The focus of this thesis is on building models by utilizing process information: from data, from our knowledge of physics, or both. The closer the model approximates reality, the better is the expected performance in forecasting, soft-sensing, process monitoring, optimization and advanced process control. In the domain of batch and continuous manufacturing, quality models can help in ensuring tightly controlled product quality, having safe and reliable operating conditions and reducing production/operation costs. To this end, first a parallel grey box model was built which makes use of a mechanistic model, and a subspace identification model for modeling a batch poly methyl methacrylate (PMMA) polymerisation process. The efficacy of such a parallel hybrid model in the context of a control problem was illustrated thereafter for reducing the volume of fines. Real-time implementation of models in many cases demand the model to be tractable and simple enough, and thus the parallel hybrid model was next adapted to have a linear representation, and then used for control computations. While the parallel hybrid modelling strategy shows great advantages in many applications, there can be other avenues of using fundamental process knowledge in conjunction with historical data. In one such approach, a unique way of adding mechanistic knowledge to improve the estimation ability of PLS models was proposed. The predictor matrix of PLS was augmented with additional trajectory information coming strategically from a mechanistic model. This augmented model was used as a soft-sensor to estimate batch end quality for a seeded batch crystallizer process. In a collaborative work with an industrial partner focusing on estimating important variables of a hydroprocessing unit, an operational data based input-output model was chosen as the right fit in the absence of available mechanistic knowledge. The usefulness of linear dynamic modeling tools for such applications was demonstrated. / Thesis / Doctor of Philosophy (PhD)

