Adaptive Sampling of Dynamic Systems for Generation of Fast and Accurate\n Surrogate Models

2021/07/29 by Torben Talis, Joris Weigert, Talis, Torben +5
Engineering · #Advanced Control Systems Optimization #FOS: Electrical engineering #Fault Detection and Control Systems #Process Optimization and Integration #Systems and Control (eess.SY) #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2107.13826

openalex publication_date 2021/07/29 · openalex created_date 2022/10/01 · openalex updated_date 2026/07/28

Abstract

For economic nonlinear model predictive control and dynamic real-time\noptimization fast and accurate models are necessary. Consequently, the use of\ndynamic surrogate models to mimic complex rigorous models is increasingly\ncoming into focus. For dynamic systems, the focus so far had been on\nidentifying a system's behavior surrounding a steady-state operation point. In\nthis contribution, we propose a novel methodology to adaptively sample rigorous\ndynamic process models to generate a dataset for building dynamic surrogate\nmodels. The goal of the developed algorithm is to cover an as large as possible\narea of the feasible region of the original model. To demonstrate the\nperformance of the presented framework it is applied on a dynamic model of a\nchlor-alkali electrolysis.\n

Related

• Task-optimal data-driven surrogate models for eNMPC via differentiable simulation and optimization
• Model fusion for efficient learning of nonlinear dynamical systems
• Learning Nonlinear Dynamics: Improving the Estimation Efficiency and Reliability of Gaussian Process State-Space Models
• Adaptive Economic Model Predictive Control: Performance Guarantees for Nonlinear Systems
• A Reinforcement Learning-based Economic Model Predictive Control\n Framework for Autonomous Operation of Chemical Reactors