This thesis presents a novel, non-simulative, probabilistic model for switching activity in sequential circuits, capturing both spatio-temporal correlations at internal nodes and higher order temporal correlations due to feedback. Switching activity, one of the key components in dynamic power dissipation, is dependent on input streams and exhibits spatio-temporal correlation amongst the signals. One can handle dependency modeling of switching activity in a combinational circuit by Bayesian Networks [2] that encapsulates the underlying joint probability distribution function exactly. We present the underlying switching model of a sequential circuit as the time coupled logic induced directed acyclic graph (TC-LiDAG), that can be constructed from the logic structure and prove it to be a dynamic Bayesian Network. Dynamic Bayesian Networks over n time slices are also minimal representation of the dependency model where nodes denote the random variable and edges either denote direct dependency between variables at one time instant or denote dependencies between the random variables at different time instants. Dynamic Bayesian Networks are extremely powerful in modeling higher order temporal as well as spatial correlations; it is an exact model for the underlying conditional independencies. The attractive feature of this graphical representation of the joint probability function is that not only does it make the dependency relationships amongst the nodes explicit but it also serves as a computational mechanism for probabilistic inference. We use stochastic inference engines for dynamic Bayesian Networks which provides any-time estimates and scales well with respect to size We observe that less than a thousand samples usually converge to the correct estimates and that three time slices are sufficient for the ISCAS benchmark circuits. The average errors in switching probability of 0.006, with errors tightly distributed around the mean error values, on ISCAS'89 benchmark circuits involving up to 10000 signals are reported.

Coastal areas are often important to economic, social, and environmental processes throughout the world. With changing climate and growing populations in these areas, coastal communities have become increasingly vulnerable to extreme flooding events, such as tsunami, storm surges, and flash floods. Within this new paradigm, there has been an effort to improve upon current methods of hazard assessment, particularly for tsunami. Recently, the American Society of Civil Engineers (ASCE) released the ASCE 7 Chapter 6 which was the world’s first standard, written in mandatory language, that addressed tsunami resilient design in a probabilistic manner for several of its prescriptions. While often the focus tends to be on mapping the hazards related to hydraulic loading conditions, post-tsunami field surveys from disaster-stricken coastal communities have also shown the importance of also considering the loads exerted by solid objects entrained within the inundating flows, commonly referred to as debris loading. Limited research has addressed debris hazard assessment in a comprehensive manner. Debris loading can be generally divided into two categories: impact and damming. Debris impact loads are caused by the rapid strike of solid objects against a structure. Debris damming loads are the result of the accumulation of debris at the face of or around a structure, causing thus an obstruction to the flow. The primary difference between these loads is the time period over which they act. The rapid loading due to debris impacts requires structural properties be considered in assessing the associated loads whereas debris damming loads are generally considered in a quasi-static manner. In assessing the hazard associated with both impact and damming loading conditions, methodologies must be developed to consider the likelihood of the load occurring and the magnitude of that load. The primary objective of this thesis was to develop a probabilistic framework for assessing debris hazards in extreme coastal flooding events. To achieve this objective, the components of the framework were split into three general categories: debris transport, debris damming, and debris impact. Several physical experimental studies were performed to address each of these components, representing the most comprehensive assessment of debris hazards in extreme flooding events to date. Debris transport was addressed to estimate the likelihood of debris loading occurring on a structure. The studies presented herein examine the different parameters that must be considered in assessing the motion of debris with the flow. The studies showed that the initial configuration of the debris and hydrodynamic conditions were critical in determining the motion of the debris. The stochastic properties of the debris motion were also assessed. It was shown that the lateral displacement of the debris could be approximated by a Gaussian distribution and the debris velocity by a Kumaraswamy (1980) distribution. The study of debris impact was further used to develop the current models used in estimating the impact force. The rigid body impact model was compared to models where the structural response was considered. The analysis showed that the effective stiffness model proposed by Haehnel and Daly (2004) was best suited to provide a conservative estimation of the impact force. Additionally, the impact geometry was taken into consideration examining the influence of various parameters on the impact force. Furthermore, debris damming was examined for the first time in transient loading conditions. This particular study examined the influence of the transient wave condition on the debris dam formation as well as the influence of different debris geometries. The influence of the debris dam geometry was correlated to increases in loading and overtopping conditions at structures. The assessment of debris hazards is critical in the development of accurate design conditions. The probabilistic framework presented within this thesis is expected to provide a basis for estimating debris hazards and inform future studies in the development of hazard assessment models.

