Differences

This shows you the differences between two versions of the page.

--- courses:b4m36uir:labs:lab07 [2019/11/10 23:22]
pragrmi1
+++ courses:b4m36uir:labs:lab07 [2019/11/10 23:27]
pragrmi1
@@ Line 13: / Line 13: @@
 \\
-===Solving TSPN with Self-organizing Map===
+==== Dubins Maneuver ====
+[[https://cw.fel.cvut.cz/wiki/_media/courses/b4m36uir/lectures/b4m36uir-lec06-slides.pdf|Lecture 06]]
-Slide 15 from the [[https://cw.fel.cvut.cz/wiki/_media/courses/b4m36uir/lectures/b4m36uir-lec05-slides.pdf|Lecture 05]]:
+\\
-{{:courses:b4m36uir:labs:som.png?700|}}
+==== Decoupled Approach ====
+Slide 36 in [[https://cw.fel.cvut.cz/wiki/_media/courses/b4m36uir/lectures/b4m36uir-lec06-slides.pdf|Lecture 06]]
-Utilize 'alternate goal' concept for solving TSP with neighborhoods (TSPN). In each epoch, the neurons are adapted towards the goals which inhibits them. But, in the TSPN, the neurons are adapted to the closes point in the specific goal neighborhood. Therefore, this concept enables to find shorter solutions, see the right image and the following GIF with SOM evolution.
+The decoupled approach comprises the following basic steps
+    - Estimate sequence of visits by Euclidean TSP connecting centers of the regions.
-{{:courses:b4m36uir:labs:som_tspn_orig.png?400|}}
+    - For each region, sample boundary points and heading angles.
-{{:courses:b4m36uir:labs:som_tspn.png?400|}}
+    - Find the shortest feasible tour comprising Dubins maneuvers connecting the regions, where the sequence of visits is estimated from the ETSP.
-Click on the following image to see the SOM evolution in GIF.
-{{:courses:b4m36uir:labs:myimage.gif?300|}}