Artificial Intelligence for Robotics by Francis X. Govers

Author:Francis X. Govers
Language: eng
Format: epub, pdf
Tags: COM067000 - COMPUTERS / Hardware / General, COM016000 - COMPUTERS / Computer Vision and Pattern Recognition, COM041000 - COMPUTERS / Microprocessors
Publisher: Packt Publishing
Published: 2018-08-30T11:44:39+00:00

Now that we have done all our processing, we add this new child path to our population, and get ready for the next generation to be evaluated. We record some data and loop back to the start:

# welcome the new baby from parent 1 (p1) and parent 2 (p2)

print("Generation: ",jj,"New population = ",len(newPop))

pop=newPop

mp.plot(trainingData)

mp.show()

So, how did we do with our mad genetic experiment? The following output chart speaks for itself:

The genetic algorithm, for all it seems like a bit of voodoo programming, works quite well as a machine learning tool for this specific case of training our robot arm. Our solution peaked at 99.76% of the goal (about 2 mm) after just 90 generations or so, which is quite fast for an AI learning process. You can see the smooth nature of the learning that shows that this approach can be used to solve problems in path planning for our robot arm. I have to admit that I was quite skeptical about this process, but it seems to work quite well for this particular problem domain.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.