Learning model predictive control with long short‐term memory networks

DOI10.1002/RNC.5519MaRDI QIDQ6060780zbMATH OpenOpenAlexFDO

Authors Enrico Terzi, Fabio Bonassi, Marcello Farina, Riccardo Scattolini

Publication date 29 November 2023

Published in International Journal of Robust and Nonlinear Control (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1910.04024

zbMATH Keywords

learning-based controllong short-term memory neural networksmachine learningnonlinear model predictive controloutput feedback predictive control

Mathematics Subject Classification ID

Model predictive control (93B45) Networked control (93B70) Nonlinear systems in control theory (93C10) Input-output approaches in control theory (93D25)

Abstract: This paper analyzes the stability-related properties of Long Short-Term Memory (LSTM) networks and investigates their use as the model of the plant in the design of Model Predictive Controllers (MPC). First, sufficient conditions guaranteeing the Input-to-State stability (ISS) and Incremental Input-to-State stability (dISS) of LSTM are derived. These properties are then exploited to design an observer with guaranteed convergence of the state estimate to the true one. Such observer is then embedded in a MPC scheme solving the tracking problem. The resulting closed-loop scheme is proved to be asymptotically stable. The training algorithm and control scheme are tested numerically on the simulator of a pH reactor, and the reported results confirm the effectiveness of the proposed approach.

Recommendations

• Advanced predictive control for GRU and LSTM networks
• Learning-based model predictive current control for synchronous machines: an LSTM approach
• Recurrent neural network-based internal model control design for stable nonlinear systems
• Nonlinear MPC design for incrementally ISS systems with application to GRU networks
• Robust constrained nonlinear model predictive control with gated recurrent unit model

Cites work

• 10.1162/153244303768966139
• Constrained state estimation for nonlinear discrete-time systems: stability and moving horizon approximations
• Data-driven control of nonlinear systems: an on-line direct approach
• Deep learning
• Discrete-time stability with perturbations: Application to model predictive control
• Dynamic recurrent neural network for system identification and control
• From model-based control to data-driven control: survey, classification and perspective
• Getting Robustness Against Unstructured Uncertainty: A Tube-Based MPC Approach
• Global asymptotic stability and stabilization of long short-term memory neural networks with constant weights and biases
• Input-to-state stability for discrete-time nonlinear systems
• Moving-horizon state estimation for nonlinear discrete-time systems: new stability results and approximation schemes
• Provably safe and robust learning-based model predictive control
• Recurrent neural network-based model predictive control for continuous pharmaceutical manufacturing
• The Vanishing Gradient Problem During Learning Recurrent Neural Nets and Problem Solutions
• Time series forecasting using a hybrid ARIMA and neural network model
• Tube-based robust nonlinear model predictive control
• Virtual reference feedback tuning: A direct method for the design of feedback controllers

Cited in

(6)

• scientific article; zbMATH DE number 5150428 (Why is no real title available?)
• Advanced predictive control for GRU and LSTM networks
• Adaptive stochastic model predictive control via network ensemble learning
• Nonlinear MPC design for incrementally ISS systems with application to GRU networks
• Deep networks for system identification: a survey
• On incremental input-to-state stability of interconnected discrete-time systems

This page was built for publication: Learning model predictive control with long short‐term memory networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6060780)