Optimal management of bio-based energy supply chains under parametric uncertainty through a data-driven decision-support framework

Medina-González, Sergio; Shokry, Ahmed; Silvente, Javier; Lupera, Gicela; Espuña, Antonio

doi:10.1016/j.cie.2018.12.008

Cited by 12 publications

(7 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Regarding the capability of the MP to manage multiple sources of uncertainty, (Charitopoulos & Dua, 2016) applied it to sustainability problems. Furthermore, recent studies successfully approve the MP combination approaches within surrogate models to promote the sustainability of industrial problems (Lupera Calahorrano, Shokry, Campanya, & Espuña, 2016;Medina-González, Shokry, Silvente, Lupera, & Espuña, 2020).…”

Section: Multi-parametric Optimization (Mp) Is a Strategy That Operat...mentioning

confidence: 99%

“…In these studies, they commonly have applied uncertainty management approaches. As a remarkable instance, a recent novel model has been proposed by (Medina-González et al, 2020) in which a multi-objective model has been applied to the bio-based energy supply chain network, subjected to multiple sources of uncertainty. However, the studies have not targeted the demand uncertainty in integrated energy/material SCs.…”

Section: Management Of Alternative Energy Resourcesmentioning

confidence: 99%

See 1 more Smart Citation

A contribution to sustainable management of integrated material/energy networks in process industries

Morakabatchiankar

View full text Add to dashboard Cite

Towards sustainability forces process industries to change their traditional patterns. Therefore, efficient retrofitting has been a significant challenge and raised several issues, motivating the process system engineering (PSE) to develop models. These models mainly aim to optimize profitability, cost reduction, energy consumption, demand satisfaction, the environmental impact associated with the production process, and social acceptance. Nevertheless, such optimization is significantly complicated if considering the presence of uncertainty and seeking compromised outcomes. This thesis aims to extend a general model to facilitate retrofitting in industrial processes and expedite optimization of the issues. Such a contribution based on developing efficient mathematical models allows coordinating many decision variables synchronizing the production and distribution tasks in terms of economic and environmental criteria. This thesis presents an overview of the retrofitting requirement towards sustainable material/energy networks, describing and analyzing the current methods, tools, and models used and identifying the most relevant open issues. The second part focused on developing current models stressing energy integration in the processing system. This part first explores how the economic performance of the network can be enhanced and environmental impacts improved simultaneously by integrating an energy generation unit into the production system. Furthermore, the network sustainability performance was explored under demand uncertainties. Additional risk indicators (including financial and environmental risk metrics) have been included to add risk management capability to the model. This part also explores the strategies that efficiently select the number of scenarios. Consequently, a novel generalized mathematical formulation that integrates equations regarding energy generation and material production decision variables is efficiently solved. The effect of uncertainty on the economic and environmental performance is analyzed by using risk analysis. Finally, the model was extended to solve multi-renewable energy generation integrated into the multi-product production process under demand uncertainty. The importance and effect of the energy/material integration over the network configuration are analyzed through sensitivity analysis. The third part of this thesis provides the conclusions and further work to be developed. El camino hacia la sostenibilidad obliga a las industrias de procesos a cambiar los patrones de trabajo tradicionales. La modernización eficiente es un importante desafío que plantea múltiples problemas, lo que ha llevado a la ingeniería de sistemas de procesos (PSE) a desarrollar modelos que intentan no solo optimizar la rentabilidad, reducir de costes o satisfacer demanda de productos y servicios, sino también afrontar el impacto ambiental asociado al proceso productivo y la aceptación social de dicho proceso, en entornos volátiles, en los que la consideración de la incertidumbre asociada al escenario de trabajo es esencial. Esta tesis tiene como objetivo ampliar y flexibilizar los modelos de gestión de cadena de suministro existentes para facilitar la retroadaptación de los procesos industriales y agilizar la optimización de los problemas de toma de decisiones en este entorno. Tal aporte está basado en el desarrollo de modelos matemáticos eficientes que permite coordinar diferentes variables de decisión, sincronizando las tareas de producción y distribución en términos de criterios económicos y ambientales. Para ello, en primer lugar se presenta una visión general del requisito de adaptación hacia redes de materiales / energía sostenibles, describiendo y analizando los métodos, herramientas y modelos actuales utilizados e identificando los problemas abiertos más relevantes. La segunda parte de esta Tesis se centra en el desarrollo de modelos que enfatizan la integración energética en el sistema de proceso. Esta parte explora primero cómo se pueden mejorar simultáneamente el desempeño económico de la red y los impactos ambientales mediante la integración de unidades de generación de energía en el sistema de producción. Además, se ha explorado el rendimiento de la sostenibilidad de la red bajo incertidumbre en la demanda. Se han incluido indicadores de riesgo adicionales (incluidas métricas de riesgo financiero y ambiental) para incorporar en el modelo la capacidad de gestión de riesgos. Esta parte también explora las estrategias que seleccionan de manera eficiente el número de escenarios a considerar. En consecuencia, se resuelve de forma eficiente una formulación matemática novedosa que generaliza e integra ecuaciones relativas a decisiones de producción energética y de materiales. El efecto de la incertidumbre sobre el desempeño económico y ambiental se analiza mediante análisis de riesgo. Finalmente, el modelo se amplía para abordar la utilización coordinada de diferentes formas de energía renovable, y su integración en el proceso de producción multi-producto bajo incertidumbre en la demanda. La importancia y el efecto de la integración de la toma de decisiones sobre la configuración de la red se discuten a través de diferentes análisis de sensibilidad. La tercera parte de esta tesis resume las conclusiones de todos estos trabajos y plantea ampliaciones y nuevas líneas de mejora que surgen a partir de los modelos y procedimientos desarrollados en esta tesis.

