The increasing market of small and affordable space systems requires fast and reliablelaunch capabilities to cover the current and future demand. This project aims to studyand implement guidance and control schemes for vertical ascent and descent phases ofa reusable launcher. Specifically, the thesis focuses on developing and applying ModelPredictive Control (MPC) and optimisation techniques to several kino-dynamic modelsof rockets. Moreover, the classical MPC method has been modified to include a decreasingfactor for the horizon used, enhancing the performance of the guidance and control.Multiple scenarios of vertical launch, landing and full fligth guidance on Earth and Mars,along with Monte Carlo analysis, were carried out to demonstrate the robustness of thealgorithm against different initial conditions. The results were promising and invite tofurther research.

Tires, being the only component of the vehicle in contact with the road surface, are responsible for generating the forces for maintaining the vehicle pose, orientation and stability of the vehicle. Additionally, the on-board advanced chassis control systems require estimation of these tire-road interaction properties for their operation. Extraction of these properties becomes extremely important in handling limit maneuvers such as Double Lane Change (DLC) and cornering wherein the lateral force transfer is dependent upon these computations. This research focuses on the development of a high-fidelity vehicle-tire model and control algorithm framework for vehicle trajectory tracking for vehicles operating in this limit handling regime. This combined vehicle-tire model places an emphasis on the lateral dynamics of the vehicle by integrating the effects of relaxation length on the contact patch force generation. The vertical dynamics of the vehicle have also been analyzed, and a novel double damper has been mathematically modeled and experimentally validated. Different control algorithms, both classical and machine learning-based, have been developed for optimizing this vertical dynamics model. Experimental data has been collected by instrumenting a vehicle with in-tire accelerometers, IMU, GPS, and encoders for slalom and lane change maneuvers. Different state estimation techniques have been developed to predict the vehicle side slip angle, tire slip angle, and normal load to further assist the developed vehicle-tire model. To make the entire framework more robust, Machine Learning algorithms have been developed to classify between different levels of tire wear. The effect of tire tread wear on the pneumatic trail of the tire has been further evaluated, which affects the aligning moment and lateral force generation. Finally, a Model Predictive Control (MPC) framework has been developed to compare the performance between the conventional vehicle models and the developed vehicle models in tracking a reference trajectory. / Doctor of Philosophy / In our rapidly advancing world, self-driving or autonomous vehicles are no longer a vision of the future but a reality of today. As we grow more reliant on these vehicles, ensuring their safety and reliability becomes increasingly critical. Unlike traditional vehicles, self-driving cars operate without human intervention. Consequently, the onus of passenger and pedestrian safety falls squarely on the vehicle's control systems. The efficiency and effectiveness of these control systems are pivotal in preventing accidents and ensuring a smooth ride. One vital aspect of these control systems lies in understanding the tires' behavior, the only parts of the vehicle that are in contact with the road surface. A tire's interaction with the road surface significantly impacts the vehicle's handling and stability. Information such as how much of the tire is in contact with the road, the forces and moments generated at this contact point, becomes valuable for optimizing the vehicle's performance. This is particularly crucial when a vehicle is turning or cornering, where the forces developed between the tires and the road are key to maintaining control and stability. In this research, a framework has been designed to improve the vehicle performance, primarily by improving the modeling of tire lag dynamics. This refers to the delay or 'lag' between a change in tire conditions (such as pressure, wear, and temperature) and the corresponding change in tire behavior. In addition, in this research a vertical dynamics model of the vehicle has also been developed incorporated with a novel double damper suspension system. To complete the entire framework, the effect of tire wear over time and how this affects its performance and safety characteristics has also been examined. By estimating and understanding this wear, we can predict how it will affect the dynamic properties of the tire, thus improving the reliability and efficiency of our autonomous vehicles. The last piece of this framework comprises the development of an MPC controller to track a reference trajectory and evaluate the performance of the developed model.

