Provably safe and robust learning-based model predictive control
From MaRDI portal
Abstract: Controller design faces a trade-off between robustness and performance, and the reliability of linear controllers has caused many practitioners to focus on the former. However, there is renewed interest in improving system performance to deal with growing energy constraints. This paper describes a learning-based model predictive control (LBMPC) scheme that provides deterministic guarantees on robustness, while statistical identification tools are used to identify richer models of the system in order to improve performance; the benefits of this framework are that it handles state and input constraints, optimizes system performance with respect to a cost function, and can be designed to use a wide variety of parametric or nonparametric statistical tools. The main insight of LBMPC is that safety and performance can be decoupled under reasonable conditions in an optimization framework by maintaining two models of the system. The first is an approximate model with bounds on its uncertainty, and the second model is updated by statistical methods. LBMPC improves performance by choosing inputs that minimize a cost subject to the learned dynamics, and it ensures safety and robustness by checking whether these same inputs keep the approximate model stable when it is subject to uncertainty. Furthermore, we show that if the system is sufficiently excited, then the LBMPC control action probabilistically converges to that of an MPC computed using the true dynamics.
Recommendations
- Learning for MPC with stability \& safety guarantees
- Robust learning-based MPC for nonlinear constrained systems
- ABC-LMPC: Safe Sample-Based Learning MPC for Stochastic Nonlinear Dynamical Systems with Adjustable Boundary Conditions
- Online learning with stability guarantees: a memory-based warm starting for real-time MPC
- Safe control of nonlinear systems in LPV framework using model-based reinforcement learning
Cites work
- scientific article; zbMATH DE number 47864 (Why is no real title available?)
- scientific article; zbMATH DE number 3551792 (Why is no real title available?)
- scientific article; zbMATH DE number 1303063 (Why is no real title available?)
- scientific article; zbMATH DE number 1956619 (Why is no real title available?)
- scientific article; zbMATH DE number 1956649 (Why is no real title available?)
- scientific article; zbMATH DE number 3997615 (Why is no real title available?)
- scientific article; zbMATH DE number 1444339 (Why is no real title available?)
- scientific article; zbMATH DE number 236540 (Why is no real title available?)
- scientific article; zbMATH DE number 6026918 (Why is no real title available?)
- scientific article; zbMATH DE number 3238721 (Why is no real title available?)
- scientific article; zbMATH DE number 3186512 (Why is no real title available?)
- A converse Lyapunov theorem for discrete-time systems with disturbances
- A distribution-free theory of nonparametric regression
- A time-dependent Hamilton-Jacobi formulation of reachable sets for continuous dynamic games
- Adaptive control of linearizable systems
- Adaptive control of unknown plants using dynamical neural networks
- Adaptive model predictive control for a class of constrained linear systems based on the comparison model
- Asymptotic Properties of Non-Linear Least Squares Estimators
- Constrained model predictive control: Stability and optimality
- Dual mode predictive tracking of piecewise constant references for constrained linear systems
- Estimating the Error Rate of a Prediction Rule: Improvement on Cross-Validation
- Estimation theory and uncertainty intervals evaluation in presence of unknown but bounded errors: Linear families of models and estimators
- Examples when nonlinear model predictive control is nonrobust
- Four encounters with system identification
- Hybrid Systems: Computation and Control
- Input-to-state stability for discrete-time nonlinear systems
- Input-to-state stability: a unifying framework for robust model predictive control
- Large‐Scale Inverse Problems and Quantification of Uncertainty
- Linear Model Selection by Cross-Validation
- Linear and nonlinear iterative learning control
- Linear time-varying model predictive control and its application to active steering systems: stability analysis and experimental validation
- MPC for tracking piecewise constant references for constrained linear systems
- Multivariate locally weighted least squares regression
- Nonlinear systems. Analysis, stability, and control
- On adaptive estimation
- Parameterized Robust Control Invariant Sets for Linear Systems: Theoretical Advances and Computational Remarks
- Problems of identification and control
- Reachability Analysis of Discrete-Time Systems With Disturbances
- Regression on manifolds: estimation of the exterior derivative
- Ridge Regression: Biased Estimation for Nonorthogonal Problems
- Robust adaptive MPC for constrained uncertain nonlinear systems
- Robust model predictive control using tubes.
