Functions: StateEstimator Internals
Augmented Model
ModelPredictiveControl.f̂!
— Functionf̂!(x̂next0, û0, estim::StateEstimator, model::SimModel, x̂0, u0, d0) -> nothing
Mutating state function $\mathbf{f̂}$ of the augmented model.
By introducing an augmented state vector $\mathbf{x̂_0}$ like in augment_model
, the function returns the next state of the augmented model, defined as:
\[\begin{aligned} \mathbf{x̂_0}(k+1) &= \mathbf{f̂}\Big(\mathbf{x̂_0}(k), \mathbf{u_0}(k), \mathbf{d_0}(k)\Big) \\ \mathbf{ŷ_0}(k) &= \mathbf{ĥ}\Big(\mathbf{x̂_0}(k), \mathbf{d_0}(k)\Big) \end{aligned}\]
where $\mathbf{x̂_0}(k+1)$ is stored in x̂next0
argument. The method mutates x̂next0
and û0
in place, the latter stores the input vector of the augmented model $\mathbf{u_0 + ŷ_{s_u}}$.
f̂!(x̂next0, _ , estim::StateEstimator, model::LinModel, x̂0, u0, d0) -> nothing
Use the augmented model matrices if model
is a LinModel
.
f̂!(x̂next0, _ , estim::InternalModel, model::NonLinModel, x̂0, u0, d0)
State function $\mathbf{f̂}$ of InternalModel
for NonLinModel
.
It calls model.f!(x̂next0, x̂0, u0 ,d0)
since this estimator does not augment the states.
ModelPredictiveControl.ĥ!
— Functionĥ!(ŷ0, estim::StateEstimator, model::SimModel, x̂0, d0) -> nothing
Mutating output function $\mathbf{ĥ}$ of the augmented model, see f̂!
.
ĥ!(ŷ0, estim::StateEstimator, model::LinModel, x̂0, d0) -> nothing
Use the augmented model matrices if model
is a LinModel
.
ĥ!(ŷ0, estim::InternalModel, model::NonLinModel, x̂0, d0)
Output function $\mathbf{ĥ}$ of InternalModel
, it calls model.h!
.
Estimator Construction
ModelPredictiveControl.init_estimstoch
— Functioninit_estimstoch(model, i_ym, nint_u, nint_ym) -> As, Cs_u, Cs_y, nxs, nint_u, nint_ym
Init stochastic model matrices from integrator specifications for state estimation.
The arguments nint_u
and nint_ym
specify how many integrators are added to each manipulated input and measured outputs. The function returns the state-space matrices As
, Cs_u
and Cs_y
of the stochastic model:
\[\begin{aligned} \mathbf{x_{s}}(k+1) &= \mathbf{A_s x_s}(k) + \mathbf{B_s e}(k) \\ \mathbf{y_{s_{u}}}(k) &= \mathbf{C_{s_{u}} x_s}(k) \\ \mathbf{y_{s_{ym}}}(k) &= \mathbf{C_{s_{ym}} x_s}(k) \end{aligned}\]
where $\mathbf{e}(k)$ is an unknown zero mean white noise and $\mathbf{A_s} = \mathrm{diag}(\mathbf{A_{s_{u}}, A_{s_{ym}}})$. The estimations does not use $\mathbf{B_s}$, it is thus ignored. The function init_integrators
builds the state-space matrices.
ModelPredictiveControl.init_integrators
— Functioninit_integrators(nint, ny, varname::String) -> A, C, nint
Calc A, C
state-space matrices from integrator specifications nint
.
This function is used to initialize the stochastic part of the augmented model for the design of state estimators. The vector nint
provides how many integrators (in series) should be incorporated for each output. The argument should have ny
element, except for nint=0
which is an alias for no integrator at all. The specific case of one integrator per output results in A = I
and C = I
. The estimation does not use the B
matrix, it is thus ignored. This function is called twice :
- for the unmeasured disturbances at manipulated inputs $\mathbf{u}$
- for the unmeasured disturbances at measured outputs $\mathbf{y^m}$
ModelPredictiveControl.augment_model
— Functionaugment_model(model::LinModel, As, Cs; verify_obsv=true) -> Â, B̂u, Ĉ, B̂d, D̂d, x̂op, f̂op
Augment LinModel
state-space matrices with the stochastic ones As
and Cs
.
If $\mathbf{x_0}$ are model.x0
states, and $\mathbf{x_s}$, the states defined at init_estimstoch
, we define an augmented state vector $\mathbf{x̂} = [ \begin{smallmatrix} \mathbf{x_0} \\ \mathbf{x_s} \end{smallmatrix} ]$. The method returns the augmented matrices Â
, B̂u
, Ĉ
, B̂d
and D̂d
:
\[\begin{aligned} \mathbf{x̂_0}(k+1) &= \mathbf{Â x̂_0}(k) + \mathbf{B̂_u u_0}(k) + \mathbf{B̂_d d_0}(k) \\ \mathbf{ŷ_0}(k) &= \mathbf{Ĉ x̂_0}(k) + \mathbf{D̂_d d_0}(k) \end{aligned}\]
An error is thrown if the augmented model is not observable and verify_obsv == true
. The augmented operating points x̂op
and f̂op
are simply $\mathbf{x_{op}}$ and $\mathbf{f_{op}}$ vectors appended with zeros (see setop!
