El control predictivo basado en modelos o Model Predictive Control (MPC), no hace referencia al diseño concreto de un controlador sino más bien a un conjunto de ideas o características para el desarrollo de estrategias de control que, aplicadas en un mayor o menor grado, dan lugar a diferentes tipos de controladores con estructuras similares. El MPC es una de las técnicas de control que más se ha desarrollado en los ámbitos académico e industrial en las últimas décadas debido sobre todo a su simplicidad y eficiencia. Sin embargo, no es fácil relacionar los parámetros de ajuste del controlador y las prestaciones del bucle cerrado. En este sentido, es importante diseñar algoritmos de control predictivo que garanticen la estabilidad nominal del bucle cerrado, con tiempos de cálculo pequeños y con un significado claro de sus parámetros sobre las prestaciones del sistema o sobre el esfuerzo de control. La aportación fundamental de esta tesis está relacionada con la definición de un nuevo tipo de controlador predictivo, el PC-GPC, versión modificada de un GPC estándar. En este controlador se ha sustituido el factor de ponderación de la acción de control por un nuevo parámetro denominado número de componentes principales (NPC). La relación entre el nuevo parámetro (NPC) y algunos indicadores numéricos, como la norma del vector de acciones de control o el número de condición de la matriz dinámica G, hacen que su elección esté basada en criterios menos subjetivos que la ponderación de las acciones de control. Además, se ha analizado este tipo de controlador tanto en el ámbito de procesos SISO como MIMO, así como sus características de robustez y estabilidad. Por otro lado, se ha deducido un método de cálculo de un controlador PC-GPC para garantizar la estabilidad nominal de bucle cerrado, cuando el modelo conocido es exacto. / Sanchís Saez, J. (2002). GPC mediante descomposición en valores singulares (SVD). Análisis de componentes principales (PCA) y criterios de selección [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/4924

yes / Batch reactive distillation is aimed at achieving a high purity product, therefore, there is a great deal to find an optimal operating condition and effective control strategy to obtain maximum of the high purity product. An off-line dynamic optimization is first performed with an objective function to provide optimal product composition for the batch reactive distillation: maximum productivity. An inferential state estimator (an extended Kalman filter, EKF) based on simplified mathematical models and on-line temperature measurements, is incorporated to estimate the compositions in the reflux drum and the reboiler. Model Predictive Control (MPC) has been implemented to provide tracking of the desired product compositions subject to simplified model equations. Simulation results demonstrate that the inferential state estimation can provide good estimates of compositions. Therefore, the control performance of the MPC with the inferential state is better than that of PID. In addition, in the presence of unknown/uncertain parameters (forward reaction rate constant), the estimator is still able to provide accurate concentrations. As a result, the MPC with the inferential state is still robust and applicable in real plants.

No / A novel threefold optimization algorithm is proposed to simultaneously solve the nonlinear model predictive control and dynamic real-time optimization for batch processes while optimizing the batch operation time. Object-oriented programming and parallel computing are exploited to make the algorithm effective to handle industrial cases. A well-known literature case is selected to validate the algorithm.

Today's Cyber-Physical Systems (CPS) are typically distributed over several computing nodes communicating by way of shared buses such as Controller Area Network (CAN). Their control performance gets degraded due to variable delays (jitters) incurred by messages on the shared CAN bus due to contention and network overhead. This work presents a novel online delay prediction approach that predicts the message delay at runtime based on real-time traffic information on CAN. It leverages the proposed method to improve control quality, by compensating for the message delay using the Model Predictive Control (MPC) algorithm in designing the controller. By simulating an automotive Cruise Control system and a DC Motor plant in a CAN environment, it goes on to demonstrate that the delay prediction is accurate, and that the MPC design which takes the message delay into consideration, performs considerably better. It also implements the proposed method on an 8-bit 16MHz ATmega328P microcontroller and measures the execution time overhead. The results clearly indicate that the method is computationally feasible for online usage. / Master of Science / In today’s world, most complicated systems such as automobiles employ a decentralized modular architecture with several nodes communicating with each other over a shared medium. The Controller Area Network (CAN) is the most widely accepted standard as far as automobiles are concerned. The performance of such systems gets degraded due to the variable delays (jitters) incurred by messages on the CAN. These delays can be caused by messages of higher importance delaying bus access to the messages of lower importance, or due to other network related issues. This work presents a novel approach that predicts the message delays in real-time based on the traffic information on CAN. This approach leverages the proposed method to improve the control quality by compensating for the message delay using an advanced controller algorithm called Model Predictive Control (MPC). By simulating an automotive Cruise Control system and a DC motor plant in a CAN environment, this work goes on to demonstrate that the delay prediction is accurate, and that the MPC design which takes the message delay into consideration, performs considerably better. It also implements the proposed approach on a low end microcontroller (8bit, 16MHz ATmega328P) and measures the time taken for predicting the delay for each message (execution overhead). The obtained results clearly indicate that the method is computationally feasible for use in a real-time scenario.

