Michael Peat, Kollin Moore, Matt Rich, Alex Reifert Advisors: Dr. - - PowerPoint PPT Presentation

SD-May1014 Team: Michael Peat, Kollin Moore, Matt Rich, Alex Reifert Advisors: Dr. Nicola Elia and Dr. Phillip Jones



Problem:

• To create a small electrically powered autonomous flying vehicle

capable of takeoff and landing from horizontal surfaces as well as stable indoor hover without human assistance. Ideally this should include onboard sensors for orientation and acceleration along all three rotational axes as well as altitude and overall position relative to known reference points.

 Solution:

• In order to achieve our goal, we researched platforms fitting of the

above outlined description.

• We also researched the most light weight sensors available to us and

cut out any that we did not absolutely need for basic takeoff, hover, and landing.

• Once a platform was firmly established and tested for capabilities we

researched and attempted to create mathematical/computer simulation models for it in order to aid us in the process of designing

• ur control system.
• We also established effective and reliable Bluetooth and RF

communication with a computer ground station for overall flight instructions.

 The operating environment for our system will be climate controlled

university buildings without any obstacles (dynamic or otherwise) in the intended flight path. The vehicle shall have enough room to takeoff, hover, and land without any probable danger of collision with its surroundings.

 Environment will be constrained by the camera limitations of the

sensor system seen below.

 The base platform we are using were to malfunction and become

unusable.

 When we are trying to write software for our control system we do

not have the necessary programming knowledge to complete it.

 The project is too large for the time constraints we were given.  Lack of funds to purchase necessary parts.  Due to the complexity of the electronics and other systems that are

needed for this project, some necessary components may be unavailable.

 Since we will be using different parts from different manufacturers,

they may not be compatible for some reason and we will not be able to integrate them.

 Mechanical Constraint:

• Our vehicle is small enough to operate indoors
• Our vehicle is able to carry all of the necessary

sensors stably

 Power Constraint:

• Our vehicle has a limited power supply for flight in the

form of a battery

• This power supply is carried within the platform

during flight

• This power supply allows at least five minutes of

stable flight with all equipment attached

• This power supply is limited in capacity due to weight

considerations

 Capability of autonomous takeoff, flight,

and landing

 Capability of making high frequency mid-

flight stability adjustments

 Capability of communicating with computer

ground station

Embry Riddle College of Engineering Carnegie Mellon University

South Dakota School of Mines and Technology (SERV Robot)

Massachusetts Institute

• f Technology

Technische Universitaet Berlin Georgia Tech 2009 Aerial Robotics Team

 Innovision Tracking Systems MaxTRAQBasic 3D

System

• Either 60 Hz or 120 Hz
• $5000 to $5500

 MaxPRO Upgraded Motion Capture Software

• Can reach 250 Hz
• $9000 -$28000

 WorldVizPPT Very Precise and Fast

• Has option of additional sensors that measure roll pitch and

yaw

• $10000 -$30000

WII Remote

• IR Camera
• Accelerometer
• Bluetooth

Transmitter

Digital to Analog Converter Ground Station Helicopter Flight Controls FPGA

Overall System Design

Bluetooth RF Transmitter Power Converter

• Converts

7.4VDC to 3.3VDC

• Provides

power to sensor system

IR LED Constellation

 The intended end use of our system will be

continued research and development into the area

• f autonomous flight systems by knowledgeable

engineering students and/or professors.

 User Interface

 Power System

• Designed a step-down buck converter circuit to

power the Wiimote sensors using the helicopter’s battery

 Required since the battery provides 7.4V and the Wiimote

• nly requires 3.3V



We decided to use the controller in an unconventional way by directly changing the control voltages with a computer program

• We first measured the control voltages and recorded how they varied
• After carefully labeling the

different control channels we stripped the controller of nonessential parts including the potentiometers

• First we tested our idea by

connecting the controller to a controlled power source

• We finally connected the

controller to a DAC and FPGA that allowed us to connect directly to our team laptop

 Reading Xbox Controller

• Rewrote and adapted software being used by Korebot Project

 Reading Data from Wii Remote

• Altered existing Cwiid library

 Manipulating values appropriately

• Mapped Xbox inputs to helicopter throttle, pitch, roll and yaw

values.

• Mapped sensor data to corresponding helicopter state values

 Autonomous Control System  Sending data using RS232 Serial port

• Sends Channel Command followed by value to FPGA state

machine.

• Found current maximum refresh rate of 2ms.
• Appropriate voltage levels are seen at the DAC outputs.

t Ir1:x Ir1:y Ir2:x Ir2:y Ir3:x Ir3:y Ir4:x Ir4:y

Sens nsor r Inputs ts

Yaw Angle θ(t) Left/Right Position Y(t)

Forward/Back Position X(t)

Up/Down Position Z(t)

Calcul ulati ations ns Θs Ys Xs Zs

Set Points nts

• Yaw Ctrl

P

Roll Ctrl

PI PI

Pitch Ctrl

PI PI

Throttle Ctrl

PD PD Con

• ntroll

roller ers

Yaw Roll Pitch

Throttle

Helico copte ter Actuat uator

• rs

Physic ical al Beha havior vior

Sensors

 Wii data analysis function completely re-

written

• Improved efficiency (run time, memory usage…) , overall analysis

capabilities (4 IR sets), expandability, and commenting significantly.

 Fabricated a cradle attaching Wii-mote to

helicopter

• Modified existing Aluminum brackets and a portion of original Wii-

mote case (This design proved difficult to work with and was discarded)

• Current mounting system utilizes Velcro to attach all additional

components to the base platform

 Completed further sensor testing

• Constellations, data resolutions, disturbance spectrum

characterization

IR Sensor Data vs Time Accelerometer Data vs Time Testing the Wii Remote sensors was a large part of determining how we could best use them to provide feedback to our control system and fly our helicopter. Above are two examples of some of the data from our sensors. The accelerometer plot also shows the effects of filtering out the disturbance/noise from the helicopter’s vibrations.

 Flight testing was performed to determine how our control

system reacted in real-time.

 Many adjustments were made in order to ensure the

smoothest flight dynamics possible.

 A presentation giving a general overview of the

current technology involved in non-ISU UAV projects, both at other universities and in the general marketplace.

 A written report detailing:

• The capabilities of our platform.
• The sensors used in our systems operation.
• The overall processes and means by which our system
• perates.
• A summary of the development process.

 Our end product itself.

15% 20% 25% 25% 15% Research Design Testing Documentation Implementation

Cost Estimate for MicroCART-Phase 5 Cost Equipment Base Platform Donated Additional Platforms $ 180.00 Replacement Parts $ 10.00 Upgraded Batteries $ 20.00 Sensors Donated Tools and hardware $ 10.00 Reporting Project Poster $ 10.00 Bound Project/Design Plans $ 15.00 Labor 1324 hours at $20 per hour $ 26,480.00 Subtotal (without labor): $ 245.00 Total: $ 26,725.00

 Position and orientation will rely less on IR

communication

 More on-board portable sensors will be

used

• accelerometers for roll and pitch
• sonic range finders for altitude
• magnetometers for heading
• GPS for absolute positioning

 Future research will be done to make the

helicopter able to move to different points

• n a grid and able to fly outside