Aerospace and Electronic Systems Magazine June 2017 - 62

wizard), and a real-time debugger. The MATLAB FL toolbox was
used for the FL controller development, with a simplified quadrotor model. Because the HCS12 instruction set included a limited
set of FL membership function types, the MATLAB FL controller development involved only the compatible types, resulting in a
controller synthesis constraint and in a benefit of code implementation to the MCU, as well as superior execution.
The quadrotor uses four rotors to produce a total thrust T = Tf +
Tb + Tl + Tr, where Tf, Tb, Tl, and Tr denote the front, back, left, and
right rotor thrust variables, respectively. The front and back rotor
pair rotates counterclockwise, while the left and right rotor pair
rotates clockwise, effectively balancing the z-axis torque forces
generated by each of the spinning rotor pairs, removing the need
for a tail rotor as used in conventional helicopters, and ensuring all
generated thrusts lift the quadrotor. Having six degrees of freedom,
i.e., yaw, pitch, and roll angles, and x, y, and z translations, the
quadrotor is in a stable hovering state when there are no net forces
and velocities in any degree of freedom. The IMU detects variations in the craft angular velocities and translational accelerations
and produces the three gyro outputs, i.e., roll, pitch, and yaw rates,
and the three accelerometer outputs, i.e., x, y, and z accelerations.
The roll angle control can be achieved by changing the relative
speed of the right and left rotors, resulting in a rolling torque, τφ
= l (Tl - Tr), where l denotes the length of the center of gravity to
the motor. The pitch angle is controlled through relative speeds of
the front and back rotors, resulting in a pitch torque τθ = l (Tf - Tb).
The yaw angle of quadrotor is controlled by varying the speeds of
clockwise rotating pair and counterclockwise rotating pair, with
associated imbalance causing a yaw torque τψ = l (Tf + Tb - Tl - Tr)
for a z-axis rotation. For clarity and intelligent controller feasibility, only the quadrotor roll- and pitch-level stabilization, i.e., a
desired position (x, y, z) and yaw ψ, was considered.

Figure 2.

The quadrotor assembly sequence.

INTELLIGENT CONTROLLER DESIGN

After successful initialization, the auto pilot module was engaged
so that the system could fly through its intended flight path. If the
flight trajectory is traversed completely, the quadrotor lands safely.
However, if there are any emergency interrupts, such as a remotely queued emergency shutdown signal, the emergency landing
module becomes active to guide the quadrotor for a safe landing.
The MCU belongs to the Freescale Semiconductor HCS12 family whose instruction set is the first to specifically accommodate
the FL implementation requirements. The integrated development
environment was CodeWarrior, an invaluable tool to construct and
compile the project code, and featured the cross-language support of C and assembler, Processor Expert (a device initialization
62

The flight stabilization test control system included four inputs
for the x-axis (pitch) and y-axis (roll) error and error differences
to the FL controller whose outputs included four throttle changes
for right, front, left, and back motors, respectively, to safely adjust
pitch and roll rates. A constant throttle reference and hard limiters
were added to each motor input to ensure the desired z-axis elevation of the quadrotor and the practical voltage values for the actual
quadrotor motor inputs. Because a system stabilization of about 0
deg pitch and roll axes is desired, i.e., a level flight, the closed-loop
system reference angles were set to zero.
The FL uses human inference system by using fuzzy sets via
membership functions and associates if-then rules. The FL con-

IEEE A&E SYSTEMS MAGAZINE

JUNE 2017

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine June 2017