Inom reglerteknik har integrationen av maskininlärningsmetoder framträtt som en central strategi för att förbättra prestanda och adaptivitet hos styrsystem. Betydande framsteg har gjorts inom flera viktiga aspekter av reglerkretsen, såsom inlärningsbaserade metoder för systemidentifiering och parameterskattning, filtrering och brusreducering samt reglersyntes. Denna avhandling fördjupar sig i området inlärning för reglerteknik med särskild betoning på inlärningsbaserade regulatorer och identifieringsmetoder.  Avhandlingens första del behandlar undersökningen av neuronnätsbaserad Modellprediktiv Reglering (MPC). Olika nätstrukturer studeras, både generella black box-nät och nät som väver in MPC-specifik information i sin struktur. Dessa nät jämförs och utvärderas med avseende på två prestandamått genom experiment på realistiska två- och fyrdimensionella system. Den huvudsakliga nyskapande aspekten är inkluderingen av gradientdata i träningsprocessen, vilket visar sig förbättra noggrannheten av de genererade styrsignalerna. Vidare påvisar de experimentella resultaten att en MPC-informerad nätstruktur leder till förbättrad prestanda när mängden träningsdata är begränsad.  Med insikt om vikten av noggranna matematiska modeller av styrsystemet, riktar den andra delen av avhandlingen sitt fokus mot inlärningsbaserade identifieringsmetoder. Denna forskningsgren behandlar karakterisering och modellering av dynamiska system med hjälp av maskininlärning. Avhandlingen bidrar till området genom att introducera kooperativa systemidentifieringsmetoder för att förbättra parameterskattningen. Specifikt utnyttjas verktyg från Optimal Transport för att introducera en ny och mer generell formulering av ramverket Correctional Learning. Detta ramverk är baserat på en mästare-lärlingsmodell, där en expertagent (mästare) observerar och modifierar den insamlade data som används av en lärande agent (lärling), med syftet att förbättra lärlingens skattningsprocess. Genom att formulera correctional learning som ett optimal transport-problem erhålls ett mer flexibelt ramverk, bättre lämpat för skattning av komplexa systemegenskaper samt anpassning till alternativa handlingsstrategier. / In the context of control systems, the integration of machine learning mechanisms has emerged as a key approach for improving performance and adaptability. Notable progress has been made across several aspects of the control loop, including learning-based techniques for system identification and estimation, filtering and denoising, and controller design. This thesis delves into the rapidly expanding domain of learning in control, with a particular focus placed on learning-based controllers and learning-based identification methods. The first part of this thesis is devoted to the investigation of Neural Network approximations of Model Predictive Control (MPC). Model-agnostic neural network structures are compared to networks employing MPC-specific information, and evaluated in terms of two performance metrics. The main novel aspect lies in the incorporation of gradient data in the training process, which is shown to enhance the accuracy of the network generated control inputs. Furthermore, experimental results reveal that MPC-informed networks outperform the agnostic counterparts in scenarios when training data is limited. In acknowledgement of the crucial role accurate system models play in in the control loop, the second part of this thesis lends its focus to learning-based identification methods. This line of work addresses the important task of characterizing and modeling dynamical systems, by introducing cooperative system identification techniques to enhance estimation performance. Specifically, it presents a novel and generalized formulation of the Correctional Learning framework, leveraging tools from Optimal Transport. The correctional learning framework centers around a teacher-student model, where an expert agent (teacher) modifies the sampled data used by the learner agent (student), to improve the student's estimation process. By formulating correctional learning as an optimal transport problem, a more adaptable framework is achieved, better suited for estimating complex system characteristics and accommodating alternative intervention strategies. / VR 2018-03438 projekt 3224

