Abstract:This study presents a method to improve the terminal components of the MPC optimization problem with the iterative execution of a control task when a system is nonlinear and initially unknown. The proposed method provides desirable properties such as robust constraint satisfaction, recursive feasibility, and input to state stability of the goal region under certain assumptions, and bound on the performance cost in each iteration is also analyzed. The results of the simulation study show that the proposed control scheme can iteratively improve the control performance. This study presents a learning‐based iterative model predictive control (MPC) scheme for unknown (Lipschitz continuous) nonlinear dynamical systems. The proposed method begins by learning the unknown part of the controlled system using a Gaussian process (GP), which helps derive multi‐step reachable sets that are guaranteed to encompass the actual system states. At each time step in each iteration, the MPC controller calculates a sequence of control inputs that robustly satisfy state and control constraints, as well as terminal constraints based on the GP‐based reachable sets. Then only the first control input is applied to the system. After the iteration, the initial state is reset, and the same procedure is executed with the MPC optimization problem defined by the updated terminal set and cost. As iteration goes on, improvement of the control performance is expected since more data is obtained and the environment is progressively explored. The proposed method provides properties such as recursive feasibility and input to state stability of the goal region under certain assumptions. Moreover, bound on the performance cost in each iteration associated with the implementation of the proposed MPC scheme is also analyzed. The results of the simulation study show that the proposed control scheme can iteratively improve the control performance.

Constraints-Informed Neural-Laguerre Approximation of Nonlinear MPC with Application in Power Electronics

Fast Linear Parameter Varying Model Predictive Control of Buck DC-DC Converters Based on FPGA.

Data-Driven Neural Predictors-Based Robust MPC for Power Converters

A Neural Network Architecture to Learn Explicit MPC Controllers from Data

Learning Model Predictive Control Law for Nonlinear Systems

Approximate non-linear model predictive control with safety-augmented neural networks

Fuzzy Approximation ARX Model-based Intelligent Two-Horizon Robust FCS-MPC for Power Converter

FCS-MPC of Power Converters: an Event-Driven Brain Emotional Learning Approach

Robust Learning-based Predictive Control for Constrained Nonlinear Systems.

Lebesgue-Approximation Model Predictive Control of Nonlinear Sampled-Data Systems

Data-Driven Iterative Learning Predictive Control for Power Converters

A Fuzzy Approximation for FCS-MPC in Power Converters

Learning an Approximate Model Predictive Controller With Guarantees

Robust learning‐based iterative model predictive control for unknown non‐linear systems

Efficient model predictive control for nonlinear systems modelled by deep neural networks

Neural Lyapunov Model Predictive Control: Learning Safe Global Controllers from Sub-optimal Examples

MPC of Uncertain Nonlinear Systems with Meta-Learning for Fast Adaptation of Neural Predictive Models

Q-Learning-Based Model Predictive Control for Nonlinear Continuous-Time Systems

Learning-based Predictive Control for Nonlinear Systems with Unknown Dynamics Subject to Safety Constraints

Safe and Near-Optimal Policy Learning for Model Predictive Control using Primal-Dual Neural Networks

PlanNetX: Learning an Efficient Neural Network Planner from MPC for Longitudinal Control