arXiv subjects

Alexander Mitsos

Publications and source records attributed to Alexander Mitsos.

63 records · Page 4Linked to original sources

arXiv2021

COMANDO: A Next-Generation Open-Source Framework for Energy Systems Optimization

Existing open-source modeling frameworks dedicated to energy systems optimization typically utilize (mixed-integer) linear programming ((MI)LP) formulations, which lack modeling freedom for technical system design and operation. We present COMANDO, an open-source Python package for component-oriented modeling and optimization for nonlinear design and operation of integrated energy systems. COMANDO allows to assemble system models from component models including nonlinear, dynamic and discrete characteristics. Based on a single system model, different deterministic and stochastic problem formulations can be obtained by varying objective function and underlying data, and by applying automatic or manual reformulations. The flexible open-source implementation allows for the integration of customized routines required to solve challenging problems, e.g., initialization, problem decomposition, or sequential solution strategies. We demonstrate features of COMANDO via case studies, including automated linearization, dynamic optimization, stochastic programming, and the use of nonlinear artificial neural networks as surrogate models in a reduced-space formulation for deterministic global optimization.

math.OC↗

arXiv2021

Using Gaussian Processes to Design Dynamic Experiments for Black-Box Model Discrimination under Uncertainty

Diverse domains of science and engineering use parameterised mechanistic models. Engineers and scientists can often hypothesise several rival models to explain a specific process or phenomenon. Consider a model discrimination setting where we wish to find the best mechanistic, dynamic model candidate and the best model parameter estimates. Typically, several rival mechanistic models can explain the available data, so design of dynamic experiments for model discrimination helps optimally collect additional data by finding experimental settings that maximise model prediction divergence. We argue there are two main approaches in the literature for solving the optimal design problem: (i) the analytical approach, using linear and Gaussian approximations to find closed-form expressions for the design objective, and (ii) the data-driven approach, which often relies on computationally intensive Monte Carlo techniques. Olofsson et al. (ICML 35, 2018) introduced Gaussian process (GP) surrogate models to hybridise the analytical and data-driven approaches, which allowed for computationally efficient design of experiments for discriminating between black-box models. In this study, we demonstrate that we can extend existing methods for optimal design of dynamic experiments to incorporate a wider range of problem uncertainty. We also extend the Olofsson et al. (2018) method of using GP surrogate models for discriminating between dynamic black-box models. We evaluate our approach on a well-known case study from literature, and explore the consequences of using GP surrogates to approximate gradient-based methods.

cs.LG↗

arXiv2021

Optimal operating policies for organic Rankine cycles for waste heat recovery under transient conditions

Waste heat recovery for trucks via organic Rankine cycle is a promising technology to reduce fuel consumption and emissions. As the vehicles are operated in street traffic, the heat source is subject to strong fluctuations. Consequently, such disturbances have to be considered to enable safe and efficient operation. Herein, we find optimal operating policies for several representative scenarios by means of dynamic optimization and discuss the implications on control strategy design. First, we optimize operation of a typical driving cycle with data from a test rig. Results indicate that operating the cycle at minimal superheat is an appropriate operating policy. Second, we consider a scenario where the permissible expander power is temporarily limited, which is realistic in street traffic. In this case, an operating policy with flexible superheat can reduce the losses associated with operation at minimal superheat by up to 53 % in the considered scenario. As the duration of power limitation increases, other constraints might become active which results in part of the exhaust gas being bypassed, hence reduced savings.

math.OC↗

arXiv2020

Adaptive Learning of Hybrid Models for Nonlinear Model Predictive Control of Distillation Columns

Nonlinear model predictive control (NMPC) requires accurate and computationally efficient plant models. Our previous work has shown that the classical compartmentalization model reduction approach for distillation columns can be enhanced by replacing parts of the system of equations by artificial neural networks (ANNs) trained on offline solved solutions to improve computational performance. In real-life applications, the absence of a high-fidelity model for data generation can, however, prevent the deployment of this approach. Therefore, we propose a method that utilizes solely plant measurement data, starting from a small initial data set and then continuously adapting to newly measured data. The efficacy of the proposed approach is examined in silico for a distillation column from literature. To this end, we first adjust our reduced hybrid mechanistic/data-driven modeling approach that originally builds on compartmentalization to a stage-aggregation procedure, tailoring it for the application within the adaptive learning framework. Second, we apply an adaptive learning algorithm that trains the ANNs replacing the stationary stage-to-stage calculations on newly available data. We apply the adaptive learning of the hybrid model within a regulatory NMPC framework and conduct closed-loop simulations. We demonstrate that by using the proposed method, the control performance can be steadily improved over time compared to a non-adaptive approach while being real-time applicable. Moreover, we show that the performance when using either a model trained on excessive amounts of offline generated data or the original high-fidelity model can be approached in the limit.

