update doc related to ŷ debug

Author: franckgaga

docs/src/manual/nonlinmpc.md

@@ -191,7 +191,7 @@ The keyword argument `Ewt` weights the economic costs relative to the other term
 objective function. The term must be large enough to be significant but a too high value can
 lead to a static error on the angle setpoint. The second element of `Mwt` is zero since the
 speed ``ω`` is not requested to track a setpoint. The closed-loop response to a 180°
-setpoint is similar:
+setpoint is:
 
 ```@example 1
 unset_time_limit_sec(empc.optim) # hide
@@ -202,7 +202,7 @@ savefig(ans, "plot5_NonLinMPC.svg"); nothing # hide
 
 ![plot5_NonLinMPC](plot5_NonLinMPC.svg)
 
-and the energy consumption is almost identical:
+and the energy consumption is slightly lower:
 
 ```@example 1
 function calcW(res)
@@ -212,7 +212,7 @@ end
 Dict(:W_nmpc => calcW(res_ry), :W_empc => calcW(res2_ry))
 ```
 
-But, for a 10° step disturbance:
+Also, for a 10° step disturbance:
 
 ```@example 1
 res2_yd = sim!(empc, N, [180; 0]; plant=plant2, x0=[π, 0], x̂0=[π, 0, 0], y_step=[10, 0])

src/predictive_control.jl

@@ -552,7 +552,7 @@ end
 @doc raw"""
     initpred!(mpc, model::LinModel, d, ym, D̂, R̂y, R̂u)
 
-Init linear model prediction matrices `F`, `q̃` and `p`.
+Init linear model prediction matrices `F, q̃, p` and current estimated output `ŷ`.
 
 See [`init_predmat`](@ref) and [`init_quadprog`](@ref) for the definition of the matrices.
 """
@@ -581,7 +581,7 @@ end
 @doc raw"""
     initpred!(mpc::PredictiveController, model::SimModel, d, ym, D̂, R̂y, R̂u)
 
-Init `Ŷop`, `d0` and `D̂0` matrices when model is not a [`LinModel`](@ref).
+Init `ŷ, Ŷop, d0, D̂0` matrices when model is not a [`LinModel`](@ref).
 
 `d0` and `D̂0` are the measured disturbances and its predictions without the operating points
 ``\mathbf{d_{op}}``. The vector `Ŷop` is kept unchanged if `mpc.estim` is not an

test/test_predictive_control.jl

@@ -372,6 +372,8 @@ end
     nmpc = NonLinMPC(linmodel, Mwt=[0], Nwt=[0], Cwt=Inf, Ewt=1, JE=JE, Hp=Hp, Hc=1)
     u = moveinput!(nmpc)
     @test u ≈ [1] atol=5e-2
+    # ensure that the current estimated output is updated for correct JE values:
+    @test nmpc.ŷ ≈ ModelPredictiveControl.evalŷ(nmpc.estim, nothing, Float64[])
     linmodel2 = LinModel([tf(5, [200, 1]) tf(7, [800,1])], 300.0, i_d=[2])
     f(x,u,d) = linmodel2.A*x + linmodel2.Bu*u + linmodel2.Bd*d
     h(x,d)   = linmodel2.C*x + linmodel2.Dd*d