Selfbalancing Handsfree InterFunctional Transport S'H'I'F'T'

1 / 48
About This Presentation
Title:

Selfbalancing Handsfree InterFunctional Transport S'H'I'F'T'

Description:

Changing centers of mass. Difficult control problem. Lessons Learned ... Steady state velocity of 4 m/s. Results. Smooth, stable operation ... – PowerPoint PPT presentation

Number of Views:78
Avg rating:3.0/5.0
Slides: 49
Provided by: And6318

less

Transcript and Presenter's Notes

Title: Selfbalancing Handsfree InterFunctional Transport S'H'I'F'T'


1
Self-balancing Hands-freeInter-Functional
Transport (S.H.I.F.T.)
  • Team MEM-11
  • Robert Ellenberg MEM/ECE
  • Andrew Moran MEM
  • John Spetrino MEM
  • Advisor Dr. Paul Oh
  • MEM Department
  • May 30th, 2007

2
Problem
  • Humanoid Mobility
  • Slow and Inefficient
  • Power Consuming
  • Wear and Tear
  • To our knowledge, an enabler does not exist that
    both Humans and Humanoids can use.

3
Thresholds and Objectives
4
Existing
  • Segway PT i2
  • nBot
  • Zappy
  • Tank Chair
  • John Deere Gator
  • Chrysler GEM
  • B.E.A.R. (Battlefield ExtractionAnd Retrieval
    Robot)
  • Electric Unicycle
  • Home-made Segway

5
GAP between Existing and Required
  • Nothing exists that is Hands Free
  • Single-Link vs. Multi-Link Pendulum
  • Labor Intensive riding
  • Unstable
  • 1 wheeled Low speed turns/idle
  • 3 wheeled High speed turns
  • Small scale
  • Can not physically carry humans, heavy loads or
    push/pull 20lbs.
  • Not designed for Human riders
  • Specifically for Robot Experimentation

6
Background
  • INSERT Sr_Design_Video_3.mpg

7
Major Design Decisions
  • 2 Parallel wheels
  • Air-filled tires
  • Adjustable foot controls
  • Off the shelf parts
  • Electrically Powered
  • Gear Drive
  • Sensor Suite Encoders / IMU / DSP

8
System Dynamics
  • Multi-link inverted pendulum
  • No practical way to measure link angles
  • Changing centers of mass
  • Difficult control problem
  • Lessons Learned
  • Know how to model single-link inverted pendulum
  • Design to reduce body movement

9
Single Link Approximation
  • Model coordinates
  • f Pitch angle
  • ? Wheel angle
  • a Ground incline

10
MATLAB Simulation for Component Requirements
  • Simulation conditions
  • Steady acceleration at 1m/s2 from t0 to t2.25
  • 8 incline from t6..10
  • Steady state velocity of 4 m/s
  • Results
  • Smooth, stable operation
  • Voltage signal oscillation, can be filtered in
    software
  • Found peak current/voltage requirements

11
MATLAB Simulation for Battery/Motor Requirements
  • Battery Simulation conditions
  • Energy Avg. Torque Total Wheel Rotation
  • Calculated minimum battery capacity 200 VAhr at
    4mph
  • Design battery capacity of 550 VAhr
  • Current Requirement 30A
  • Peak Current 120A
  • Motor Simulation
  • Incline Torque 220 oz.-in.
  • Motor Speed 4000rpm
  • Peak current 60A

12
Power System Components
  • Magmotor Inc.S28-400-E4
  • 200 oz-in continuous Torque
  • 1700 oz-in peak Torque
  • 10.54 krpm/V
  • Thunderpower Li-Poly batteries
  • 3.85 A-hr / cell
  • 48V (11V37V)
  • 22C continuous current capacity

13
Open Source Motor Controller (OSMC)
  • Low price (200/unit)
  • Low on-resistance
  • Analog PWM Input
  • Manufacturer Rating
  • 13-50 VDC
  • 160A Continuous
  • 400A Surge Current

14
Sparkfun IMU
  • 6DOF ADXRS Gyro
  • Pitch rate of 150 deg/s
  • 10 bit data resolution
  • Low price 359
  • DSP capable of running simple filtering/calibratio
    n algorithms
  • Accelerometers to measurestatic angle

15
Processor
  • Texas Instruments F2812 DSP
  • Processing Power
  • Multiple Peripheral Interfaces (QEP, SCI, ADC,
    PWM)
  • Floating Point Processing
  • Programmable in C/C
  • Extensive code base
  • In house knowledge

16
Mechanical Design
  • Minimized moments of inertia
  • 3 DOF Adjustable User Interface
  • Drive shaft alignment 0.002
  • Dowel pins for precise alignment
  • Battery isolation
  • Accessible electrical components.

