Lyapunov-Based Model Predictive Control of Nonlinear Systems Subject to Data Losses

Peña, David Muñoz de la; Christofides, Panagiotis D.

doi:10.1109/tac.2008.929401

Cited by 261 publications

(117 citation statements)

References 43 publications

Supporting

Mentioning

116

Contrasting

Order By: Relevance

“…For completeness of presentation, the stability properties are summarized below (the proofs of Propositions 1-2 can be found in [29] and the proof of Theorem 1 in [20]). The results utilize the following properties which follow from the fact that g(·, ·, ·) is a locally Lipschitz vector function in its arguments, the Lyapunov function V (·) is continuously differentiable, and the state vector z, the input vector u m , and the disturbance vector w are all bounded in a compact set.…”

Section: Stability Analysismentioning

confidence: 99%

Integrated Design of Control Actuator Layer and Economic Model Predictive Control for Nonlinear Processes

Durand

2014

Ind. Eng. Chem. Res.

Self Cite

View full text Add to dashboard Cite

In the present work, an economic model predictive control (EMPC) system is designed that accounts for the dynamics of the control actuators. A combined process-actuator dynamic model is developed to describe the process and control actuator dynamics and is used within the EMPC system. Integrating the design of the regulatory control layer, which controls the control actuators, and the supervisory control layer consisting of an EMPC system is an important consideration given the fact that EMPC may force an unsteady-state operating policy to optimize the process economics and the dynamics of the control actuator layer may affect the closed-loop process-actuator dynamics. Moreover, integral or average input constraints are often imposed within the EMPC solution. However, if the actuator layer is not accounted for in the EMPC system, the actuator output trajectory may not satisfy the integral input constraints. To address closed-loop stability of the combined process-actuator closed-loop system, stability constraints, designed via Lyapunov-based techniques, are imposed on the EMPC problem to guarantee closed-loop stability of the process system under the EMPC. An EMPC system accounting for the control actuator dynamics is applied to ii a benchmark chemical process example to study the impact of the actuator dynamics on closed-loop economic performance and reactant material constraint satisfaction.iii The thesis of Helen Elaine Durand is approved.

show abstract

Section: Stability Analysismentioning

confidence: 99%

Integrated Design of Control Actuator Layer and Economic Model Predictive Control for Nonlinear Processes

Durand

2014

Ind. Eng. Chem. Res.

Self Cite

View full text Add to dashboard Cite

show abstract

“…To guarantee that the resulting closedloop system is stable, we follow an LMPC approach for the design of the upper-tier controller, see Mhaskar et al (2005Mhaskar et al ( , 2006 and Muñoz de la Peña and Christofides (2008). LMPC is based on including a contractive constraint that allows to prove practical stability of the closed-loop system using an auxiliary Lyapunov-based controller.…”

Section: Upper-tier Lmpc Designmentioning

confidence: 99%

“…LMPC is based on including a contractive constraint that allows to prove practical stability of the closed-loop system using an auxiliary Lyapunov-based controller. In the previous LMPC schemes (Mhaskar et al, 2005(Mhaskar et al, , 2006Muñoz de la Peña and Christofides, 2008), the contractive constraints are defined based on a known Lyapunov-based state feedback controller. In the present work, the contractive constraint of the proposed upper-tier LMPC design is based on the Lyapunov function of the closed-loop system under the lower-tier controller k s .…”

Section: Upper-tier Lmpc Designmentioning

confidence: 99%

“…This class of systems arises in the context of process control systems utilizing hybrid communication networks consisting of point-to-point wired links integrated with wireless actuator/sensor networks. Assuming that there exists a lower-tier control system which relies on point-to-point communication and continuous measurements to stabilize the closed-loop system, we proposed to use Lyapunov-based model predictive control (LMPC) theory (Mhaskar et al, 2005(Mhaskar et al, , 2006; Muñoz de la Peña and Christofides, 2008) to design an upper-tier networked control system which profits from both the continuous and the asynchronous measurements as well as from additional networked control actuators. LMPC is based on the concept of uniting model predictive control with control Lyapunov functions as a way of guaranteeing closed-loop stability.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A two-tier architecture for networked process control

