4.6 Article

Formulating data-driven surrogate models for process optimization

相关参考文献

注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article Energy & Fuels

Multi-objective constrained optimization for energy applications via tree ensembles

Alexander Thebelt et al.

Summary: This paper proposes a novel data-driven strategy using tree ensembles for constrained multi-objective optimization of black-box problems with heterogeneous variable spaces. It demonstrates competitive performance and sampling efficiency in real-world applications with limited evaluation budgets compared to other state-of-the-art tools.

APPLIED ENERGY (2022)

Article Engineering, Chemical

Maximizing information from chemical engineering data sets: Applications to machine learning

Alexander Thebelt et al.

Summary: This review discusses the challenging data characteristics in chemical engineering applications and explores how current chemical engineering research is expanding the fields of data science and machine learning.

CHEMICAL ENGINEERING SCIENCE (2022)

Article Computer Science, Interdisciplinary Applications

Data-driven strategies for extractive distillation unit optimization

Kaiwen Ma et al.

Summary: This study provides insights into the development of a fast, dynamic, and adaptive way of modeling and optimizing chemical processes. Two data-driven methodologies are investigated to optimize the energy cost of an extractive distillation process. The study finds that ALAMO performs well for less complex systems, while the ReLU network performs better for more complex systems. BBO with a smooth penalty function is the more effective approach for the problems considered in this study.

COMPUTERS & CHEMICAL ENGINEERING (2022)

Article Computer Science, Software Engineering

Deterministic global optimization with Gaussian processes embedded

Artur M. Schweidtmann et al.

Summary: The proposed method simplifies the space and branches only on free variables for faster convergence, leading to significantly reduced computational burden. Tight relaxations of common covariance functions and acquisition functions in Bayesian optimization are derived to further accelerate the process.

MATHEMATICAL PROGRAMMING COMPUTATION (2021)

Review Engineering, Chemical

Machine Learning in Chemical Engineering: A Perspective

Artur M. Schweidtmann et al.

Summary: The transformation of the chemical industry towards renewable energy and feedstock supply requires new paradigms for design. Recent breakthroughs in machine learning offer unique opportunities, but interdisciplinary research between machine learning and chemical engineering is essential to unlock the full potential. Six challenges have been identified to open up new methods in chemical engineering and formulate new problems for machine learning.

CHEMIE INGENIEUR TECHNIK (2021)

Article Computer Science, Interdisciplinary Applications

Real-time refinery optimization with re duce d-order fluidize d catalytic cracker model and surrogate-based trust region filter method

Xinhe Chen et al.

Summary: Real-Time Optimization (RTO) is widely appreciated for optimizing process decision variables in refineries. With increasing demand for improved optimization strategies, surrogate models and the trust region filter (TRF) optimization strategy have been embedded within the RTO framework. This approach has been shown to improve efficiency and reduce computational burden in real-world refinery applications.

COMPUTERS & CHEMICAL ENGINEERING (2021)

Article Computer Science, Interdisciplinary Applications

Sobolev trained neural network surrogate models for optimization

Calvin Tsay

Summary: This paper examines the direct impacts of Sobolev training on neural network surrogate models embedded in optimization problems and proposes a systematic strategy for scaling Sobolev-space targets during NN training. It is shown that Sobolev training results in surrogate models with more accurate derivatives, with direct benefits in gradient-based optimization. The advantages of Sobolev training are especially significant in cases of low data volume and/or optimal points near the boundary of the training dataset.

COMPUTERS & CHEMICAL ENGINEERING (2021)

Review Computer Science, Interdisciplinary Applications

Process systems engineering - The generation next?

E. N. Pistikopoulos et al.

Summary: Process Systems Engineering (PSE) is a scientific discipline that integrates scales and components to describe the behavior of physicochemical systems through mathematical modeling, data analytics, design, optimization, and control. It provides a scientific basis and computational tools to address contemporary and future challenges in energy, environment, future industries, and sustainability. This perspective article offers a guide to the next generation of PSE developments by examining its history, core competencies, current status, and ongoing trends.

COMPUTERS & CHEMICAL ENGINEERING (2021)

Article Computer Science, Interdisciplinary Applications

ENTMOOT: A framework for optimization over ensemble tree models

Alexander Thebelt et al.

Summary: Gradient boosted trees and other regression tree models are effective in real-world applications, providing insight into important prediction features and managing sparse data effectively. However, they are generally unpopular for decision-making tasks and black-box optimization due to their difficult-to-optimize structure and lack of a reliable uncertainty measure.

COMPUTERS & CHEMICAL ENGINEERING (2021)

Article Computer Science, Interdisciplinary Applications

Mixed-Integer Convex Nonlinear Optimization with Gradient-Boosted Trees Embedded

Miten Mistry et al.

Summary: This paper discusses using decision trees to represent sparse, high-dimensional, and noisy data, as well as how to integrate the learned function into larger decision-making problems. It proposes a mixed-integer optimization problem for optimizing tree models, and develops heuristic methods and an exact branch-and-bound algorithm to find feasible solutions.

INFORMS JOURNAL ON COMPUTING (2021)

Article Mathematics

Fit without fear: remarkable mathematical phenomena of deep learning through the prism of interpolation

Mikhail Belkin

Summary: This paper explores the gap between the mathematical theory of machine learning and practical challenges, focusing on the key themes of interpolation and over-parametrization in understanding the foundations of deep learning.

ACTA NUMERICA (2021)

Article Computer Science, Interdisciplinary Applications

Integrating deep learning models and multiparametric programming

Justin Katz et al.

