Practical control problems are subject to dealing with instrumentation noise and inaccurate models. These can be modelled as measurement and state noise, respectively. Nonlinear state estimators, for example a particle filter, can be used to mitigate these effects. However, they are usually computationally expensive which makes them impractical for industrial use. This text investigates using General Purpose Graphics Processing Units (GPGPU) to improve the performance particle and Gaussian sum filters by parallelizing their prediction, update and resampling steps. GPGPU accelerated filters are found to outperform non-accelerated filters as the number of particle increases. GPGPU acceleration also allows particle filters with 2^19.5 particles to be used on systems with dynamic time constants on the order of 0.1 second and for Gaussian sum filters with 2^18.5 particles to be used with time constants on the order of 1 second. The filters are applied to a bioreactor system containing R. Oryzae, where MPC control is applied to the production phase fumaric acid and glucose concentrations. The bioreactor is modelled using results from Iplik (2017) and Swart (2019). It is found that the GPGPU filters improved run times allow for more particles to be used which provides increased filter accuracy and thus better performance. This improved performance comes at the cost of consuming more energy. Thus, it is believed that the GPGPU implementations should be used for applications with complex dynamics/noise that require large numbers of particles and/or high sampling rates. / Dissertation (MEng (Control Engineering))--University of Pretoria, 2021. / Chemical Engineering / MEng (Control Engineering) / Unrestricted

Electric vehicles present an opportunity to reduce the substantial global footprint of road transportation. Cost and range anxiety issues, however, remain major roadblocks to their widespread adoption. One of the simplest ways to reduce cost is to remove components from the vehicle via novel topologies, estimation and control; to reduce range anxiety, charging infrastructure needs to be simplified and the power electronics in the vehicle made more efficient. This thesis proposes a bidirectional non-isolated fast charger integrated in the traction drivetrain of an electric vehicle that is enabled by a modular power electronic converter topology called the autoconverter module. The autoconverter module is an evolution of previous modular power electronic concepts with the goal of a highly integrated, high performance converter capable of being used in a number of applications through simple parallelization. By simplifying system design through the use of one base power conversion block, overall system cost can be reduced. Key to the realization of the power module is state estimation. To enable high performance operation of the system, low noise state information must be provided to the controller. State estimation is capable of filtering measurement noise to achieve this goal. However, conventional estimation techniques typically have low bandwidth and a convergence time associated with them, limiting the overall control system's performance. Higher performance techniques, such as receding horizon estimation, offer near-instantaneous estimation with noise rejection capabilities, which makes it an attractive solution. State estimators can also realize a cost reduction through the removal of sensors with little to no performance degradation. Using high performance state estimation and three autoconverter modules in parallel, a novel three-phase inverter/rectifier topology is conceived. Using this topology, a bidirectional non-isolated integrated fast charger capable of universal, i.e. single- and three-phase AC and DC, electric vehicle charging is realized. To interface with the AC power grid and AC traction motor, a novel three-phase common mode voltage controller is developed. By controlling the common mode voltage, leakage currents, which are generated by the fluctuation of the common mode voltage across a parasitic capacitance, can be attenuated and the transformer safely removed from the system. The removal of the transformer presents a significant cost and efficiency gain for both on-board chargers and dedicated charging units. With no transformer, the need for a dedicated on-board charger is obviated; instead, the existing high power traction inverter can be used as the primary charging interface, significantly reducing the cost, size and weight of on-board charging. High efficiency in charging mode is demonstrated, with a peak efficiency of 99.4% and an efficiency at rated power of 11kW of 98.4% shown. Traction mode efficiency with the proposed integrated charger is increased by 0.6% relative to a standard drive at the motor's rated power of 5kW. Damaging leakage currents and shaft voltages are reduced by over 90% because of the common mode voltage control, which will increase drive reliability and lifetime. The topology can be applied to motor drive applications outside automotive to increase efficiency and reliability. State estimation theory for permanent magnet synchronous machine drives is expanded upon and guarantees for estimatability and stability of the estimators are provided. Two estimation schemes, a Luenberger observer and a receding horizon estimator, are studied for sensor removal and the development of a failsafe operating mode involving one phase current sensor. Both estimators function equivalently in the steady state with the receding horizon estimator having slightly better transient performance. The Luenberger observer has conditions on estimatability, whereas the receding horizon estimator has none. Both estimators permit the removal of one current sensor for cost reduction. In regular operation, there is no performance degradation.

