v0.21.0

ModelPredictiveControl v0.21.0

Diff since v0.20.2

BREAKING CHANGE
All the keyword arguments related to initial values e.g. σP0, x0 and x̂0 now require an underscore e.g. σP_0, x_0, x̂_0 (to differentiate from operating point deviation vectors)

• Added: setmodel! for runtime model adaptation of controller/estimator based on LinModel
• Added: linearize and setop! now support non-equilibrium points
• Added: successive linearization MPC with the new setmodel! and linearize functions
• Added: successive linearization MHE with the new setmodel! and linearize functions
• Added: linearize! method for in-place model linearization (to reduce allocations)
• Added: 6 args. LinModel constructor now support scalars (similarly to ss function)
• Added: ExtendedKalmanFilter now compute the Jacobians in-place (to reduce allocations)
• Changed: struct state data x and state estimate x̂ renamed to x0 and x̂0
• Debug: ExplicitMPC with non-Float64 now works
• Debug: accept integers in linearize arguments
• Debug: call empty! on JuMP.Model to support re-construction of MPC instances
• Doc: new setmodel!, setop! and linearize function documentation
• Doc: example of model adaptation with successive linearization on the pendulum (very efficient!)

Merged pull requests:

• Added: adaptatation of LinMPC model through setmodel! (#52) (@franckgaga)
• Added: adaptation of controller/estimator based on LinModel (#54) (@franckgaga)
• Added: setmodel! for MovingHorizonEstimator (#55) (@franckgaga)

v0.20.2

ModelPredictiveControl v0.20.2

Diff since v0.20.1

• added: print info on controller and estimator constraint softening (slack var. ϵ)
• reduce allocations LinMPC, NonLinMPC, MovingHorizonEstimator
• cleanup namespace with imports instead of usings

Merged pull requests:

• Changed: avoid collect in NonLinMPC and MovingHorizonEstimator objective and constraints (#46) (@franckgaga)
• changed: error handling one less alloc (#47) (@franckgaga)
• added: print info on controller and estimator constraints softening (#49) (@franckgaga)
• cleaning the namespace with imports instead of usings (#50) (@franckgaga)
• Namespace, NonLinMPC and MovingHorizonEsitimator cleanup (#51) (@franckgaga)

v0.20.1

ModelPredictiveControl v0.20.1

Diff since v0.20.0

• Reduce allocation for estimator based on NonLinModel
• Reduce allocations for LinMPC and KalmanFilter
• Improve performance of LinMPC and MovingHorizonEstimator with new JuMP batch update methods

Merged pull requests:

• doc: correct errors in EKF equations (#43) (@franckgaga)
• Reduce allocation for estimators based on augmented NonLinModel (#44) (@franckgaga)
• Improve performance and reduce allocations of LinMPC and KalmanFilter (#45) (@franckgaga)

v0.20.0

ModelPredictiveControl v0.20.0

Diff since v0.19.2

• Added: custom estimator for the approximation of the arrival covariance in the MHE
• Changed: MHE based on NonLinModel now defaults to UnscentedKalmanFilter for arrival covariance estimation
• Added: tests with custom covariance estimator in the MHE

Merged pull requests:

• Added: MHE supports custom estimator for the arrival covariance (#41) (@franckgaga)
• added: tests for MHE custom covariance estimator (#42) (@franckgaga)

v0.19.2

ModelPredictiveControl v0.19.2

Diff since v0.19.1

Merged pull requests:

• Debug: MHE with NonLinModel update covariance with the correct state estimate (#40) (@franckgaga)

v0.19.1

ModelPredictiveControl v0.19.1

Diff since v0.19.0

Merged pull requests:

• Minor doc correction and code cleanup (#39) (@franckgaga)

v0.19.0

ModelPredictiveControl v0.19.0

Diff since v0.18.1

BREAKING CHANGE
NonLinModel constructor now supposes continuous dynamics by default. Use solver=nothing for discrete-time models.

changelog:

• added: support for NonLinModel with continuous dynamics
• added: 4th order Runge-Kutta solver with 0 allocation
• doc: the pendulum example now use the built-in Runge-Kutta solve
• added some tests with continuous NonLinModel

Merged pull requests:

• starting support of continuous NonLinModel (RK4 only for now) (#38) (@baggepinnen)

v0.18.1

ModelPredictiveControl v0.18.1

Diff since v0.18.0

• remove useless allocations for mutating NonLinModels

Merged pull requests:

• minor doc correction (#36) (@franckgaga)
• Remove useless allocations for in-place NonLinModel (#37) (@franckgaga)

v0.18.0

ModelPredictiveControl v0.18.0

Diff since v0.17.0

• added: support for mutating f! and h! functions in NonLinModel to reduce allocations
• doc: added simple example of the MovingHorizonEstimator for the CSTR in the manual

Merged pull requests:

• Update README.md (#27) (@franckgaga)
• Added a MHE example on the CSTR in the manual (#28) (@franckgaga)
• fix wrong function signature ared (#29) (@baggepinnen)
• dummy commit to trigger doc deploy (#30) (@franckgaga)
• doc: minor correction MHE manual (#32) (@franckgaga)
• Added: mutating NonLinModel (#33) (@franckgaga)
• Re-enable precompilation (#34) (@franckgaga)
• bump version (#35) (@franckgaga)

Closed issues:

• Docs questions (#22)
• Documentation no longer deployed (#26)

v0.17.0

ModelPredictiveControl v0.17.0

Diff since v0.16.1

• added: support for non-diagonal symmetric M_Hp, N_Hc and L_Hp weights, allowing terminal costs and LQR tuning
• doc: added comment on a terminal weight
• reduce allocations for PredictiveControllers and MovingHorizonEstimator (with in-place operations)

Merged pull requests:

• Update README links (#21) (@baggepinnen)
• todo: add terminal costs (#23) (@franckgaga)
• Update README.md (#24) (@franckgaga)
• Added: terminal cost support (#25) (@franckgaga)