Neste trabalho é abordado o problema de controle e de otimização de unidades industriais de hidrotratamento de diesel (UHDT) por controladores MPC (Model Predictive Control). É apresentado um breve histórico dos controladores MPC convencionais e de horizonte infinito (IHMPC), bem como uma breve descrição do processo de Hidrotratamento de Diesel e das particularidades da aplicação do controle de processos a este tipo de planta industrial. Em seguida foi gerado, passo a passo, um algoritmo de controle que sumarizou e agregou características de vários controladores MPC disponíveis na literatura aberta, em especial os que foram desenvolvidos ao longo dos últimos anos pelo laboratório de simulação e controle da USP (LSCP), a fim de se obter um algoritmo adequado para a solução do problema de controle abordado. Em ambiente computacional de simulação, o algoritmo resultante possibilitou controlar e otimizar simultaneamente processos contínuos, sendo capaz de estabilizar a planta industrial de forma robusta e, ao mesmo tempo, aumentar a lucratividade de sua operação. Para tanto, foi desenvolvida uma função objetivo econômica que aumentou a conversão da carga bruta em produtos hidrotratados e minimizou o consumo de insumos, sendo que essa correlação foi agregada ao algoritmo de controle. As simulações permitiram que as estratégias de controle previamente discutidas pudessem ser testadas e seus resultados apresentados e debatidos. / This work addresses the control and optimization problem of industrial diesel hydrotreating units (UHDT) by MPC controllers (Model Predictive Control). It is presented a brief historical of conventional MPC controllers and infinite horizon controllers (IHMPC), as well as a brief description of the Diesel Hydrotreating process and the particulars of the application of process control for this type of industrial plant. It was then generated, step by step, one algorithm that summarized and aggregated control characteristics of various MPC controllers available in the open literature, in particular those that have been developed over the past few years by USPs laboratory of simulation and control of (LSCP), in order to obtain an algorithm suitable for solving the addressed control problem. In a computational simulation environment, the resulting algorithm allowed to simultaneously control and optimize continuous processes, being able to robustly stabilize the industrial plant and at the same time increase the profitability of its operation. For this purpose, an \"objective function\" was developed which increased the economic conversion of crude feed to hydrotreated product and minimized the consumption of raw materials, and this correlation was added to the control algorithm. The simulations allowed that the previously discussed control strategies could be tested and the results presented and discussed.

O objetivo do trabalho é apresentar dois algoritmos para a solução de problemas de controle ótimo quadrático robusto com restrições, dentro de um contexto de controladores preditivos (MPC do inglês Model Predictive Control). Inicialmente apresentamos uma breve introdução aos algoritmos MPC, com ênfase na abordagem do controlador linear quadrático. Em seguida são apresentados os dois algoritmos de interesse, que utilizam técnicas de otimização LMI. Dessa forma as restrições e as incertezas podem ser colocadas em formas computacionalmente tratáveis. Por fim são realizadas simulações e comparações entre esses algoritmos, bem como com técnicas de MPC encontradas na literatura atual. / The goal of the work is to present two algorithms for the solution of robust quadratic optimal control problems with restrictions, within a model predictive control (MPC) setup. Initially we present a brief introduction of the MPC algorithms, emphasizing the linear quadratic controller approach. Next the two algorithms of interest, using LMI optimization techniques, are presented. By using this technique the restrictions and uncertainties can be written in a computational way. Finally some simulations and comparisons between these algorithms, as well as with MPC techniques found in the current literature, are performed.

