Transcript Detailed Design Review ppt

UAV Automated Flight &
Seeded Fault Control
Detailed Design Review
Aurora Kiehl
Scott Neuman
Jeremie Snyder
Dennis Vega
Stephen Wess
•
•
•
•
•
•
•
•
•
•
•
•
•
Agenda
Project Goals
Aircraft Selection
ArduPilot Overview
Data Acquisition
Data Logging Capabilities
Seeded Fault Design
o Mechanical Implementation
o Electrical Implementation
Video System Cost Analysis
Imaging System
System Integration
Preliminary Test Plans
Bill of Materials
Risks
Future Plan of Action
Project Goals
UAV project consists of three overarching goals:
1. Demonstrate the capability of automated flight
between GPS waypoints for future use with proposed
Imaging Science project
2. Incorporate ability to initiate seeded faults and detect
that they have occurred
3. Log flight data, fault status, & on board accelerometer
data
4. Display live video feed to users on the ground and
allow user to capture still images in flight and store
these images for viewing after landing
RC Aircraft Selection
Hobbico Nexstar Mini EP
Price: $170.00
Wingspan: 3.7ft
(Meets spec: wingspan < 5ft)
Image Credit: modelairplanenews.com
Electric Powered
Balsa Wood Construction
(Easier to modify than foam
construction)
Independent Aileron Control
(Meets spec allowing for loss of control for one aileron only)
ArduPilot Overview
3DRobotics ArduPilot w/ ArduPlane Software
Price: $310.00
Allows for automated flight via GPS waypoints
(Meets need for automated flight capabilities)
Includes instrumentation for measuring roll, pitch, yaw, altitude,
and ground speed
(3-axis gyros/accelerometers/magnetometers, barometer, GPS unit)
Collects measurements at either 10Hz or 50Hz (GPS data @
5Hz)(Meets data refresh rate spec)
Automatic data logging w/ 4MB of onboard memory
Relay for aileron fault
ArduPilot
Cross Section
Top View
ArduPilot Simulation
By using X-Plane Flight Simulator, a hardwarein-the-loop (HIL) simulation can be performed
on ArduPilot.
X-Plane provides ArduPilot with GPS and
sensor data similar to a realistic flight and
ArduPilot flies the plane.
ArduPilot Simulation
ArduPilot Simulation
ArduPilot Simulation
Data Acquisition
3-axis accelerometer x3
Sensitivity Range:
Selectable ±1.5 or ±6 g
Analog output
Data Logging Capabilities
12 Channels of 10-bit ADC
15ksps ADC capability, >50 Hz sampling based
on available CPU time
~330kB accelerometer data for 10 minute flight
@ 50 Hz
Rudder Failure
Servo pulls pin connecting upper and lower sections of
rudder
Loss of control of upper rudder section
Open circuit indicates fault has been successfully seeded
Fault Detection Circuit
Wing Section Failure
Servo released spring loaded portion of wing section
Lower section of wing protrudes into the airflow
Broken electrical connection pulls down fault line
Fault Detection Circuit
Fault door
Wing Fault - Door
Aileron Failure
Relay allows aileron to be deactivated, thus limiting the
aircraft to using a single aileron for roll control.
Video System Analysis
CMOS 26N/P
- Less risk in integration than
Keychain #16
- Bandwidth
Source: 3D Robotics
5.8 GHz Tx/Rx Kit
- Lower risk of signal loss beyond "line of sight"
ArduPilot Mega MinimOSD R1.1
Goggles too costly for current scope (~ $250)
Separate battery pack for video system
- 11.1V, 1250 mAh
Imaging System
Camera capable of capturing still images to be installed on
aircraft
Command sent through ArduPilot will trigger camera to take
image
Low mass, high resolution camera desirable
Option #1
HD Mini Camera
Cost: $30
Image Resolution: 12Mp?
USB Charger
Micro SD card storage
Ships from China
Image Credit: www.k-ding.cn
Imaging System
Option #2
Smile Button Hidden Camera
Cost: $60
Image Resolution: 3Mp
USB Charger
Micro SD card storage
Image Credit: www.internetsiao.com
Camera to be reversed engineered, allowing a voltage
signal to emulate user action to capture image
Remote Camera Trigger Circuit
The Switch on the camera can be
replaced with an NMOS pass
transistor, which acts like an
open circuit when the input is
'low' and a short circuit when the
input is 'high.'
System Integration
Testing: Ardupilot
Ensure signals pass through Ardupilot in Manual Mode
Ensure all data of interest is collected and stored
properly.
Manually fly UAV to certain altitude, switch to fly-bywire A mode and verify it flies to waypoints
Datalogging Capabilities can be tested by running
Ardupilot and collecting data for 10 minutes and
determining if it fills up the 4MB onboard flash memory.
Testing: Ground Station
Ensure that ground station can communicate
necessary information with the UAV remotely on
ground
Verify that all servos can be controlled either using the
laptop or controller
Modify Ardupilot Mission GUI to add fault and imaging
features, verify that these features perform as required
on ground
Testing: Fault Seeding/Detection
Test that all faults can be triggered on the ground and
occur as expected
Run vibration test on fault detection system and make
sure the accelerometer data is stored and looks as
expected
Create a circuit that "open circuits" when a fault
occurs, obtain timestamp when this occurs for
correlation with accelerometer data by future groups
Testing: Video/Imaging System
Use the video transmitter to verify that video is sent to
laptop remotely on the ground.
Verify remote triggering of ‘Take Photo’ command
works and saves the photo on the UAV.
Testing: Power Subsystems
To ensure our ideal flight time of 10 minutes, All
batteries should last at least this long
The current draw for all components can be thoroughly
tested on the ground: the battery life will be equal to
the strength of the battery (mAh) divided by the total
current draw
Bill of Materials
Updated Risks
Future Plan of Action