).
augment_model(model::SimModel, As, _ , _ ) -> Â, B̂u, Ĉ, B̂d, D̂d, x̂op, f̂op
Return empty matrices, and x̂op
& f̂op
vectors, if model
is not a LinModel
.
ModelPredictiveControl.init_ukf
— Functioninit_ukf(model, nx̂, α, β, κ) -> nσ, γ, m̂, Ŝ
Compute the UnscentedKalmanFilter
constants from $α, β$ and $κ$.
With $n_\mathbf{x̂}$ elements in the state vector $\mathbf{x̂}$ and $n_σ = 2 n_\mathbf{x̂} + 1$ sigma points, the scaling factor applied on standard deviation matrices $\sqrt{\mathbf{P̂}}$ is:
\[ γ = α \sqrt{ n_\mathbf{x̂} + κ }\]
The weight vector $(n_σ × 1)$ for the mean and the weight matrix $(n_σ × n_σ)$ for the covariance are respectively:
\[\begin{aligned} \mathbf{m̂} &= \begin{bmatrix} 1 - \tfrac{n_\mathbf{x̂}}{γ^2} & \tfrac{1}{2γ^2} & \tfrac{1}{2γ^2} & \cdots & \tfrac{1}{2γ^2} \end{bmatrix}' \\ \mathbf{Ŝ} &= \mathrm{diag}\big( 2 - α^2 + β - \tfrac{n_\mathbf{x̂}}{γ^2} \:,\; \tfrac{1}{2γ^2} \:,\; \tfrac{1}{2γ^2} \:,\; \cdots \:,\; \tfrac{1}{2γ^2} \big) \end{aligned}\]
See update_estimate!(::UnscentedKalmanFilter)
for other details.
ModelPredictiveControl.init_internalmodel
— Functioninit_internalmodel(As, Bs, Cs, Ds) -> Âs, B̂s
Calc stochastic model update matrices Âs
and B̂s
for InternalModel
estimator.
As
, Bs
, Cs
and Ds
are the stochastic model matrices :
\[\begin{aligned} \mathbf{x_s}(k+1) &= \mathbf{A_s x_s}(k) + \mathbf{B_s e}(k) \\ \mathbf{y_s}(k) &= \mathbf{C_s x_s}(k) + \mathbf{D_s e}(k) \end{aligned}\]
where $\mathbf{e}(k)$ is conceptual and unknown zero mean white noise. Its optimal estimation is $\mathbf{ê=0}$, the expected value. Thus, the Âs
and B̂s
matrices that optimally update the stochastic estimate $\mathbf{x̂_s}$ are:
\[\begin{aligned} \mathbf{x̂_s}(k+1) &= \mathbf{(A_s - B_s D_s^{-1} C_s) x̂_s}(k) + \mathbf{(B_s D_s^{-1}) ŷ_s}(k) \\ &= \mathbf{Â_s x̂_s}(k) + \mathbf{B̂_s ŷ_s}(k) \end{aligned}\]
with current stochastic outputs estimation $\mathbf{ŷ_s}(k)$, composed of the measured $\mathbf{ŷ_s^m}(k) = \mathbf{y^m}(k) - \mathbf{ŷ_d^m}(k)$ and unmeasured $\mathbf{ŷ_s^u = 0}$ outputs. See [1].
ModelPredictiveControl.init_predmat_mhe
— Functioninit_predmat_mhe(
model::LinModel, He, i_ym, Â, B̂u, Ĉ, B̂d, D̂d
) -> E, F, G, J, ex̄, fx̄, Ex̂, Fx̂, Gx̂, Jx̂
Construct the MHE prediction matrices for LinModel
model
.
Introducing the vector $\mathbf{Z} = [\begin{smallmatrix} \mathbf{x̂}_k(k-N_k+1) \\ \mathbf{Ŵ} \end{smallmatrix}]$ with the decision variables, the estimated sensor noises from time $k-N_k+1$ to $k$ are computed by:
\[\begin{aligned} \mathbf{V̂} = \mathbf{Y^m - Ŷ^m} &= \mathbf{E Z + G U + J D + Y^m} \\ &= \mathbf{E Z + F} \end{aligned}\]
in which $\mathbf{U, D}$ and $\mathbf{Y^m}$ contains respectively the manipulated inputs, measured disturbances and measured outputs from time $k-N_k+1$ to $k$. The method also returns similar matrices but for the estimation error at arrival:
\[\mathbf{x̄} = \mathbf{x̂}_{k-N_k}(k-N_k+1) - \mathbf{x̂}_{k}(k-N_k+1) = \mathbf{e_x̄ Z + f_x̄}\]
Lastly, the estimated states from time $k-N_k+2$ to $k+1$ are given by the equation:
\[\begin{aligned} \mathbf{X̂} &= \mathbf{E_x̂ Z + G_x̂ U + J_x̂ D} \\ &= \mathbf{E_x̂ Z + F_x̂} \end{aligned}\]
All these equations omit the operating points $\mathbf{u_{op}, y_{op}, d_{op}}$.