A Identificação de Sistemas é uma tarefa significativa em termos de tempo e custo no trabalho de implementação de sistemas de controle que usam Controle Preditivo baseado em Modelos (MPC). Após a implementação, o controlador tende a permanecer com o mesmo modelo por muito tempo, ignorando mudanças que tenham ocorrido com o processo, perdendo qualidade e podendo até ser abandonado. Este trabalho propõe uma metodologia simples e eficaz para se proceder à reidentificação de uma planta industrial que use MPC mantendo o processo em malha fechada. Os principais aspectos deste problema são discutidos, e as escolhas que foram feitas para a realização dos experimentos e obtenção dos modelos são explicadas. Apresenta-se um caso em Matlab sobre um sistema 2x2 cobrindo diferentes situações, e é feita uma comparação de identificação realizada através de sinais PRBS e de testes com degraus, sempre em malha fechada. Aplica-se a metodologia a um controlador industrial, e os modelos identificados são introduzidos no controlador. O princípio básico desta metodologia consiste em efetuar perturbações multivariáveis nos set-points ou restrições ativas das controladas e determinar o modelo através da estrutura ARX. Entre as vantagens da metodologia proposta, estão a facilidade de automatizar a identificação do processo e a garantia de manter o processo sob controle durante os testes. / System identification is a major task in the process of implementing Model-based Predictive Control (MPC) algorithms in industrial applications. Once the controller is working, there is a tendency to leave it with the original model for a long time, neglecting changes to the process during this time, leading to performance degradation. This work proposes a simple and effective methodology to re-identify plants under MPC in closed loop. The main issues concerning this problem are discussed, and choices for experiments are made. A Matlab case involving a 2x2 problem is presented, covering a range of different situations, and a comparison between identification using PRBS reference signals and standard step tests is shown. An industrial case is studied, applying the proposed method to a real situation, re-identifying an existing MPC model and reconfiguring it afterwards. This methodology is based on the application of multivariable perturbations on the controlled variables set-points or active restrictions, obtaining an ARX model structure. It uses an automatic process identification proceeding, keeping the process under control along the tests.

Platooning is an interesting area which involve the possibility of decreasing the fuel consumption of heavy-duty vehicles. By reducing the inter-vehicle spacing in the platoon we can reduce air drag, which in turn reduces fuel consumption. Two fuel-efficient model predictive controllers for HDVs in a platoon has been formulated in this master thesis, both utilizing road grade preview information. The first controller is based on linear programming (LP) algorithms and the second on quadratic programming (QP). These two platooning controllers are compared with each other and with generic controllers from Scania. The LP controller proved to be more fuel-efficient than the QP controller, the Scania controllers are however more fuel-efficient than the LP controller.

O objetivo do trabalho é apresentar dois algoritmos para a solução de problemas de controle ótimo quadrático robusto com restrições, dentro de um contexto de controladores preditivos (MPC do inglês Model Predictive Control). Inicialmente apresentamos uma breve introdução aos algoritmos MPC, com ênfase na abordagem do controlador linear quadrático. Em seguida são apresentados os dois algoritmos de interesse, que utilizam técnicas de otimização LMI. Dessa forma as restrições e as incertezas podem ser colocadas em formas computacionalmente tratáveis. Por fim são realizadas simulações e comparações entre esses algoritmos, bem como com técnicas de MPC encontradas na literatura atual. / The goal of the work is to present two algorithms for the solution of robust quadratic optimal control problems with restrictions, within a model predictive control (MPC) setup. Initially we present a brief introduction of the MPC algorithms, emphasizing the linear quadratic controller approach. Next the two algorithms of interest, using LMI optimization techniques, are presented. By using this technique the restrictions and uncertainties can be written in a computational way. Finally some simulations and comparisons between these algorithms, as well as with MPC techniques found in the current literature, are performed.

A Identificação de Sistemas é uma tarefa significativa em termos de tempo e custo no trabalho de implementação de sistemas de controle que usam Controle Preditivo baseado em Modelos (MPC). Após a implementação, o controlador tende a permanecer com o mesmo modelo por muito tempo, ignorando mudanças que tenham ocorrido com o processo, perdendo qualidade e podendo até ser abandonado. Este trabalho propõe uma metodologia simples e eficaz para se proceder à reidentificação de uma planta industrial que use MPC mantendo o processo em malha fechada. Os principais aspectos deste problema são discutidos, e as escolhas que foram feitas para a realização dos experimentos e obtenção dos modelos são explicadas. Apresenta-se um caso em Matlab sobre um sistema 2x2 cobrindo diferentes situações, e é feita uma comparação de identificação realizada através de sinais PRBS e de testes com degraus, sempre em malha fechada. Aplica-se a metodologia a um controlador industrial, e os modelos identificados são introduzidos no controlador. O princípio básico desta metodologia consiste em efetuar perturbações multivariáveis nos set-points ou restrições ativas das controladas e determinar o modelo através da estrutura ARX. Entre as vantagens da metodologia proposta, estão a facilidade de automatizar a identificação do processo e a garantia de manter o processo sob controle durante os testes. / System identification is a major task in the process of implementing Model-based Predictive Control (MPC) algorithms in industrial applications. Once the controller is working, there is a tendency to leave it with the original model for a long time, neglecting changes to the process during this time, leading to performance degradation. This work proposes a simple and effective methodology to re-identify plants under MPC in closed loop. The main issues concerning this problem are discussed, and choices for experiments are made. A Matlab case involving a 2x2 problem is presented, covering a range of different situations, and a comparison between identification using PRBS reference signals and standard step tests is shown. An industrial case is studied, applying the proposed method to a real situation, re-identifying an existing MPC model and reconfiguring it afterwards. This methodology is based on the application of multivariable perturbations on the controlled variables set-points or active restrictions, obtaining an ARX model structure. It uses an automatic process identification proceeding, keeping the process under control along the tests.