京都大学 / 新制・課程博士 / 博士(情報学) / 甲第25437号 / 情博第875号 / 京都大学大学院情報学研究科数理工学専攻 / (主査)准教授 加嶋 健司, 教授 山下 信雄, 教授 大塚 敏之 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DGAM

<p dir="ltr">The automotive industry has incorporated controls since the 1970s, starting with the pioneering application of an air-to-fuel ratio feedback control carburetor. Over time, significant advancements have been made in control strategies to meet industry standards for reduced fuel consumption, exhaust emissions, and enhanced safety. This thesis focuses on the implementation of advanced control strategies in heavy-duty diesel powertrains and their advantages over traditional control methods commonly employed in the automotive industry.</p><p dir="ltr">The initial part of the thesis demonstrates the utilization of model predictive control (MPC) to generate an optimized velocity profile for class 8 trucks. These velocity profiles are designed to minimize fuel consumption along a given route with known grade conditions, while adhering to the time constraints comparable to those of standard commercial cruise controllers. This methodology is further expanded to include the platooning of two trucks, with the rear truck following a desired gap (variable or fixed), resulting in additional fuel savings throughout the designated route. Through collaborative efforts involving Cummins, Peloton Technology, and Purdue University, these control strategies were implemented and validated through simulation, hardware-in-the-loop testing, and ultimately, in demonstration vehicles.</p><p dir="ltr">MPC is highly effective for vehicle-level controls due to the accurate plant model used for optimization. However, when it comes to engine controls, the plant model becomes highly nonlinear and loses accuracy when linearized [20]. To address this issue, robust control techniques are introduced to account for the inherent inaccuracies in the plant model, which can be represented as uncertainties.</p><p dir="ltr">The second study showcases the application of robust controllers in diesel engine operations, focusing on a 4.5L John Deere diesel engine equipped with an electrified intake boosting system. The intake boosting system is selectively activated during transient operations to mitigate drops in the air-to-fuel ratio (AFR), which can result in smoke emissions. Initially, a two-degree-of-freedom robustsingle-input single-output (SISO) eBooster controller is synthesized to control the eBooster during load transients. Although the robust SISO controller yields improvements, the eBooster alone does not encompass all factors affecting the gas exchange process. Other actuators, such as the exhaust throttle and EGR valve, need to be considered to enhance the air handling system. To achieve this, a robust model-basedmultiple-input multiple-output (MIMO) controller is developed to regulate the desired AFR, engine speed, and diluent air ratio (DAR) using various air handling actuators and fueling strategies. The robust MIMO controller is synthesized based on a physics-based mean value engine model, which has been calibrated to accurately reflect high-fidelity engine simulation software. The robust SISO and MIMO controllers are implemented in simulation using the high-fidelity engine simulation software. Following the simulation, the controllers are validated through experimental testing conducted in an engine dynamometer at University of Wisconsin. Results from these controllers are compared against a non-eBoosted engine, which serves as the baseline. While both the SISO and MIMO controllers show improvements in AFR (Air-Fuel Ratio), DAR (Diluent Air Ratio), and engine speed recovery during the load transients, the robust MIMO controller outperforms them by demonstrating the best overall engine performance. This superiority is attributed to its comprehensive understanding of the coupling between each actuator input and the model output. When the MIMO controller operates alongside the electrified intake boosting system, the engine exhibits remarkable enhancements. Not only does it recover back to a steady state 70% faster than the baseline, but it also reduces engine speed droop by 45%. Consequently, the engine's ability to accept load torque increases significantly.</p><p dir="ltr">As a result, a single robust MIMO controller can efficiently perform the same task instead of employing multiple PIDs or look-up tables for each actuator.</p>

