Vision and infrared sensors are very common in surveillance and security applications, and there are numerous examples where a critical infrastructure, e.g. a harbor, an airport, or a military camp, is monitored by video surveillance systems. There is a need for automatic processing of sensor data and intelligent control of the sensor in order to obtain efficient and high performance solutions that can support a human operator. This thesis considers two subparts of the complex sensor fusion system; namely target tracking and sensor control.The multiple target tracking problem using particle filtering is studied. In particular, applications where road constrained targets are tracked with an airborne video or infrared camera are considered. By utilizing the information about the road network map it is possible to enhance the target tracking and prediction performance. A dynamic model suitable for on-road target tracking with a camera is proposed and the computational load of the particle filter is treated by a Rao-Blackwellized particle filter. Moreover, a pedestrian tracking framework is developed and evaluated in a real world experiment. The exploitation of contextual information, such as road network information, is highly desirable not only to enhance the tracking performance, but also for track analysis, anomaly detection and efficient sensor management. Planning for surveillance and reconnaissance is a broad field with numerous problem definitions and applications. Two types of surveillance and reconnaissance problems are considered in this thesis. The first problem is a multi-target search and tracking problem. Here, the task is to control the trajectory of an aerial sensor platform and the pointing direction of its camera to be able to keep track of discovered targets and at the same time search for new ones. The key to successful planning is a measure that makes it possible to compare different tracking and searching tasks in a unified framework and this thesis suggests one such measure. An algorithm based on this measure is developed and simulation results of a multi-target search and tracking scenario in an urban area are given. The second problem is aerial information exploration for single target estimation and area surveillance. In the single target case the problem is to control the trajectory of a sensor platform with a vision or infrared camera such that the estimation performance of the target is maximized. The problem is treated both from an information filtering and from a particle filtering point of view. In area exploration the task is to gather useful image data of the area of interest by controlling the trajectory of the sensor platform and the pointing direction of the camera. Good exploration of a point of interest is characterized by several images from different viewpoints. A method based on multiple information filters is developed and simulation results from area and road exploration scenarios are presented.

We investigate sensor control and scheduling strategies to most effectively use the limited resources of an ad hoc network or closed-loop sensor network. In this context, we examine the following three problems. Where to focus sensing? Certain types of sensors, such as cameras or radars, are unable to simultaneously collect high fidelity data from all environmental locations, and thus require some sort of sensing strategy. Considering a meteorological radar network, we show that the main benefits of optimizing sensing over expected future states of the environment are when there are multiple small phenomena in the environment. Considering multiple users, we show that the problem of call admission control (i.e., deciding which sensing requests to satisfy) in the context of a virtualized private sensor network can be solved in polynomial time when sensor requests are divisible or fixed in time. When sensor requests are indivisible but may be shifted in time, we show that the call admission control problem is NP-complete. How to make sensing robust to delayed and dropped packets? In a closed-loop sensor network, data collected by the sensors determines each sensor's future data collection strategy. Network delays, however, constrain the quantity of data received by the time a control decision must be made, and consequently affect the quality of the computed sensor control. We investigate the value of separate handling of sensor control and data traffc, during times of congestion, in a closed-loop sensor network. Grounding our analysis in a meteorological radar network, we show that prioritizing sensor control traffc decreases the round-trip control-loop delay, and thus increases the quantity and quality of the collected data and improves application performance. How to make routing robust to network changes? In wireless sensor and mobile ad-hoc networks, variable link characteristics and node mobility give rise to changing network conditions. We propose a routing algorithm that selects a type of routing subgraph (a braid) that is robust to changes in the network topology. We analytically characterize the reliability of a class of braids and their optimality properties, and give counter-examples to other conjectured optimality properties in a well-structured (grid) network. Comparing with dynamic source routing, we show that braid routing can significantly decrease control overhead while only minimally degrading the number of packets delivered, with gains dependent on node density.

