Computerized Labyrinth Solver - PowerPoint PPT Presentation

1 / 22
About This Presentation
Title:

Computerized Labyrinth Solver

Description:

Computerized Labyrinth Solver. Gregory Schallert. Chad Craw. System Overview. Motors Control Playing Platform by Video feedback. Video processing is done by the PC ... – PowerPoint PPT presentation

Number of Views:30
Avg rating:3.0/5.0
Slides: 23
Provided by: ceR4
Category:

less

Transcript and Presenter's Notes

Title: Computerized Labyrinth Solver


1
Computerized Labyrinth Solver
  • Gregory Schallert
  • Chad Craw

2
System Overview
  • Motors Control Playing Platform by Video
    feedback.
  • Video processing is done by the PC
  • Motor control is done by the HC12

3
User Interface
  • Minimal amount of buttons and LEDs
  • The interface shall report current operating
    status, and any encountered errors through LED
    codes

4
User Interface
5
System Test on Startup
  • The system will rotate the maze surface a maximum
    deflection on both axes
  • All status and Error LEDs will flash in unison to
    make sure the are operational
  • The system will analyze the current maze
    structure, and output a binary stream
    representing the wall placements as well as
    values of the current X and Y axis rotation to
    the Serial interface.

6
Dynamic Maze Construction
  • Maze walls are placed in arbitrary positions to
    create a new maze each time.
  • Limited Start and End Tags are placed along the
    outer edge of the wall

7
Maze Construction Diagrams
Wall Slot
Maze Surface
Contact Nodes
8
Maze Scanning Phase
  • An array of multiplexers are attached to the
    contact nodes of each wall slot.
  • These connections are scanned synchronously by
    the HC12 board to determine positions of walls.
  • Once this scan is complete, the virtual maze is
    passed to the connected PC via the serial
    interface to start the maze solving algorithm.

0010000111 0101001111 0111101101 0110111100 101010
0001 0001000101 1111001010 0010100111 1001011100 A
ng X 1.2 Ang Y 5.5 X Pos01 Y Pos54
9
Lee Moore Algorithm
  • Inputs Walls, Start and End Points.
  • Output Finds the a path from start to finish.
  • Each cell is given a value of -1.
  • Starting Cell is given a value of 0
  • If we have not reached the end, set the current
    cell to the highest value not yet used.
  • Each cell touching the current cell with a -1 has
    the value of this cell 1
  • Check each cell to see if we are at the end
  • If not, continue

10
Video Capture
  • Use the CCD Camera to find ball location
  • Compare the current location to the desired path
  • Send the offset to the next goal to the serial
    interface

11
The Camera
  • Logitech QuickCam CCD Camera
  • 352x288 Maximum True Resolution
  • 30 Frames/sec
  • 25cm/120 2mm grid
  • 25cm/288 lt 1mm grid
  • We want 4mm resolution
  • MEANS WE ARE GOOD TO GO

12
The Software
  • OpenCV
  • Video Tracking algorithms
  • DirectShow
  • Video Capture Filters
  • Logitech SDK
  • Camera Calibration and Windows Drivers
  • All of the Software is Free and Documented

13
The Process
  • DirectShow captures the video input
  • The OpenCV algorithms are applied to track the
    object
  • The information is sent to the HC12
  • Any additional text overlays are added and
    rendered to the screen

14
A System Model
  • Sum the masses
  • Sum the Torques
  • Solve for acceleration
  • This relates the acceleration to the angle of the
    platform

15
A System Model (cont)
  • Take two snapshots
  • One at the zero position
  • One after some time
  • Note R and d stay constant
  • Use law of cosines to solve for the angle
  • This relates the acceleration to the distance the
    control axle at r has turned
  • Can also solve for vertical displacement

16
Control System
  • The algorithms for the control system shall
    operate on the HC12 board.
  • The system will collect feedback data from the
    Video tracking algorithm on the PC via the Serial
    Interface of the HC12 board
  • This data shall be sent as a 3-byte packet at
    every sample interval of the tracking system
  • It shall contain a signal start byte (to maintain
    alignment), followed by 2 data bytes for the
    current X and Y offsets
  • The position of the ball while traversing the
    maze shall be maintained with a Fuzzy Logic
    Control system.
  • The only data that needs to be collected from the
    PC Video tracking system is the balls current
    offset from its target position

17
Control System
  • The FLC will use a rule set based on 5 states of
    the system
  • The ball is far to the left(top) of its target
    position
  • The ball is near to the left(top) of its target
    position
  • The ball is at its target position
  • The ball is near to the right(bottom) of its
    target position
  • The ball is far to the right(bottom) of its
    target position
  • These rules shall be applied symmetrically to
    both the X and Y axes of the balls position.
  • Using this rule-set, the control system shall
    derive the angular rotation necessary to move the
    ball in the desired direction at the desired
    speed.

18
Control System (Pathfinder)
Y offset 0 X offset 10
Y offset 0 X offset 0
Y
Y offset 10 X offset 0
X
Y offset 20 X offset 0
End
Start
19
Motor Control
  • Each motor shall be attached to the actuators of
    each axis, and shall operate independently of one
    another.
  • The operation of each motor is identical due to
    the symmetrical nature of the maze surface.

20
Motor Specs
  • SL561 Standard, BB
  • Size
  • Length 1.51
  • Width 0.73
  • Height 1.37
  • Torque 46.1 oz
  • Speed 0.18 sec/60
  • Weight 1.50 oz
  • Voltage 4.8V 6V

21
Possible Errors
  • Nonlinear elements of the system
  • Fuzzy Logic Control
  • Keep speed Down
  • Video Resolution/timing
  • Damp the system
  • Keep speed Down

22
Conclusion
Write a Comment
User Comments (0)
About PowerShow.com