17
(No Transcript)
18
Mechanical Design
19
Mechanical DesignFinite Element Analysis
  • Aluminum Chassis
  • Ease of machining
  • Easy to hold tolerances
  • Lightweight
  • FEA
  • Shaft bending and shear load

20
Manufacturing and Testing
21
Manufacturing and Assembly
  • Machining and Assembly
  • Over 100 shop hours per team members
  • Extra time required for tight tolerances
  • Shaft and drivetrain tolerances /- .002
  • Lessons learned
  • Underestimated machining time
  • Limited access to precision tools
  • Limited manpower

22
Lessons Learned OSMC
  • OSMC did not perform as specified by manufacturer
  • Official Rating 50VDC max
  • Actual Rating 36VDC max
  • Lacked Over-voltage Protection
  • Lessons Learned
  • Proper over-voltage protection
  • Emergency motor disconnect
  • Voltage safety margin
  • Reputable manufacturer

23
Lessons Learned Electrical Power System
  • Separate motor relay/switch
  • Separate motor and battery overcurrent protection
  • Battery quick disconnect
  • Wire harness for battery tray

24
Lessons Learned Sparkfun IMU
  • Sparkfun IMU
  • Required extensive characterization
  • Damaged during initial testing
  • Lessons learned
  • IMU data processing exceeded project scope
  • Reliability of manufacturer
  • DMU 300

25
Timeline
26
Thresholds and ObjectivesAchieved
27
Thresholds and ObjectivesAchieved
28
Economic Analysis
29
The Next Steps
  • Optimize Balancing
  • Improve sensor feedback
  • Improve Data processing
  • Achieve Top Speed
  • Robust 48V motor drivers
  • Additional safety measures
  • Endurance
  • Eliminate testing casters
  • Use full battery loadout
  • Travel to KAIST to study in Humanoid Robotics
    lab

30
Societal Impact
  • Users
  • Reduced fatigue
  • Improved mobility
  • User becomes less active
  • Peers
  • Slight risk of collision/injury
  • Society
  • Enabling device that will allow humanoids to play
    more crucial role

31
Environmental Impact
  • Costs of Production and Disposal
  • Motor
  • Batteries
  • Electronics
  • Operation
  • Generation of electricity and associated
    emissions (CO / NOx / SOx)

32
Deliverables
  • S.H.I.F.T. Prototype
  • Source Code
  • Design drawings and wire layouts
  • Associated research documentation
  • Motor Specifications, QFD analysis, etc.

33
Acknowledgements
  • Dr. Paul Oh
  • Dr. B.C. Chang
  • DASL Students
  • Ellenberg Family
  • MEM Department

34
QUESTIONS ?
35
(No Transcript)
36
Motors
  • Magmotor Inc.S28-400-E4Servomotor
  • Low terminal resistance/inductance
  • Light weight (6 lb)
  • 42 commutator bars
  • Rotor winding for 48V supply
  • NEMA 34 Mount

37
Gearboxes and Drivetrain
  • Danaher MotionNEMA-True 34 (NT34-010)
  • 13 arc. Min precision
  • 93 efficient
  • 700 in.-lbs. output Torque
  • Common 101 Ratio
  • largest reduction without 2nd stage
  • Bolt directly to S28-400-E4
  • 530 each

38
Gearboxes and Drivetrain cont.
  • 3/4 Belts and Pulleys to act as clutch.
  • Protect motor
  • Isolates external loads from motor and gearbox
  • Change gear ratios cheaply and easily.
  • 1.441 pulley ratio gives final drive ratio
    14.41
  • Max speed of about 15mph

39
Tires and Wheels
  • Kevlar Belted Tires
  • Skyway Mags
  • 20 Utility Configuration
  • ¾ Hub 3/16 Keyway
  • Discounted to 30 Shipping
  • No Non-Disclosure Agreement
  • Lightweight
  • 3lbs vs. 12-14lbs.

40
Processor
  • Texas Instruments F2812 DSP
  • Complete development package
  • Programmable in C/C
  • Extensive code base
  • In house knowledge

40
41
IMU
  • Sparkfun IMU, 6DOF ADXRS Gyro
  • Pitch rate of 150 deg/s sufficient
  • 10 bit data resolution sufficient
  • Low price 359
  • DSP capable of running simple filtering/calibratio
    n algorithms

42
Optical Encoders
  • Grayhill 63R256
  • Small form factor
  • Sufficient angular resolution
  • Best performance/
  • Best response

43
Motor Drivers
  • Open Source Motor Controller (OSMC)
  • Low price (200/unit)
  • Simple, robust, H-Bridge
  • Low on-resistance
  • Compatible w/ DSP PWM output

44
Project Risk Items - High
45
Project Risk Items - Medium
46
Project Risk Items - Low
47
Software Development
  • Reading sensors
  • Encoders (QEP)
  • IMU (Serial)
  • Analog input and gain adjustment
  • Calculate states
  • Apply gains/negative feedback
  • Generate PWM waveform

48
Original Economic Analysis
  • Should we delete this?
Write a Comment
User Comments (0)