Data-driven predictive control in a stochastic setting: a unified framework

DOI10.1016/J.AUTOMATICA.2023.110961zbMATH Open1519.93068arXiv2203.10846OpenAlexW4360745415MaRDI QIDQ6160766FDOQ6160766

Authors: Valentina Breschi, Alessandro Chiuso, Simone Formentin 

Publication date: 26 June 2023

Published in: Automatica (Search for Journal in Brave)

Abstract: Data-driven predictive control (DDPC) has been recently proposed as an effective alternative to traditional model-predictive control (MPC) for its unique features of being time-efficient and unbiased with respect to the oracle solution. Nonetheless, it has also been observed that noise may strongly jeopardize the final closed-loop performance since it affects both the data-based system representation and the control update computed from the online measurements. Recent studies have shown that regularization is potentially a successful tool to counteract the effect of noise. At the same time, regularization requires the tuning of a set of penalty terms, whose choice might be practically difficult without closed-loop experiments. In this paper, by means of subspace identification tools, we pursue a three-fold goal: $(i)$ we set up a unified framework for the existing regularized data-driven predictive control schemes for stochastic systems; $(ii)$ we introduce $gamma$-DDPC, an efficient two-stage scheme that splits the optimization problem into two parts: fitting the initial conditions and optimizing the future performance, while guaranteeing constraint satisfaction; $(iii)$ we discuss the role of regularization for data-driven predictive control, providing new insight on $when$ and $how$ it should be applied. A benchmark numerical case study finally illustrates the performance of $gamma$-DDPC, showing how controller design can be simplified in terms of tuning effort and computational complexity when benefiting from the insights coming from the subspace identification realm.

Full work available at URL: https://arxiv.org/abs/2203.10846

Recommendations

• Data-driven subspace predictive control: stability and horizon tuning
• Robust data-driven predictive control using reachability analysis
• A novel robust predictive controller design based on data-driven subspace identification
• Data-driven predictive control for networked control systems
• Adaptive predictive control: a data-driven closed-loop subspace identification approach

zbMATH Keywords

regularizationcontrol of constrained systemsdata-based controlidentification for control

Mathematics Subject Classification ID

Linear systems in control theory (93C05) Discrete-time control/observation systems (93C55) Model predictive control (93B45)

Cites Work

• Asymptotic Statistics
• Fitting autoregressive models for prediction
• Graph implementations for nonsmooth convex programs
• The explicit linear quadratic regulator for constrained systems
• Consistency analysis of some closed-loop subspace identification methods
• The role of vector autoregressive modeling in predictor-based subspace identification
• A note on persistency of excitation
• Asymptotic Variance of Closed-Loop Subspace Identification Methods
• 4SID: Subspace algorithms for the identification of combined deterministic-stochastic systems
• Robust data-driven predictive control using reachability analysis
• Identification for control: from the early achievements to the revival of experiment design
• From experiment design to closed-loop control
• Asymptotic properties of subspace estimators
• Analysis of the asymptotic properties of the MOESP type of subspace algorithms
• On the Relation Between CCA and Predictor-Based Subspace Identification
• On- and off-line identification of linear state-space models
• The relation of the CCA subspace method to a balanced reduction of an autoregressive model.
• Unit canonical correlations between future and past
• On the Asymptotic Properties of Closed-Loop CCA-Type Subspace Algorithms: Equivalence Results and Role of the Future Horizon
• A comparison of model-based and data-driven controller tuning
• A data-driven approach to robust control of multivariable systems by convex optimization
• Data-Driven Model Predictive Control With Stability and Robustness Guarantees
• Deterministic continuous-time virtual reference feedback tuning (VRFT) with application to PID design
• Direct data‐driven design of switching controllers

Cited In (14)

• A Sample-Wise Data Driven Control Solver for the Stochastic Optimal Control Problem with Unknown Model Parameters
• Online data-enabled predictive control
• Data-driven control via Petersen's lemma
• Output‐feedback robust control of systems with uncertain dynamics via data‐driven policy learning
• Output prediction under scarce data operation: Control applications
• Robust Data-Enabled Predictive Control: Tractable Formulations and Performance Guarantees
• Data-driven uncertainty propagation for stochastic predictive control of multi-energy systems
• Data-driven output prediction and control of stochastic systems: an innovation-based approach
• Data-driven stabilization of switched and constrained linear systems
• Extensions to ``Output prediction under scarce data operation: Control applications
• Stochastic predictive control under intermittent observations and unreliable actions
• Robust data-driven predictive control using reachability analysis
• Data-driven control: a behavioral approach
• Data-driven subspace predictive control: stability and horizon tuning

This page was built for publication: Data-driven predictive control in a stochastic setting: a unified framework

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6160766)