math.OC↗

arXiv2020

Obey validity limits of data-driven models

Data-driven models are becoming increasingly popular in engineering, on their own or in combination with mechanistic models. Commonly, the trained models are subsequently used in model-based optimization of design and/or operation of processes. Thus, it is critical to ensure that data-driven models are not evaluated outside their validity domain during process optimization. We propose a method to learn this validity domain and encode it as constraints in process optimization. We first perform a topological data analysis using persistent homology identifying potential holes or separated clusters in the training data. In case clusters or holes are identified, we train a one-class classifier, i.e., a one-class support vector machine, on the training data domain and encode it as constraints in the subsequent process optimization. Otherwise, we construct the convex hull of the data and encode it as constraints. We finally perform deterministic global process optimization with the data-driven models subject to their respective validity constraints. To ensure computational tractability, we develop a reduced-space formulation for trained one-class support vector machines and show that our formulation outperforms common full-space formulations by a factor of over 3,000, making it a viable tool for engineering applications. The method is ready-to-use and available open-source as part of our MeLOn toolbox (https://git.rwth-aachen.de/avt.svt/public/MeLOn).

math.OC↗

arXiv2020

Global Optimization of Gaussian processes

Gaussian processes~(Kriging) are interpolating data-driven models that are frequently applied in various disciplines. Often, Gaussian processes are trained on datasets and are subsequently embedded as surrogate models in optimization problems. These optimization problems are nonconvex and global optimization is desired. However, previous literature observed computational burdens limiting deterministic global optimization to Gaussian processes trained on few data points. We propose a reduced-space formulation for deterministic global optimization with trained Gaussian processes embedded. For optimization, the branch-and-bound solver branches only on the degrees of freedom and McCormick relaxations are propagated through explicit Gaussian process models. The approach also leads to significantly smaller and computationally cheaper subproblems for lower and upper bounding. To further accelerate convergence, we derive envelopes of common covariance functions for GPs and tight relaxations of acquisition functions used in Bayesian optimization including expected improvement, probability of improvement, and lower confidence bound. In total, we reduce computational time by orders of magnitude compared to state-of-the-art methods, thus overcoming previous computational burdens. We demonstrate the performance and scaling of the proposed method and apply it to Bayesian optimization with global optimization of the acquisition function and chance-constrained programming. The Gaussian process models, acquisition functions, and training scripts are available open-source within the "MeLOn - Machine Learning Models for Optimization" toolbox~(https://git.rwth-aachen.de/avt.svt/public/MeLOn).

math.OC↗

arXiv2018

Global Deterministic Optimization with Artificial Neural Networks Embedded

Artificial neural networks (ANNs) are used in various applications for data-driven black-box modeling and subsequent optimization. Herein, we present an efficient method for deterministic global optimization of ANN embedded optimization problems. The proposed method is based on relaxations of algorithms using McCormick relaxations in a reduced-space [\textit{SIOPT}, 20 (2009), pp. 573-601] including the convex and concave envelopes of the nonlinear activation function of ANNs. The optimization problem is solved using our in-house global deterministic solver MAiNGO. The performance of the proposed method is shown in four optimization examples: an illustrative function, a fermentation process, a compressor plant and a chemical process optimization. The results show that computational solution time is favorable compared to the global general-purpose optimization solver BARON.

math.OC↗

arXiv2017

Tighter McCormick Relaxations through Subgradient Propagation

Tight convex and concave relaxations are of high importance in the field of deterministic global optimization. We present a heuristic to tighten relaxations obtained by the McCormick technique. We use the McCormick subgradient propagation (Mitsos et al., SIAM J. Optim., 2009) to construct simple affine under- and overestimators of each factor of the original factorable function. Then, we minimize and maximize these affine relaxations in order to obtain possibly improved range bounds for every factor resulting in possibly tighter final McCormick relaxations. We discuss the heuristic and its limitations, in particular the lack of guarantee for improvement. Subsequently, we provide numerical results for benchmark cases found in the COCONUT library and case studies presented in previous works and discuss computational efficiency. We see that the presented heuristic provides a significant improvement in tightness and decrease in computational time in many cases.

math.OC↗

arXiv2016

Optimal Deterministic Algorithm Generation

A formulation for the automated generation of algorithms via mathematical programming (optimization) is proposed. The formulation is based on the concept of optimizing within a parameterized family of algorithms, or equivalently a family of functions describing the algorithmic steps. The optimization variables are the parameters -within this family of algorithms- that encode algorithm design: the computational steps of which the selected algorithms consists. The objective function of the optimization problem encodes the merit function of the algorithm, e.g., the computational cost (possibly also including a cost component for memory requirements) of the algorithm execution. The constraints of the optimization problem ensure convergence of the algorithm, i.e., solution of the problem at hand. The formulation is described prototypically for algorithms used in solving nonlinear equations and in performing unconstrained optimization; the parametrized algorithm family considered is that of monomials in function and derivative evaluation (including negative powers). A prototype implementation in GAMS is developed along with illustrative results demonstrating cases for which well-known algorithms are shown to be optimal. The formulation is a mixed-integer nonlinear program (MINLP). To overcome the multimodality arising from nonconvexity in the optimization problem, a combination of brute force and general-purpose deterministic global algorithms is employed to guarantee the optimality of the algorithm devised. We then discuss several directions towards which this methodology can be extended, their scope and limitations.

math.OC↗