This thesis will present the design, implementation, and testing of a high voltage Charge-Coupled Device (CCD) controller ASIC for the Large Synoptic Survey Telescope (LSST), which will be used to study dark energy and dark matter. The LSST observatory, which includes a 3.2-gigapixel camera, will cover the entire sky every three nights by taking continuous 15-second exposures. The CCD controller ASIC, or Sensor Control Chip (SCC), will provide five CCD driver channels that are capable of generating serial or parallel clock signals for the LSST’s imaging sensors during readout mode. The SCC will also provide three programmable bias voltages for the CCDs along with eight supplementary programmable voltages and currents for the CCD’s output drain terminals. Additionally, the controller ASIC includes eight control signals for a separate Analog Signal Processing Integrated Circuit (ASPIC) that is designed as the readout chip for LSST. The SCC is designed to operate down to 153 K. Fabricated in a commercially available 0.8-micron Bipolar-CMOS-DMOS Silicon-On-Insulator (BCD-SOI) process, the SCC has been verified to meet all design requirements.

This thesis will present the design, implementation, and testing of a high voltage Charge-Coupled Device (CCD) controller ASIC for the Large Synoptic Survey Telescope (LSST), which will be used to study dark energy and dark matter. The LSST observatory, which includes a 3.2-gigapixel camera, will cover the entire sky every three nights by taking continuous 15-second exposures. The CCD controller ASIC, or Sensor Control Chip (SCC), will provide five CCD driver channels that are capable of generating serial or parallel clock signals for the LSST’s imaging sensors during readout mode. The SCC will also provide three programmable bias voltages for the CCDs along with eight supplementary programmable voltages and currents for the CCD’s output drain terminals. Additionally, the controller ASIC includes eight control signals for a separate Analog Signal Processing Integrated Circuit (ASPIC) that is designed as the readout chip for LSST. The SCC is designed to operate down to 153 K. Fabricated in a commercially available 0.8-micron Bipolar-CMOS-DMOS Silicon-On-Insulator (BCD-SOI) process, the SCC has been verified to meet all design requirements.

<p>Sensor path planning and control refer to the problems of determining the trajectory and feedback control law that best support sensing objectives, such as monitoring, detection, classification, and tracking. Many autonomous systems developed, for example, to conduct environmental monitoring, search-and-rescue operations, demining, or surveillance, consist of a mobile vehicle instrumented with a suite of proprioceptive and exteroceptive sensors characterized by a bounded field-of-view (FOV) and a performance that is highly dependent on target and environmental conditions and, thus, on the vehicle position and orientation relative to the target and the environment. As a result, the sensor performance can be significantly improved by planning the vehicle motion and attitude in concert with the measurement sequence. This dissertation develops a general and systematic approach for deriving information-driven path planning and control methods that maximize the expected utility of the sensor measurements subject to the vehicle kinodynamic constraints.</p><p>The approach is used to develop three path planning and control methods: the information potential method (IP) for integrated path planning and control, the optimized coverage planning based on the Dirichlet process-Gaussian process (DP-GP) expected Kullback-Leibler (KL) divergence, and the optimized visibility planning for simultaneous target tracking and localization. The IP method is demonstrated on a benchmark problem, referred to as treasure hunt, in which an active vision sensor is mounted on a mobile unicycle platform and is deployed to classify stationary targets characterized by discrete random variables, in an obstacle-populated environment. In the IP method, an artificial potential function is generated from the expected conditional mutual information of the targets and is used to design a closed-loop switched controller. The information potential is also used to construct an information roadmap for escaping local minima. Theoretical analysis shows that the closed-loop robotic system is asymptotically stable and that an escaping path can be found when the robotic sensor is trapped in a local minimum. Numerical simulation results show that this method outperforms rapidly-exploring random trees and classical potential methods. The optimized coverage planning method maximizes the DP-GP expected KL divergence approximated by Monte Carlo integration in order to optimize the information value of a vision sensor deployed to track and model multiple moving targets. The variance of the KL approximation error is proven to decrease linearly with the inverse of the number of samples. This approach is demonstrated through a camera-intruder problem, in which the camera pan, tilt, and zoom variables are controlled to model multiple moving targets with unknown kinematics by nonparametric DP-GP mixture models. Numerical simulations as well as physical experiments show that the optimized coverage planning approach outperforms other applicable algorithms, such as methods based on mutual information, rule-based systems, and randomized planning. The third approach developed in this dissertation, referred to as optimized visibility motion planning, uses the output of an extended Kalman filter (EKF) algorithm to optimize the simultaneous tracking and localization performance of a robot equipped with proprioceptive and exteroceptive sensors, that is deployed to track a moving target in a global positioning system (GPS) denied environment.