Neste trabalho é abordado o problema de controle e de otimização de unidades industriais de hidrotratamento de diesel (UHDT) por controladores MPC (Model Predictive Control). É apresentado um breve histórico dos controladores MPC convencionais e de horizonte infinito (IHMPC), bem como uma breve descrição do processo de Hidrotratamento de Diesel e das particularidades da aplicação do controle de processos a este tipo de planta industrial. Em seguida foi gerado, passo a passo, um algoritmo de controle que sumarizou e agregou características de vários controladores MPC disponíveis na literatura aberta, em especial os que foram desenvolvidos ao longo dos últimos anos pelo laboratório de simulação e controle da USP (LSCP), a fim de se obter um algoritmo adequado para a solução do problema de controle abordado. Em ambiente computacional de simulação, o algoritmo resultante possibilitou controlar e otimizar simultaneamente processos contínuos, sendo capaz de estabilizar a planta industrial de forma robusta e, ao mesmo tempo, aumentar a lucratividade de sua operação. Para tanto, foi desenvolvida uma função objetivo econômica que aumentou a conversão da carga bruta em produtos hidrotratados e minimizou o consumo de insumos, sendo que essa correlação foi agregada ao algoritmo de controle. As simulações permitiram que as estratégias de controle previamente discutidas pudessem ser testadas e seus resultados apresentados e debatidos. / This work addresses the control and optimization problem of industrial diesel hydrotreating units (UHDT) by MPC controllers (Model Predictive Control). It is presented a brief historical of conventional MPC controllers and infinite horizon controllers (IHMPC), as well as a brief description of the Diesel Hydrotreating process and the particulars of the application of process control for this type of industrial plant. It was then generated, step by step, one algorithm that summarized and aggregated control characteristics of various MPC controllers available in the open literature, in particular those that have been developed over the past few years by USPs laboratory of simulation and control of (LSCP), in order to obtain an algorithm suitable for solving the addressed control problem. In a computational simulation environment, the resulting algorithm allowed to simultaneously control and optimize continuous processes, being able to robustly stabilize the industrial plant and at the same time increase the profitability of its operation. For this purpose, an \"objective function\" was developed which increased the economic conversion of crude feed to hydrotreated product and minimized the consumption of raw materials, and this correlation was added to the control algorithm. The simulations allowed that the previously discussed control strategies could be tested and the results presented and discussed.

