codecov[bot]
commented
2 years ago Codecov Report
Merging #76 (df7d7a6) into main (96c4a3c) will decrease coverage by
16.13%. The diff coverage is29.11%.
@@ Coverage Diff @@
## main #76 +/- ##
===========================================
- Coverage 79.50% 63.36% -16.14%
===========================================
Files 15 16 +1
Lines 1361 1736 +375
===========================================
+ Hits 1082 1100 +18
- Misses 279 636 +357 | Impacted Files | Coverage Δ | |
|---|---|---|
| src/edmd/edmd.jl | 100.00% <ø> (ø) |
|
| src/linear_model.jl | 0.00% <0.00%> (ø) |
|
| src/utils.jl | 80.95% <ø> (-1.98%) |
:arrow_down: |
| src/trajopt_qp.jl | 28.26% <28.26%> (ø) |
|
| src/edmd/regression.jl | 46.03% <40.90%> (-23.54%) |
:arrow_down: |
| src/edmd/eigenfunctions.jl | 37.68% <75.00%> (-2.60%) |
:arrow_down: |
| src/lqr_solver.jl | 100.00% <100.00%> (ø) |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact),ø = not affected,? = missing dataPowered by Codecov. Last update 739a976...df7d7a6. Read the comment docs.
Adds an example that learns a bilinear model for the pendulum about the top equilibrium. A convex MPC controller is used to stabilize the real nonlinear system about the upright position (the unstable equilibrium).
Check out the script in
examples/learned_models/pendulum_equilibrium.jl.As a part of this, I added a few new controllers, most importantly the
BilinearMPCcontroller that takes a bilinear model and forms a MPC controller for tracking a reference trajectory of the original system.