Helwan faculty of engineering
Mechatronics departure
4th Year
Applications of fuzzy logic
‘’Unmanned Helicopter’’
By
Abdurrahman Omer Fawzi
Youssef Gharib Imam
Abstract
This work presents a horizontal velocity controller for the unmanned helicopter
APID MKI11 by Scandicraft AB in Sweden. We use a novel approach to the design
consisting of two steps: first, a Mamdani-type of fuzzy rules compute for each
desired horizontal velocity the corresponding desired values for the attitude
angles and the main rotor collective pitch; second, a TakagiSugeno controller is
used to regulate the attitude angles so that the helicopter achieves its desired
horizontal velocities at a desired altitude. The performance of the combined
linguistic/model based controller is evaluated in simulation and shows that the
proposed design method achieves its intended purpose.
1-The APID MK-I11 model
The mathematical model used for attitude and altitude control of APID MK-I11 is
of the form for details see (2)
2-Takagi-Sugeno controller
First, the nonlinearities in the control inputs of the nonlinear model from Sect. 2
are decoupled by adding first-order actuator transfer functions - as a result, these
nonlinearities are moved into the state. The transformed model is then given as:
The above model is transformed into a TS model by a novel technique called
linear bounding which not only approximates this nonlinear model exactly, but
also drastically reduces the number of linear sub-models that constitute the TS
model. In what follows we describe briefly this step of the design. Consider again
the model described in (2). The nonlinear terms to be linearized, so that the fuzzy
system represents exactly the nonlinear.
3-The Mamdani-type controller
As already mentioned, the Mamdani-type of linguistic controller is used to
generate desired values for attitude angles given desired horizontal velocities at, a
given altitude. This type of controller has a heuristic nature which reflects the
experience of a human “pilot” who is an expert in remotely controlling the
vehicle. The motivation for resorting to such a heuristic approach is as follows:
The available equations describing the dynamics of horizontal motion do
not take into account aerodynamic effects related to the main rotor. Also
the contributions of the tail rotor torque and force are neglected. Thus
using these equations to derive desired attitude angles, given desired
horizontal velocities, is not a reliable option. Instead, the Mamdani-type of
linguistic controller uses the magnitude of acceleration arid velocity error
to infer attitude angles that if achieved will reduce the velocity error to
aero. Thus they “mimic ” a human “pilot’s” behavior when trying to achieve
certain desired velocities via remote control.
is higher than the desired one and we are reducing it, i.e.: we are
moving closer to the desired heading which is caused by certain orientation of the
horizontal velocity. In this case we maintain the current yaw. Furthermore, Neg.
Pos, and Zero are linguistic labels for the magnitudes of e,, i,. and the current yaw.
Figure 5 illustrates the above rule in terms of membership functions
corresponding to these linguistic labels. All in all there arc 9 rules describing the
relationship between ey, P, and the yaw. The first two types of rules neglect the
cross-couplings between pitch and roll angles in the dynamics of longitudinal and
lateral motions. However, these couplings are taken care by the heading rules
that in addition also prevent sideslip by rebtricting the yaw to be always equal to
the heading. Furthermore, the pitch and roll angles affect the dynamics of vertical
motion so that they cause a drop in altitude. Preventing this is taken care of at the
level of the TS controller. The control scheme coxnputing desired attitude angles
given desired horizontal velocities at a given altitude is presented in Fig. 6.
4-Simulation results
The integration between the Mamdani and TS controllers is illustrated in Fig. 7.
The numerical experiments are performed with the controllers designed in the
previous sections and acting on the nonlinear model from Sect. 3.
The
first experiment, depicted in Fig. 8, shows the results from set-point. regulation
around a desired low and high longitudinal Velocities.
The second experiment,
depicted in Fig. 9, shows the results from set-point regulation around a desired
low and high lateral velocities.
The last experiment, depicted in Fig. 10, shows the result-s from tracking a
desired heading computed from desired horizontal velocities.
horizontal high-speed velocities and altitude using attitude angles within much
larger ranges than the ones currently available on the APID MK-111 platform.
Future work will address the use of the approach presented here for position
control and for the purpose of behavior-based helicopter control.
5-References
• Autonomous helicopter control using fuzzy gain scheduling. (By Bourhane
Kadmiry , P. Bergesten , Dimitir Driankov)