Extended Help
Extended Help
Using the augmented matrices $\mathbf{Â, B̂_u, Ĉ, B̂_d, D̂_d}$, the prediction matrices for the sensor noises are computed by (notice the minus signs after the equalities):
\[\begin{aligned} \mathbf{E} &= - \begin{bmatrix} \mathbf{Ĉ^m}\mathbf{Â}^{0} & \mathbf{0} & \cdots & \mathbf{0} \\ \mathbf{Ĉ^m}\mathbf{Â}^{1} & \mathbf{Ĉ^m}\mathbf{Â}^{0} & \cdots & \mathbf{0} \\ \vdots & \vdots & \ddots & \vdots \\ \mathbf{Ĉ^m}\mathbf{Â}^{H_e-1} & \mathbf{Ĉ^m}\mathbf{Â}^{H_e-2} & \cdots & \mathbf{0} \end{bmatrix} \\ \mathbf{G} &= - \begin{bmatrix} \mathbf{0} & \mathbf{0} & \cdots & \mathbf{0} \\ \mathbf{Ĉ^m}\mathbf{Â}^{0}\mathbf{B̂_u} & \mathbf{0} & \cdots & \mathbf{0} \\ \vdots & \vdots & \ddots & \vdots \\ \mathbf{Ĉ^m}\mathbf{Â}^{H_e-2}\mathbf{B̂_u} & \mathbf{Ĉ^m}\mathbf{Â}^{H_e-3}\mathbf{B̂_u} & \cdots & \mathbf{0} \end{bmatrix} \\ \mathbf{J} &= - \begin{bmatrix} \mathbf{D̂^m} & \mathbf{0} & \cdots & \mathbf{0} \\ \mathbf{Ĉ^m}\mathbf{Â}^{0}\mathbf{B̂_d} & \mathbf{D̂^m} & \cdots & \mathbf{0} \\ \vdots & \vdots & \ddots & \vdots \\ \mathbf{Ĉ^m}\mathbf{Â}^{H_e-2}\mathbf{B̂_d} & \mathbf{Ĉ^m}\mathbf{Â}^{H_e-3}\mathbf{B̂_d} & \cdots & \mathbf{D̂^m} \end{bmatrix} \end{aligned}\]
for the estimation error at arrival:
\[\mathbf{e_x̄} = \begin{bmatrix} -\mathbf{I} & \mathbf{0} & \cdots & \mathbf{0} \end{bmatrix}\]
and, for the estimated states:
\[\begin{aligned} \mathbf{E_x̂} &= \begin{bmatrix} \mathbf{Â}^{1} & \mathbf{I} & \cdots & \mathbf{0} \\ \mathbf{Â}^{2} & \mathbf{Â}^{1} & \cdots & \mathbf{0} \\ \vdots & \vdots & \ddots & \vdots \\ \mathbf{Â}^{H_e} & \mathbf{Â}^{H_e-1} & \cdots & \mathbf{Â}^{1} \end{bmatrix} \\ \mathbf{G_x̂} &= \begin{bmatrix} \mathbf{Â}^{0}\mathbf{B̂_u} & \mathbf{0} & \cdots & \mathbf{0} \\ \mathbf{Â}^{1}\mathbf{B̂_u} & \mathbf{Â}^{0}\mathbf{B̂_u} & \cdots & \mathbf{0} \\ \vdots & \vdots & \ddots & \vdots \\ \mathbf{Â}^{H_e-1}\mathbf{B̂_u} & \mathbf{Â}^{H_e-2}\mathbf{B̂_u} & \cdots & \mathbf{Â}^{0}\mathbf{B̂_u} \end{bmatrix} \\ \mathbf{J_x̂} &= \begin{bmatrix} \mathbf{Â}^{0}\mathbf{B̂_d} & \mathbf{0} & \cdots & \mathbf{0} \\ \mathbf{Â}^{1}\mathbf{B̂_d} & \mathbf{Â}^{0}\mathbf{B̂_d} & \cdots & \mathbf{0} \\ \vdots & \vdots & \ddots & \vdots \\ \mathbf{Â}^{H_e-1}\mathbf{B̂_d} & \mathbf{Â}^{H_e-2}\mathbf{B̂_d} & \cdots & \mathbf{Â}^{0}\mathbf{B̂_d} \end{bmatrix} \end{aligned}\]
All these matrices are truncated when $N_k < H_e$ (at the beginning).
Return empty matrices if model
is not a LinModel
, except for ex̄
.
ModelPredictiveControl.relaxarrival
— Functionrelaxarrival(
model::SimModel, C, c_x̂min, c_x̂max, x̂min, x̂max, ex̄
) -> A_x̃min, A_x̃max, x̃min, x̃max, ẽx̄
Augment arrival state constraints with slack variable ϵ for softening the MHE.
