IMPLEMENTATION_PROGRAMMING_LIST = (

Constant tuple programming list.

AbstractController

An abstract type that should be subtyped for controllers extensions.

AbstractImplementaiton

An abstract type that should be subtyped for activation function extensions, mainly for relu.

AbstractReferences

An abstract type that should be subtyped for references extensions.

AbstractSolvers

An abstract type that should be subtyped for solver extensions.

AbstractTerminal

An abstract type that should be subtyped for terminal extensions.

AbstractWeights

An abstract type that should be subtyped for weights coefficients extensions.

FuzzyProgramming

Linear fuzzy tool for modeler implementation of neural networks.

LinearProgramming

Linear tool for modeler implementation of neural networks.

MixedIntegerLinearProgramming

Milp tool for modeler implementation of Fnn and Resnet with relu activation function.

ModelPredictiveControlController

Model predictive control main struct parameters. The controller as all the necessary before optimization.

** Fields **

• system: mathematical system type of the dynamical system (f and constraints).
• tuning: model predictive control implementation according to reference.
• initialization: initialization of the model predictive control before computation.
• inputs_command: model predictive control input after computation, which are sent to dynamical system.

ModelPredictiveControlResults

Model predictive control results after computation.

** Fields **

• x: state computed.
• e_x: state deviation computed.
• u: input computed.
• e_u: input deviation computed.

ModelPredictiveControlTuning

Model predictive control tuning implementation according to parameters and references.

** Fields **

• modeler: model predictive control implementation acocrding to JuMP.
• reference: model predictive control references.
• horizon: model predictive control horizon.
• weights: model predictive control weighting coefficients.
• terminal_ingredient: model predictive control terminal ingredients.
• sample_time: model predictive control sample time.
• max_time: model predictive control maximum time computation.

NonLinearProgramming

Nl tool for modeler implementation of Fnn and Resnet abd DenseNet with any activation function.

ReferencesStateInput

State and input references for model predictive control.

** Fields **

• x: state references.
• u: input references.

WeightsCoefficient

State, input and input rate weighting coefficient for the cost function.

** Fields **

• Q: state weight coefficient.
• R: input weight coefficient.
• S: input rate weight coefficient.

auto

Automatique selection of the solver according to the method.

highs

Linear quadratic solver.

ipopt

Linear and non-linear quadratic solver.

osqp

Linear quadratic solver.

SCIP

Linear and integer and non-linear quadratic solver.

_create_quadratic_cost_function

Function that create the quadratic cost of the model predictive control.

** Required fields **

• model_mpc: the JuMP struct.
• weights: the weighing coefficient struct.
• P_cost: the terminal ingredient cost.

_create_terminal_ingredient

Function that create the terminal ingredient for model predictive control. The terminal ingredients are the terminal weight and terminal constraints. The terminal constraints could be optional.

** Required fields **

• model_mpc: the JuMP from the modeler design.
• terminal_ingredient: a string to set the terminal constraints.
• system: the dynamical system mathemacal systems.
• references: the states and inputs references.
• weights: the weighting coefficients.
• kws optional parameters.

_create_weights_coefficients

Function that create the weighting struct for model predictive control.

** Required fields **

• system: the mathematical system from the dynamical system.
• kws: optional parameters.

_create_weights_coefficients

Function that create the weighting struct for model predictive control.

** Required fields **

• system: the mathematical system from the dynamical system.
• kws: optional parameters.

_design_reference_mpc

A function for references for model predictive control and linearization point for economic model predictive control.

The following variables are mendatories:

• state_reference: the state reference.
• input_reference: the input reference.
• horizon: the horizon for the mpc or empc.

_model_predictive_control_design

Function that tunes a model predictive control.

** Required fields **

• system: mathematical system type.
• horizon: horizon length of the model predictive control.
• method: implementation method (linear, non linear, mixed integer linear, fuzzy rules)
• sample_time: time sample of the model predictive control.
• kws optional fields

_model_predictive_control_design

Function that tunes a model predictive control.

** Required fields **

• system: mathematical system type.
• horizon: horizon length of the model predictive control.
• method: implementation method (linear, non linear, mixed integer linear, fuzzy rules)
• sample_time: time sample of the model predictive control.
• kws optional fields

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• model_mlj: model of the dynamical system from AutomationLabsIdentification package.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

_model_predictive_control_modeler_implementation

Modeler implementation of a Model Predictive Control.

** Required fields **

• method: method which is multipled dispatched type. Type are LinearProgramming, TakagiSugeno, NonLinearProgramming and MixedIntegerLinearProgramming.
• system: system from MethematicalSystems.
• horizon : model predictive control horizon parameter.
• references : model predictive control references.
• solver: model predictive control solver selection.

calculate!(C::ModelPredictiveControlController)

Model predictive control computation function. The controller is computed and optimization results values are written on mutable struct controller.

** Required fields **

• C: the model predictive control controller.

proceed_controller

A function for model predictive control and economic model predictive control.

The following variables are mendatories:

• system: a mathematical systems from AutomationLabsSystems.
• mpc_controller_type: a string for model predictive control or economic model predictive control.
• mpc_horizon: a horizon for the predictive controller.
• mpc_sample_time: a sample time for the predictive controller.
• mpc_state_reference: state reference for mpc or linearization point for empc.
• mpc_input_reference: input reference for mpc or linearization point for empc.
• kws optional argument.

update_initialization!(C::ModelPredictiveControlController, initialization::Vector)

Update initialization (also known as state measure) of Model predictive control controller.

** Required fields **

• C: the design model predictive control controller.
• initialization: initialization of the model predictive control befre computation (also now as the state measures).