NonLinMPC with LinModel uses SteadyKalmanFilter by default

Author: franckgaga

docs/src/public/predictive_control.md

@@ -6,10 +6,10 @@ Pages = ["predictive_control.md"]
 
 All the predictive controllers in this module rely on a state estimator to compute the
 predictions. The default [`LinMPC`](@ref) estimator is a [`SteadyKalmanFilter`](@ref), and
-[`NonLinMPC`](@ref), an [`UnscentedKalmanFilter`](@ref). For simpler and more classical
-designs, an [`InternalModel`](@ref) structure is also available, that assumes by default
-that the current model mismatch estimation is constant in the future (same approach than
-dynamic matrix control, DMC).
+[`NonLinMPC`](@ref) with nonlinear models, an [`UnscentedKalmanFilter`](@ref). For simpler
+and more classical designs, an [`InternalModel`](@ref) structure is also available, that
+assumes by default that the current model mismatch estimation is constant in the future
+(same approach than dynamic matrix control, DMC).
 
 !!! info
     The nomenclature uses hats for the predictions (or estimations, for the state

example/juMPC.jl

@@ -151,6 +151,7 @@ test_mpc(linModel4, mpc)
 
 test_mpc(linModel4, nmpc)
 @time u_data, y_data, r_data, d_data = test_mpc(linModel4, nmpc)
+@profview test_mpc(linModel4, nmpc)
 
 test_mpc(nonLinModel2, nmpc2)
 @time u_data, y_data, r_data, d_data = test_mpc(nonLinModel2, nmpc2)

src/controller/nonlinmpc.jl

@@ -105,8 +105,10 @@ since ``H_c ≤ H_p`` implies that ``\mathbf{u}(k+H_p) = \mathbf{u}(k+H_p-1)``.
 !!! tip
     Replace any of the 3 arguments with `_` if not needed (see `JE` default value below).
 
-This method uses the default state estimator, an [`UnscentedKalmanFilter`](@ref) with 
-default arguments. 
+This method uses the default state estimator :
+
+- if `model` is a [`LinModel`](@ref), a [`SteadyKalmanFilter`](@ref) with default arguments;
+- else, an [`UnscentedKalmanFilter`](@ref) with default arguments. 
 
 !!! warning
     See Extended Help if you get an error like `MethodError: no method matching 
@@ -152,6 +154,7 @@ and `h` functions must be compatible with this feature. See [Automatic different
 for common mistakes when writing these functions.
 """
 NonLinMPC(model::SimModel; kwargs...) = NonLinMPC(UnscentedKalmanFilter(model); kwargs...)
+NonLinMPC(model::LinModel; kwargs...) = NonLinMPC(SteadyKalmanFilter(model); kwargs...)
 
 
 """

test/test_predictive_control.jl

@@ -76,9 +76,10 @@ end
     @test mpc1.estim.x̂ ≈ [1,2,3,4]
 end
 
-
 @testset "NonLinMPC construction" begin
     linmodel1 = LinModel(sys,Ts,i_d=[3])
+    nmpc0 = NonLinMPC(linmodel1, Hp=15)
+    @test isa(nmpc0.estim, SteadyKalmanFilter)
     f(x,u,d) = linmodel1.A*x + linmodel1.Bu*u + linmodel1.Bd*d
     h(x,d)   = linmodel1.C*x + linmodel1.Du*d
     nonlinmodel = NonLinModel(f, h, Ts, 2, 4, 2, 1)