehicle it is possible to obtain images at low altitudes, thereby resulting in images with high resolution and quality, which may have levels of detail far superior to the ones that are obtained by satellites or manned aircraft, with a much lower cost.
To control this system we opted a design built around a high performance ARM Cortex-M3 microcontroller, thus making possible the use of control techniques that may require a higher processing.
In our work, the identification for the mathematical model of the UAV was conducted to provide data of the prototype to the stage of simulation, in order to obtain more realistic results, thus facilitating the phase of tuning the controller. We identified the thrust coefficient of the propellers, the mathematical model of the four motors and model of the UAV's complete structure.
The identification of some parameters was made with the aid of a test platform built throughout the project. This platform gives information on the rotation speed of the engines (using Hall-effect sensors), the thrust of the engines (using a load cell based on strain gauges) and also the electric input power. These parameters are acquired at 100 Hz sample rate and transferred to a computer for further processing and analyses.
Brushless electric motors were used in the prototype. This type of engine has a higher energy efficiency and durability compared to brushed motors. However its control system is more complex thus requiring an electronic module dedicated to this task, called ESC (Electronic Speed Controller).
This module basically provides a sequence of current pulses to the motor windings to produce a rotating field. The rotational speed and electrical power given to the motor is controlled by varying the duty cycle of a PWM signal at the ESC input.
Most of the ESCs modules have a commercial 8-bit microprocessor (ATMEGA8), which performs functions of speed control of the motors and also some special functions, such as the process of braking and acceleration curve.
The control system performs a key role in the quad-rotor’s stability, making possible to control precisely the attitude and altitude states of the aircraft. It's main goal is to make the quad-rotor moves to a new desired position (called reference) and also react to external disturbances quickly and in a controlled way.
Attitude control is the key element to maintain stability during flight so we decoded to implement a PID based algorithm, because it is widely used in the industry on practical applications with good results, and also has an easy implementation.
As the results presented in this paper show, it was possible to demonstrate the stability of the UAV using a PID controller. These results refer only to the pitch and roll movements, showing that the identified model is, in fact, very close to reality, considering the range of rotational speeds of motors used. There are still components that were not considered for model in the identification stage, which can later be taken into account to improve the reliability.
Due to the nonlinearity present in the system, the controller cannot meet the requirements when the velocities of all motors are simultaneously increased. One example is when there is the need to gain altitude.
This problem can be solved by applying an adaptive control where the component PID would work in conjunction with a fuzzy controller which, depending of the current speed of the motors, would adjust the coefficients of the PID automatically.
Sensor fusion techniques are being implemented to improve the quality of the angle measurements of the model, thereby also enabling the control of the UAV based on latitude, longitude and altitude.
To read this external content in full, download the complete paper from the author open online archives at Federal University of Parana.