La industria química ha experimentado en las últimas décadas un aumento en la competencia por la cual las empresas se ven obligadas a adaptarse a un mercado cambiante y cada vez más exigente. Aunque la globalización ha abierto nuevos mercados, ha incrementado también el número de competidores, de tal manera que sólo las empresas que usen las plantas más integradas y eficientes podrán mantenerse en el negocio. En este contexto global, el principal propósito de esta tesis es desarrollar métodos que exploten la flexibilidad de los procesos, con el objetivo de aumentar la eficiencia de las plantas y asegurar los requerimientos de seguridad y calidad de los productos. Esta tesis contribuye a la optimización y a la gestión de la producción desde pequeñas plantas que usen procesos discontinuos hasta grandes plantas de procesado continuo.En primer lugar, esta tesis trata la gestión de los procesos continuos en los que suelen fabricar productos muy similares a gran escala. La gran ventaja de los procesos continuos es que pueden conseguir mayor consistencia en la calidad de los productos y que pueden aprovechar las economías de escala que reducen los costes y residuos. Sin embargo, la industria química para mantenerse competitiva necesita adaptar continuamente sus procesos a las condiciones del mercado y de operación. El sistema de control supervisor presentado en esta parte de la tesis disminuye el tiempo de reacción frente a incidentes en los procesos continuos y re-optimiza la producción en tiempo real, si existe posibilidad de mejora.A continuación, esta tesis trata la gestión de los procesos semicontinuos que permiten una operación más flexible y personalizada. Los procesos semicontinuos operan con puestas en marcha y paradas periódicas para acomodar las frecuentes transiciones entre diferentes productos. Esta tesis presenta un nuevo concepto de fabricación flexible que permite programar perfiles variables de velocidad de producción dentro de cada campaña de producción.La mayor parte del trabajo de investigación de esta tesis se dedica a la planificación de la producción en los procesos discontinuos por lotes, utilizados principalmente en la producción de productos químicos con alto valor añadido. Estos procesos ofrecen varias ventajas respecto a los procesos continuos y semicontinuos debido a la mayor flexibilidad para acomodar diversos productos, diferentes capacidades de producción, y la posibilidad de realizar operaciones completamente diferentes en los mismos equipos. Sin embargo, la obtención del plan de producción óptimo usando se complica al aumentar la complejidad de la planta y/o el número de lotes a planificar. La simplificación de considerar tiempos de transferencia despreciables es generalmente aceptada en la literatura para evitar la complejidad del manejo de las operaciones de transferencia. En cambio, esta tesis pretende resaltar el papel crítico que juegan las operaciones de transferencia en la sincronización de tareas, y en la consiguiente determinación de planes de producción factibles.Siguiendo con los procesos por lotes, esta tesis demuestra que el uso del concepto de recetas flexibles mejora la operación de los procesos en ambientes de producción con mucha incertidumbre. La flexibilidad de las receta se considera como una oportunidad adicional, tanto para la planificación de la producción reactiva como preactiva, reduciendo el riesgo de llegar a resultados económicamente desfavorables.Finalmente, esta tesis presenta las plantas discontinuas sin tuberías como una alternativas a las plantas por lotes clásicas. En la búsqueda de formas más competitivas y efectivas de producción, la flexibilidad para producir un elevado número de productos en plantas por lotes es limitada debido a la necesidad de equipos fijos conectados por tuberías y frecuentes tareas de limpieza. Las plantas sin tuberías presentan una mayor flexibilidad ya que el material se transfiere entre estaciones de procesamiento usando equipos que se mueven dentro de la planta. El trabajo presentado en esta parte de la tesis contribuye a la mejora en la gestión de este tipo de plantas proponiendo una formulación más eficiente a las encontradas en la literatura que resuelve el problema de la planificación de la producción.En resumen, esta tesis desarrolla nuevas estrategias de modelado y métodos de resolución encaminados al soporte de la toma de decisiones que explotan la flexibilidad intrínseca de los procesos químicos. Las principales ventajas de cada una de las contribuciones de esta tesis se demuestran mediante su aplicación a diferentes casos de estudio. / The chemical industry has become increasingly competitive over the past decades. Companies are required to adapt to changing market conditions and meet stricter product specifications. While globalization has opened new markets for the chemical industry, it has also increased the competitor pool, giving an advantage to companies with more efficient and highly integrated plants.In this context, the main aim of this thesis is to demonstrate new concepts and computational methods that exploit process flexibility to enhance plant profitability under transient operating conditions. These methods ensure that safety and product quality requirements are consistently met. This thesis makes contributions to the optimization and management of production in plants ranging from small batch plants to large capacity continuous processes.First, this thesis addresses the management of continuous processes, in which similar products are mass produced. Continuous processes can achieve the highest consistency and product quality by taking advantage of economies of scale and reduced manufacturing costs and waste. However, in order to remain competitive in the market, plants are required to dynamically adapt their processes to fit the continuously changing market and operating conditions. The supervisory control system presented in this part of the thesis decreases the system reaction time to incidences and re-optimizes the production in real time if the opportunity for improved performance exists.Next, this thesis addresses the management of semicontinuous processes, which allow more customized and flexible operation. Semicontinuous processes run with periodic start-ups and shutdowns to accommodate frequent product transitions. This thesis proposes an optimization model that creates improved production schedules by introducing a new concept of flexible manufacturing that allows production rate profiles to be programmed within each operation campaign.The major part of the research work of this thesis deals with the operational management of batch processes, which are mainly used for the production of high value-added chemicals. Batch processing offers the advantage of increased flexibility in product variety, production volume, and the assortment of operations that can be processed by a particular piece of equipment. However, the trade-off is that production scheduling is significantly complicated by the large number of batches involved with different production paths. In order to avoid the complexity of managing transfer operations, the assumption of negligible transfer times is generally accepted in batch scheduling. Conversely, this thesis highlights the critical role that transfer operations play in the synchronization of tasks and in determining the feasibility of production schedules.Continuing to focus on batch plant operation, this thesis demonstrates that the use of the concept of flexible recipes enhances the operation batch plants within an uncertain environment. Recipe flexibility is considered as an additional opportunity for reactive scheduling as well as a proactive way to reduce the risk of meeting unfavorable scenarios.Finally, this thesis examines pipeless plants as an alternative to batch plants. In the search for more competitive and effective ways of production, flexibility of batch plants for producing a large number of products is limited due to the need for equipment, piping and frequent cleaning tasks. Pipeless plants have enhanced flexibility over batch plants, because the material is moved along its production path through moveable vessels. This part of the thesis contributes to the optimization of the management of pipeless plants by proposing an alternative formulation for solving short-term scheduling problems.In summary, this thesis provides novel modeling approaches and solution methods aimed at supporting the decision-making process in plant production scheduling which exploit the existing flexibility in chemical processes. The main advantages of each contribution are highlighted through case studies.

