Loading...

High Precision Localization by Optical Flow and INS Sensor data Fusion

Azizi, Arash | 2014

475 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 45891 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Vossoughi, Gholamreza
  7. Abstract:
  8. Robot navigation and guidance is an important issue in robotic. Navigation by itself includes localization, path planning and guidance. For having a successful navigation it is necessary to succeed in all of those parts. The localization precision as basic part of localization is important and having more accurate information of robot position is important. Recently, sensors which use optical flow technic and are widely used in optical mouses are so common and they are able to measure displacement with resolution of 3 to 63 micron. In this research, a localization method for micro robot localization based on optical flow and INS sensor data fusion is presented.
    Two data fusion method based on extended kalman filter as first and the second method based on attenuating stochastic noise of sensors by kalman filter and then fusing sensed data by other equations are presented. Their performance were evaluated by simulation and the second method was choose based on the localization error.
    We have used common INS and two optical flow sensor for experiment of the second method. At first we did tests to recognize the sensors performance and model their error. Then localization was done for two path including one two degree of freedom and one three degree of freedom movement. The localization error was about 2 mili meter which was approximately equal the simulation method. Besides, some tests for rcognizing the properties of suitable surface, from the perspective of grain size were done and the results showed that surface with grain size bigger than sensor resolution and high variation in shape and position are suitable and will lead to lower error for optical flow sensor measurement
  9. Keywords:
  10. Location ; Grain Size ; Kalman Filters ; Optical Flow ; Inertial Sensor ; Inertial Navigation System

 Digital Object List

 Bookmark

No TOC