With the growing use and complexity of proton therapy, the safety and accuracy of the machines becomes increasing important. This, to be able to deliver the prescribed dose to the target while minimizing the dose to healthy tissue. In this project, machine quality assurance data are analyzed to quantify the existing positional machine uncertainties in the form of deviations from expected value and their effect on the dose accuracy in order to improve precision. The method consisted of two main parts. In the first part, two systems to monitor the measured deviations variations from the machine quality assurance tests were implemented. In the second part, two ways to measure the impact of the positional machine uncertainties were developed. The monitoring systems showed that the uncertainties had shrunken over time or were stable, and that the tolerance limits currently used for the machine quality assurance can be lowered. The measured impact of the positional machine uncertainties showed that a margin of 0.61 mm for treatment room 1 and a margin of 1.02 mm for treatment room 2 was required to compensated for the machine uncertainties. When the uncertainties we reincorporated into a clinical approved robust optimized plan, the result showed no significant change in dose to the different treatment volumes. The result gives the Scandion clinic insight and tools to minimize the impact of machine uncertainties and to be able to improve the precision of future treatments.

The concept of portfolio optimization has been widely studied in the academy and implemented in the financial markets since its introduction by Markowitz 70 years ago. The problem of the mean-variance optimization framework caused by input uncertainty has been one of the foci in the previous research. In this study, several models (linear shrinkage and Black-Litterman) based on Bayesian approaches are studied to improve the estimation of inputs. Moreover, a new framework based on robust optimization is presented to mitigate the input uncertainty further.  An out-of-sample test is specially designed, and the results show that Bayesian models in this study can improve the optimization results in terms of higher Sharpe ratios (the quotient between portfolio returns and their risks). Both covariance matrix estimators based on the linear shrinkage method contain less error and provide better optimization results, i.e. higher Sharpe ratios. The Black-Litterman model with a proper choice of inputs can significantly improve the portfolio return. The new framework based on the combination of shrinkage estimators, Black-Litterman, and robust optimization presents a better way for portfolio optimization than the classical framework of mean-variance optimization.

In modern manufacturing industries, many applications require precision motion control of multi-agent systems, like multi-joint robot arms and multi-axis machine tools. Cutter (end effector) should stay as close as possible to the reference trajectory to ensure the quality of the final products. In conventional computer numerical control (CNC), the control unit of each axis is independently designed to achieve the best individual tracking performance. However, this becomes less effective when dealing with multi-axis contour following tasks because of the lack of coordination among axes. This dissertation studies the control of multi-axis machine tools with focus on reducing the contour error. The proposed research explicitly addresses the minimization of contour error and treats the multi-axis machine tool as a multi-input-multi-output (MIMO) system instead of several decoupled single-input-single-output (SISO) systems. New control schemes are developed to achieve superior contour following performance even in the presence of disturbances. This study also extends the applications of the proposed control system from plane contours to regular contours in R3. The effectiveness of the developed control systems is experimentally verified on a micro milling machine.

This dissertation presents design challenges and promising solutions for temperature and pressure sensors which are highly desirable for harsh-environment applications, such as turbine engines. To survive the harsh environment consisting of high temperatures above 1000°C, high pressures around 300 psi, and corrosive gases, the sensors are required to be robust both electrically and mechanically. In addition, wire connection of the sensors is a challenging packaging problem, which remains unresolved as of today. In this dissertation, robust ceramic sensors are demonstrated for both high temperature and pressure measurements. Also, the wireless sensors are achieved based on microwave resonators. Two types of temperature sensors are realized using integrated resonator/antennas and reflective patches, respectively. Both types of the sensors utilize alumina substrate which has a temperature-dependent dielectric constant. The temperature in the harsh environment is wirelessly detected by measuring the resonant frequency of the microwave resonator, which is dependent on the substrate permittivity. The integrated resonator/antenna structure minimizes the sensor dimension by adopting a seamless design between the resonator sensor and antenna. This integration technique can be also used to achieve an antenna array integrated with cavity filters. Alternatively, the aforementioned reflective patch sensor works simultaneously as a resonator sensor and a radiation element. Due to its planar structure, the reflective patch sensor is easy for design and fabrication. Both temperature sensors are measured above 1000°C. A pressure sensor is also demonstrated for high-temperature applications. Pressure is detected via the change in resonant frequency of an evanescent-mode resonator which corresponds to cavity deformation under gas pressure. A compact sensor size is achieved with a post loading the cavity resonator and a low-profile antenna connecting to the sensor. Polymer-Derived-Ceramic (PDC) is developed and used for the sensor fabrication. The pressure sensor is characterized under various pressures at high temperatures up to 800°C. In addition, to facilitate sensor characterizations, a robust antenna is developed in order to wirelessly interrogate the sensors. This specially-developed antenna is able to survive a record-setting temperature of 1300°C.

