Real-Time Implementation of Explicit Model Predictive Control

Kvasnica, Michal; Jones, Colin N.; Pejcic, Ivan; Holaza, Juraj; Korda, Milan; Bakaráč, Peter

doi:10.1007/978-3-319-77489-3_17

Cited by 6 publications

(2 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One of the main advantages of neural networks is to emulate, with minimal errors, the behaviour of any mathematical function, given sufficiently many neurons in sufficiently many layers. Since MPC strategies can be expressed as closed explicit functions (as shown in [10]), the neural network can be easily used in this setup as well. Let us denote the neural network controller as…”

Section: Neural Network Controllersmentioning

confidence: 99%

Nearly-optimal Explicit MPC-based Reference Governors with Long Prediction Horizons Generated with Machine Learning

Kiš,

Klaučo

2023

2023 24th International Conference on Process Control (PC)

View full text Add to dashboard Cite

The paper shows a procedure for constructing an approximated explicit form of the MPC-based reference governor. MPC-based reference governors are often setup up with long prediction horizons with a significant number of constraints, which forbids using conventional parametric optimisation to obtain the explicit solution. This paper explores the approach of mimicking the behaviour of the MPC-based reference governor with a neural network. The paper shows methods that ensure pointwise satisfaction of process constraints during neural network training. A demonstration using a well-known MIMO process is offered to evaluate control performance.

show abstract

Section: Neural Network Controllersmentioning

confidence: 99%

Nearly-optimal Explicit MPC-based Reference Governors with Long Prediction Horizons Generated with Machine Learning

Kiš,

Klaučo

2023

2023 24th International Conference on Process Control (PC)

View full text Add to dashboard Cite

show abstract

“…The complexity of the explicit control increases very rapidly with both the state-space dimension of the plant and the prediction horizon, leading to challenges with storing the explicit control in a data structure that facilitates retrieving the optimal control [7], [8], [9]. Kvasnica and coauthors have studied complexity reduction in explicit MPC; see the brief survey [10]. In [11] and [12], the behavior of the partitioning of the state space was investigated, resulting in substantial new knowledge on the structure of the explicit MPC controller.…”

Section: Introductionmentioning

confidence: 99%

Region-free explicit model predictive control for linear systems on Hilbert spaces

Kurula¹,

Humaloja²,

Dubljević³

2022

Preprint

View full text Add to dashboard Cite

We extend explicit model predictive control (MPC) rigorously to linear distributed parameter systems (DPS). In a theoretical discussion on the foundations of explicit MPC with Hilbert-space optimization, we sort out some of the frequent misconceptions in the present explicit MPC literature. Based on our main result, and a result that it is enough to search among candidate active sets that satisfy the linear independence constraint qualification (LICQ), we give a novel active-set algorithm that implements very fast region-free explicit MPC for highdimensional plants with a relatively long optimization horizon. If the Slater condition is met, then there exists an active set which characterizes optimality, and in this case the algorithm is guaranteed to find it in finite time. We use a Timoshenko beam with input and state constraints to demonstrate the efficacy of the proposed design and the capability of controlling a continuoustime hyperbolic PDE with a discrete-time explicit MPC controller.

show abstract

Parametric optimization and control toward the design of a smart metal hydride refueling system

Ogumerem¹,

Pistikopoulos²

2019

AIChE Journal

View full text Add to dashboard Cite

In this paper, we present an extension of the parametric optimization and control (PAROC) framework toward the design of a novel smart metal hydride storage system. Both in silico and laboratory implementation are considered toward a smart prototype system application. In silico PAROC considerations include the development and validation of high‐fidelity model and the derivation of explicit optimal feedback law through the solution of a receding horizon parametric optimization problem formulation. The derived explicit parametric control strategy is validated first in silico then in real time. A laboratory scale metal hydride hydrogen storage system is designed and built as a prototype to replicate the refueling process of a fuel cell electric vehicle (FCEV) in a hydrogen gas station. Integration of the explicit model predictive control feedback law into the prototype system creates a smart metal hydride storage technology that can be deployed for onboard hydrogen storage in an FCEV. The results obtained from the real‐time implementation of the explicit feedback law demonstrate an effective thermal management of the refueling operation in a metal hydride system.

show abstract

Real-Time Implementation of Explicit Model Predictive Control

Cited by 6 publications

References 32 publications

Nearly-optimal Explicit MPC-based Reference Governors with Long Prediction Horizons Generated with Machine Learning

Nearly-optimal Explicit MPC-based Reference Governors with Long Prediction Horizons Generated with Machine Learning

Region-free explicit model predictive control for linear systems on Hilbert spaces

Parametric optimization and control toward the design of a smart metal hydride refueling system

Contact Info

Product

Resources

About