Bibliography: leaves 90-93.

This thesis presents a means for representing and computing beliefs in the form of arbitrary probability density functions with a guarantee for the ongoing validity of such beliefs over indefinte time frames. The foremost aspect of this proposal is the introduction of a general, theoretical, solution to the guaranteed state estimation problem from within the recursive Bayesian estimation framework. The solution presented here determines the minimum space required, at each stage of the estimation process, to represent the belief with limited, or no, loss of information. Beyond this purely theoretical aspect, a number of numerical techniques, capable of determining the required space and performing the appropriate spatial reconfiguration, whilst also computing and representing the belief functions, are developed. This includes a new, hybrid particle-element approach to recursive Bayesian estimation. The advantage of spatial reconfiguration as presented here is that it ensures that the belief functions consider all plausible states of the target system, without altering the recursive Bayesian estimation equations used to form those beliefs. Furthermore, spatial reconfiguration as proposed in this dissertation enhances the estimation process since it allows computational resources to be concentrated on only those states considered plausible. Autonomous maritime search and rescue is used as a focus application throughout this dissertation since the searching-and-tracking requirements of the problem involve uncertainty, the use of arbitrary belief functions and dynamic target systems. Nevertheless, the theoretical development in this dissertation has been kept general and independent of an application, and as such the theory and techniques presented here may be applied to any problem involving dynamic Bayesian beliefs. A number of numerical experiments and simulations show the efficacy of the proposed spatially reconfigurable representations, not only in ensuring the validity of the belief functions over indefinite time frames, but also in reducing computation time and improving the accuracy of function approximation. Improvements of an order of magnitude were achieved when compared with traditional, spatially static representations.

Doctor of Philosophy (PhD) / This thesis addresses the problem of building statistical models for multi-sensor perception in unstructured outdoor environments. The perception problem is divided into three distinct tasks: recognition, representation and association. Recognition is cast as a statistical classification problem where inputs are images or a combination of images and ranging information. Given the complexity and variability of natural environments, this thesis investigates the use of Bayesian statistics and supervised dimensionality reduction to incorporate prior information and fuse sensory data. A compact probabilistic representation of natural objects is essential for many problems in field robotics. This thesis presents techniques for combining non-linear dimensionality reduction with parametric learning through Expectation Maximisation to build general representations of natural features. Once created these models need to be rapidly processed to account for incoming information. To this end, techniques for efficient probabilistic inference are proposed. The robustness of localisation and mapping algorithms is directly related to reliable data association. Conventional algorithms employ only geometric information which can become inconsistent for large trajectories. A new data association algorithm incorporating visual and geometric information is proposed to improve the reliability of this task. The method uses a compact probabilistic representation of objects to fuse visual and geometric information for the association decision. The main contributions of this thesis are: 1) a stochastic representation of objects through non-linear dimensionality reduction; 2) a landmark recognition system using a visual and ranging sensors; 3) a data association algorithm combining appearance and position properties; 4) a real-time algorithm for detection and segmentation of natural objects from few training images and 5) a real-time place recognition system combining dimensionality reduction and Bayesian learning. The theoretical contributions of this thesis are demonstrated with a series of experiments in unstructured environments. In particular, the combination of recognition, representation and association algorithms is applied to the Simultaneous Localisation and Mapping problem (SLAM) to close large loops in outdoor trajectories, proving the benefits of the proposed methodology.

<p>This report develops a method for probabilistic conceptual sensor modeling. The idea is to generate probabilities for detection, recognition and identification based on a few simple factors. The</p><p>focus lies on FLIR sensors and thermal radiation, even if discussions of other wavelength bands are made. The model can be used as a hole or some or several parts can be used to create a simpler model. The core of the model is based on the Johnson criteria that uses resolution as the input parameter. Some extensions that models other factors are also implemented. In the end a short discussion of the possibility to use this model for other sensors than FLIR is made.</p>

We give a necessary and sufficient condition for the existence of an increasing coupling of N (N >= 2) synchronous dynamics on S-Zd (PCA). Increasing means the coupling preserves stochastic ordering. We first present our main construction theorem in the case where S is totally ordered; applications to attractive PCAs are given. When S is only partially ordered, we show on two examples that a coupling of more than two synchronous dynamics may not exist. We also prove an extension of our main result for a particular class of partially ordered spaces.