</p><p>Because active sensors with multiple modes can be modeled as a switched hierarchical system, the sensor path planning problem can be viewed as a hybrid optimal control problem involving both discrete and continuous state and control variables. For example, several authors have shown that a sensor with multiple modalities is a switched hybrid system that can be modeled by a hierarchical control architecture with components of mission planning, trajectory planning, and robot control. Then, the sensor performance can be represented by two Lagrangian functions, one function of the discrete state and control variables, and one function of the continuous state and control variables. Because information value functions are typically nonlinear, this dissertation also presents an adaptive dynamic programming approach for the model-free control of nonlinear switched systems (hybrid ADP), which is capable of learning the optimal continuous and discrete controllers online. The hybrid ADP approach is based on new recursive relationships derived in this dissertation and is proven to converge to the solution of the hybrid optimal control problem. Simulation results show that the hybrid ADP approach is capable of converging to the optimal controllers by minimizing the cost-to-go online based on a fully observable state vector.</p> / Dissertation

An intelligent sensor system has the potential of providing its operator with relevant information, lowering the risk of human errors, and easing the operator's workload. One way of creating such a system is by using reinforcement learning, and this thesis studies how reinforcement learning can be applied to a simple sensor control task within a detailed 3D rendered environment. The studied agent controls a stationary camera (pan, tilt, zoom) and has the task of finding stationary targets in its surrounding environment. The agent is end-to-end, meaning that it only uses its sensory input, in this case images, to derive its actions. The aim was to study how an agent using a simple neural network performs on the given task and whether behavior cloning can be used to improve the agent's performance. The best-performing agents in this thesis developed a behavior of rotating until a target came into their view. Then they directed their camera to place the target at the image center. The performance of these agents was not perfect, their movement contained quite a bit of randomness and sometimes they failed their task. But even though the performance was not perfect, the results were positive since the developed behavior would be able to solve the task efficiently given that it is refined. This indicates that the problem is solvable using methods similar to ours. The best agent using behavior cloning performed on par with the best agent that did not use behavior cloning. Therefore, behavior cloning did not lead to improved performance.

The ability to search their environment is one of the most fundamental skills for any living creature. Visual search in particular is abundantly common for almost all animals. This act of searching is generally active in nature, with vision not simply reacting to incoming stimuli but also actively searching the environment for potential stimuli (such as by moving their head or eyes). Automatic visual search, likewise, is a crucial and powerful tool within a wide variety of different fields. However, performing such an active search is a nontrivial issue for many machine learning approaches. The added complexity of choosing which area to observe, as well as the common case of having a camera with adaptive field-of-view capabilities further complicates the problem. Hierarchical Reinforcement Learning have in recent years proven to be a particularly powerful means of solving hard machine learning problems by a divide-and-conquer methodology, where one highly complex task can be broken down into smaller sub-tasks which on their own may be more easily learnable. In this thesis, we present a hierarchical reinforcement learning system for solving a visual search problem in a stationary camera environment with adjustable pan, tilt and field-of-view capabilities. This hierarchical model also incorporates non-reinforcement learning agents in its workflow to better utilize the strengths of different agents and form a more powerful overall model. This model is then compared to a non-hierarchical baseline as well as some learning-free approaches.