Colloidal particles provide the ideal building blocks for the next generation of microdevices, such as advanced sensors and precision drug delivery systems. However, many such applications require the use of top-down (i.e. humanly controllable) forces to manipulate colloidal particles with single-particle precision, and current methods can only achieve such precision for small numbers of particles at a time. To address this challenge, we propose using chemical forces in combination with existing top-down techniques to enable the control of larger numbers of particles simultaneously. Controlling colloids using chemical reactions is a novel technique not typically utilized. Due to its distinct difference from other control methods, it provides new degrees of freedom to work with which offer new opportunities for designing colloidal devices. In this dissertation, we show how modern control theory can be used to implement the control of colloidal particles using chemical forces. We use Brownian dynamics simulations to test control strategies for three different situations: directly controlling chemical reactions to produce a desired concentration gradient, controlling a reactive colloidal particle that interacts chemically with other colloids to move them to desired locations, and controlling the dynamics of active colloidal particles to manipulate their collective behavior. The results obtained in this work will demonstrate the plausibility of each of these three control strategies and provide insights into the choices of physical parameters that can be used in future experiments.

no / A milk pasteurization process is a nonlinear process and multivariable interacting system. This makes it difficultly to control by the conventional on-off controllers. Even if the on-off controller can managed the milk temperatures in the holding tube and the cooling stage of the plate pasteurizer according to the plant's requirements, the dynamic profiles of the milk temperature are oscillating around a desired value. Consequently, this work is aimed at improving the control performance by a multi-variables control approach with model predictive control (MPC). The proposed algorithm was tested in the case of set point tracking under nominal condition gathered by the real observation. To compare the performance of the MPC controller, a model-based control approach of generic model control (GMC) coupled with cascade control strategy is taken into account. The simulation results demonstrated that a proposed control algorithm performed well in keeping both the milk and water temperatures at the desired set points without any oscillation and overshoot. Because of the predictive control strategy, the control response for MPC was less drastic control action compared to the GMC.