Process control and monitoring is a common problem in high value added chemical manufacturing industries where batch processes are used to produce wide range of products on the same piece of equipment. This results in frequent adjustments on control and monitoring schemes. A chemical toner manufacturing process is representative of an industrial case which is used in this thesis. Process control and monitoring problem of batch processes have been researched, mostly through the simulation, and published in the past . However, the concept of applying the subject to chemical toner manufacturing process or to use a single indicator for multiple pieces of equipment have never been visited previously. In the case study of this research, there are many different factors that may affect the final quality of the products including reactor batch temperature, jacket temperature, impeller speed, rate of the addition of material to the reactor, or process variable associated with the pre-weight tank. One of the challenging tasks for engineers is monitoring of these process variables and to make necessary adjustments during the progression of a batch and change controls strategy of future batches upon completion of an existing batch. Another objective of the proposed research is the establishment of the operational boundaries to monitor the process through the usage of process trajectories of the history of the past successful batches. In this research, process measurements and product quality values of the past successful batches were collected and projected into matrix of data; and preprocessed through time alignment, centering, and scaling. Then the preprocessed data was projected into lower dimensions (latent variables) to produce latent variables and their trajectories during successful batches. Following the identification of latent variables, an empirical model was built through a 4-fold cross validation that can represent the operation of a successful batch. The behavior of two abnormal batches, batch 517 and 629, is then compared to the model by testing its statistical properties. Once the abnormal batches were flagged, their data set were folded back to original dimension to form a localization path for the time of abnormality and process variables that contributed to the abnormality. In each case the process measurement were used to establish operational boundaries on the latent variable space.

No description available.

Process control and monitoring is a common problem in high value added chemical manufacturing industries where batch processes are used to produce wide range of products on the same piece of equipment. This results in frequent adjustments on control and monitoring schemes. A chemical toner manufacturing process is representative of an industrial case which is used in this thesis. Process control and monitoring problem of batch processes have been researched, mostly through the simulation, and published in the past . However, the concept of applying the subject to chemical toner manufacturing process or to use a single indicator for multiple pieces of equipment have never been visited previously. In the case study of this research, there are many different factors that may affect the final quality of the products including reactor batch temperature, jacket temperature, impeller speed, rate of the addition of material to the reactor, or process variable associated with the pre-weight tank. One of the challenging tasks for engineers is monitoring of these process variables and to make necessary adjustments during the progression of a batch and change controls strategy of future batches upon completion of an existing batch. Another objective of the proposed research is the establishment of the operational boundaries to monitor the process through the usage of process trajectories of the history of the past successful batches. In this research, process measurements and product quality values of the past successful batches were collected and projected into matrix of data; and preprocessed through time alignment, centering, and scaling. Then the preprocessed data was projected into lower dimensions (latent variables) to produce latent variables and their trajectories during successful batches. Following the identification of latent variables, an empirical model was built through a 4-fold cross validation that can represent the operation of a successful batch. The behavior of two abnormal batches, batch 517 and 629, is then compared to the model by testing its statistical properties. Once the abnormal batches were flagged, their data set were folded back to original dimension to form a localization path for the time of abnormality and process variables that contributed to the abnormality. In each case the process measurement were used to establish operational boundaries on the latent variable space.