Denoting the MHE decision variable augmented with the slack variable $\mathbf{Z̃} = [\begin{smallmatrix} ϵ \\ \mathbf{Z} \end{smallmatrix}]$, it returns the $\mathbf{ẽ_x̄}$ matrix that appears in the estimation error at arrival equation $\mathbf{x̄} = \mathbf{ẽ_x̄ Z̃ + f_x̄}$. It also returns the augmented constraints $\mathbf{x̃_{min}}$ and $\mathbf{x̃_{max}}$, and the $\mathbf{A}$ matrices for the inequality constraints:
\[\begin{bmatrix} \mathbf{A_{x̃_{min}}} \\ \mathbf{A_{x̃_{max}}} \end{bmatrix} \mathbf{Z̃} ≤ \begin{bmatrix} - \mathbf{x̃_{min} + f_x̄} \\ + \mathbf{x̃_{max} - f_x̄} \end{bmatrix}\]
ModelPredictiveControl.relaxX̂
— FunctionrelaxX̂(model::SimModel, C, C_x̂min, C_x̂max, Ex̂) -> A_X̂min, A_X̂max, Ẽx̂
Augment estimated state constraints with slack variable ϵ for softening the MHE.
Denoting the MHE decision variable augmented with the slack variable $\mathbf{Z̃} = [\begin{smallmatrix} ϵ \\ \mathbf{Z} \end{smallmatrix}]$, it returns the $\mathbf{Ẽ_x̂}$ matrix that appears in estimated states equation $\mathbf{X̂} = \mathbf{Ẽ_x̂ Z̃ + F_x̂}$. It also returns the $\mathbf{A}$ matrices for the inequality constraints:
\[\begin{bmatrix} \mathbf{A_{X̂_{min}}} \\ \mathbf{A_{X̂_{max}}} \end{bmatrix} \mathbf{Z̃} ≤ \begin{bmatrix} - \mathbf{X̂_{min} + F_x̂} \\ + \mathbf{X̂_{max} - F_x̂} \end{bmatrix}\]
Return empty matrices if model is not a LinModel
ModelPredictiveControl.relaxŴ
— FunctionrelaxŴ(model::SimModel, C, C_ŵmin, C_ŵmax, nx̂) -> A_Ŵmin, A_Ŵmax
Augment estimated process noise constraints with slack variable ϵ for softening the MHE.
Denoting the MHE decision variable augmented with the slack variable $\mathbf{Z̃} = [\begin{smallmatrix} ϵ \\ \mathbf{Z} \end{smallmatrix}]$, it returns the $\mathbf{A}$ matrices for the inequality constraints:
\[\begin{bmatrix} \mathbf{A_{Ŵ_{min}}} \\ \mathbf{A_{Ŵ_{max}}} \end{bmatrix} \mathbf{Z̃} ≤ \begin{bmatrix} - \mathbf{Ŵ_{min}} \\ + \mathbf{Ŵ_{max}} \end{bmatrix}\]
ModelPredictiveControl.relaxV̂
— FunctionrelaxV̂(model::SimModel, C, C_v̂min, C_v̂max, E) -> A_V̂min, A_V̂max, Ẽ
Augment estimated sensor noise constraints with slack variable ϵ for softening the MHE.
Denoting the MHE decision variable augmented with the slack variable $\mathbf{Z̃} = [\begin{smallmatrix} ϵ \\ \mathbf{Z} \end{smallmatrix}]$, it returns the $\mathbf{Ẽ}$ matrix that appears in estimated sensor noise equation $\mathbf{V̂} = \mathbf{Ẽ Z̃ + F}$. It also returns the $\mathbf{A}$ matrices for the inequality constraints:
\[\begin{bmatrix} \mathbf{A_{V̂_{min}}} \\ \mathbf{A_{V̂_{max}}} \end{bmatrix} \mathbf{Z̃} ≤ \begin{bmatrix} - \mathbf{V̂_{min} + F} \\ + \mathbf{V̂_{max} - F} \end{bmatrix}\]
Return empty matrices if model is not a LinModel
ModelPredictiveControl.init_matconstraint_mhe
— Functioninit_matconstraint_mhe(model::LinModel,
i_x̃min, i_x̃max, i_X̂min, i_X̂max, i_Ŵmin, i_Ŵmax, i_V̂min, i_V̂max, args...
) -> i_b, i_g, A
Init i_b
, i_g
and A
matrices for the MHE linear inequality constraints.
The linear and nonlinear inequality constraints are respectively defined as:
\[\begin{aligned} \mathbf{A Z̃ } &≤ \mathbf{b} \\ \mathbf{g(Z̃)} &≤ \mathbf{0} \end{aligned}\]
i_b
is a BitVector
including the indices of $\mathbf{b}$ that are finite numbers. i_g
is a similar vector but for the indices of $\mathbf{g}$ (empty if model
is a LinModel
). The method also returns the $\mathbf{A}$ matrix if args
is provided. In such a case, args
needs to contain all the inequality constraint matrices: A_x̃min, A_x̃max, A_X̂min, A_X̂max, A_Ŵmin, A_Ŵmax, A_V̂min, A_V̂max
.