show abstract

Section: Multi-parametric Optimization (Mp) Is a Strategy That Operat...mentioning

confidence: 99%

Section: Management Of Alternative Energy Resourcesmentioning

confidence: 99%

A contribution to sustainable management of integrated material/energy networks in process industries

Morakabatchiankar

View full text Add to dashboard Cite

show abstract

“…In contrast, the latter two types of uncertainty sources (i.e., process-inherent and external uncertainty) may occur in a sudden and unexpected way. Hence, many methods have been developed for handling these two types of uncertainty in optimization problems, most of them can be categorized into two main approaches: proactive and reactive (Medina-González, et al, 2020). The proactive approach aims at providing conservative optimal decisions minimizing the consequences of the uncertainty and variability on the performance measure(s)…”

Section: Steady-state Optimization and Uncertainty Handlingmentioning

confidence: 99%

“…However, again, further to the complex mathematical knowledge required to develop the MPP analysis (Rivotti, et al, 2012), the availability of a dynamic discrete-time linear statespace model of the process is usually a necessity for the practical application of the explicit MPC (Pistikopoulos, et al, 2002;Kouramas, et al, 2011). This, again, may hinder the MP-MPC usage in cases where the available process dynamic FPM is highly nonlinear, high dimensional, with a complicated structure (e.g., sequential simulation models) and/or black box (Rivotti, et al, 2012;Medina-González, et al, 2020). Model approximation and order reduction techniques have been proposed (Rivotti, et al, 2012); however, this may oversimplify the processes behavior and, consequently, degrade the controller performance.…”

Section: Model Predictive Controlmentioning

confidence: 99%

“…This is typically encountered in the upper level decision making layers, (e.g., supply-chain management, planning) where "managemental-like" decisions are to be made. In these situations, a model of the enterprise, in which the units and production processes are roughly represent by linear or slightly nonlinear shortcut models (e.g., fabrication of 1 kg of product requires 2 kg of raw materials and 3 liters of fuel/energy) (Medina-González, et al, 2020). Here the evaluation of the model may be cheap, but a large number of variables and constraints exist (e.g., to express the mass and energy balances of the system/process in the form of explicit optimization constraints).…”

Section: Optimization Algorithmsmentioning

confidence: 99%

See 1 more Smart Citation

A contribution to chemical process operation support: new machine learning and surrogate models based approaches for process optimization, supervision and control