Processos industriais que ocorrem em bateladas são empregados com freqüência na produção de alguns itens. Tais processos disponibilizam uma estrutura de dados bastante peculiar e, diante disso, existe um crescente interesse no desenvolvimento de cartas de controle multivariadas mais apropriadas para seu monitoramento. Destaca-se aqui uma abordagem recente que utiliza cartas de controle baseadas no método Statis. O Statis constitui-se numa técnica exploratória que permite avaliar similaridade entre matrizes de dados. Entretanto, esta técnica avalia a similaridade no contexto linear, isto é, investiga estruturas de correlação lineares nos dados. Propõe-se nesta tese a utilização de cartas de controle baseadas no Statis em conjunto com um kernel para monitoramento de processos com presença de nãolinearidades fortes. Através dos kernels, definem-se funções não lineares dos dados para melhor representação da estrutura a ser caracterizada pelo método Statis. Esta nova abordagem, denominada Kernel-Statis, é desenvolvida e avaliada utilizando dados de um processo simulado. / Industrial batch processes are widely used in the production of some items. Such processes provide a peculiar data structure; therefore, there is a growing interest in the development of customized multivariate control charts for their monitoring. We investigate a recent approach that uses control charts based on the Statis method. Statis is an exploratory technique for measuring similarities between data matrices. However, the technique only assesses similarities in a linear context, i.e. investigating structures of linear correlation in the data. In this thesis we propose control charts based on the Statis method in conjunction with a kernel for monitoring processes in the presence of strong non-linearities. Through the kernels we define non-linear functions of data for better representing the structure to be characterized by the Statis method. The new approach, named Kernel-Statis, is developed and illustrated using simulated data.

Processos industriais que ocorrem em bateladas são empregados com freqüência na produção de alguns itens. Tais processos disponibilizam uma estrutura de dados bastante peculiar e, diante disso, existe um crescente interesse no desenvolvimento de cartas de controle multivariadas mais apropriadas para seu monitoramento. Destaca-se aqui uma abordagem recente que utiliza cartas de controle baseadas no método Statis. O Statis constitui-se numa técnica exploratória que permite avaliar similaridade entre matrizes de dados. Entretanto, esta técnica avalia a similaridade no contexto linear, isto é, investiga estruturas de correlação lineares nos dados. Propõe-se nesta tese a utilização de cartas de controle baseadas no Statis em conjunto com um kernel para monitoramento de processos com presença de nãolinearidades fortes. Através dos kernels, definem-se funções não lineares dos dados para melhor representação da estrutura a ser caracterizada pelo método Statis. Esta nova abordagem, denominada Kernel-Statis, é desenvolvida e avaliada utilizando dados de um processo simulado. / Industrial batch processes are widely used in the production of some items. Such processes provide a peculiar data structure; therefore, there is a growing interest in the development of customized multivariate control charts for their monitoring. We investigate a recent approach that uses control charts based on the Statis method. Statis is an exploratory technique for measuring similarities between data matrices. However, the technique only assesses similarities in a linear context, i.e. investigating structures of linear correlation in the data. In this thesis we propose control charts based on the Statis method in conjunction with a kernel for monitoring processes in the presence of strong non-linearities. Through the kernels we define non-linear functions of data for better representing the structure to be characterized by the Statis method. The new approach, named Kernel-Statis, is developed and illustrated using simulated data.

Processos industriais que ocorrem em bateladas são empregados com freqüência na produção de alguns itens. Tais processos disponibilizam uma estrutura de dados bastante peculiar e, diante disso, existe um crescente interesse no desenvolvimento de cartas de controle multivariadas mais apropriadas para seu monitoramento. Destaca-se aqui uma abordagem recente que utiliza cartas de controle baseadas no método Statis. O Statis constitui-se numa técnica exploratória que permite avaliar similaridade entre matrizes de dados. Entretanto, esta técnica avalia a similaridade no contexto linear, isto é, investiga estruturas de correlação lineares nos dados. Propõe-se nesta tese a utilização de cartas de controle baseadas no Statis em conjunto com um kernel para monitoramento de processos com presença de nãolinearidades fortes. Através dos kernels, definem-se funções não lineares dos dados para melhor representação da estrutura a ser caracterizada pelo método Statis. Esta nova abordagem, denominada Kernel-Statis, é desenvolvida e avaliada utilizando dados de um processo simulado. / Industrial batch processes are widely used in the production of some items. Such processes provide a peculiar data structure; therefore, there is a growing interest in the development of customized multivariate control charts for their monitoring. We investigate a recent approach that uses control charts based on the Statis method. Statis is an exploratory technique for measuring similarities between data matrices. However, the technique only assesses similarities in a linear context, i.e. investigating structures of linear correlation in the data. In this thesis we propose control charts based on the Statis method in conjunction with a kernel for monitoring processes in the presence of strong non-linearities. Through the kernels we define non-linear functions of data for better representing the structure to be characterized by the Statis method. The new approach, named Kernel-Statis, is developed and illustrated using simulated data.