Init i_b, A
without state and sensor noise constraints if model
is not a LinModel
.
Update Quadratic Optimization
ModelPredictiveControl.initpred!
— Methodinitpred!(estim::MovingHorizonEstimator, model::LinModel) -> nothing
Init quadratic optimization matrices F, fx̄, H̃, q̃, p
for MovingHorizonEstimator
.
See init_predmat_mhe
for the definition of the vectors $\mathbf{F, f_x̄}$. It also inits estim.optim
objective function, expressed as the quadratic general form:
\[ J = \min_{\mathbf{Z̃}} \frac{1}{2}\mathbf{Z̃' H̃ Z̃} + \mathbf{q̃' Z̃} + p \]
in which $\mathbf{Z̃} = [\begin{smallmatrix} ϵ \\ \mathbf{Z} \end{smallmatrix}]$. Note that $p$ is useless at optimization but required to evaluate the objective minima $J$. The Hessian $\mathbf{H̃}$ matrix of the quadratic general form is not constant here because of the time-varying $\mathbf{P̄}$ covariance . The computed variables are:
\[\begin{aligned} \mathbf{F} &= \mathbf{G U} + \mathbf{J D} + \mathbf{Y^m} \\ \mathbf{f_x̄} &= \mathbf{x̂}_{k-N_k}(k-N_k+1) \\ \mathbf{F_Z̃} &= [\begin{smallmatrix}\mathbf{f_x̄} \\ \mathbf{F} \end{smallmatrix}] \\ \mathbf{Ẽ_Z̃} &= [\begin{smallmatrix}\mathbf{ẽ_x̄} \\ \mathbf{Ẽ} \end{smallmatrix}] \\ \mathbf{M}_{N_k} &= \mathrm{diag}(\mathbf{P̄}^{-1}, \mathbf{R̂}_{N_k}^{-1}) \\ \mathbf{Ñ}_{N_k} &= \mathrm{diag}(C, \mathbf{0}, \mathbf{Q̂}_{N_k}^{-1}) \\ \mathbf{H̃} &= 2(\mathbf{Ẽ_Z̃}' \mathbf{M}_{N_k} \mathbf{Ẽ_Z̃} + \mathbf{Ñ}_{N_k}) \\ \mathbf{q̃} &= 2(\mathbf{M}_{N_k} \mathbf{Ẽ_Z̃})' \mathbf{F_Z̃} \\ p &= \mathbf{F_Z̃}' \mathbf{M}_{N_k} \mathbf{F_Z̃} \end{aligned}\]
ModelPredictiveControl.linconstraint!
— Methodlinconstraint!(estim::MovingHorizonEstimator, model::LinModel)
Set b
vector for the linear model inequality constraints ($\mathbf{A Z̃ ≤ b}$) of MHE.
Also init $\mathbf{F_x̂ = G_x̂ U + J_x̂ D}$ vector for the state constraints, see init_predmat_mhe
.
Evaluate Estimated Output
ModelPredictiveControl.evalŷ
— Functionevalŷ(estim::InternalModel, ym, d) -> ŷ
Get InternalModel
output ŷ
from current measured outputs ym
and dist. d
.
InternalModel
estimator needs current measured outputs $\mathbf{y^m}(k)$ to estimate its outputs $\mathbf{ŷ}(k)$, since the strategy imposes that $\mathbf{ŷ^m}(k) = \mathbf{y^m}(k)$ is always true.
evalŷ(estim::StateEstimator, _ , d) -> ŷ
Evaluate StateEstimator
output ŷ
from measured disturbance d
and estim.x̂0
.
Second argument is ignored, except for InternalModel
.
Remove Operating Points
ModelPredictiveControl.remove_op!
— Functionremove_op!(estim::StateEstimator, u, ym, d) -> u0, ym0, d0
Remove operating points on inputs u
, measured outputs ym
and disturbances d
.
Also store current inputs without operating points u0
in estim.lastu0
. This field is used for PredictiveController
computations.
Init Estimate
ModelPredictiveControl.init_estimate!
— Functioninit_estimate!(estim::StateEstimator, model::LinModel, u0, ym0, d0)
Init estim.x̂0
estimate with the steady-state solution if model
is a LinModel
.
Using u0
, ym0
and d0
arguments (deviation values, see setop!