Es ist ein politisch erklärtes Ziel, den Ausstoß von klimaschädlichen Treibhausgasen weltweit zu verringern. Eine wesentliche Stellschraube im Gebiet des Bauwesens stellt hierbei die Einsparung von Energien zur Raumkonditionierung dar. Diese wird unter anderem über das Lüftungskonzept beeinflusst. Die Belüftung von Gebäuden ist zwingend notwendig, um die Emissionen der Baustoffe und die der Menschen, beispielsweise ihren CO2-Ausstoß über die Atmung, abzuführen und der Schimmelbildung vorzubeugen. Erfolgt die Belüftung über öffenbare Fenster – natürliche Lüftung – wird so allerdings energetisch aufwändig temperierte Raumluft mit untemperierter Außenluft ausgetauscht. Daraus können Wärmeverluste und thermisches Unbehagen resultieren. Energieeffiziente Technologien sind ventilatorgestützte Lüftungssysteme mit Wärmerückgewinnung. Doch nicht für alle Gebäudekonzepte und Nutzer stellen diese Lüftungskonzepte einen hohen Nutzerkomfort dar. Korrelationen zwischen Gebäuden mit ventilatorgestützten Lüftungsystemen und dem Sick-Building-Syndrom sind in der Literatur beschrieben, während hier für natürliche Lüftungskonzepte keine Korrelation besteht. Stattdessen wird in Nutzerbefragungen der natürlichen Lüftung eine hohe Akzeptanz zugeschrieben. Mit elektrisch angetriebenen Fenstern kann die natürliche Lüftung nutzerunabhängig gesteuert und so Wärmeverluste und thermisches Unbehagen kontrolliert werden. Bisher sind die Auslegungen solcher kontrollierten natürlichen Lüftungskonzepte noch sehr planungsintensiv. Das Ziel der Arbeit ist es, für Büro- und Verwaltungsgebäude Öffnungs- und Schließsignale einer kontrollierten natürlichen Lüftung zu geben. Diese zeichnen sich darüber aus, dass sie ein gesundes Raumklima, eine hohe Nutzerbehaglichkeit und Energieeffizienz über den Jahresverlauf schaffen und auf ihre Robustheit gegenüber Änderungen von Gebäuderandbedingungen überprüft sind. Für das Ziel wird ein über CO2- und Temperatursensoren gesteuertes Fenstersystem mittels dynamisch thermischer Gebäudesimulationen in vier Varianten von Schließsignalen auf thermische Behaglichkeit und Energiebedarf untersucht. Die Grundlage dazu stellt die bezüglich Entwurf, Konstruktion und Nutzung allgemeingültige Entwicklung eines Büroraums dar. Der Büroraum wird im Simulationsmodell abgebildet und in Realität errichtet. Die Kombination von Simulationsmodell und realem, als experimentellem Teststand ausgeführtem Büroraum ermöglicht verifizierte Ergebnisse. So werden vier Berechnungsmodelle für Luftvolumenströme von Fenstern über den Teststand verifiziert. Dazu dienen Luftwechselmessungen nach der Konstantinjektionsmethode an 173 Fensteröffnungen für fünf Außentemperatur- und elf Windgeschwindigkeitsbereiche. Das Berechnungsmodell nach DIN EN 16798-7 zeigt sich als realitätsnah. Da dieses Berechnungsmodell nicht im Gebäudesimulationsprogramm implementiert ist, wird eine Methode zur Implementierung entwickelt. Über das entwickelte Simulationsmodell zeigt sich, dass eine kombinierte CO2- und temperaturgesteuerte kontrollierte natürliche Lüftung nur zweimal im Jahr ihre Grenzwerte zur Fensteröffnung und -schließung variieren muss, um ganzjährig eine hohe Energieeffizienz und Nutzerbehaglichkeit zu schaffen. Die Schließsignale des sensorgesteuerten Fenstersystems werden in eine Zeitsteuerung überführt. Es zeigt sich, dass für die kühlen Monate jede Öffnung mit identischer Dauer angesetzt werden darf. In wärmeren Monaten muss die Öffnungsdauer in Abhängigkeit der Außentemperatur angepasst werden, so dass eine Zeitsteuerung mit einer Außentemperaturmessung gekoppelt werden muss. Die Ergebnisse zeigen, dass über eine Variation der Schließsignale einer kontrollierten natürlichen Lüftung die Energieeffizienz und die thermische Behaglichkeit wesentlich gesteigert werden und dass selbst bei geringen Windgeschwindigkeiten und Temperaturdifferenzen die Raumluftqualität stets gewährleistet ist. Für nahezu alle Standorte in Deutschland kann die kontrollierte natürliche Lüftung so den Kühlbedarf der untersuchten Büroräume eliminieren, ohne in einer sommerlichen Überhitzung der Räume zu resultieren. Die entwickelten und bezüglich Raumluftqualität und thermischer Behaglichkeit charakterisierten Sensor- und Zeitsteuerungen tragen dazu bei, die kontrollierte natürliche Lüftung als wartungsarme, technikreduzierte Alternative zu der ventilatorgestützten Lüftung zu etablieren.:1 Einleitung 2 Natürliche Lüftung 3 Kontrollmöglichkeiten der natürlichen Lüftung 4 Entwicklung der Untersuchungsmodelle 5 Voruntersuchungen 6 Sensorsteuerung für den Basisraum 7 Zeitsteuerung für den Basisraum 8 Übertragung auf unterschiedliche Gebäuderandbedingungen 9 Diskussion und Empfehlungen 10 Zusammenfassung