This Master's thesis focuses on improving the understanding and monitoring of the Hot Isostatic Pressing (HIP) sintering process used by Seco Tools AB to manufacture cemented carbides for cutting tools. While essential for producing cutting tools with superior hardness and toughness the HIP sintering process introduces a complex relationship between the selected process parameters and the achieved materials properties. With the goal of establishing batch process monitoring capabilities, this master thesis employs Multivariate Statistical Process Control (MSPC) strategies through the creation of Batch Evolution Models (BEMs) and Batch Level Models (BLMs) to monitor, predict end-product quality, and analyze the batch production HIP sintering process.  The developed models effectively account for significant variation in the HIP sintering process and demonstrate potential in identifying deviant batches. Enhancements to the models' performance are achieved through the incorporation of preprocessing, phase-specific variable selection, and specialized model training. These proposed enhancements yield discernible improvements, as evidenced by enhanced model fit and other statistical metrics.  Challenges arise when the models are tested with real-time data due to progressive changes in some tracked process variables. Block-scaling is applied to restore the real-time monitoring capabilities, but also introduces additional complexity to the models. In addition, this master thesis highlights the need for continuous and regular maintenance of these models to ensure real-time monitoring and anomaly detection capabilities. The models demonstrate varied effectiveness in predicting final product quality. For instance, they exhibit some potential in predicting Magnetic Saturation (MS), but their ability to predict Magnetic Coercivity (HC) seems nonexistent. Despite attempts to improve the predictive abilities the models are still not able to confidently predict these metrics. The master’s thesis highlights variability in powder contents and access to data of known quality nonconformities as potential areas for improving the predictive models.

Predictive maintenance is a topic that has been researched and theorized for decades. With the advent of Industry 4.0 and greater technological capabilities in the form of advanced AI, the concept of predicting the need for maintenance in a system or its components is quickly becoming more of a reality for complex processes. The possibility of estimating remaining useful lifetimes would help businesses with maintenance scheduling to avoid unnecessary maintenance actions, but also process failures. Predicting when maintenance is needed would ensure system or component reparation or replacement before they are degraded to the point of negative product quality impact and production losses. While there are many studies on predictive maintenance and how it can be implemented and used in continuous processes, the research on complex batch processes is minimal. Therefore, this thesis aims to construct a framework based on literature for implementing predictive maintenance in batch processes. Parts of the framework are then applied and validated on a complex batch process in the form of sintering at Seco Tools AB. Recommendations are given on how to implement predictive maintenance and what is required in the company’s specific case based on the sintering process’ agreement with the framework. The framework consists of two main parts with several underlying requirements: Data collection and pre-processing and Predictive models. Evaluating the sintering process based on these requirements reveals that many parts of the framework are already in place or possible to implement, while other areas are lacking. There is a need for data cleaning and data related to component health and issues, while the amount of specific parameter data on temperatures, pressures, and similar variables is large. It is possible to predict these parameters accurately through building, training, and validating linear regression models. These predictions can be used as inputs in future models to predict the Remaining Useful Life (RUL) of components or the entire system. Due to the inherent complexity of the sintering process and similar industrial manufacturing processes, which involve numerous interdependent variables affecting product quality and component health, it is imperative to develop machine learning models and neural networks for future predictive maintenance algorithms. Moreover, as highlighted in this thesis, the attainment of predictive maintenance in an industrial environment necessitates prioritizing augmented data collection on component conditions, investing in hardware to bolster computational power, and acquiring the essential expertise to design and implement tailored predictive maintenance algorithms for dedicated manufacturing processes.