), the steadystate problem combined to the equality constraint $\mathbf{ŷ_0^m} = \mathbf{y_0^m}$ engenders the following system to solve:
\[\begin{bmatrix} \mathbf{I} - \mathbf{Â} \\ \mathbf{Ĉ^m} \end{bmatrix} \mathbf{x̂_0} = \begin{bmatrix} \mathbf{B̂_u u_0 + B̂_d d_0 + f̂_{op} - x̂_{op}} \\ \mathbf{y_0^m - D̂_d^m d_0} \end{bmatrix}\]
in which $\mathbf{Ĉ^m, D̂_d^m}$ are the rows of estim.Ĉ, estim.D̂d
that correspond to measured outputs $\mathbf{y^m}$.
init_estimate!(estim::StateEstimator, model::SimModel, _ , _ , _ )
Left estim.x̂0
estimate unchanged if model
is not a LinModel
.
init_estimate!(estim::InternalModel, model::LinModel, u0, ym0, d0)
Init estim.x̂0
/x̂d
/x̂s
estimate at steady-state for InternalModel
.
The deterministic estimates estim.x̂d
start at steady-state using u0
and d0
arguments:
\[ \mathbf{x̂_d} = \mathbf{(I - A)^{-1} (B_u u_0 + B_d d_0 + f_{op} - x_{op})}\]
Based on ym0
argument and current stochastic outputs estimation $\mathbf{ŷ_s}$, composed of the measured $\mathbf{ŷ_s^m} = \mathbf{y_0^m} - \mathbf{ŷ_{d0}^m}$ and unmeasured $\mathbf{ŷ_s^u = 0}$ outputs, the stochastic estimates also start at steady-state:
\[ \mathbf{x̂_s} = \mathbf{(I - Â_s)^{-1} B̂_s ŷ_s}\]
This estimator does not augment the state vector, thus $\mathbf{x̂ = x̂_d}$. See init_internalmodel
for details.
Update Estimate
All these methods assume that the operating points are already removed in u
, ym
and d
arguments. Strickly speaking, the arguments should be called u0
, ym0
and d0
, following setop!
notation. The 0
is dropped to simplify the notation.
ModelPredictiveControl.update_estimate!
— Functionupdate_estimate!(estim::SteadyKalmanFilter, u, ym, d=[])
Update estim.x̂0
estimate with current inputs u
, measured outputs ym
and dist. d
.
The SteadyKalmanFilter
updates it with the precomputed Kalman gain $\mathbf{K̂}$:
\[\mathbf{x̂}_{k}(k+1) = \mathbf{Â x̂}_{k-1}(k) + \mathbf{B̂_u u}(k) + \mathbf{B̂_d d}(k) + \mathbf{K̂}[\mathbf{y^m}(k) - \mathbf{Ĉ^m x̂}_{k-1}(k) - \mathbf{D̂_d^m d}(k)]\]
update_estimate!(estim::KalmanFilter, u, ym, d=[])
Update KalmanFilter
state estim.x̂0
and estimation error covariance estim.P̂
.
It implements the time-varying Kalman Filter in its predictor (observer) form :
\[\begin{aligned} \mathbf{M̂}(k) &= \mathbf{P̂}_{k-1}(k)\mathbf{Ĉ^m}' [\mathbf{Ĉ^m P̂}_{k-1}(k)\mathbf{Ĉ^m}' + \mathbf{R̂}]^{-1} \\ \mathbf{K̂}(k) &= \mathbf{Â M̂}(k) \\ \mathbf{ŷ^m}(k) &= \mathbf{Ĉ^m x̂}_{k-1}(k) + \mathbf{D̂_d^m d}(k) \\ \mathbf{x̂}_{k}(k+1) &= \mathbf{Â x̂}_{k-1}(k) + \mathbf{B̂_u u}(k) + \mathbf{B̂_d d}(k) + \mathbf{K̂}(k)[\mathbf{y^m}(k) - \mathbf{ŷ^m}(k)] \\ \mathbf{P̂}_{k}(k+1) &= \mathbf{Â}[\mathbf{P̂}_{k-1}(k) - \mathbf{M̂}(k)\mathbf{Ĉ^m P̂}_{k-1}(k)]\mathbf{Â}' + \mathbf{Q̂} \end{aligned}\]
based on the process model described in SteadyKalmanFilter
. The notation $\mathbf{x̂}_{k-1}(k)$ refers to the state for the current time $k$ estimated at the last control period $k-1$. See [2] for details.
update_estimate!(estim::UnscentedKalmanFilter, u, ym, d=[])
Update UnscentedKalmanFilter
state estim.x̂0
and covariance estimate estim.P̂
.
It implements the unscented Kalman Filter in its predictor (observer) form, based on the generalized unscented transform[3]. See init_ukf
for the definition of the constants $\mathbf{m̂, Ŝ}$ and $γ$.