Liu

Peña

Ohran

et al. 2008

Chemical Engineering Science

Self Cite

View full text Add to dashboard Cite

Traditionally, process control systems utilize dedicated, point-to-point wired communication links using a small number of sensors and actuators to regulate appropriate process variables at desired values. While this paradigm to process control has been successful, chemical plant operation could substantially benefit from an efficient integration of the existing, point-to-point control networks (wired connections from each actuator/sensor to the control system using dedicated local area networks) with additional networked (wired or wireless) actuator/sensor devices. However, augmenting existing control networks with real-time wired/wireless sensor and actuator networks challenges many of the assumptions made in the development of traditional process control methods dealing with dynamical systems linked through ideal channels with flawless, continuous communication. In the context of control systems which utilize networked sensors and actuators, key issues that need to be carefully handled at the control system design level include data losses due to field interference and time delays due to network traffic. Motivated by the above technological advances and the lack of methods to design control systems that utilize hybrid communication networks, in the present work, we present a novel two-tier control architecture for networked process control problems that involve nonlinear processes and heterogeneous measurements consisting of continuous measurements and asynchronous, delayed measurements. This class of control problems arises naturally when nonlinear processes are controlled via control systems based on hybrid communication networks (i.e., point-to-point wired links integrated with networked wired/wireless communication) or utilizing multiple heterogeneous measurements (e.g., temperature measurements which can be taken to be continuous and species concentration measurements which are fed to the control system at asynchronous time instants and frequently involve delays). While point-to-point wired links are very reliable, the presence of a shared communication network in the closed-loop system introduces additional delays and data losses and these issues should be handled at the controller design level. In the two-tier control architecture presented in this work, a lower-tier control system, which relies on point-topoint communication and continuous measurements, is first designed to stabilize the closed-loop system, and an upper-tier networked control system is subsequently designed, using Lyapunov-based model predictive control theory, to profit from both the continuous and the asynchronous, delayed measurements as well as from additional networked control actuators to improve the closed-loop system performance. The proposed two-tier control architecture preserves the stability properties of the lower-tier controller while improving the closed-loop performance. The applicability and effectiveness of the proposed control method is demonstrated using two chemical process examples.

show abstract

“…An effective strategy for combating packet dropouts is to compensate them. The works [16,17] use the model predictive controller to send the current and future control inputs in a single packet. The authors of [18] consider the compensation for dropped feedback measurements, while the authors of [19] propose a compensation scheme for the control packet dropout at the actuator using the past control signals.…”

Section: Introductionmentioning

confidence: 99%

Separation principle for networked control systems with multiple-packet transmission

Chen

2011

IET Control Theory Appl.

View full text Add to dashboard Cite

Abstract:The authors investigate a class of observer-based discrete-time networked control systems (NCSs) with multiple-packet transmission where random packet dropouts occur independently in both the sensor-to-controller (S/C) and controller-to-actuator (C/A) channels. The authors first propose and prove the separation principle for the NCSs where packet dropouts in the C/A and S/C channels are governed by two independent Markov chains, respectively. Secondly, the authors derive a sufficient condition, in terms of linear matrix inequalities (LMIs), for stabilisation control of the Markov chain-driven NCSs. The authors also derive the necessary and sufficient condition for stabilisation control of the memoryless process-driven NCSs as a special case. A numerical example is provided to illustrate the effectiveness of our method.

show abstract

Lyapunov-Based Model Predictive Control of Nonlinear Systems Subject to Data Losses

Cited by 261 publications

References 43 publications

Integrated Design of Control Actuator Layer and Economic Model Predictive Control for Nonlinear Processes

Integrated Design of Control Actuator Layer and Economic Model Predictive Control for Nonlinear Processes

A two-tier architecture for networked process control

Separation principle for networked control systems with multiple-packet transmission

Contact Info

Product

Resources

About