Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Adaptive Model Predictive Control for Deployable Control Systems with Constraints
Structural control in the context of extreme hazard mitigation has been the subject of research for nearly two decades. Despite this, two issues have not received the attention they deserve. The first issue is one of actuator constraints. In many applications, a controller must abide by a number of physical operating constraints such as actuator stroke or force limitations. The second issue is one of adaptability in order to maximize utilization of the control device and improve control performance. Popular active structural control formulations such as linear quadratic methods or frequency domain approaches do not directly account for operating constraints and lack the ability to tune or modify the objective function once defined. This paper proposes an adaptive model predictive control (MPC) scheme that addresses both these crucial aspects. The controller was implemented in real time on a deployable control system, which was recently proposed as a means for immediate, short-term control applications. This system presents a natural test bed to experimentally validate the controller formulation because both operating constraints and utilization issues arise naturally in its implementation. MPC is an advanced control approach that relies explicitly on a mathematical model of the process to predict future behaviors and compute control actions that minimize the structural response. The optimization was performed at every time step which—despite posing a serious computation challenge—provides the framework to systematically address physical operating constraints and facilitate the desired adaptability by allowing the objective function to be redefined with specified penalty weights. Real-time hybrid simulation (RTHS) was used to successfully demonstrate real-time implementation of the proposed controller on full-scale structures. Results show that MPC can be an effective tool to address the issues of actuator constraints and controller adaptability.
Adaptive Model Predictive Control for Deployable Control Systems with Constraints
Structural control in the context of extreme hazard mitigation has been the subject of research for nearly two decades. Despite this, two issues have not received the attention they deserve. The first issue is one of actuator constraints. In many applications, a controller must abide by a number of physical operating constraints such as actuator stroke or force limitations. The second issue is one of adaptability in order to maximize utilization of the control device and improve control performance. Popular active structural control formulations such as linear quadratic methods or frequency domain approaches do not directly account for operating constraints and lack the ability to tune or modify the objective function once defined. This paper proposes an adaptive model predictive control (MPC) scheme that addresses both these crucial aspects. The controller was implemented in real time on a deployable control system, which was recently proposed as a means for immediate, short-term control applications. This system presents a natural test bed to experimentally validate the controller formulation because both operating constraints and utilization issues arise naturally in its implementation. MPC is an advanced control approach that relies explicitly on a mathematical model of the process to predict future behaviors and compute control actions that minimize the structural response. The optimization was performed at every time step which—despite posing a serious computation challenge—provides the framework to systematically address physical operating constraints and facilitate the desired adaptability by allowing the objective function to be redefined with specified penalty weights. Real-time hybrid simulation (RTHS) was used to successfully demonstrate real-time implementation of the proposed controller on full-scale structures. Results show that MPC can be an effective tool to address the issues of actuator constraints and controller adaptability.
Adaptive Model Predictive Control for Deployable Control Systems with Constraints
Goorts, K. (Autor:in) / Narasimhan, S. (Autor:in)
09.08.2019
Aufsatz (Zeitschrift)
Elektronische Ressource
Unbekannt
Transient thermal comfort constraints for model predictive heating control
| UB Braunschweig | 2018
Adaptive Control System for Biogas Power Plant Using Model Predictive Control
| DOAJ | 2021
Adaptive Intelligent Model Predictive Control for Microgrid Load Frequency
| DOAJ | 2022