Denoting $\mathbf{x̂}_{k-1}(k)$ as the state for the current time $k$ estimated at the last period $k-1$, $\mathbf{0}$, a null vector, $n_σ = 2 n_\mathbf{x̂} + 1$, the number of sigma points, and $\mathbf{X̂}_{k-1}^j(k)$, the vector at the $j$th column of $\mathbf{X̂}_{k-1}(k)$, the estimator updates the states with:
\[\begin{aligned} \mathbf{X̂}_{k-1}(k) &= \bigg[\begin{matrix} \mathbf{x̂}_{k-1}(k) & \mathbf{x̂}_{k-1}(k) & \cdots & \mathbf{x̂}_{k-1}(k) \end{matrix}\bigg] + \bigg[\begin{matrix} \mathbf{0} & γ \sqrt{\mathbf{P̂}_{k-1}(k)} & -γ \sqrt{\mathbf{P̂}_{k-1}(k)} \end{matrix}\bigg] \\ \mathbf{Ŷ^m}(k) &= \bigg[\begin{matrix} \mathbf{ĥ^m}\Big( \mathbf{X̂}_{k-1}^{1}(k) \Big) & \mathbf{ĥ^m}\Big( \mathbf{X̂}_{k-1}^{2}(k) \Big) & \cdots & \mathbf{ĥ^m}\Big( \mathbf{X̂}_{k-1}^{n_σ}(k) \Big) \end{matrix}\bigg] \\ \mathbf{ŷ^m}(k) &= \mathbf{Ŷ^m}(k) \mathbf{m̂} \\ \mathbf{X̄}_{k-1}(k) &= \begin{bmatrix} \mathbf{X̂}_{k-1}^{1}(k) - \mathbf{x̂}_{k-1}(k) & \mathbf{X̂}_{k-1}^{2}(k) - \mathbf{x̂}_{k-1}(k) & \cdots & \mathbf{X̂}_{k-1}^{n_σ}(k) - \mathbf{x̂}_{k-1}(k) \end{bmatrix} \\ \mathbf{Ȳ^m}(k) &= \begin{bmatrix} \mathbf{Ŷ^m}^{1}(k) - \mathbf{ŷ^m}(k) & \mathbf{Ŷ^m}^{2}(k) - \mathbf{ŷ^m}(k) & \cdots & \mathbf{Ŷ^m}^{n_σ}(k) - \mathbf{ŷ^m}(k) \end{bmatrix} \\ \mathbf{M̂}(k) &= \mathbf{Ȳ^m}(k) \mathbf{Ŝ} \mathbf{Ȳ^m}'(k) + \mathbf{R̂} \\ \mathbf{K̂}(k) &= \mathbf{X̄}_{k-1}(k) \mathbf{Ŝ} \mathbf{Ȳ^m}'(k) \mathbf{M̂^{-1}}(k) \\ \mathbf{x̂}_k(k) &= \mathbf{x̂}_{k-1}(k) + \mathbf{K̂}(k) \big[ \mathbf{y^m}(k) - \mathbf{ŷ^m}(k) \big] \\ \mathbf{P̂}_k(k) &= \mathbf{P̂}_{k-1}(k) - \mathbf{K̂}(k) \mathbf{M̂}(k) \mathbf{K̂}'(k) \\ \mathbf{X̂}_k(k) &= \bigg[\begin{matrix} \mathbf{x̂}_{k}(k) & \mathbf{x̂}_{k}(k) & \cdots & \mathbf{x̂}_{k}(k) \end{matrix}\bigg] + \bigg[\begin{matrix} \mathbf{0} & \gamma \sqrt{\mathbf{P̂}_{k}(k)} & - \gamma \sqrt{\mathbf{P̂}_{k}(k)} \end{matrix}\bigg] \\ \mathbf{X̂}_{k}(k+1) &= \bigg[\begin{matrix} \mathbf{f̂}\Big( \mathbf{X̂}_{k}^{1}(k), \mathbf{u}(k), \mathbf{d}(k) \Big) & \mathbf{f̂}\Big( \mathbf{X̂}_{k}^{2}(k), \mathbf{u}(k), \mathbf{d}(k) \Big) & \cdots & \mathbf{f̂}\Big( \mathbf{X̂}_{k}^{n_σ}(k), \mathbf{u}(k), \mathbf{d}(k) \Big) \end{matrix}\bigg] \\ \mathbf{x̂}_{k}(k+1) &= \mathbf{X̂}_{k}(k+1)\mathbf{m̂} \\ \mathbf{X̄}_k(k+1) &= \begin{bmatrix} \mathbf{X̂}_{k}^{1}(k+1) - \mathbf{x̂}_{k}(k+1) & \mathbf{X̂}_{k}^{2}(k+1) - \mathbf{x̂}_{k}(k+1) & \cdots &\, \mathbf{X̂}_{k}^{n_σ}(k+1) - \mathbf{x̂}_{k}(k+1) \end{bmatrix} \\ \mathbf{P̂}_k(k+1) &= \mathbf{X̄}_k(k+1) \mathbf{Ŝ} \mathbf{X̄}_k'(k+1) + \mathbf{Q̂} \end{aligned} \]
by using the lower triangular factor of cholesky
to compute $\sqrt{\mathbf{P̂}_{k-1}(k)}$ and $\sqrt{\mathbf{P̂}_{k}(k)}$. The matrices $\mathbf{P̂, Q̂, R̂}$ are the covariance of the estimation error, process noise and sensor noise, respectively.
update_estimate!(estim::ExtendedKalmanFilter, u, ym, d=[])
Update ExtendedKalmanFilter
state estim.x̂0
and covariance estim.P̂
.