The control of systems with uncertain nonlinear dynamics is an important field of control science attracting decades of focus. In this dissertation, four different control strategies are presented using sliding mode control, adaptive control, dynamic compensation, and neural network for a nonlinear aeroelastic system with bounded uncertainties and external disturbance. In Chapter 2, partial state feedback adaptive control designs are proposed for two different aeroelastic systems operating in unsteady flow. In Chapter 3, a continuous robust control design is proposed for a class of single input and single output system with uncertainties. An aeroelastic system with a trailingedge flap as its control input will be considered as the plant for demonstration of effectiveness of the controller. The controller is proved to be robust by both athematical proof and simulation results. In Chapter 3, a robust output feedback control strategy is discussed for the vibration suppression of an aeroelastic system operating in an unsteady incompressible flowfield. The aeroelastic system is actuated using a combination of leading-edge (LE) and trailing-edge (TE) flaps in the presence of different kinds of gust disturbances. In Chapter 5, a neural-network based model-free controller is designed for an aeroelastic system operating at supersonic speed. The controller is shown to be able to effectively asymptotically stabilize the system via both a Lyapunov-based stability proof and numerical simulation results.

In this dissertation. we document the progress in the control design for a class of MIMO nonlinear uncertain system from five papers. In the first part, we address the problem of adaptive control design for a class of multi-input multi-output (MIMO) nonlinear systems. A Lypaunov based singularity free control law, which compensates for parametric uncertainty in both the drift vector and the input gain matrix, is proposed under the mild assumption that the signs of the leading minors of the control input gain matrix are known. Lyapunov analysis shows global uniform ultimate boundedness (GUUB) result for the tracking error under full state feedback (FSFB). Under the restriction that only the output vector is available for measurement, an output feedback (OFB) controller is designed based on a standard high gain observer (HGO) stability under OFB is fostered by the uniformity of the FSFB solution. Simulation results for both FSFB and OFB controllers demonstrate the efcacy of the MIMO control design in the classical 2-DOF robot manipulator model. In the second part, an adaptive feedback control is designed for a class of MIMO nonlinear systems containing parametric uncertainty in both the drift vector and the input gain matrix, which is assumed to be full-rank and non-symmetric in general. Based on an SDU decomposition of the gain matrix, a singularity-free adaptive tracking control law is proposed that is shown to be globally asymptotically stable (GAS) under full-state feedback. iii Output feedback results are facilitated via the use of a high-gain observer (HGO). Under output feedback control, ultimate boundedness of the error signals is obtained the size of the bound is related to the size of the uncertainty in the parameters. An explicit upper bound is also provided on the size of the HGO gain constant. In third part, a class of aeroelastic systems with an unmodeled nonlinearity and external disturbance is considered. By using leading- and trailing-edge control surface actuations, a full-state feedforward/feedback controller is designed to suppress the aeroelastic vibrations of a nonlinear wing section subject to external disturbance. The full-state feedback control yields a uniformly ultimately bounded result for two-axis vibration suppression. With the restriction that only pitching and plunging displacements are measurable while their rates are not, a high-gain observer is used to modify the full-state feedback control design to an output feedback design. Simulation results demonstrate the ef cacy of the multi-input multioutput control toward suppressing aeroelastic vibration and limit cycle oscillations occurring in pre and post utter velocity regimes when the system is subjected to a variety of external disturbance signals. Comparisons are drawn with a previously designed adaptive multi-input multi-output controller. In the fourth part, a continuous robust feedback control is designed for a class of high-order multi-input multi-output (MIMO) nonlinear systems with two degrees of freedom containing unstructured nonlinear uncertainties in the drift vector and parametric uncertainties in the high frequency gain matrix, which is allowed to be non-symmetric in general. Given some mild assumptions on the system model, a singularity-free continuous robust tracking coniv trol law is designed that is shown to be semi-globally asymptotically stable under full-state feedback through a Lyapunov stability analysis. The performance of the proposed algorithm have been verified on a two-link robot manipulator model and 2-DOF aeroelastic model.