Les travaux de cette thèse portent sur l’investigation d’approches de commande et de supervision permettant d’aborder la problématique de gestion d’énergie des réseaux électriques multi-sources. l’objectif souhaité était de proposer une démarche de conception de lois de commande pour ce type de système en vue de réguler la tension de sortie et de gérer d’une manière optimale les flux d’énergie entre les différentes sources et les consommateurs et au vu de minimiser la consommation d’hydrogène.A cette fin, deux configurations ont été envisagées :l’application d’approches à base d’un modèle statique et des stratégies à base d’un modèle dynamique de la PàC. Dans un premier temps, trois approches de gestion énergétique ont été appliquées au système visant à minimiser la consommation de masse d’hydrogène tout en respectant les contraintes physiques du système.Tout d’abord, l’optimisation est réalisée en utilisant une méthode d’optimisation hors ligne appelée la programmation dynamique. Deuxièmement, deux approches d’optimisation en ligne sont utilisées : stratégies ECMS et MPC. Une comparaison en termes de consommation d’hydrogène et de temps de calcul est réalisée.Dans un deuxième temps, une approche décentralisée de commande a été envisagée afin de tenir compte du modèle dynamique de la PàC dans la conception du superviseur. L’avantage de cette architecture réside dans sa capacité a aborder séparément chacune des problématiques dans l’optique de répondre aux différents objectifs de commande. Dans ce cadre, la régulation du système PàC et de l’état de charge de l’ESS est réalisée séparément avec deux contrôleurs différents, tous deux conçus en utilisant l’approche (MPC-LTV). Les troisième et quatrième niveaux de la structure de contrôle décentralisée consistent en des boucles de locale des courants de la PàC et de SC et un contrôle de tension du bus continu, conçu à l’aide de contrôleurs PI. La validation de la structure de contrôle est réalisée en simulation en utilisant un modèle non linéaire du système PàC.A la fin, les approches de commande conçues à base du modèle statique de la PàC (DP, ECMS et MPC) seront appliquées sur le modèle dynamique de cette dernière afin d’évaluer les performances de ces approches et d’en déduire des conclusions sur l’apport de l’intégration de la dynamique de la PàC dans la synthèse des lois de commande. / This thesis focuses on the investigation of control approaches to treat the issue of energy management of multi-source electrical networks. The considered electric motor supply system consists on a fuel cell as a main energy source and an additional element that supplies peak power and charges by regenerative braking. At first, three energy management strategies have been applied to the sypply system aiming to minimize the fuel cell hydrogen mass consumption while satisfying the system physical constraints. First, the optimization is realized using dynamic programming,an off-line optimization method that requires the knowledge of the entire power load profile. Secondly, twoon-line optimization approaches are used : ECMS and MPC strategies, for which only the current power demand is demanded.The second part of this thesis presents a decentralized control strategy applied to the power system. The dedicated control structure aims to assure an optimal operation of the FC system while respecting the compressor physical limits and to control the converter current sand network output voltage. To attain these objectives, a dynamic model of the FC system is used,in addition to the SSE and electric network dynamics. The FC system regulation and the control of the SSE state of energy are performed separately with two different controllers, both designed using (MPC-LTV) approach. The third and fourth levels of the decentralized control structure consists on inner control loops for fuel cell/supercapacitor currents and a DC bus voltage control loop, designed using PI controllers. The validation of the control structure is performed in simulation using a nonlinear models of the FC system and the SSE. To validate and compare the performance of different control methods based on a fuel cell static model, these approaches have been applied to the dynamic model of the FC and compared to the results obtained by applying the approched designed and based on an FC dynamic model. A comparison in terms of network efficiency and hydrogen consumption has been done.

Privacy is a crucial aspect of any system involving user-supplied data. There exist multiple approaches to protecting the identity and secrecy of users in data submission systems. In this thesis, I consider the case of privacy-enhancing of data reporting in Participatory Sensing systems. I conducted an extensive literature overview to explore privacy-oriented enhancements to data submission that are applicable in the PS systems. I designed a protocol for proximity-based data aggregation that utilizes Multi-party Secure Computations over Bluetooth Low Energy. Users are divided into groups that perform sub-aggregations and report results to central entities, protecting themselves from honest-but-curious adversary threats. I present a mobile app and web servers for central entities that follow the design of the protocol. I evaluated the achieved effectiveness and discuss the utility and privacy trade-offs. The implementation performs typically for an MPC system with high communication overhead, and is implemented over Bluetooth, with the additional time needed for discovering and connecting devices. The overall performance of the system suggests that deployments targeting 1-second intervals of data submission are feasible. Main use cases are sensitive measurements, such as medical data or highly private user information. / Sekretess är en avgörande aspekt av alla system som involverar data som tillhandahålls av användare. Det finns flera tillvägagångssätt för att skydda användarnas identitet och sekretess i datainlämningssystem. I den här avhandlingen behandlar jag fallet med integritetsförbättrande datarapportering i Participatory Sensing-system. Jag genomförde en omfattande litteraturöversikt för att utforska integritetsorienterade förbättringar av datainlämning som är tillämpliga i PS-systemen. Jag designade ett protokoll för närhetsbaserad dataaggregering som använder flerpartssäkra beräkningar över Bluetooth Low Energy. Användare är indelade i grupper som utför sub-aggregeringar och rapporterar resultat till centrala enheter, och skyddar sig själva från ärliga men nyfikna motståndarhot. Jag presenterar en mobilapp och webbservrar för centrala enheter som följer protokollets design. Jag utvärderade den uppnådda effektiviteten och diskuterade nytta och sekretessavvägningar. Implementeringen fungerar som man kan förvänta sig för ett MPC-system med höga kommunikationskostnader, och implementeras över Bluetooth, med den extra tid som krävs för att upptäcka och ansluta enheter. Systemets övergripande prestanda tyder på att implementeringar som är inriktade på 1-sekunds intervaller för datainlämning är genomförbara. Huvudsakliga användningsfall är känsliga mätningar, såsom medicinska data eller mycket privat användarinformation.

