Self-Reflective Model Predictive Control

Houska, Boris; Telen, Dries; Logist, Filip; Impe, Jan Van

doi:10.1137/15m1049865

Cited by 9 publications

(9 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, tuning these weighting matrices by hand may be cumbersome and is, to a certain extent, ambiguous. An alternative is to automatically compute more natural tradeoff weights by using the self-reflective model predictive control scheme [21] that is reviewed next.…”

Section: Optimal Experiments Design Based Mpc Objectivesmentioning

confidence: 99%

“…As measurement errors and process noise cannot be predicted, they cause an inevitable loss of control performance when compared to an utopia feedforward controller, which can predict all future measurement errors and process noise. Self-reflective MPC [21] is a controller that minimizes the sum of its nominal performance and a second order approximation of its own expected future loss of optimality compared to an utopia feedback controller that can predict everything.…”

Section: Self-reflective Model Predictive Controlmentioning

confidence: 99%

“…However, notice that in the self-reflective problem (10) the weights α k = 1 2 Φ(x k , u k , Ω k+1 ) are non-trivial functions of x k and u k and require the computation of the ancillary backward recursion for the variables Ω k , i.e., the tradeoff between the nominal tracking performance and the additive learning term is optimized. Notice that the term 1 2 Tr (Φ(x k , u k , Ω k+1 )Σ k ) approximates the self-reflective MPC controller's own inherent expected loss of optimality in the presence of measurement errors and process noise up to terms of order O γ 3 , as established in [21]; see also [40] for a dicussion and interpretation of this term in the context of unconstrained linear stochastic control. Other advantages of the self-reflective MPC formulation as well as general economic optimal experiment design criteria are discussed in [20,21].…”

Section: Self-reflective Model Predictive Controlmentioning

confidence: 99%

“…An even more recent trend is to extend the concept of application oriented optimal experiment design [17] by associated terms in the objective or constraints of an MPC problem [22,23]. Moreover, a recent paper on self-reflective model predictive control [21] proposes a model predictive controller that minimizes a prediction of its own expected loss of control performance in the presence of measurement noise. Notice that all these developments are closely related to the research on output-feedback MPC [11].…”

Section: Introductionmentioning

confidence: 99%

“…Therefore, a principal goal of this paper is to develop a real-time algorithm that can exploit the structure of such problems. Here, we focus on the self-reflective model predictive control formulation, which has been proposed in [21] and which is based on augmenting a nominal MPC objective with an economic optimal experiment design criterion [20]. However, the methods in this paper can also be extended for most of the other integrated experiment design MPC methods that have been reviewed above.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Real-time algorithm for self-reflective model predictive control

Feng

Houska

2018

Journal of Process Control

Self Cite

View full text Add to dashboard Cite

This paper is about a real-time model predictive control (MPC) algorithm for a particular class of model based controllers, whose objective consists of a nominal tracking objective and an additional learning objective. Here, the construction of the learning term is based on economic optimal experiment design criteria. It is added to the MPC objective in order to excite the system from time-to-time on purpose in order to improve the accuracy of the state and parameter estimates in the presence of incomplete or noise affected measurements. A particular focus of this paper is on so-called self-reflective model predictive control schemes, which have the property that the additional learning term can be interpreted as the expected loss of optimality of the controller in the presence of random measurement errors. The main contribution of this paper is a formulation-tailored algorithm, which exploits the particular structure of self-reflective MPC problems in order to speed-up the online computation. It is shown that, in contrast to generic state-ofthe-art optimal control problem solvers, the proposed algorithm can solve the selfreflective optimization problems with reasonable additional computational effort and in real-time. The advantages of the proposed real-time scheme are illustrated by applying the algorithm to a nonlinear process control problem in the presence of measurement errors and process noise.

show abstract

Section: Optimal Experiments Design Based Mpc Objectivesmentioning

confidence: 99%

Section: Self-reflective Model Predictive Controlmentioning

confidence: 99%

Section: Self-reflective Model Predictive Controlmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Real-time algorithm for self-reflective model predictive control

Feng

Houska

2018

Journal of Process Control

Self Cite

View full text Add to dashboard Cite

show abstract

Model predictive control with active learning under model uncertainty: Why, when, and how

et al. 2018

View full text Add to dashboard Cite

Optimal control relies on a model, which is generally uncertain because of incomplete knowledge of the system and changes in the dynamics over time. Probing the system under closed‐loop control can reduce the model uncertainty through generating input‐output data that is more informative than the data generated from normal operation. This paper addresses the problem of model predictive control (MPC) with active learning, with a particular focus on how incorporating probing in the control action can reduce model uncertainty. We discuss some of the central theoretical questions in this problem, and demonstrate the potential of active learning for maintaining MPC performance in the presence of uncertainty in model parameters and structure. Simulation results show that active learning is particularly beneficial when a system undergoes abrupt changes (such as the sudden occurrence of a fault) that can compromise operational safety, reliability, and profitability. © 2018 American Institute of Chemical Engineers AIChE J, 64: 3071–3081, 2018

show abstract

A nonlinear model predictive control scheme for sensor fault tolerance in observation processes

Knudsen

Alessandretti

Jones

et al. 2020

Intl J Robust & Nonlinear

View full text Add to dashboard Cite

SummaryThis article addresses the problem of designing a sensor fault‐tolerant controller for an observation process where a primary, controlled system observes, through a set of measurements, an exogenous system to estimate the state of this system. We consider sensor faults captured by a change in a set of sensor parameters affecting the measurements. Using this parametrization, we present a nonlinear model predictive control (NMPC) scheme to control the observation process and actively detect and estimate possible sensor faults, with adaptive controller reconfiguration to optimize the use of the remaining sensing capabilities. A key feature of the proposed scheme is the design of observability indices for the NMPC stage cost to improve the observability of both the state of the exogenous system and the sensor fault parameters. The effectiveness of the proposed scheme is illustrated through numerical simulations.

show abstract

Self-Reflective Model Predictive Control

Cited by 9 publications

References 52 publications

Real-time algorithm for self-reflective model predictive control

Real-time algorithm for self-reflective model predictive control

Model predictive control with active learning under model uncertainty: Why, when, and how

A nonlinear model predictive control scheme for sensor fault tolerance in observation processes

Contact Info

Product

Resources

About