In this thesis, a Model Predictive Control (MPC) based re-planning and control system is proposed. The MPC re-planner will generate a collision-free path for the controlled vehicle when obstacles are detected, and the controller will make the vehicle move along the reference or re-planned path by adjusting its velocity and steering angle. The MPC re-planner and controller are built based on different vehicle dynamic models, i.e., the bicycle model and point-mass model, respectively. Simulation results show that the trajectory tracking performance when the velocity and steering are controlled simultaneously are better than using steering MPC alone. Then the effects on computational time of two critical parameters, prediction horizon and control horizon, are studied to find reasonable horizons that can enable real-time control. The robustness of the obstacle avoidance function is tested using obstacles with increasing sizes and the results show that the controlled vehicle is able to avoid a 6 m obstacle so that it can overtake other car-like vehicles in the driving process. Finally, a closed-loop one-lane road with some moving vehicles is built as a test scenario for the MPC-based re-planning and control system. According to the results, the controlled vehicle can successfully follow the centerline of the road and overtake other vehicles. / I detta examensarbete föreslås ett fordonsreglersystem baserat på modellprediktiv reglering (MPC) med en omplaneringsfunktion. Det MPC-baserade omplaneringssystemet ska hitta en kollisionsfri väg åt det reglerade fordonet när hinder upptäcks på vägen och hastighet samt styrvinkel anpassas med hjälp av reglersystemet så att fordonet kan köra längs referensvägen eller den nya vägen efter omplaneringen. Det MPC-baserade fordonsregler- och omplaneringssystemet är uppbyggt baserat på olika fordonsdynamiska modeller, cykelmodellen och en punktmassemodell. Simuleringsresultaten visar att prestandan för trajektorieföljningen är avsevärt bättre i fallet då både fordonshastighet och styrvinkel regleras jämfört med att enbart reglera styrvinkeln. Därefter studerades vilken inverkan två kritiska parametrar, förutsägelsehorisont och reglerhorisont, har på simuleringstiden för att hitta rimliga horisonter som kan möjliggöra realtidsreglering. Omplaneringsfunktionens robusthet utvärderades med hjälp av olika hinder med ökande storlekar. Resultaten visar att det reglerade fordonet har förmåga att undvika hinder upp till sex meters storlek, vilket betyder att fordonet kan passera andra billiknande fordon under körning. Slutligen, för att utvärdera det MPC-baserade regler- och omplaneringssystemet skapas ett testscenario där fordonet kör på en enfilig väg och där det finns andra fordon i rörelse samtidigt. Enligt simuleringsresultaten så kan det reglerade fordonet följa vägens mittlinje samt köra om de andra fordonen som färdas på samma väg.