Shokry Abdelaleem Taha Zied

View full text Add to dashboard Cite

In the chemical process industry, the decision-making hierarchy is inherently model-based. The scale and complexity of the considered models (e.g., enterprise, plant or unit model) depend on the decision-making level (e.g., supply-chain management, planning, scheduling, operation) and the allowable time slot (weeks, hours, seconds) within which model simulation runs must be performed and their output is analyzed to support the decision making. The use of high-fidelity models, which include detailed physics-based description of the process, is attracting wide interests of the process engineers. Since, these First Principle Model (FPMs) are able to accurately predict the real behavior of the process, leading to realistic optimal decisions. However, their use is hindered by practical challenges as the high computational time required for their simulation and the unguaranteed reliability of their consistent convergence. The challenges become prohibitive at lower levels of the decision-making hierarchy (i.e., operation), where decisions are required online within time slots of minutes or seconds entailing lots of simulation runs using such complex and highly nonlinear FPMs. Surrogate modelling techniques are potential solution for these challenges, which relies on developing simplified, but accurate, data-driven or machine learning models using data generated by a FPM simulations, or collected from a real process. Although, there are progressive developments of surrogate-based methods in the chemical engineering area, they are concentrated in process design and steady-state optimization areas. This Thesis presents a framework for the proper and effective use of surrogate models and machine learning techniques in different phases of the process operation. The objective is to provide efficient methodologies, each supports the decision making in a specific phase of the process operation, namely; steady-state operation optimization, Model Predictive Control (MPC), multivariate system identification and multistep-ahead predictions, dynamic optimization, Fault Detection and Diagnosis (FDD) and soft-sensing. Each developed methodology is designated according to careful State-Of-Art (SOA) review that identifies the gaps and missing requirements to be covered. The SOA, identified gaps and the contributions of each methodology are summarized in Chapter 1 and detailed in the introduction of each of the following chapters. In this context, Chapter 3 presents a surrogate-based methodology for steady-state operation optimization of complex nonlinear chemical processes modelled by black-box functions. Chapter 4 proposes machine learning-based methodologies for multiparametric solution of complex operation optimization problems subjected to uncertainty. Chapter 5 presents a data-based multiparametric MPC methodology that enables simple implementations of explicit MPC for nonlinear chemical processes. Chapter 6 proposes a data-driven methodology for multivariate dynamic modelling of nonlinear chemical processes and for multistep-ahead prediction. Chapter 7 suggests a dynamic optimization methodology for solving optimal control problems of complex nonlinear processes based on data-driven dynamic models. Chapter 8 shows a hybrid methodology to improve FDD of chemical processes run under time-varying inputs based on multivariate data-driven dynamic models and classification techniques. Chapter 9 presents data-driven soft-sensing methodologies for batch processes operated under changeable initial conditions. The effectiveness of the developed methodologies is proved by comparing their performances to those of classical solution procedures existing in the SOA, via their applications to different benchmark examples and case studies. The promising results and their sound analysis allowed to publish many papers in top-ranked journals and proceedings, and to present them at several top-ranked international conferences including two Keynote presentations. En la Industria de Proceso, como en otros ámbitos, la toma de decisiones se basa en la valoración de las consecuencias de dichas decisiones a través de modelos (implícitos o explícitos). La escala y complejidad de los modelos necesarios dependen de la complejidad del proceso, del nivel jerárquico al que se toman las decisiones (p. ej.: gestión de la cadena de suministro, planificación de proceso, programación de operaciones, control) y del horizonte de tiempo considerado. El uso de modelos basados en principios básicos (First Principle Models (FPM) habitualmente permite predecir con precisión el comportamiento de un sistema y llevar así a decisiones fundamentadas y explicables. Sin embargo, su uso se ve obstaculizado por problemas prácticos, dado que en ocasiones requiere cálculos iterativos aún sin tener garantizada su convergencia a una solución factible. Estos problemas son más frecuentes a medida que se desciende en la jerarquía de toma de decisiones (p. ej.: control supervisor), especialmente si la resolución (optimización) del sistema implica muchos cálculos de simulación utilizando un FPM complejo (p. ej.: altamente no lineal, involucrando variables enteras, etc.). Una forma de superar estas dificultades consiste en aplicar técnicas basadas en “modelos subrogados” o sustitutos, construidos a partir de datos recopilados del proceso real, de datos previamente simulados (utilizando un FPM), o de una combinación de ambos. Aunque estos modelos se utilizan en muchas áreas, en el ámbito de la ingeniería química habitualmente se emplean solo para el diseño de procesos y en sistemas de optimización de estado estacionario. Esta tesis presenta un marco para el uso eficaz y eficiente de modelos subrogados, construidos mediante técnicas de aprendizaje automático, en la toma de decisiones en diferentes fases de la operación, el control y la optimización de un proceso. En este contexto, el Capítulo 3 presenta una metodología para la optimización de la operación en estado estacionario de procesos no lineales. El Capítulo 4 propone la utilización de metodologías basadas en el aprendizaje automático en problemas de optimización de operaciones sujetas a incertidumbre (optimización multiparamétrica). El Capítulo 5 extiende este planteamiento a la construcción de sistemas de control predictivo (MPC) explícito de procesos no lineales. El Capítulo 6 propone una metodología para la construcción sistemática de modelos subrogados en sistemas dinámicos no-lineales multivariable, metodología que se aplica en el capítulo 7 a la de optimización de procesos dinámicos (control óptimo de sistemas no lineales en estado no estacionario). Esta misma metodología se integra en el Capítulo 8 con técnicas de clasificación para su aplicación a la detección y diagnosis de fallos (Fault Detection and Diagnosis - FDD) de sistemas dinámicos multivariable. Finalmente, en el Capítulo 9 se presenta la aplicación de estas metodologías para el entrenamiento de sensores virtuales (“soft-sensors”) y su aplicación a procesos de producción por lotes que trabajan con condiciones iniciales cambiantes. Cada una de estas aplicaciones, y los prototipos resultantes, se han plateado después de una cuidadosa revisión de las aportaciones más recientes en estos campos, que ha permitido identificar las dificultades para la implementación de las técnicas existentes en sistemas prácticos de soporte a la toma de decisiones, y la forma de superar estas dificultades mediante la utilización de modelos alternativos, que se resumen en el Capítulo 1. La eficacia de las metodologías desarrolladas se ilustra a través del análisis de su aplicación a diferentes casos, tanto propuestos en esta Tesis como de referencia en los diferentes ámbitos de aplicación. Estos resultados han merecido su publicación en diferentes revistas científicas de primer nivel, así como su difusión a través de congresos internacionales de primer nivel, incluidas dos presentaciones de Keynote.

