Artificial Intelligence for Robotics

Q-learning implementation

Now in this next part of the computer code, we will be evolving our arm learning code to include something that you may have noted was missing from the first program in the preceding section. It did not do any path planning; it just provided the combination of three motor positions that resulted in the the hand being in the proper position. It did not try to find a good path from the starting point to the ending point. Our next program will do just that, and will utilize the Q-learning technique we discussed.

I'm going to leave out the utility functions that we developed in the first program that are reused in the second program. The complete code will be available for download from the GitHub website. This lets ...