In the era of Industry 4.0, conventional supply chains are undergoing a transformation into digital supply chains with the wide application of digital technologies such as big data, cloud computing, and Internet of Things. A digital supply chain is an intelligent and value-driven process that has superior features such as speed, flexibility, transparency, and real-time inventory monitoring and management. This concept is further included in the framework of Supply Chain 4.0, which emphasizes the connection between supply chain and Industry 4.0. In this context, data analytics for supply chain management presents a promising research opportunity. This thesis aims to investigate the use of data analytics in supply chain decision-making, including modelling, monitoring, and optimization. First, this thesis investigates supply chain monitoring (SCMo) using data analytics. The goal of SCMo is to raise an alarm when abnormal supply chain events occur and identify the potential reason. We propose a framework of SCMo based on a data-driven method, principal component analysis (PCA). Within this framework, supply chain data such as inventory levels and customer demand are collected, and the normal operating conditions of a supply chain are characterized using PCA. Fault detection and diagnosis are implemented by examining the monitoring statistics and variable contributions. A supply chain simulation model is developed to carry out the case studies. The results show that dynamic PCA (DPCA) successfully detected abnormal behaviour of the supply chain, such as transportation delay, low production rate, and supply shortage. Moreover, the contribution plot is shown to be effective in interpreting the abnormality and identify the fault-related variables. The method of using data-driven methods for SCMo is named data-driven SCMo in this work. Then, a further investigation of data-driven SCMo based on another statistical process monitoring method, canonical variate analysis (CVA), is conducted. CVA utilizes the state-space model of a system and determines the canonical states by maximizing the correlation between the combination of past system outputs and inputs and the combination of future outputs. A state-space model of supply chain is developed, which forms the basis of applying CVA to detect supply chain faults. The performance of CVA and PCA are assessed and compared in terms of dimensionality reduction, false alarm rate, missed detection rate, and detection delay. Case studies show that CVA identifies a smaller system order than PCA and achieves comparable performance to PCA in a lower-dimensional latent space. Next, we investigate data-driven supply chain control under uncertainty with risk taken into account. The method under investigation is reinforcement learning (RL). Within the RL framework, an agent learns an optimal policy that maps the state to action during the process of interacting with the non-deterministic environment, such that a numerical reward is maximized. The current literature regarding supply chain control focuses on conventional RL that maximizes the expected return. However, this may be not the best option for risk-averse decision makers. In this work, we explore the use of safe RL, which takes into account the concept of risk in the learning process. Two safe RL algorithms, Q-hat-learning and Beta-pessimistic Q-learning, are investigated. Case studies are carried out based on the supply chain simulator developed using agent-based modelling. Results show that Q-learning has the best performance under normal scenarios, while safe RL algorithms perform better under abnormal scenarios and are more robust to changes in the environment. Moreover, we find that the benefits of safe RL are more pronounced in a closed-loop supply chain. Finally, we investigate real-time supply chain optimization. The operational optimization problems for supply chains of realistic size are often large and complex, and solving them in real time can be challenging. This work aims to address the problem by using a deep learning-based model predictive control (MPC) technique. The MPC problem for supply chain operation is formulated based on the state space model of a supply chain, and the optimal state-input pairs are precomputed in the offline phase. Then, a deep neural network is built to map the state to input, which is then used in the online phase to reduce solution time. We propose an approach to implement the deep learning-based MPC method when there are delayed terms in the system, and a heuristic approach to feasibility recovery for mixed-integer MPC, with binary decision variables taken into account. Case studies show that compared with solving the nominal MPC problem online, deep learning-based MPC can provide near-optimal solution at a lower computational cost. / Thesis / Doctor of Philosophy (PhD)

Distributed generators (DGs) with integration of renewable resources (RRs) such as photovoltaic (PV) and wind turbine have been widely considered to reduce the dependency on conventional power generation systems along with enhancement of the quality and sustainability of the power system. Recently, DC microgrid has gained popularity in many real-world applications such as rural electrification due to its simplicity and low power losses. However, the power variability of renewable resources and continuous change in load demand imposes risks of power mismatch in standalone DC systems that increase the chances of stability and reliability issues. Therefore, complementary generation and/or storage systems are coupled with standalone DC microgrid to mitigate the power fluctuations and maintain a power balance in the system. This dissertation presents a power management strategy (PMS) based on model predictive control (MPC) for a standalone DC microgrid. A control scheme for a standalone DC microgrid system with RRs, storage, and load is desired to have the capability of effective power management that maximizes the extraction of energy from renewable generators, minimizes the transients in the system during disturbances, and protects the storage from over/under charging conditions. As a part of the proposed MPC, an optimization problem is formulated to meet the voltage performance in the system with respect to operating conditions and constraints. The proposed PMS uses the ARIMA prediction method to forecast the load and environmental parameters. The predicted parameters are utilized to estimate the future performance of the system by solving the dynamic model of the system, and a cost function is optimized to generate suitable control sequences. This research also presents detailed mathematical models of the considered systems. This dissertation presents an extensive simulation-based analysis of the proposed approach. With the proposed control, maximum utilization of the renewable generators has been achieved, and the DC bus voltage is regulated at nominal value with minimum transients under various load/environmental disturbances. Moreover, the research investigates the proposed MPC based on ARIMA prediction by comparing the performance of different types of prediction methods. The dissertation also measures the effectiveness of the proposed MPC by comparing its performance with a conventional PI controller.

No description available.

No description available.

No description available.

No description available.