L'évolution de l'électronique de puissance depuis ces dernières années est le résultat des enjeux énergétiques actuels qui exigent, entre autres, des architectures de conversion d'énergie capables de traiter des puissances de plus en plus importantes. Parmi les éléments les plus caractéristiques de cette évolution, l'avancement technologique des composants semi-conducteurs (nouveaux composants SiC ou GaN) ainsi que la conception de nouvelles architectures de convertisseurs statiques jouent un rôle important. Parmi ces architectures, différentes associations basées sur la connexion en série et en parallèle de cellules de commutation classiques ont été proposées. Ces associations permettent d'augmenter la puissance traitée par les convertisseurs sans accroitre les contraintes au niveau des interrupteurs. Elles permettent également l'obtention de signaux de sortie d'une meilleure qualité avec des fréquences apparentes de découpage plus importantes. Ces architectures utilisent des éléments de stockage d'énergie qui diminuent les contraintes au niveau des interrupteurs mais qui exigent, en revanche, une régulation précise des grandeurs de tension ou de courant propres à ces éléments. Pour l'association en série, les tensions des condensateurs doivent rester autour d'une fraction de la tension du bus d'entrée. Pour l'association en parallèle, le courant de sortie doit être réparti équitablement entre les différents bras afin d'éviter les phénomènes non linéaires propres aux éléments magnétiques utilisés dans les inductances (séparées ou magnétiquement couplées). Dans la première partie de cette thèse, nous présentons les généralités de l'association en série et parallèle des cellules de commutation. La modélisation des éléments magnétiques utilisés pour la mise en parallèle est également détaillée dans le but d'identifier de possibles sources de déséquilibre sur la répartition du courant de sortie. Une modélisation matricielle est appliquée pour simplifier la relation entre les variables propres à chaque association et les ordres de commande de toutes les cellules. Cette modélisation matricielle sera la base des stratégies de commande que nous avons développées dans la suite de nos travaux. Dans la deuxième partie de cette thèse, nous présentons les différentes stratégies de commande pouvant être appliquées sur ces convertisseurs. Les premières stratégies sont basées sur une approche classique utilisant un modulateur, un générateur d'ordres de commande et des régulateurs de type linéaire pour la régulation des variables internes et externes de chaque association. En termes de modulateurs, nous présentons principalement un modulateur de type PS (Phase Shifted), tandis que quelques applications et résultats sont présentés pour un modulateur de type PD (Phase Disposition). D'autres stratégies basées sur la commande prédictive sont également présentées. La première est la stratégie MPC qui utilise une fonction de coût pour choisir l'état optimal du convertisseur pour chaque période d'échantillonnage. Cette stratégie a été introduite récemment dans le domaine des convertisseurs statiques et présente des avantages liées à la facilité de sa mise en place ainsi qu'aux réponses du système lors des régimes transitoires. La deuxième stratégie, basée sur la commande prédictive, utilise des instants de commutation variables, une fonction de coût simplifiée et une machine d'état. Cette dernière permet de gérer les ordres de commande de toutes les cellules de commutation en fonction des variables à réguler. En plus des avantages liés à la stratégie MPC, sa mise en place est bien plus simple car elle fonctionne à une fréquence de découpage fixe et s'adapte facilement à différents points de fonctionnement. Dans la dernière partie de cette thèse, nous présentons l'implantation expérimentale de ces stratégies afin de valider leur performance sur les convertisseurs multicellulaires. / In the last years, the development in the power electronics field is the result of the current energy challenges. These challenges require power converters able to work with increasingly important powers. Among the most characteristic elements of this development, we can find the technological advancements of the semiconductor devices (based principally in SiC and GaN) and the conception of new power converters topologies. These new power converter topologies are principally based on the serial and parallel association of classical commutation cells. With these associations, the energy treated by the converter can be increased using the current semiconductor technology. The quality of the output signals can also be improved with higher apparent switching frequencies. These associations use elements for storing energy, such as inductors or capacitors. They equally allow the reduction of the constraints on the switches given the higher voltages and currents. However, the use of these elements requires a good control of the capacitors' voltage for the serial connection and a good distribution of the output current among the different phases for the parallel connection. In the parallel connection, when we use Inter Cells Transformers (ICT) instead of classical inductors, all the phase currents reduce their ripples while their frequency is reduced. Nevertheless, some differences between all the phases' currents can entail non-linear phenomena, producing perturbations and instabilities in the system. In the serial connection, the capacitor voltages must stay around a fraction of the input voltage in order to get an optimal multilevel output voltage. In the first part of this thesis, we present the generalities of the serial and parallel association of classical commutation cells. Different models of magnetic elements are used for getting a better representation of an ICT; these models are used for finding possible sources of currents imbalances. A matrix model is used to simplify the relationship between the control variables with the control of each commutation cell. In the second part of this thesis, some control strategies that can be applied with these converters are presented. The first strategy is based on a conventional approach that uses a modulator, a state machine for generating the commands of each cell and linear regulators for controlling the internal and external variables (output voltage and currents, capacitors in the serial association and the distribution of the current for the parallel connection). In terms of modulators, we present primarily a PS (Phase Shifted) modulator while some applications and results are presented for a PD (Phase Disposition) modulator. Other strategies based on predictive control are also presented. The first of these strategies is the classical MPC (model predictive control) strategy that uses a cost function to select the optimal state of the converter for each sampling period. This strategy has recently been introduced in the field of static converters and it has several advantages related to the facility of its implementation and the optimal transient responses. The second strategy uses variable switching instants, a simplified cost function and a state machine. The state machine is used to manage the capacitors' voltages and the differential currents (differences between the phase currents) while the cost function is used for controlling the output voltage and current. This strategy is simpler to be implemented, presents fast transient responses and works with a fixed switching frequency in different operating points. In the last part of this thesis, we present the experimental implementation of these strategies in order to validate their performance in the power converters based in the serial and parallel association of classical commutation cells.