The inherited rotation in a caster wheel allows movement in any direction, but pays at the expense of reaction torques. When implemented in a mobile robot, these forces have a negative impact in its performance. One approach is to restrict rotations on the spot by attaching a filter to the output of the motion planner. However, this formulation compromises the navigation’s completion in critical scenarios, such as parking, taking curves in narrow corridors or navigating at the presence of a high density of obstacles. Therefore, in this thesis we consider the influence of caster wheels in the motion planning stage, commonly presented as local planning. This work proposes a Model Predictive Control (MPC) based local planner that integrates the caster wheel physics into the motion planning stage. A caster wheel aware term is combined with a reference tracking based navigation, which leads to the formulation of the Caster Wheel Aware Local Planner (CWAWLP). Since this method requires knowing the caster wheel’s state and there is no sensor that provides this information, a caster wheel state observer is also formulated. In order to evaluate the impact of the caster wheel aware term, CWAWLP is compared to a Caster Wheel based Agnostic Local Planner (CWAGLP) and a Caster Wheel based Agnostic Planner Local Planner with Path Filter (CWPFLP). After running simulations for three case studies in a virtual framework, two experimental case studies are conducted in an intra-logistics robot. These are evaluated according to the navigation’s quality, motor torque usage and energy consumption. According to the patterns observed in the evaluation, CWAWLP covers a longer distance than CWAGLP wihout decreasing the navigation’s quality. At the same time, its motor torques are similar to the ones of CWPFLP. Therefore, CWAWLP is capable of considering caster wheel physics without sacrificing navigation capabilities. The formulated caster wheel aware term is compatible with any MPC based navigation algorithm and inherits the derivation of an observer capable of estimating caster wheel rotation angles and rolling speeds. Even if the caster wheel awareness has been implemented in a differential driven robot, this approach is also applicable to vehicles with an alternative drivetrain, such as car-like robots. / Den ärvda rotationen i ett hjul möjliggör rörelse i vilken riktning som helst, men fås på bekostnad av reaktionsmoment. När de implementeras i en mobil robot har dessa krafter en negativ inverkan på dess prestanda. Ett tillvägagångssätt är att begränsa rotationer på plats genom att applicera ett filter på rörelseplannerns utgång. Denna formulering komprometterar dock navigeringens slutförande i kritiska scenarier, såsom parkering, kurvor i smala korridorer eller navigering i närheten av höga hinder. Därför beaktar vi i denna avhandling påverkan av hjul på hjulplaneringen, som ofta presenteras som lokal planering. Detta arbete föreslår en Model Predictive Control (MPC) -baserad lokal planerare som integrerar svängbara länkhjuls fysik i rörelseplaneringsstadiet. En kugghjulmedveten term kombineras med en referensspårningsbaserad navigering, vilket leder till formuleringen av Caster Wheel Aware Local Planner (CWAWLP). Eftersom denna metod kräver kunskap om svängbara länkhjuls tillstånd och det inte finns någon sensor som ger denna information, formuleras också en hjulhjulstillståndsobservatör. För att utvärdera effekten av det medvetna begreppet svängbara änkhjul jämförs CWAWLP med en Caster Wheel-baserad Agnostic Local Planner (CWAGLP) och en Caster Wheel-baserad Agnostic Planner Local Planner with Path Filter (CWPFLP). Efter att ha kört simuleringar för tre fallstudier i ett virtuellt ramverk genomförs två experimentella fallstudier i en intra-logistikrobot. Dessa utvärderas enligt navigeringens kvalitet, vridmomentanvändning och energiförbrukning. Enligt de mönster som observerats i utvärderingen når CWAWLP ett längre avstånd än CWAGLP utan att sänka navigeringens kvalitet. Samtidigt liknar motorns vridmoment dem som CWPFLP. Därför kan CWAWLP ta hänsyn till svängbara länkhjuls fysik utan att offra navigationsfunktionerna. Den formulerade medhjulningsmedveten termen är kompatibel med vilken MPC-baserad navigationsalgoritm som helst och ärver härledningen av en observatör som kan uppskatta hjulets rotationsvinklar och rullningshastigheter. Även om hjulhjälpmedvetenheten har implementerats i en differentierad robot, är detta tillvägagångssätt också tillämpligt på fordon med ett alternativt drivsystem, såsom billiknande robotar.