Le sujet principal de cette thèse est le mouvement des systèmes anthropomorphes, et plus particulièrement la locomotion bipède des humains et des robots humanoïdes. Pour caractériser et comprendre la locomotion bipède, il est instructif d'en étudier les causes, qui résident dans le contrôle et l'organisation du mouvement, et les conséquences qui en résultent, que sont le mouvement et les interactions physiques avec l'environnement. Concernant les causes, par exemple, quels sont les principes qui régissent l'organisation des ordres moteurs pour élaborer une stratégie de déplacement spécifique ? Puis, quelles grandeurs physiques pouvons-nous calculer pour décrire au mieux le mouvement résultant de ces commandes motrices ? Ces questions sont en partie abordées par la proposition d'une extension mathématique de l'approche du Uncontrolled Manifold au contrôle moteur de tâches dynamiques, puis par la présentation d'un nouveau descripteur de la locomotion anthropomorphe. En lien avec ce travail analytique vient le problème de l'estimation de l'état pour les systèmes anthropomorphes. La difficulté d'un tel problème vient du fait que les mesures apportent un bruit qui n'est pas toujours séparable des données informatives, et que l'état du système n'est pas nécessairement observable. Pour se débarrasser du bruit, des techniques de filtrage classiques peuvent être employées, mais elles sont susceptibles d'altérer le contenu des signaux d'intérêt. Pour faire face à ce problème, nous présentons une méthode récursive, basée sur le filtrage complémentaire, pour estimer la position du centre de masse et la variation du moment cinétique d'un système en contact, deux quantités centrales de la locomotion bipède. Une autre idée pour se débarrasser du bruit de mesure est de réaliser qu'il résulte en une estimation irréaliste de la dynamique du système. En exploitant les équations du mouvement, qui dictent la dynamique temporelle du système, et en estimant une trajectoire plutôt qu'un point unique, nous présentons ensuite une estimation du maximum de vraisemblance en utilisant l'algorithme de programmation différentielle dynamique pour effectuer une estimation optimale de l'état centroidal des systèmes en contact. Finalement, une réflexion pluridisciplinaire est présentée, sur le rôle fonctionnel et computationnel joué par la tête chez les animaux. La pertinence de son utilisation en robotique mobile y est discutée, pour l'estimation d'état et la perception multisensorielle. / The substantive subject of this thesis is the motion of anthropomorphic systems, and more particularly the bipedal locomotion of humans and humanoid robots. To characterize and understand bipedal locomotion, it is instructive to study its motor causes and its resulting physical consequences, namely, the interactions with the environment. Concerning the causes, for instance, what are the principles that govern the organization of motor orders in humans for elaborating a specific displacement strategy? And then, which physical quantities can we compute for best describing the motion resulting from these motor orders ? These questions are in part addressed by the proposal of a mathematical extension of the Uncontrolled Manifold approach for the motor control of dynamic tasks and through the presentation of a new descriptor of anthropomorphic locomotion. In connection with this analytical work, comes the problem of state estimation in anthropomorphic systems. The difficulty of such a problem comes from the fact that the measurements carry noise which is not always separable from the informative data, and that the state of the system is not necessarily observable. To get rid of the noise, classical filtering techniques can be employed but they are likely to distort the signals. To cope with this issue, we present a recursive method, based on complementary filtering, to estimate the position of the center of mass and the angular momentum variation of the human body, two central quantities of human locomotion. Another idea to get rid of the measurements noise is to acknowledge the fact that it results in an unrealistic estimation of the motion dynamics. By exploiting the equations of motion, which dictate the temporal dynamics of the system, and by estimating a trajectory versus a single point, we then present maximum likelihood estimation using the dynamic differential programming algorithm to perform optimal centroidal state estimation for systems in contact. Finally, a multidisciplinary reflection on the functional and computational role played by the head in animals is presented. The relevance of using this solution in mobile robotics is discussed, particularly for state estimation and multisensory perception.