Syftet med studien är att undersöka pedagogers syn på konflikter mellan kollegor i förskolan. Genom studien önskar vi belysa konflikter mellan pedagoger i förskolan samt vilken inverkan konflikter kan ha på pedagogers samspel, trivsel och välmående. Vi fann här ett GAP i den forskning som vi hittat inom området konflikt och konflikthantering.   Vi använder oss av en kvantitativ metod genom enkäter för att samla in empirisk data. Enkäterna besvarades av 171 pedagoger i 28 skånska kommuner. Den insamlade empirin analyseras genom deskriptiv statistik och tolkas utifrån Jordans (2015) begrepp: robust samarbetskultur och självinsikt, konfliktnivåerna: individnivå, relationsnivå och systemnivå samt Galtungs ABC-modell.   Resultatet visar att det i stor utsträckning förekommer konflikter mellan kollegor i förskolan samt att dessa kan komma att inverka på samspel, trivseln och välmåendet om de förblir olösta. Trivsel anses vara högt prioriterat av respondenterna och trots förekomsten av konflikter verkar respondenterna trivas på arbetsplatsen. Utifrån resultatet drar vi slutsatsen att majoriteten av respondenterna uppfattar begreppet konflikt som negativt betingat. En annan slutsats är att det behöver finnas kunskap om konflikter och konflikthantering kollegor emellan för att kunna skapa ett gott sampel samt ökad trivseln och välmåendet hos pedagoger på arbetsplatsen.

Localization is a fundamental part of achieving fully autonomous vehicles. A localization system needs to constantly provide accurate information about the position of the vehicle and failure could lead to catastrophic consequences. Global Navigation Satellite Systems (GNSS) can supply accurate positional measurements but are susceptible to disturbances and outages in environments such as indoors, in tunnels, or nearby tall buildings. A common method called simultaneous localization and mapping (SLAM) creates a spatial map and simultaneously determines the position of a robot or vehicle. Utilizing different sensors for localization can increase the accuracy and robustness of such a system if used correctly. This thesis uses a graph-based version of SLAM called graph SLAM which stores previous measurements in a factor graph, making it possible to adjust the trajectory and map as new information is gained. The best position state estimation is gained by optimizing the graph representing the log-likelihood of the data. To treat GNSS outliers in a graph SLAM system, robust optimization techniques can be used, and this thesis investigates two techniques called realizing, reversing, recovering (RRR), and dynamic covariance scaling (DCS). High-end GNSS and Lidar sensors are used to gather a data set on a suburban public road. Information about the position and orientation of the vehicle are inferred from the data set using graph SLAM together with robust techniques in three different scenarios. The scenarios contain disturbances called multipathing, Gaussian disturbances, and outages. A parameter study examines the free parameters Φ in DCS and the p-value in the RRR method. The localization performance varies less when changing the free parameter in RRR than in DCS. The localization performance from RRR is consistent for most values of p. DCS shows greater variation in the localization performance for different values of Φ. In the tested cases, results conclude that Φ should be set to 2.5 for the most consistent localization across all states. RRR performed best with a p-value set to 0.85. A lower value led to too many discarded measurements which decreased performance. DCS outperforms RRR across the tested scenarios but further testing is needed to determine whether RRR is better suited for handling larger errors. / Lokalisering är en fundamental del i att uppnå självkörande fordon. Lokaliseringssytemets uppgift är att kontinuerligt förse exakt information om fordonets position, och vid fel kan detta leda till katastrofala följder. Global Navigation Satellite Systems (GNSS) används ofta i ett lokaliseringssystem för att uppnå exakta positionsmätningar, men i vissa miljöer så som parkeringshus, tunnlar eller storstäder kan störningar uppstå. Genom att förlita sig på fler typer av sensorer kan lokaliseringen bli mer noggrann och robust mot störningar. En vanlig metod som kan skatta ett fordons position och samtidigt skapa en karta över omgivningen är simultaneous localization and mapping SLAM. I detta examensarbete används graph SLAM, en version av SLAM som utnyttjar en faktorgraf för att representera mätvärden och sedan estimera position av fordonet. Robusta metoder kan användas inom SLAM för hantering av felaktiga mätningar i ett grafbaserat SLAM-nätverk, och här undersöks två metoder, realizing, reversing, recovering (RRR) och dynamic covariance scaling DCS. Data från GNSS och Lidarsensorer av hög kvalitet samlades in på en offentlig väg i stadsmiljö. I tre olika scenarion beräknas testfordonets position och orientering med graph SLAM tillsammans med de två robusta metoderna som undersöks. Scenarion utgör fall med olika typer av störningar som agerar på gnss-mätningarna. Störningarna är av typerna multipath, Gaussiskt brus, samt avbrott. DCS presterar bättre jämfört med RRR under de tester som utförts. En parameterstudie har utförts som undersöker parametern Φ i DCS och p i RRR. När Φ varieras i DCS ger det en större skillnad på resultatet än när p varieras i RRR. Detta indikerar att det är lättare att hantera och använda RRR optimalt. Trots att DCS presterar bättre än RRR i de testade fallen, krävs vidare undersökning för att besluta om RRR hanterar stora fel bättre än DCS. De bästa inställningarna visades vara 2,5 för Φ i DCS och större än 0,85 för p i RRR.