COMPUTERS & CHEMICAL ENGINEERING (2020)

Article Computer Science, Software Engineering

Strong mixed-integer programming formulations for trained neural networks

Ross Anderson et al.

MATHEMATICAL PROGRAMMING (2020)

Article Computer Science, Interdisciplinary Applications

Surrogate-based optimization for mixed-integer nonlinear problems

Sun Hye Kim et al.

COMPUTERS & CHEMICAL ENGINEERING (2020)

Article Management

Optimization of Tree Ensembles

Velibor V. Misic

OPERATIONS RESEARCH (2020)

Article Computer Science, Interdisciplinary Applications

Optimization under uncertainty in the era of big data and deep learning: When machine learning meets mathematical programming

Chao Ning et al.

COMPUTERS & CHEMICAL ENGINEERING (2019)

Article Computer Science, Interdisciplinary Applications

Reduced order models for dynamic molecular weight distribution in polymerization processes

Jiayuan Kang et al.

COMPUTERS & CHEMICAL ENGINEERING (2019)

Article Computer Science, Interdisciplinary Applications

Advances and opportunities in machine learning for process data analytics

S. Joe Qin et al.

COMPUTERS & CHEMICAL ENGINEERING (2019)

Article Engineering, Chemical

110th Anniversary: Using Data to Bridge the Time and Length Scales of Process Systems

Calvin Tsay et al.

INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH (2019)

Review Engineering, Chemical

Overview of Surrogate Modeling in Chemical Process Engineering

Kevin McBride et al.

CHEMIE INGENIEUR TECHNIK (2019)

Article Computer Science, Interdisciplinary Applications

ReLU networks as surrogate models in mixed-integer linear programs

Bjarne Grimstad et al.

COMPUTERS & CHEMICAL ENGINEERING (2019)

Article Operations Research & Management Science

Deterministic Global Optimization with Artificial Neural Networks Embedded

Artur M. Schweidtmann et al.

JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS (2019)

Review Computer Science, Interdisciplinary Applications

Advances in surrogate based modeling, feasibility analysis, and optimization: A review

Atharv Bhosekar et al.

COMPUTERS & CHEMICAL ENGINEERING (2018)

Article Computer Science, Artificial Intelligence

Deep neural networks and mixed integer linear optimization

Matteo Fischetti et al.

CONSTRAINTS (2018)

Article Engineering, Chemical

Advanced Trust Region Optimization Strategies for Glass Box/Black Box Models

John P. Eason et al.

AICHE JOURNAL (2018)

Article Computer Science, Interdisciplinary Applications

A trust region-based two phase algorithm for constrained black-box and grey-box optimization with infeasible initial point

Ishan Bajaj et al.

COMPUTERS & CHEMICAL ENGINEERING (2018)

Article Operations Research & Management Science

ARGONAUT: AlgoRithms for Global Optimization of coNstrAined grey-box compUTational problems

Fani Boukouvala et al.

OPTIMIZATION LETTERS (2017)

Article Computer Science, Artificial Intelligence

Optimal classification trees

Dimitris Bertsimas et al.

MACHINE LEARNING (2017)

Article Computer Science, Interdisciplinary Applications

The ALAMO approach to machine learning

Zachary T. Wilson et al.

COMPUTERS & CHEMICAL ENGINEERING (2017)

Article Engineering, Chemical

A Trust Region Filter Method for Glass Box/Black Box Optimization

John P. Eason et al.

AICHE JOURNAL (2016)

Review Computer Science, Interdisciplinary Applications

Hybrid semi-parametric modeling in process systems engineering: Past, present and future

Moritz von Stosch et al.

COMPUTERS & CHEMICAL ENGINEERING (2014)

Article Operations Research & Management Science

Derivative-free optimization: a review of algorithms and comparison of software implementations

Luis Miguel Rios et al.

JOURNAL OF GLOBAL OPTIMIZATION (2013)

Article Engineering, Multidisciplinary

A trust-region framework for constrained optimization using reduced order modeling

Anshul Agarwal et al.

OPTIMIZATION AND ENGINEERING (2013)

Article Engineering, Chemical

Surrogate-Based Superstructure Optimization Framework

Carlos A. Henao et al.

AICHE JOURNAL (2011)

Article Management

Robust Optimization for Unconstrained Simulation-Based Problems

Dimitris Bertsimas et al.

OPERATIONS RESEARCH (2010)

Article Computer Science, Interdisciplinary Applications

Nonconvex Robust Optimization for Problems with Constraints

Dimitris Bertsimas et al.

INFORMS JOURNAL ON COMPUTING (2010)

Article Computer Science, Software Engineering

SNOBFIT - Stable noisy optimization by branch and fit

Waltraud Huyer et al.

ACM TRANSACTIONS ON MATHEMATICAL SOFTWARE (2008)

Article Engineering, Chemical

An algorithm for the use of surrogate models in modular flowsheet optimization

Jose A. Caballero et al.

AICHE JOURNAL (2008)

Article Engineering, Geological

Drilling response of drag bits: Theory and experiment

Emmanuel Detournay et al.

INTERNATIONAL JOURNAL OF ROCK MECHANICS AND MINING SCIENCES (2008)

Article Computer Science, Software Engineering

On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming

A Wachter et al.

MATHEMATICAL PROGRAMMING (2006)

Article Engineering, Chemical

Mathematical programming model for heat-exchanger network synthesis including detailed heat-exchanger designs. 2. Network synthesis

FT Mizutani et al.

INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH (2003)

Article Engineering, Chemical

Metamodeling approach to optimization of steady-state flowsheet simulations - Model generation

K Palmer et al.

CHEMICAL ENGINEERING RESEARCH & DESIGN (2002)