The equations are similar to update_estimate!(::KalmanFilter)
but with the substitutions $\mathbf{Â = F̂}(k)$ and $\mathbf{Ĉ^m = Ĥ^m}(k)$:
\[\begin{aligned} \mathbf{M̂}(k) &= \mathbf{P̂}_{k-1}(k)\mathbf{Ĥ^m}'(k) [\mathbf{Ĥ^m}(k)\mathbf{P̂}_{k-1}(k)\mathbf{Ĥ^m}'(k) + \mathbf{R̂}]^{-1} \\ \mathbf{K̂}(k) &= \mathbf{F̂}(k) \mathbf{M̂}(k) \\ \mathbf{ŷ^m}(k) &= \mathbf{ĥ^m}\Big( \mathbf{x̂}_{k-1}(k), \mathbf{d}(k) \Big) \\ \mathbf{x̂}_{k}(k+1) &= \mathbf{f̂}\Big( \mathbf{x̂}_{k-1}(k), \mathbf{u}(k), \mathbf{d}(k) \Big) + \mathbf{K̂}(k)[\mathbf{y^m}(k) - \mathbf{ŷ^m}(k)] \\ \mathbf{P̂}_{k}(k+1) &= \mathbf{F̂}(k)[\mathbf{P̂}_{k-1}(k) - \mathbf{M̂}(k)\mathbf{Ĥ^m}(k)\mathbf{P̂}_{k-1}(k)]\mathbf{F̂}'(k) + \mathbf{Q̂} \end{aligned}\]
ForwardDiff.jacobian
automatically computes the Jacobians:
\[\begin{aligned} \mathbf{F̂}(k) &= \left. \frac{∂\mathbf{f̂}(\mathbf{x̂}, \mathbf{u}, \mathbf{d})}{∂\mathbf{x̂}} \right|_{\mathbf{x̂ = x̂}_{k-1}(k),\, \mathbf{u = u}(k),\, \mathbf{d = d}(k)} \\ \mathbf{Ĥ}(k) &= \left. \frac{∂\mathbf{ĥ}(\mathbf{x̂}, \mathbf{d})}{∂\mathbf{x̂}} \right|_{\mathbf{x̂ = x̂}_{k-1}(k),\, \mathbf{d = d}(k)} \end{aligned}\]
The matrix $\mathbf{Ĥ^m}$ is the rows of $\mathbf{Ĥ}$ that are measured outputs.
update_estimate!(estim::Luenberger, u, ym, d=[])
Same than update_estimate!(::SteadyKalmanFilter)
but using Luenberger
.
update_estimate!(estim::MovingHorizonEstimator, u, ym, d=[])
Update MovingHorizonEstimator
state estim.x̂
.
The optimization problem of MovingHorizonEstimator
documentation is solved at each discrete time $k$. Once solved, the next estimate $\mathbf{x̂}_k(k+1)$ is computed by inserting the optimal values of $\mathbf{x̂}_k(k-N_k+1)$ and $\mathbf{Ŵ}$ in the augmented model from $j = N_k-1$ to $0$ inclusively. Afterward, if $k ≥ H_e$, the arrival covariance for the next time step $\mathbf{P̂}_{k-N_k+1}(k-N_k+2)$ is estimated with the equations of update_estimate!(::ExtendedKalmanFilter)
, or KalmanFilter
, for LinModel
.
update_estimate!(estim::InternalModel, u, ym, d=[])
Update estim.x̂0
/x̂d
/x̂s
with current inputs u
, measured outputs ym
and dist. d
.
The InternalModel
updates the deterministic x̂d
and stochastic x̂s
estimates with:
\[\begin{aligned} \mathbf{x̂_d}(k+1) &= \mathbf{f}\Big( \mathbf{x̂_d}(k), \mathbf{u}(k), \mathbf{d}(k) \Big) \\ \mathbf{x̂_s}(k+1) &= \mathbf{Â_s x̂_s}(k) + \mathbf{B̂_s ŷ_s}(k) \end{aligned}\]
This estimator does not augment the state vector, thus $\mathbf{x̂ = x̂_d}$. See init_internalmodel
for details.
- 1Desbiens, A., D. Hodouin & É. Plamondon. 2000, "Global predictive control : a unified control structure for decoupling setpoint tracking, feedforward compensation and disturbance rejection dynamics", IEE Proceedings - Control Theory and Applications, vol. 147, no 4, https://doi.org/10.1049/ip-cta:20000443, p. 465–475, ISSN 1350-2379.
- 2Boyd S., "Lecture 8 : The Kalman Filter" (Winter 2008-09) [course slides], EE363: Linear Dynamical Systems, https://web.stanford.edu/class/ee363/lectures/kf.pdf.
- 3Simon, D. 2006, "Chapter 14: The unscented Kalman filter" in "Optimal State Estimation: Kalman, H∞, and Nonlinear Approaches", John Wiley & Sons, p. 433–459, https://doi.org/10.1002/0470045345.ch14, ISBN9780470045343.