This dissertation develops a comprehensive statistical learning framework that is robust to (distributional) perturbations in the data using Distributionally Robust Optimization (DRO) under the Wasserstein metric. The learning problems that are studied include: (i) Distributionally Robust Linear Regression (DRLR), which estimates a robustified linear regression plane by minimizing the worst-case expected absolute loss over a probabilistic ambiguity set characterized by the Wasserstein metric; (ii) Groupwise Wasserstein Grouped LASSO (GWGL), which aims at inducing sparsity at a group level when there exists a predefined grouping structure for the predictors, through defining a specially structured Wasserstein metric for DRO; (iii) Optimal decision making using DRLR informed K-Nearest Neighbors (K-NN) estimation, which selects among a set of actions the optimal one through predicting the outcome under each action using K-NN with a distance metric weighted by the DRLR solution; and (iv) Distributionally Robust Multivariate Learning, which solves a DRO problem with a multi-dimensional response/label vector, as in Multivariate Linear Regression (MLR) and Multiclass Logistic Regression (MLG), generalizing the univariate response model addressed in DRLR. A tractable DRO relaxation for each problem is being derived, establishing a connection between robustness and regularization, and obtaining upper bounds on the prediction and estimation errors of the solution. The accuracy and robustness of the estimator is verified through a series of synthetic and real data experiments. The experiments with real data are all associated with various health informatics applications, an application area which motivated the work in this dissertation. In addition to estimation (regression and classification), this dissertation also considers outlier detection applications.

<p dir="ltr">Engine downsizing with the assistance of turbomachinery and/or energy storage system has been realized to be one of the most promising and cost-effective solutions in pursuit of cleaner and more efficient engine products. Fundamental challenges however, exist in terms of control and energy management of advanced powertrain featuring turbocharged engines due to their complex dynamics, inherent coupling nature, and strict emission regulations concerning environmental preservation. For the purpose of addressing those challenges, this dissertation develops an integrated framework for modeling, robust coordinated control, and power management of advanced powertrains featuring turbocharged engines.</p><p dir="ltr">This dissertation first studies an advanced turbocharged lean-burn SI natural gas engine manufactured by Caterpillar, and develops an intuitive physics-based, control-oriented model. The obtained control-oriented model is validated against a high-fidelity truth-reference model and serves as the basis on which a robust coordinated control system is developed. The dissertation then proposes a comprehensive procedure for synthesizing a robust coordinated control system applying optimization-based H_infinity control theory. Specifically, this framework outlines a methodology of modeling uncertainties to account for system robustness, and providing valuable insights into the tuning of general coordinated control system design. For performance testing, the synthesized robust coordinated control system is implemented on the high-fidelity truth-reference model. A parallel closed-loop simulation strategy is adopted so that direct comparison between the robust coordinated control system and benchmark production control system (composed of multiple fine-tuned PID controllers) developed by Caterpillar can be carried out. Simulation results manage to demonstrate the merit of utilizing the robust coordinated control system, with better performances observed in terms of steady-state tracking, transient response, and disturbance attenuation.</p><p dir="ltr">The second part of this dissertation focuses on the development of a proposed novel hybrid electric wheel loader which features a downsized engine assisted by turbocharger and an energy storage system. Research efforts documented in this dissertation involve system configuration, controller design (both component-level and supervisory-level), simulation development (both software-in-the-loop and hardware-in-the-loop) and simulated validation for the proposed novel wheel loader. Inspired by the successful simulation results, John Deere assembled a real demo vehicle with the proposed powertrain and conducted some in-field testing, from which encouraging experimental results are observed.</p>