show abstract

Data science and big data analytics: a systematic review of methodologies used in the supply chain and logistics research

2023

View full text Add to dashboard Cite

Data science and big data analytics (DS &BDA) methodologies and tools are used extensively in supply chains and logistics (SC &L). However, the existing insights are scattered over different literature sources and there is a lack of a structured and unbiased review methodology to systematise DS &BDA application areas in the SC &L comprehensively covering efficiency, resilience and sustainability paradigms. In this study, we first propose an unique systematic review methodology for the field of DS &BDA in SC &L. Second, we use the methodology proposed for a systematic literature review on DS &BDA techniques in the SC &L fields aiming at classifying the existing DS &BDA models/techniques employed, structuring their practical application areas, identifying the research gaps and potential future research directions. We analyse 364 publications which use a variety of DS &BDA-driven modelling methods for SC &L processes across different decision-making levels. Our analysis is triangulated across efficiency, resilience, and sustainability perspectives. The developed review methodology and proposed novel classifications and categorisations can be used by researchers and practitioners alike for a structured analysis and applications of DS &BDA in SC &L.

show abstract

Optimal management of bio-based energy supply chains under parametric uncertainty through a data-driven decision-support framework

Cited by 12 publications

References 35 publications

A contribution to sustainable management of integrated material/energy networks in process industries

A contribution to sustainable management of integrated material/energy networks in process industries

A contribution to chemical process operation support: new machine learning and surrogate models based approaches for process optimization, supervision and control

Data science and big data analytics: a systematic review of methodologies used in the supply chain and logistics research

Contact Info

Product

Resources

About