In this thesis we have investigated different numerical filters for joint input and state estimation, with the aim of designing a robust algorithm capable of monitoring the continuous motion and loading in a truck chassis. The algorithm has to be able to use sparse measurements of the motion on different parts of the truck as it is excited by road induced vibrations, and transform this data into knowledge of the state in the entire system. To do this, the algorithm has to be supplied with information about the dynamic properties of the current system. In Paper A we have developed and implemented a fixed-lag smoother for joint input and state estimation in linear time-invariant dynamic structures. A fixed-lag smoother maximizes the use of information available in the measurements by allowing a small time lag in the estimation. As input, external forces as well as support motions can be computed. Furthermore, both measurement noise and model errors are accounted for and simulated as stochastic processes. The filter is firstly verified with straightforward numerical simulations of a simply supported beam, followed by a more involved simulation of a truck fuel tank. It is shown that the fixed-lag smoother performs very well, it estimates both input and states with a high accuracy even though the signals are contaminated with noise and the model contains errors. In Paper B the fixed-lag smoother is applied on real measurements. We investigate the capabilities of the proposed filter by analysing acceleration measurements from a truck side skirt excited by road induced vibrations. In this study, we focus on estimating the state in the side skirt body from a minimum number of measurement sensors. The dynamic properties of the side skirt are obtained experimentally from an operational modal analysis. It is shown that the fixed-lag smoother estimates the state very well. The results also shows that the smoothing effect is larger when fewer measurement sensors are used. / <p>QC 20160928</p>

In vehicle control practice, there are some variables, such as lateral tire force, body slip angle and yaw rate, that cannot or is hard to be measured directly and accurately. Vehicle model, like the bicycle model, offers an alternative way to get them indirectly, however due to the widely existent simplification and inaccuracy of vehicle models, there are always biases and errors in prediction from them. When developing advanced vehicle control functions, it is necessary and significant to know these variables in relatively high precision. Kalman filter offers a choice to estimate these variables accurately with measurable variables and with vehicle model together. In this thesis, estimation models based on Extended Kalman Filter (EKF) and Uncented Kalman Filter (UKF) are built separately to evaluate the lateral tire force, body slip angel and yaw rate of two typical passenger vehicles. Matlab toolbox EKF/UKF developed by Simo Särkkä, et al. is used to implement the estimation models. By comparing their principle, algorithm and results, the better one for vehicle state estimation will be chosen and justified. The thesis is organized in the following 4 parts: First, EKF and UKF are studied from their theory and features. Second, vehicle model used for prediction in Kalman filter is build and justified. Third, algorithms of EKF and UKF for this specific case are analysed. EKF and UKF are then implemented based on the algorithms with the help of Matlab toolbox EKF/UKF. Finally, comparisons between EKF and UKF are presented and discussed.

