06.03.2013 Views

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

The SOM algorithm has been successfully applied by Heikkonen et al.<br />

[8] to online navigational planning. An outline of the algorithm for online<br />

navigation is presented below.<br />

Sensor Inputs<br />

Training Vectors<br />

Next direction<br />

Current direction<br />

Fig. 24.17: Architecture of SOM showing the association between<br />

the sensory inputs <strong>and</strong> the control outputs.<br />

Algorithm<br />

Input: A set of l vector of 10 dimension, 8 of which denote the sensory<br />

reading for range of obstacles, 1 for current direction of movement<br />

<strong>and</strong> 1 for the next direction of movement<br />

A set of such patterns is used as inputs;<br />

A set of n 2 neurons with r<strong>and</strong>om weights on a 2D plane<br />

Output: The weight corresponding to the next direction of movement in the<br />

previous vector.<br />

The algorithm consists of two modules: Procedure training <strong>and</strong> Find-direction.<br />

Procedure Training<br />

Begin<br />

For each input pattern<br />

Call algorithm for SOM<br />

End For;<br />

End.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!