Escaping the Maze

Click here to load reader

  • date post

    07-Jan-2016
  • Category

    Documents

  • view

    26
  • download

    0

Embed Size (px)

description

Signals , Instruments and Systems. Escaping the Maze. Ezzat Leïla & Orthlieb Camille June 2010. Steps of the p roject. Description of the environment. Simplification of the world E- puck turns always left E- puck follows the wall at equal distance - PowerPoint PPT Presentation

Transcript of Escaping the Maze

Diapositive 1

Escaping the Maze

Ezzat Lela & Orthlieb CamilleJune 2010

Signals, Instruments and Systems1Steps of the project

Description of the environmentSimplification of the worldE-puck turns always leftE-puck follows the wall at equal distanceUse of front and side sensors (0,2 and 5)Use of compass and GPS datas E-puck starts updating its position after a corner

Detection of changes of colorsResponse value sent after each time stepNumber of time steps stocked before each change of colorDetection and storage of change of color

Link frequencies and locationDetect frequencyUpdate believed positions from the cornerDeduce the wallKnow from which corner he startsFind the good believed position

Update his believed positions after the first cornerwe compute the four coordonates for each starting cornerOn the same time, he reads the frequencies and after 3 changes of color, he deduces the right frequencyOnce he found the frequency, he knows on which wall he is, so he knows from which corner he starts. And he deduces which believed position is right.Finally, we can compare the position found with the GPS data.

Like you see in this schema, Like in mazeFrom a known position, find the final point the mazes escape

To approach a little bite more the initial goal, that is escaping a maze, we introduce a box on the center of the environment. Finally, the aim of the e-puck is to find his position and from this position deduce the trajectory to enter in the red box. Expliquer la trajectoire du robotEffect of noise on the IR sensor Front Sensor : to move and turn 1% of noise doesnt change anything 10% the robot cant follow the wall

Right sensor : to detect frequency 30% of noise doesnt affect the measures 50% doesnt detect any change of color

We tried to add noise on the IR sensor. We had to change our threshold (like to turn)First, On the front sensor used for the displacement, we add from 1% of noise, it doesnt change very much, we just see that the robot doesnt go very straightWith 10% the robot cant follow the wall, and driftOn the right sensor (detecting the frequency), with 30% he even manages to detect the frequency and his position because the difference of values between two colors is large, but not with 50%... 7The movieConclusionMain objectives reachedOur findings have to be improved in order to escape a real maze Make a code enought flexible to reproduce many times the lecture of frequencies and update positionsRevise the stability of the detect valuesFace more random situationsAppreciations