Renewable energy policies, such as the Renewable Portfolio Standard, arising from increasing environmental concerns have set very ambitious targets for wind power penetration in electric power systems throughout the world. In many cases, the geographical locations with good wind resources are not close to the main load centers. It becomes extremely important to assess adequate transmission facility to deliver wind power to the power grid. <p>Wind is a highly variable energy source, and therefore, transmission system planning for wind delivery is very different from conventional transmission planning. Most electric power utilities use a deterministic n-1 criterion in transmission system planning. Deterministic methods cannot recognize the random nature of wind variation that dictates the power generated from wind power sources. This thesis presents probabilistic method to evaluate the contribution of a wind power delivery system to the overall system reliability. The effects of site-specific wind regime, system load, transmission line unavailability, and redundancy on system reliability were studied using a basic system model. The developed method responds to the various system parameters and is capable of assessing the actual system risks. <p>Modern power system aims to provide reliable as well as cost effective power supply to its consumers. Reliability benefits, environmental benefits and operating cost savings from wind power integration should be compared with the associated investment costs in order to determine optimum transmission facility for wind power delivery. This thesis presents the reliability/cost techniques for determining appropriate transmission line capacity to connect a wind farm to a power grid. The effect of transmission system cost, line length, wind regime, wind penetration and customer interruption cost on the optimum transmission line sizing were studied using a basic system model. The methodology and results presented in this thesis should be useful in transmission system planning for delivering wind power to a power system.

The reliability of supply in a bulk electricity system is directly related to the availability of the generation and transmission facilities. In a conventional vertically integrated system these facilities are usually owned and operated by a single company. In the new deregulated utility environment, these facilities could be owned and operated by a number of independent organizations. In this case, the overall system reliability is the responsibility of an independent system operator (ISO). The load point and system reliabilities are a function of the capacities and availabilities of the generation and transmission facilities and the system topology. This research examines the effect of equipment unavailability on the load point and system reliability of two test systems. The unavailabilities of specific generation and transmission facilities have major impacts on the load point and system reliabilities. These impacts are not uniform throughout the system and are highly dependent on the overall system topology and the operational philosophy of the system. Contingency evaluation is a basic planning and operating procedure and different contingencies can have quite different system and load point impacts. The risk levels associated with a given contingency cannot be estimated using deterministic criteria. The studies presented in this thesis estimate the risk associated with each case using probability techniques and rank the cases based on the predicted risk levels. This information should assist power system managers and planners to make objective decisions regarding reliability and cost. Composite system preventive maintenance scheduling is a challenging task. The functional separation of generation and transmission in the new market environment creates operational and scheduling problems related to maintenance. Maintenance schedules must be coordinated through an independent entity (ISO) to assure reliable and economical service. The methods adopted by an ISO to coordinate planned outages are normally based on traditional load flow and stability analysis and deterministic operating criteria. A new method designated as the maintenance coordination technique (MCT) is proposed in this thesis to coordinate maintenance scheduling. The research work illustrated in this thesis indicates that probabilistic criteria and techniques for composite power system analysis can be effectively utilized in both vertically integrated and deregulated utility systems. The conclusions and the techniques presented in this thesis should prove valuable to those responsible for system planning and maintenance coordination.

The current method of determining component failure rates for probabilistic risk assessment (PRA) in the nuclear industry is to take the total number of failures divided by the time over which the failures occurred. The method proposed in this study is the reliability growth method and involves taking into account the fact that the amount of failures per additional year of operation generally decreases yearly because the operational staff becomes familiar with the equipment. The reliability growth method will result in lower component failure rates which when used in PRA studies could result in a lower core melt frequency value. The component failure rate would be expected to be higher in the early stages and should gradually decrease as time goes on. This study will compare the final core melt frequency of the Trojan Nuclear Power Plant using both methods. The Nuclear Power Reactor Data System (NPRDS) data base from the Institute of Nuclear Power Operations (INPO) was used in this study. The components which were examined for the reliability growth method are motor operated valves, service water pump/motors and emergency diesel generator air chargers. These data were screened to ensure that only true failures were reported. A comparison was made of the overall core melt frequency between the conventional failure rate method and reliability growth method for the motor operated valves. The overall core melt frequency was decreased by 1.8 % when using the reliability growth method compared to the conventional method. / Graduation date: 1993