Normalizing Each Objective Class's motionCost()

[akash-venkateshwaran]

Description

• Resolves Custom Weights for Different Optimization Objective sailbot_workspace#290

• Added two functions in the parent class: (i) Creates random sample states; (ii) Compares every pair in that sample and picks the maximum motionCost()
• Inherited these functions in each Objective class to find the max_motionCost for each objective function
• Modified the motionCost() function of each Objective class to normalise its cost using max_motionCost

Verification

• Verified using a simple setup example problem
• Haven't included unit tests

Resources

[patrick-5546]

I added some tests in test_distance_objective that we should use for the other objectives:

• sample_states(): commented out checking that all states are valid, which I think is what we want right?
• find_maximum_motion_cost()
  • Implicitly test by comparing max_motion_cost to a hardcoded value
  • Tried to test find_maximum_motion_cost() directly but was not able to get the type of states correct so I commented it out for now
    • Update: fixed and uncommented

If you want to merge this PR today and aren't able to finish the tests, I'm ok with merging this PR and working on the tests in another PR. Just leave some # TODO: <message> comments

[akash-venkateshwaran]

I added some tests in test_distance_objective that we should use for the other objectives:

• sample_states(): commented out checking that all states are valid, which I think is what we want right?

• find_maximum_motion_cost()

  • Implicitly test by comparing max_motion_cost to a hardcoded value

  • Tried to test find_maximum_motion_cost() directly but was not able to get the type of states correct so I commented it out for now

    • Update: fixed and uncommented
      If you want to merge this PR today and aren't able to finish the tests, I'm ok with merging this PR and working on the tests in another PR. Just leave some # TODO: <message> comments

Thanks for adding the unit tests.
Yes we want to make sure all the sampled states are valid which means they are not in the collision area
Great! So now I have to repeat these tests for other objective functions right?

[patrick-5546]

Great! So now I have to repeat these tests for other objective functions right?

Yes and ensure that they are the sampled states are valid. The validity checker currently is just a placeholder but it will be good to set this up now anyways.

[patrick-5546]

It seems like sometimes the normalized cost is greater than 1: https://github.com/UBCSailbot/local_pathfinding/actions/runs/7688802043/job/20950435771

[patrick-5546]

As discussed, just need to resolve merge conflicts:

• git merge origin main
• Add speed objective
• Normalize speed objective
• Update weights