Tato diplomová práce se zabývá revizí elektroniky robotického vozidla car4 se zaměřením na měření rychlosti otáčení kol, výkonovou a řídicí elektroniku. Car4 posloužilo jako základ pro již přes 20 diplomových prací od roku 2010. To znamenalo, že některé základní aspekty car4 byly zastaralé. Bylo vytvořeno schéma elektroniky vozidla, které bylo dále využito pro vývoj a výrobu nového hardwaru. Komponenty byly poté otestovány a implementovány na vozidlo. Dále byl vytvořen kinematický model 4WS vozidla za použití Ackermanovy geometrie, který byl implenetován a otestován na car4. Dále by měl sloužit jako základ řídicího algoritmu pro budoucí vývoj car4.

This thesis deals with the design and theoretical description of the hybrid control system for quadrocopter. First, a mathematical model is presented, sensors needed for development and their principles. This work shows the system architecture design and scheme of communication channels between the blocks. Implementation section discusses specific uses of computing nodes/sensors and their settings. In this thesis there are illustrated messages sent between computing nodes and the control loop machine. It also describes how to test the sensors, measurement results and techniques for improvement.

The goal of this master's thesis is design and built of platform for three-rotor helicopter development. The helicopter is also known as tricopter. Theoretical part describes principle of tricopter's flight and stabilization. There is also described basics inertial navigation system and sensors which are required for correct functionality. Practical part is dedicated to development of tricopter's frame, schematics diagram, communication between subsystems and stabilization system development. Flight stablization system is base on ATmega128A an using PID Controller. In the end is described testing of developed platform.

Le mélange de poudres est une opération essentielle dans divers domaines industriels, tels que les produits pharmaceutiques, agro-alimentaires, ciments, etc. Le mélange de poudres en continu, comme alternative intéressante au mélange conventionnel en batch, suscite beaucoup d'intérêt surtout en raison de sa capacité à gérer de grands volumes de production. Ce travail vise à contribuer à l'implémentation des applications de contrôle-commande. Avant de développer des stratégies de contrôle-commande, de nouveaux développements ont été présentés pour mieux comprendre le mélange continu de deux composants. La masse retenue et la répartition relative de la masse retenue de chaque composant dans le mélangeur ont été étudiées et prédites sous les différentes conditions opératoires. L'amélioration d'un modèle basé sur les chaînes de Markov a été finalement présentée pour prédire la concentration moyenne des mélanges obtenus à la sortie du mélangeur. L'implémentation d'un contrôleur PID a été expérimentalement réalisée comme une première tentative pour contrôler en temps réel l'homogénéité du mélange produit. La vitesse de rotation du mobile d'agitation, identifiée comme un facteur important influençant l'efficacité du mixer, a été considérée comme la variable manipulée. La commande en boucle fermée est basée, soit sur la concentration moyenne, soit sur le coefficient de variation. Les performances des boucles fermées proposées ont été évaluées pour le mélange continu subi à des changements d'échelon dans les débits d'alimentation du mélangeur. Quatre études de cas ont été définies et présentées. Le défi principal dans le système de contrôle-commande est le réglage des paramètres PID. La performance de commande en boucle fermée soit de la concentration moyenne, soit du coefficient de variation a été comparée à l'opération en boucle ouverte. / Powder mixing is an essential operation in various industrial fields, such as pharmaceuticals, agro-food, cements, etc. Continuous powder mixing, as an alternative to conventional batch mixing, has attracted a lot of interest mainly due to its capacity in handling high volume manufacturing. This work aims at the contribution to the implementation of process control applications for powder mixing in a pilot-scale continuous mixer. Prior to developing process control strategies, new developments have been presented for better understanding continuous mixing of two components. Hold-up weight and relative hold-up weight distribution of each component in the whole mixer have been experimentally investigated under different operating conditions. An improved Markov chain model has been finally presented to predict the mean concentration of the mixtures obtained at the mixer's outlet. The implementation of a proportional-integral-derivative (PID) controller has been experimentally performed as an initial attempt to real-time control the homogeneity of the mixture produced. The rotational speed of the stirrer, identified as an important deciding factor towards the mixer's efficiency, has been considered as the manipulated variable. The closed-loop control is based on either the mean concentration or the relative standard deviation. The performances of the proposed closed-loops have been evaluated for continuous mixing subjected to step changes in feed rates of the mixer. Four case studies have been defined and presented. The main challenge in the process control system is the tuning of PID parameters. The performance of closed-loop control of either the mean concentration or the relative standard deviation has been compared to open-loop operation.

<p>  </p> <p>Cavitation, liquid slosh, and splashes are ubiquitous in science and engineering. However, these phenomena are not fully understood. Yet to date, we do not understand when or why sometimes the splash seals, and other times does not. Regarding cavitation, a high temporal resolution method is needed to characterize this phenomenon. The low temporal resolution of experimental data suggests a model-based analysis of this problem. However, high-fidelity models are not always available, and even for these models, the sensitivity of the model outputs to the initial input parameters makes this method less reliable since some initial inputs are not experimentally measurable. As for sloshing, the air-liquid interface area and hydrodynamic stress for the liquid slosh inside a confined accelerating cylinder have not been experimentally measured due to the challenges for direct measurement.</p>

This thesis focuses on a control system for theater scenes. The system is based on technology of industrial computers and can handle theatre scene equipment, such as slot. The asynchronous AC engines and computer controlled frequency changers are used along with various sensors. The control is done via CANopen protocol and the software was written using Control Web.

