Conversation
…h-testbed into feature/wpts_mpc
…nto feature/wpts_mpc
…h-testbed into feature/wpts_mpc
AaronYoung5
left a comment
AaronYoung5
left a comment
There was a problem hiding this comment.
@harryzhang1018 let's clean this up sooner rather than later
atk.yml
Outdated
| USER_GID: "@{gid}" | ||
| APT_DEPENDENCIES: "bash zsh vim git git-lfs python3-pip python3-tk python3-opencv rapidjson-dev" | ||
| PIP_REQUIREMENTS: "wa_simulator pandas matplotlib numpy>=1.19 opencv-python tornado black Pillow torch torchvision pyserial nvidia-pyindex" | ||
| PIP_REQUIREMENTS: "wa_simulator pandas matplotlib numpy>=1.19 opencv-python tornado black Pillow torch torchvision pyserial nvidia-pyindex casadi cvxpy scipy osqp" |
There was a problem hiding this comment.
do we still need these?
There was a problem hiding this comment.
Yes, "casadi cvxpy scipy osqp" is for MPC, I also need to add pip and apt dependencies for RTK/GPS driver to run
| vehicle.SetTireType(tire_model) | ||
| vehicle.SetInitPosition(chrono.ChCoordsysD(initLoc, initRot)) | ||
| vehicle.SetTireStepSize(tire_step_size) | ||
| vehicle.SetMaxMotorVoltageRatio(0.16) |
There was a problem hiding this comment.
needed? is this only in chrono-internal?
There was a problem hiding this comment.
I don't think this needed to be merged into main cuz we are not gonna run simulation on Jetson right?
There was a problem hiding this comment.
@harryzhang1018 your job is to figure out what we need here (if anything)
There was a problem hiding this comment.
Pretty much all of the keras files can be deleted... We have a seperate repo (which should end up somewhere on SBEL github) for training, and it takes <20 min to train each model.
There was a problem hiding this comment.
Might just be nice to have at least one either on the github or loaded (similar to Yolo)... but that should come from a control node PR not from here
|
|
||
|
|
||
|
|
||
| def mpc_cvxpy_solver_v2(xref,yref,v_current,u_current): |
There was a problem hiding this comment.
Keep? if yes, please add docstrings
There was a problem hiding this comment.
can we just create a data folder somewhere that has all these paths?
There was a problem hiding this comment.
This also doesn't need to be merged, part of Path Planning PR
|
|
||
| self.go = False | ||
| self.error_state = VehicleState() | ||
| self.file = open("/home/art-ishaan/art-ishaan/workspace/src/path_planning/path_planning/lot17_sinsquare_sv.csv") |
There was a problem hiding this comment.
let's not use hardcoded paths
|
i'd recommend cloning locally and deleting stuff you don't need and then pushing back to this branch. Do we need all these datafiles? Can we have a datafolder on box or something that houses all of them? |
|
@AaronYoung5 @StefanCaldararu |
StefanCaldararu
left a comment
StefanCaldararu
left a comment
There was a problem hiding this comment.
@harryzhang1018 Looked through this. IMO, best way to go about things is as follows:
- start by making sure everything that we want to keep is backed up. From what I can tell, these are path files and keras files. (I think you've already done this). I think it's fine to have these backed up externally for now, until we find a home for them (we should really take care of #39).
- Test the real vehicle with the new changes Aaron has made, and try and get that working.
- Delete this branch/pr once nothing is needed anymore.
3 participants
Don't merge into master branch