- Robustness and robust design of MPC for nonlinear discrete-time systems
- SNOPT: An SQP Algorithm for Large-Scale Constrained Optimization
- Sensitivity analysis for nonlinear programming using penalty methods
- Strong consistency of least squares estimates in multiple regression II
- Systems with persistent disturbances: Predictive control with restricted constraints
- The Consistency of Nonlinear Regressions
- Theory and computation of disturbance invariant sets for discrete-time linear systems
- Variational Analysis
- Weighted Local Regression and Kernel Methods for Nonparametric Curve Fitting
Cited in
(57)- Constrained robust model predictive control embedded with a new data-driven technique
- Combining learning and control in linear systems
- Adaptive learning-based model predictive control for uncertain interconnected systems: a set membership identification approach
- Almost surely safe exploration and exploitation for deep reinforcement learning with state safety estimation
- Model reference safety-critical adaptive control for nonlinear and switched systems
- A learning- and scenario-based MPC design for nonlinear systems in LPV framework with safety and stability guarantees
- Data-driven safe gain-scheduling control
- Robust adaptive tube tracking model predictive control for piece‐wise constant reference signals
- Sampled-data funnel control and its use for safe continual learning
- Adaptive stochastic model predictive control via network ensemble learning
- Robust learning-based MPC for nonlinear constrained systems
- Performance bounds of adaptive MPC with bounded parameter uncertainties
- Separation of learning and control for cyber-physical systems
- Rolling self-triggered distributed MPC for dynamically coupled nonlinear systems
- Learning-based predictive control for linear systems: a unitary approach
- Safe Formation Control with Constrained Linear Model Predictive Control
- Stability-constrained Markov decision processes using MPC
- Learning model predictive control with long short‐term memory networks
- Adaptive model predictive control for linear time varying MIMO systems
- Safe reward‐based deep reinforcement learning control for an electro‐hydraulic servo system
- Online learning with stability guarantees: a memory-based warm starting for real-time MPC
- Robust adaptive model predictive control: Performance and parameter estimation
- Model-based Reinforcement Learning: A Survey
- Indirect adaptive model predictive control and its application to uncertain linear systems
- Lazily adapted constant kinky inference for nonparametric regression and model-reference adaptive control
- Safe learning-based model predictive control using the compatible models approach
- Adaptive receding horizon control for constrained MIMO systems
- Adaptive control Lyapunov function based model predictive control for continuous nonlinear systems
- Stability of controllers for Gaussian process dynamics
- Robust adaptive model predictive control with persistent excitation conditions
- Adaptive‐optimal control under time‐varying stochastic uncertainty using past learning
- Data‐enabled predictive control for quadcopters
- Training classifiers for feedback control with safety in mind
- A General Safety Framework for Learning-Based Control in Uncertain Robotic Systems
- Online learning constrained model predictive control based on double prediction
- A Q-learning predictive control scheme with guaranteed stability
- Bayesian optimization with safety constraints: safe and automatic parameter tuning in robotics
- Learning-based state estimation and control using MHE and MPC schemes with imperfect models
- Safe control of nonlinear systems in LPV framework using model-based reinforcement learning
- On the construction of safe controllable regions for affine systems with applications to robotics
- Learning-based MPC-architecture satisfying constraints during open loop identification
- Run-time optimization for learned controllers through quantitative games
- Online learning based risk-averse stochastic MPC of constrained linear uncertain systems
- ABC-LMPC: Safe Sample-Based Learning MPC for Stochastic Nonlinear Dynamical Systems with Adjustable Boundary Conditions
- Inverse optimization with noisy data
- SAMBA: safe model-based \& active reinforcement learning
- Model predictive control for nonlinear systems with time‐varying dynamics and guaranteed Lyapunov stability
- Economic Lyapunov‐based model predictive control with event‐triggered parametric identification
- Stabilizing predictive control with persistence of excitation for constrained linear systems
- Nonlinear learning‐based model predictive control supporting state and input dependent model uncertainty estimates
- Behavioral analytics for myopic agents
- Learning for MPC with stability \& safety guarantees
- Robust MPC with recursive model update
- Online learning‐based model predictive control with Gaussian process models and stability guarantees
- Minimum time learning model predictive control
- Learning‐based adaptive‐scenario‐tree model predictive control with improved probabilistic safety using robust Bayesian neural networks
- Statistics with set-valued functions: applications to inverse approximate optimization
This page was built for publication: Provably safe and robust learning-based model predictive control
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q490551)