This report will cover the work involved in building a quadruped robot which should to a certain extent resemeble a four-legged mammal. The introduction will present information related to what has inspired the project, purpose/aim, specifications, limitations and research questions. Most important in the introduction are probably the purpose/aim and research questions. Mainly, the goal was to have the robot be able to self-balance and later also be able to walk to some degree if possible. The first research question concerns how well a PID controller would affect the stability of the robot and the second question is about answering if referencing a fourlegged mammal is a good idea when building a quadruped robot. Theory and methods were combined and written as one chapter. This way relevant information could be provided in the appropriate places as the method for creating the robot was described. The chapter dives into the primary parts of the robot which are the choice of components, construction, inverse kinematics and last but not least the code including the implementation of a PID. The results show that PID regulation improves stability and performance, but PI regulation actually performed the best. Furthermore, the question regarding referencing a four-legged mammal remains inconclusive even though the model was sufficient for the task.

Self-stabilizing systems can be found in many contexts. They are used in aircraft and camera gimbals to name a few. In this project, a self-stabilizing container was constructed. The construction consists of three parts. An inner ring which rotates around the Z-axis, an outer ring which rotates around the Y-axis and a handle with space for three DC motors and a microcontroller. In this project an Arduino Nano was used. To detect inclination an IMU (Inertial Measurement Unit) was deployed. An IMU is a sensor consisting of three gyroscopes and three accelerometers, one for each coordinate axis. The software for the construction consists of four parts; angle reading, a Kalman filter, two PID-controllers and a motor controller. When a container is inserted into the construction the four-part system keeps the container horizontal and stable. Experimental data shows that in 84% of the tests the construction could stabilize the container. / Självstabiliserande system kan man finna i många olika sammanhang. Några exempel på självstabiliserande system är flygplan och kamerastabilisatorer. I detta projekt konstruerades en självstabiliserande behållare. Konstruktionen består av tre delar. En ring som kan rotera runt Z-axeln, en ring som kan rotera runt Y-axeln och ett handtag med plats för likströmsmotorer och mikrokontroller. I detta projekt användes Arduino Nano. För att avläsa vinklarna användes en tröghetsmäatare. En tröghetsmätare är en sensor som består av tre gyroskop och tre accelerometrar, en för varje axel. Mjukvaran i konstruktionen består av fyra delar; vinkelavläsning, ett Kalmanfilter, två PID-regulatorer och motorkontroller. Beroende på vilken vinkel konstruktionen har kommer någon av motorerna att korrigera vinkeln på behållaren. Testerna visade att konstruktionen kunde stabilisera behållaren i 84% av alla tester.

Recent advancements in processing power, storage capabilities, and availability of data, has led to improvements in many applications through the use of machine learning. Using machine learning in control systems was first suggested in the 1990s, but is more recently being implemented. In this thesis, an embedded climate system, which is a type of control system, will be looked at. The ways in which machine learning can be used to replicate portions of the climate system is looked at. Deep Belief Networks are the machine learning models of choice. Firstly, the functionality of a PID controller is replicated using a Deep Belief Network. Then, the functionality of a more complex control path is replicated. The performance of the Deep Belief Networks are evaluated at how they compare to the original control portions, and the performance in hardware. It is found that the Deep Belief Network can quite accurately replicate the behaviour of a PID controller, whilst the performance is worse for the more complex control path. It was seen that the use of delays in input features gave better results than without. A climate system with a Deep Belief Network was also loaded onto hardware. The minimum requirements of memory usage and CPU usage were met. However, the CPU usage was greatly affected, and if this was to be used in practice, work should be done to decrease it. / Många applikationer har förbättras genom användningen av maskininlärning. Maskininlärning för reglersystem föreslogs redan på 1990-talet och har nu börjat tillämpas, eftersom processorkraft, lagringsmöjligheter och tillgänglighet till rådata ökat. I detta examensarbete användes ett inbäddat klimatsystem, som är en typ av reglersystem. Maskininlärningsmodellen Deep Belief Network användes för att undersöka hur delar av klimatsystemet skulle kunna återskapas. Först återskapades funktionaliteten hos en PID-regulator och sedan funktionaliteten av en mer komplex del av reglersystemet Prestandan hos nätverken utvärderades i jämförelse med prestandan i de ursprungliga kontrolldelarna och hårdvaran. Det visade sig att Deep Belief Network utmärkt kunde replikera PID-regulatorns beteende, medan prestandan var lägre för den komplexa delen av reglersystemet. Användningen av fördröjningar i indata till nätverken gav bättre resultat än utan. Ett klimatsystem med ett Deep Belief Network laddades också över på hårdvaran. Minimikrav för minnesanvändning och CPU- användning var uppfyllda, men CPU- användningen påverkades kraftigt. Detta gör, att om maskininlärning ska kunna användas i verkligheten, bör CPU-användningen minskas.

One of the things augmented reality depends on is object tracking, which is a problem classically found in cinematography and security. However, the algorithms designed for the classical application are often too expensive computationally or too complex to run on simpler mobile hardware. One of the methods to do object tracking is with a trained neural network, this has already led to great results but is unfortunately still running into some of the same problems as the classical algorithms. For this reason a neural network designed specifically for object tracking on mobile hardware needs to be developed. This thesis will propose two di erent neural networks designed for object tracking on mobile hardware. Both are based on a siamese network structure and methods to improve their accuracy using filtering are also introduced. The first network is a modified version of “CNN architecture for geometric matching” that utilizes an a ne regression to perform object tracking. This network was shown to underperform in the MOT benchmark as-well as the VOT benchmark and therefore not further developed. The second network is an object detector based on “SqueezeDet” in a siamese network structure utilizing the performance optimized layers of “MobileNets”. The accuracy of the object detector network is shown to be competitive in the VOT benchmark, placing at the 16th place compared to trackers from the 2016 challenge. It was also shown to run in real-time on mobile hardware. Thus the one shot object detection network used for a tracking application can improve the experience of augmented reality applications on mobile hardware.