changed: Ipopt nlp_scaling_max_gradient default to 10.0/C

Author: franckgaga

src/controller/nonlinmpc.jl

@@ -276,7 +276,7 @@ Init the nonlinear optimization for [`NonLinMPC`](@ref) controllers.
 """
 function init_optimization!(mpc::NonLinMPC, optim::JuMP.GenericModel{JNT}) where JNT<:Real
     # --- variables and linear constraints ---
-    con = mpc.con
+    C, con = mpc.C, mpc.con
     nΔŨ = length(mpc.ΔŨ)
     set_silent(optim)
     limit_solve_time(mpc.optim, mpc.estim.model.Ts)
@@ -285,6 +285,14 @@ function init_optimization!(mpc::NonLinMPC, optim::JuMP.GenericModel{JNT}) where
     b = con.b[con.i_b]
     @constraint(optim, linconstraint, A*ΔŨvar .≤ b)
     # --- nonlinear optimization init ---
+    if !isinf(C) && solver_name(optim) == "Ipopt"
+        try
+            get_attribute(optim, "nlp_scaling_max_gradient")
+        catch
+            # default "nlp_scaling_max_gradient" to `10.0/C` if not already set:
+            set_attribute(optim, "nlp_scaling_max_gradient", 10.0/C)
+        end
+    end
     model = mpc.estim.model
     ny, nx̂, Hp, ng = model.ny, mpc.estim.nx̂, mpc.Hp, length(con.i_g)
     # inspired from https://jump.dev/JuMP.jl/stable/tutorials/nonlinear/tips_and_tricks/#User-defined-operators-with-vector-outputs

src/estimator/mhe/construct.jl

@@ -964,17 +964,26 @@ Init the nonlinear optimization of [`MovingHorizonEstimator`](@ref).
 function init_optimization!(
     estim::MovingHorizonEstimator, model::SimModel, optim::JuMP.GenericModel{JNT},
 ) where JNT<:Real
-    He, con = estim.He, estim.con
-    nV̂, nX̂, ng = He*estim.nym, He*estim.nx̂, length(con.i_g)
-    # --- variables and linear constraints ---
+    C, con = estim.C, estim.con
     nZ̃ = length(estim.Z̃)
+    # --- variables and linear constraints ---
     set_silent(optim)
     limit_solve_time(estim.optim, estim.model.Ts)
     @variable(optim, Z̃var[1:nZ̃])
     A = estim.con.A[con.i_b, :]
     b = estim.con.b[con.i_b]
     @constraint(optim, linconstraint, A*Z̃var .≤ b)
     # --- nonlinear optimization init ---
+    if !isinf(C) && solver_name(optim) == "Ipopt"
+        try
+            get_attribute(optim, "nlp_scaling_max_gradient")
+        catch
+            # default "nlp_scaling_max_gradient" to `10.0/C` if not already set:
+            set_attribute(optim, "nlp_scaling_max_gradient", 10.0/C)
+        end
+    end
+    He = estim.He
+    nV̂, nX̂, ng = He*estim.nym, He*estim.nx̂, length(con.i_g)
     # see init_optimization!(mpc::NonLinMPC, optim) for details on the inspiration
     Jfunc, gfunc = let estim=estim, model=model, nZ̃=nZ̃ , nV̂=nV̂, nX̂=nX̂, ng=ng
         last_Z̃tup_float, last_Z̃tup_dual = nothing, nothing

src/precompile.jl

@@ -1,8 +1,8 @@
 sys = [ 
-    tf(1.90, [18, 1]) tf(1.90, [18, 1]);
-    tf(-0.74,[8, 1])  tf(0.74, [8, 1]) 
+    tf(1.90, [1800, 1]) tf(1.90, [1800, 1]);
+    tf(-0.74,[800, 1])  tf(0.74, [800, 1]) 
 ]
-Ts = 4.0
+Ts = 400.0
 model = setop!(LinModel(sys, Ts), uop=[10, 10], yop=[50, 30])
 y = model()
 
@@ -74,7 +74,7 @@ initstate!(nmpc_im, nlmodel.uop, y)
 u = nmpc_im([55, 30], ym=y)
 sim!(nmpc_im, 3, [55, 30])
 
-nmpc_ukf = setconstraint!(NonLinMPC(UnscentedKalmanFilter(nlmodel), Hp=10, Cwt=10), ymin=[45, -Inf])
+nmpc_ukf = setconstraint!(NonLinMPC(UnscentedKalmanFilter(nlmodel), Hp=10, Cwt=1e3), ymin=[45, -Inf])
 initstate!(nmpc_ukf, nlmodel.uop, y)
 u = nmpc_ukf([55, 30])
 sim!(nmpc_ukf, 3, [55, 30])