und Ausblick 11 Literatur 12 Abbildungsnachweis 13 Bezeichnungen 14 Anhang / It is a politically declared goal to reduce the emission of climate-damaging greenhouse gases worldwide. To support this goal by the building industry a key driver is the saving of energy for room conditioning. Among other factors, this is influenced by the ventilation concept. Also the ventilation of buildings is absolutely necessary in order to remove the emissions of the building materials and those of the people, for example their CO2 emissions through breathing as well as to prevent mould. However, if ventilation is carried out via openable windows - natural ventilation - then energetically expensive tempered room air is exchanged with cold outside air. This could result in heat loss and thermal discomfort. Mechanical ventilation systems with heat recovery are energy-efficient technologies. However, these ventilation concepts do not represent a high level of user comfort for all building concepts and users. Correlations between buildings with mechanical ventilation systems and sick building syndrome are described in the literature, while there is no such correlation for natural ventilation concepts. Instead, a high level of acceptance is attributed to it in user surveys. With electrically driven and controlled windows, natural ventilation can be controlled independently from the user, thus minimizing heat loss and thermal discomfort. So far, the design of such controlled natural ventilation concepts is still very planning-intensive. The aim of this work is to provide opening and closing signals for controlled natural ventilation in office buildings. These are characterized for their capability to create a high indoor air quality, high user comfort and high energy efficiency over the course of the year and are tested for their robustness against changes in building characteristics. To achieve this goal, a window system controlled by CO2 and temperature sensors is examined for its impact on thermal comfort and energy demand by means of building simulation tools with four variants of closing signals. As a basis for this examination an office room is utilized that conforms to the current standards in terms of design, construction and use. The office space is transferred to a simulation model and constructed in reality. The combination of the simulation model and the real office space, which is designed as an experimental test rig, enables verified results. Thus, four calculation models for air flow volumes of windows are verified via the test rig. Air exchange measurements according to the constant injection method on 173 window openings for five outdoor temperature and eleven wind speed ranges are used for this purpose. The calculation model according to DIN EN 16798-7 proves to be close to reality. Since this calculation model is not implemented in the building simulation program, a method for its implementation is developed. Using the developed simulation model, it is shown that a combined CO2- and temperature-controlled natural ventilation creates a high energy efficiency and user comfort throughout the year by varying its limit values for window opening and closing only twice a year. The closing signals of the sensor controlled window system are transferred to a time control system. It turns out that for the cold months, each opening could be set to the same opening time. In warmer months, the opening time must be adjusted depending on the outside temperature. Thus, a time control should be coupled with an outside air temperature measurement. The results show that by varying the closing signals of a controlled natural ventilation system, the energy efficiency and thermal comfort is significantly increased and that a high indoor air quality is always guaranteed even at low wind speeds and low temperature differences. For almost all locations in Germany, controlled natural ventilation can thus eliminate the cooling requirements in the office spaces studied without overheating in the summer. The developed sensor and time control systems are characterized by high indoor air quality and good thermal comfort. Thus, these systems are a contribution to promote controlled natural ventilation as a low-maintenance and technically reduced alternative to mechanical ventilation.:1 Einleitung 2 Natürliche Lüftung 3 Kontrollmöglichkeiten der natürlichen Lüftung 4 Entwicklung der Untersuchungsmodelle 5 Voruntersuchungen 6 Sensorsteuerung für den Basisraum 7 Zeitsteuerung für den Basisraum 8 Übertragung auf unterschiedliche Gebäuderandbedingungen 9 Diskussion und Empfehlungen 10 Zusammenfassung und Ausblick 11 Literatur 12 Abbildungsnachweis 13 Bezeichnungen 14 Anhang