Klaus-Peter Neitzke – FH Nordhausen
MEMS in Aktion: Aufbau, Funktionsweise und Grenzen von Micro Air Vehicles (MAVs) 17. September 2014 Jahrestagung des mst-Netzwerk Rhein-Main e.V. 1
Index • From AIRBUS to MAV • Measurement techniques • Algorithms • Autopilots
• Details MAV
• Limits MAV • Outlook 2
Index
Measurement techniques
X
X
X
Algorithms X Autopilots X 3
Measurements techniques • Force measurements • Moment measurements • Pressure distribution measurements • Tuft visualisation & oil-flow visualisation • Attitude measurements
standard
• Hot wire technique • Hot film technique • Particle image velocimetry • Boundary layer measurements • Infra red measurements • Geometric measurements • Acoustic measurements
advanced
= MEMS
4
Measurement - Acoustic Delay and sum beam former
Array with 144 microphones, 120kHz sample rate
5
Measurement - Acoustic NLR measurement at Amsterdam airport
~ 68 m/s
0.1 s
1.0 s
~ 43 m main array 243 microphones
2 lateral arrays 2 x 40 microphones
6
Measurement - Attitude Application of inclinometers
7
Measurement - Attitude Data fusion of acceleration and gyroscope data Measurement of Alpha
Problem: Additional Acceleration
Low pass
Acceleration sensor
Corrected Alpha Complementary Filter
High pass Integration
Measurement of Alpha/Time
Problem: Drift
Gyroscope sensor 8
Algorithms - Axis coupling Stabilization – Not only 3 x PID closed loop control Example: Stable flight, then the sensors measure: + 90 deg pitch (nose down) + 90 deg roll (to the right) - 90 deg pitch (nose up) Question: What should the controller answer to stabilize the aircraft?
3 Acceleration sensors 3 Gyroscope sensors
Stabilization of: • Pitch angle • Roll angle • Yaw angle PID - closed loop control
Answer: - 90 deg yaw Necessary: Application of rotation matrix (or quaternion).
Horizontal tail planes Ailerons Vertical tail planes Engines 9
Algorithms - Axis coupling Simplification of the rotation matrix Unit vektor
Rotation angle
Low pass for alpha= f (acceleration)
High pass for alpha = f (gyroscope + rotation)
Sum Rotation matrix for small angles 10
Global view: Aircraft - MAV
Measurement techniques Algorithms
Autopilots
• Measurement technique is very similar for different aircraft dimension. • An university is able to participate at international competitions and research projects. This is possible with a limited budged. 11
Algorithms 3 Layer closed loop control for autonomous aircrafts
Route control:
Pressure sensor Magnet sensor GPS
3 Acceleration sensors 3 Gyroscope sensors
Navigation of:
Stabilization of:
Nordhausen
• Course
• Pitch angle
Bremen
• Speed
• Roll angle
Berlin
• Altitude
• Yaw angle
PID - closed loop control
Target • Latitude • Longitude • Altitude • Speed
(Pilot)
Telemetry
PID - closed loop control Target • Pitch • Roll • Yaw • Thrust
Horizontal tail planes Ailerons Vertical tail planes Engines 12
Algorithms Parameter adjustment in manual flight tests
Route control:
Pressure sensor Magnet sensor GPS
3 Acceleration sensors 3 Gyroscope sensors
Navigation of:
Stabilization of:
Nordhausen
• Course
• Pitch angle
Bremen
• Speed
• Roll angle
Berlin
• Altitude
• Yaw angle
PID - closed loop control
Target • Latitude • Longitude • Altitude • Speed
PID - closed loop control Pilot • Pitch • Roll • Yaw • Thrust
Telemetry
Horizontal tail planes Ailerons Vertical tail planes Engines 13
Example - Autonomous System Aircraft Lerche (Lark) 500 • Wingspan 490 mm • Take off weight 250 gram • Flight time 20 minutes • Control Motor, Aileron, Pitch, Drop • Motor Controller 6 A • Telemetry 868 MHz • RC – Receiver 35 MHz • Camera 380 lines, 2.4 GHz • GPS 4 Hz • Autopilot 30 x 40 mm
3rd place for autonomous aircrafts at IMAV 2010
14
Example - Autonomous System Quadcopter Ninja + Zecke + Bolt
• Max dimension 185 .. 500 mm • Take off weight 180 .. 500 gram • Flight time up to 15 minutes • Telemetry 868 MHz • RC – Receiver 2.4 GHz • IMU 3x ADXRS610 + 3x ADXL322 • Camera 480 lines, 5.8 GHz • GPS 4 Hz • Autopilots 20 x 30 mm .. 22 x 40 mm 3 x 1st place for manual / autonomous aircrafts at IMAV 2011
15
Drehflügler im Detail Entwurf einer Luftschraube von Leonardo da Vinci aus dem 15. Jahrhundert.
16
Quadcopter im Detail Curtiss-Wright VZ-7, etwa 1958.
17
Quadcopter im Detail Neuanfang 1990. Zuerst Keyenche (Japan).
Dann 1991 Nordhausen (Deutschland).
18
Quadcopter im Detail Boom der Multicopter (Tri, Quad, Hexa, Octo) ab 2002 (Industrie) und 2006 (Hobby).
19
Quadcopter im Detail Funktionsweise: Drehrichtung links/rechts :
16.09.2014 20
Autopilot 2011 - MEMS = 200€ 3 Gyroscope GPS
Magnet
Pressure 3 Acceleration
Route & Navigation
Stabilization Motor controler + Motor
Telemetry
RC
Servo Motor 21
Autopilot 2014 - MEMS = 10€ • Autopilots for IMAV (Competition 4 weeks ago)
22
Example – Current development • Autopilots for UAV “Songbird” from Geo-Technic (www.geo-technic.de)
23
https://www.youtube.com/watch?v=NzPSuOdyn4M http://www.youtube.com/watch?v=O3Fn-l5DoBs
24
Flight Time Copter ‚Ninja‘ and ‚Wanze‘ D = 200 mm
D = 100 mm
The vehicles were used during IMAV 2011 and IMAV 2012. The ‚Ninja‘ dimension is two times the ‚Wanze‘ dimension. They are very similar from the technical point of view. 25
Propeller Theorie Equation to calculate the necessary power for the hover flight.
Bla bla The induced flow velocity near to the propeller plane
Induced velocity 26
The necessary power is a function of the mass of the vehicle, the acceleration of gravity, the density of the air and the rotor area.
Possible flight time Comparison to flight tests.
The results fits well if we consider an efficiency factor for the ‚Ninja‘ of 0.35. 27
Wettbewerb IMAV 2014
28
Zusammenfassung • Die Theorie hinter großen Flugzeugen und MAVs ist gleich. • Man kann fast alles an MAVs erproben. • Die spannende Zukunft bei den MAVs ist eingebaute Intelligenz und Flugleistung. • Die Sensoren (MEMS) haben einen sehr guten Status (+ Kosten) erreicht. 29
Büro
Flugfeld
Ich bedanke mich für Ihre Aufmerksamkeit. Kontakt:
[email protected]