Made available in DSpace on 2014-12-17T14:56:17Z (GMT). No. of bitstreams: 1 GasparFDJ_DISSERT.pdf: 3388304 bytes, checksum: 086a8f61099f69978a8b9f477f351d24 (MD5) Previous issue date: 2014-01-23 / Petr?leo Brasileiro SA - PETROBRAS / Slugging is a well-known slugging phenomenon in multiphase flow, which may cause problems such as vibration in pipeline and high liquid level in the separator. It can be classified according to the place of its occurrence. The most severe, known as slugging in the riser, occurs in the vertical pipe which feeds the platform. Also known as severe slugging, it is capable of causing severe pressure fluctuations in the flow of the process, excessive vibration, flooding in separator tanks, limited production, nonscheduled stop of production, among other negative aspects that motivated the production of this work . A feasible solution to deal with this problem would be to design an effective method for the removal or reduction of the system, a controller. According to the literature, a conventional PID controller did not produce good results due to the high degree of nonlinearity of the process, fueling the development of advanced control techniques. Among these, the model predictive controller (MPC), where the control action results from the solution of an optimization problem, it is robust, can incorporate physical and /or security constraints. The objective of this work is to apply a non-conventional non-linear model predictive control technique to severe slugging, where the amount of liquid mass in the riser is controlled by the production valve and, indirectly, the oscillation of flow and pressure is suppressed, while looking for environmental and economic benefits. The proposed strategy is based on the use of the model linear approximations and repeatedly solving of a quadratic optimization problem, providing solutions that improve at each iteration. In the event where the convergence of this algorithm is satisfied, the predicted values of the process variables are the same as to those obtained by the original nonlinear model, ensuring that the constraints are satisfied for them along the prediction horizon. A mathematical model recently published in the literature, capable of representing characteristics of severe slugging in a real oil well, is used both for simulation and for the project of the proposed controller, whose performance is compared to a linear MPC / A golfada ? um regime inst?vel do fluxo multif?sico, com oscila??es de press?o e vaz?o abruptas no processo de produ??o de petr?leo, podendo ocasionar problemas tais como vibra??o na tubula??o e alto n?vel de l?quido nos separadores. Pode ser classificada de acordo com seu local de ocorr?ncia. A mais severa destas, conhecida como golfada no riser, ocorre na tubula??o vertical que alimenta a plataforma. Conhecida tamb?m como golfada severa, ela ? capaz de causar bruscas oscila??es na press?o, nas vaz?es do processo, vibra??o excessiva, inunda??o dos tanques separadores, produ??o limitada, parada n?o programada da plataforma, entre outros aspectos negativos que motivaram a produ??o deste trabalho. Uma solu??o vi?vel para lidar com tal problema seria projetar um m?todo efetivo para a remo??o ou diminui??o deste regime, como um controlador. De acordo com a literatura, o controlador convencional PID n?o apresenta bons resultados devido ao alto grau de n?o linearidade do processo, o que impulsionou o desenvolvimento de t?cnicas avan?adas de controle. Dentre estas, o controlador preditivo, cuja a??o de controle resulta da solu??o de um problema de otimiza??o, al?m de ser uma t?cnica que apresenta robustez e pode incorporar restri??es f?sicas e/ou de seguran?a. O objetivo deste trabalho ? estudar a aplica??o de uma t?cnica de controle preditivo n?o linear ao controle de golfada severa, visando controlar a quantidade de massa l?quida no riser atuando na v?lvula de produ??o e, indiretamente, suprimir as oscila??es de vaz?o e press?o. Com a finalidade de obter benef?cios ambientais e econ?micos. A t?cnica de controle preditivo proposta baseia-se no uso de aproxima??es lineares do modelo e na resolu??o repetida de um problema de otimiza??o quadr?tica que proporciona solu??es que melhoram a cada itera??o. No caso em que a converg?ncia desse algoritmo ? satisfeita, os valores preditos das vari?veis do processo s?o iguais ?queles que seriam obtidos pelo modelo n?o linear original, garantindo que as restri??es nessas vari?veis sejam satisfeitas ao longo do horizonte de predi??o. Um modelo matem?tico publicado recentemente na literatura, capaz de representar caracter?sticas da golfada severa em um po?o real, ? utilizado tanto para a simula??o, quanto para projeto do controlador proposto, cujo desempenho ? comparado ao de um controlador preditivo linear