v1.8.0

ModelPredictiveControl v1.8.0

Diff since v1.7.0

• added: move blocking feature in LinMPC, ExplicitMPC and NonLinMPC (see Hc argument)

Merged pull requests:

• added: move blocking feature in LinMPC, ExplicitMPC and NonLinMPC (#201) (@franckgaga)

Closed issues:

• Implement manipulated input blocking (#177)

v1.7.0

ModelPredictiveControl v1.7.0

Diff since v1.6.2

• added: ManualEstimator to turn off built-in state estimation and provide your own estimate

Merged pull requests:

• added: ManualEstimator to turn off built-in state estimation (#206) (@franckgaga)
• changed: moved lastu0 inside PredictiveController objects (#208) (@franckgaga)
• Added tests and improve doc for ManualEstimator (#209) (@franckgaga)

Closed issues:

• Create a CustomEstimator to ease interfacing with Low-level particle filter package (#142)
• Support nonlinear estimator with linear model predictive control (#205)

v1.6.2

ModelPredictiveControl v1.6.2

Diff since v1.6.1

Merged pull requests:

• CompatHelper: bump compat for DifferentiationInterface to 0.7, (keep existing compat) (#203) (@github-actions[bot])
• changed: compat of DAQP to 0.6, 0.7.1 (#204) (@franckgaga)
• debug: force update of gradient/jacobian in MovingHorzionEstimator when window not filled (#207) (@franckgaga)

v1.6.1

ModelPredictiveControl v1.6.1

Diff since v1.6.0

• added: improve NonLinMPC performances with specialized conversion and weight matrices
• changed: M_Hp, N_Hc and L_Hp keyword arguments now default to Diagonal instead of diagm matrices for all PredictiveController constructors

Merged pull requests:

• Cherry pick doc and general improvements commits from hessian_objective (#200) (@franckgaga)
• added: be more efficient with weights and conversion matrices for NonLinMPC (#202) (@franckgaga)

Closed issues:

• Support exact Hessian of objective function J in NonLinMPC and MovingHorizonEstimator (#193)
• Use one function in linearization (#199)

v1.6.0

ModelPredictiveControl v1.6.0

Diff since v1.5.3

• added: general performance boost of NonLinMPC and MovingHorizonEstimator using value_and_gradient!/jacobian! of DifferentiationInterface.jl instead of individual calls (similar to DiffResults.jl goal) 🚀🚀🚀
• added: setstate! now allows manual modifications of the estimation error covariance $\mathbf{\hat{P}}$ (if computed by the estimator)

Merged pull requests:

• added: setstate! now allows estimation covariance P̂ modification (if applicable) (#192) (@franckgaga)
• doc: minor corrections (#196) (@franckgaga)
• Added: NonLinMPC and MovingHorizonEstimator now use value_and_gradient!/jacobian! of DI.jl (#197) (@franckgaga)

Closed issues:

• Precompilation is failing (#195)

v1.5.3

ModelPredictiveControl v1.5.3

Diff since v1.5.2

Merged pull requests:

• Test: added allocations tests for types that are known to be allocation-free (SKIP THEM FOR NOW) (#187) (@franckgaga)
• test: fix CI badge (#188) (@franckgaga)
• debug: remove .data in KalmanFilter matrix products (#189) (@franckgaga)
• test: adapt tests for the new automatically balancing minreal function (#191) (@franckgaga)

v1.5.2

ModelPredictiveControl v1.5.2

Diff since v1.5.1

Merged pull requests:

• debug: do not call jacobian! if nd==0 in linearize (#186) (@franckgaga)

Closed issues:

• Linearize is failing when using AutoFiniteDiff() (#185)

v1.5.1

ModelPredictiveControl v1.5.1

Diff since v1.5.0

• removed: DiffCaches in RungeKutta solver
• debug: no more noisy @warn about DiffCache chunk size
• changed: compat for ForwardDiff to 0.10, 1
• test: added compat entries for the test-only dependencies in [extras] section

Merged pull requests:

• CompatHelper: bump compat for ForwardDiff to 1, (keep existing compat) (#182) (@github-actions[bot])
• Changed: RungeKutta no longer uses DiffCache (#183) (@franckgaga)

v1.5.0

ModelPredictiveControl v1.5.0

Diff since v1.4.4

• added: migration to DifferentiationInterface.jl
• added: new gradient and jacobian keyword arguments for NonLinMPC
• added: new gradient and jacobian keyword arguments for MovingHorizonEstimator
• added: new jacobian keyword argument for NonLinModel (for linearization)
• added: new jacobian keyword argument for ExtendedKalmanFilter
• added: ExtendedKalmanFilter is now allocation-free at runtime
• test: many new test with AutoFiniteDiff backend

Merged pull requests:

• added: NonLinMPC and MovingHorizonEstimator integration with DI.jl (#174) (@franckgaga)
• added: linearize! now uses DifferentiationInterface.jl (#179) (@franckgaga)
• Added: ExtendedKalmanFilter now uses DifferentiationInterface.jl + now allocation-free (#180) (@franckgaga)

Closed issues:

• Use DifferentiationInterface.jl (#162)

v1.4.4

ModelPredictiveControl v1.4.4

Diff since v1.4.3

Merged pull requests:

• Debug: ineq. constraint for NonLinModel and MultipleShooting (#178) (@franckgaga)