August 4
-
So to optimise these variable we start with Gradient Descent method.
-
The idea is to take one variable, for example, the attraction force, and the rest with a fixed value. We then look for the value that minimizes the error.
-
Error is calculate based on the difference between the reference path and the path that is travelled by the robot.
Task for tomorrow:
-
Developing an analytical tool to optimise the variables
-
devise an optimizing strategy
-
write down the steps in automating the workflow