Au sein des systèmes avancés d’aide à la conduite (Advanced Driver Assistance Systems - ADAS) pour les systèmes de transport intelligents (Intelligent Transport Systems - ITS), les systèmes de positionnement, ou de localisation, du véhicule jouent un rôle primordial. Le système GPS (Global Positioning System) largement employé ne peut donner seul un résultat précis à cause de facteurs extérieurs comme un environnement contraint ou l’affaiblissement des signaux. Ces erreurs peuvent être en partie corrigées en fusionnant les données GPS avec des informations supplémentaires provenant d'autres capteurs. La multiplication des systèmes d’aide à la conduite disponibles dans les véhicules nécessite de plus en plus de capteurs installés et augmente le volume de données utilisables. Dans ce cadre, nous nous sommes intéressés à la fusion des données provenant de capteurs bas cout pour améliorer le positionnement du véhicule. Parmi ces sources d’information, en parallèle au GPS, nous avons considérés les caméras disponibles sur les véhicules dans le but de faire de l’odométrie visuelle (Visual Odometry - VO), couplée à une carte de l’environnement. Nous avons étudié les caractéristiques de cette trajectoire reconstituée dans le but d’améliorer la qualité du positionnement latéral et longitudinal du véhicule sur la route, et de détecter les changements de voies possibles. Après avoir été fusionnée avec les données GPS, cette trajectoire générée est couplée avec la carte de l’environnement provenant d’Open-StreetMap (OSM). L'erreur de positionnement latérale est réduite en utilisant les informations de distribution de voie fournies par OSM, tandis que le positionnement longitudinal est optimisé avec une correspondance de courbes entre la trajectoire provenant de l’odométrie visuelle et les routes segmentées décrites dans OSM. Pour vérifier la robustesse du système, la méthode a été validée avec des jeux de données KITTI en considérant des données GPS bruitées par des modèles de bruits usuels. Plusieurs méthodes d’odométrie visuelle ont été utilisées pour comparer l’influence de la méthode sur le niveau d'amélioration du résultat après fusion des données. En utilisant la technique d’appariement des courbes que nous proposons, la précision du positionnement connait une amélioration significative, en particulier pour l’erreur longitudinale. Les performances de localisation sont comparables à celles des techniques SLAM (Simultaneous Localization And Mapping), corrigeant l’erreur d’orientation initiale provenant de l’odométrie visuelle. Nous avons ensuite employé la trajectoire provenant de l’odométrie visuelle dans le cadre de la détection de changement de voie. Cette indication est utile dans pour les systèmes de navigation des véhicules. La détection de changement de voie a été réalisée par une somme cumulative et une technique d’ajustement de courbe et obtient de très bon taux de réussite. Des perspectives de recherche sur la stratégie de détection sont proposées pour déterminer la voie initiale du véhicule. En conclusion, les résultats obtenus lors de ces travaux montrent l’intérêt de l’utilisation de la trajectoire provenant de l’odométrie visuelle comme source d’information pour la fusion de données à faible coût pour la localisation des véhicules. Cette source d’information provenant de la caméra est complémentaire aux données d’images traitées qui pourront par ailleurs être utilisées pour les différentes taches visée par les systèmes d’aides à la conduite. / With the growing research on Advanced Driver Assistance Systems (ADAS) for Intelligent Transport Systems (ITS), accurate vehicle localization plays an important role in intelligent vehicles. The Global Positioning System (GPS) has been widely used but its accuracy deteriorates and susceptible to positioning error due to factors such as the restricting environments that results in signal weakening. This problem can be addressed by integrating the GPS data with additional information from other sensors. Meanwhile, nowadays, we can find vehicles equipped with sensors for ADAS applications. In this research, fusion of GPS with visual odometry (VO) and digital map is proposed as a solution to localization improvement with low-cost data fusion. From the published works on VO, it is interesting to know how the generated trajectory can further improve vehicle localization. By integrating the VO output with GPS and OpenStreetMap (OSM) data, estimates of vehicle position on the map can be obtained. The lateral positioning error is reduced by utilizing lane distribution information provided by OSM while the longitudinal positioning is optimized with curve matching between VO trajectory trail and segmented roads. To observe the system robustness, the method was validated with KITTI datasets tested with different common GPS noise. Several published VO methods were also used to compare improvement level after data fusion. Validation results show that the positioning accuracy achieved significant improvement especially for the longitudinal error with curve matching technique. The localization performance is on par with Simultaneous Localization and Mapping (SLAM) SLAM techniques despite the drift in VO trajectory input. The research on employability of VO trajectory is extended for a deterministic task in lane-change detection. This is to assist the routing service for lane-level direction in navigation. The lane-change detection was conducted by CUSUM and curve fitting technique that resulted in 100% successful detection for stereo VO. Further study for the detection strategy is however required to obtain the current true lane of the vehicle for lane-level accurate localization. With the results obtained from the proposed low-cost data fusion for localization, we see a bright prospect of utilizing VO trajectory with information from OSM to improve the performance. In addition to obtain VO trajectory, the camera mounted on the vehicle can also be used for other image processing applications to complement the system. This research will continue to develop with future works concluded in the last chapter of this thesis.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第20024号 / 情博第619号 / 新制||情||108(附属図書館) / 33120 / 京都大学大学院情報学研究科知能情報学専攻 / (主査)教授 松山 隆司, 教授 熊田 孝恒, 准教授 川嶋 宏彰 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

No description available.

No description available.

Distribution load flow and state estimation are two important functions in distribution energy management systems (DEMS) and advanced distribution automation (ADA) systems. Distribution load flow analysis is a tool which helps to analyze the status of a power distribution system under steady-state operating conditions. In this research, an effective and comprehensive load flow algorithm is developed to extensively incorporate the distribution system components. Distribution system state estimation is a mathematical procedure which aims to estimate the operating states of a power distribution system by utilizing the information collected from available measurement devices in real-time. An efficient and computationally effective state estimation algorithm adapting the weighted-least-squares (WLS) method has been developed in this research. Both the developed algorithms are tested on different I testeeders and the results obtained are justified.