migration to JuMP + OSQP packages for LinMPC

Author: franckgaga

Project.toml

@@ -1,7 +1,7 @@
 name = "ModelPredictiveControl"
 uuid = "61f9bdb8-6ae4-484a-811f-bbf86720c31c"
 authors = ["Francis Gagnon"]
-version = "0.2.1"
+version = "0.3.0"
 
 [deps]
 ControlSystemsBase = "aaaaaaaa-a6ca-5380-bf3e-84a91bcd477e"

src/predictive_control.jl

@@ -196,8 +196,8 @@ arguments.
 - `Lwt=fill(0.0,model.nu)` : main diagonal of ``\mathbf{L}`` weight matrix (vector)
 - `Cwt=1e5` : slack variable weight ``C`` (scalar), use `Cwt=Inf` for hard constraints only
 - `ru=model.uop` : manipulated input setpoints ``\mathbf{r_u}`` (vector)
-- `optim=JuMP.Model(OSQP.MathOptInterfaceOSQP.Optimizer)` : the MPC quadratic optimizer, 
-  provided as a [`JuMP.Model`](https://jump.dev/JuMP.jl/stable/reference/models/#JuMP.Model)
+- `optim=JuMP.Model(OSQP.MathOptInterfaceOSQP.Optimizer)` : quadratic optimizer used in
+  the predictive controller, provided as a [`JuMP.Model`](https://jump.dev/JuMP.jl/stable/reference/models/#JuMP.Model)
 
 # Extended Help
 Manipulated inputs setpoints ``\mathbf{r_u}`` are not common but they can be interesting