The ability to navigate in an unknown environment is an enabler for truly utonomous systems. Such a system must be aware of its relative position to the surroundings using sensor measurements. It is instrumental that these measurements are monitored for disturbances and faults. Having correct measurements, the challenging problem for a robot is to estimate its own position and simultaneously build a map of the environment. This problem is referred to as the Simultaneous Localization and Mapping (SLAM) problem. This thesis studies several topics related to SLAM, on-board sensor processing, exploration and disturbance detection. The particle filter (PF) solution to the SLAM problem is commonly referred to as FastSLAM and has been used extensively for ground robot applications. Having more complex vehicle models using for example flying robots extends the state dimension of the vehicle model and makes the existing solution computationally infeasible. The factorization of the problem made in this thesis allows for a computationally tractable solution. Disturbance detection for magnetometers and detection of spurious features in image sensors must be done before these sensor measurements can be used for estimation. Disturbance detection based on comparing a batch of data with a model of the system using the generalized likelihood ratio test is considered. There are two approaches to this problem. One is based on the traditional parity space method, where the influence of the initial state is removed by projection, and the other on combining prior information with data in the batch. An efficient parameterization of incipient faults is given which is shown to improve the results considerably. Another common situation in robotics is to have different sampling rates of the sensors. More complex sensors such as cameras often have slower update rate than accelerometers and gyroscopes. An algorithm for this situation is derived for a class of models with linear Gaussian dynamic model and sensors with different sampling rates, one slow with a nonlinear and/or non-Gaussian measurement relation and one fast with a linear Gaussian measurement relation. For this case, the Kalman filter is used to process the information from the fast sensor and the information from the slow sensor is processed using the PF. The problem formulation covers the important special case of fast dynamics and one slow sensor, which appears in many navigation and tracking problems. Vision based target tracking is another important estimation problem in robotics. Distributed exploration with multi-aircraft flight experiments has demonstrated localization of a stationary target with estimate covariance on the order of meters. Grid-based estimation as well as the PF have been examined. / The third article in this thesis is included with permission of the IEEE. Such permission of the IEEE does not in any way imply IEEE endorsement of any of Linköping University's products or services. Internal or personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by writing to pubs-permissions@ieee.org. By choosing to view this material, you agree to all provisions of the copyright laws protecting it.Please be advised that wherever a copyright notice from another organization is displayed beneath a figure, a photo, a videotape or a Powerpoint presentation, you must get permission from that organization, as IEEE would not be the copyright holder.

No description available.

Removing the effects of the atmosphere from remote sensing data requires accurate knowledge of the physical properties of the atmosphere during the time of measurement. There is a nonlinear relationship that maps atmospheric composition to emitted spectra, but it cannot be easily inverted. The time evolution of atmospheric composition is approximately Markovian, and can be estimated using hyperspectral measurements of the atmosphere with particle filters. The difficulties associated with particle filtering high-dimension data can be mitigated by incorporating future measurement data with the proposal density.

We consider the filtering problem in linear state space models with heavy tailed process and measurement noise. Our work is based on Student's t distribution, for which we give a number of useful results. The derived filtering algorithm is a generalization of the ubiquitous Kalman filter, and reduces to it as special case. Both Kalman filter and the new algorithm are compared on a challenging tracking example where a maneuvering target is observed in clutter. / MC Impulse

Microwave multiplexing networks are widely used in satellite communication systems. Their performances, such as the insertion loss and rejection between channels, are extremely critical. This research aims at effective design of multiplexing networks for applications in communication satellites exploring different design methodologies. First, the enhanced type of multiplexer has been proven to be able to significantly improve its performances by forming an extra pole in the pass-band of each channel using the existing manifold. For completeness of the design methodology, the design of the enhanced type of multiplexers with H-plane T-junctions is investigated in this thesis. Design equations are given. By changing the distance between the channel filters and the distance between manifold waveguide and channel filter, the resonances and the magnitude of coupling can be readily controlled to satisfy the design requirement. Examples are used to demonstrate the feasibility and effectiveness of the new design using H-plane T-junctions. Secondly, Artificial Neural Network (ANN) is applied in the spurious mode compensation in the multiplexer design. Channel filter tuning, circuit model simplification, and neural network training processes are detailed in the thesis. Space-Mapping optimization technology has been applied in the tuning of a channel filter. The artificial neural network is applied as the method to compensate disadvantages of both electromagnetic (EM) simulation, which is accurate but extremely time consuming, and circuit model, which is fast but with limited accuracy. The neural network model of a simplified circuit can rapidly and accurately simulate the spurious mode of a channel filter in its out-of-band. Design examples are used to demonstrate the modeling steps. Good agreements are observed between EM simulation and the results from the developed model. / UOIT

The web is getting more and more characterized by personalization. Big socialnetworks like Facebook as well as the leading search engine Google increasingly usepersonalization algorithms to tailor the information that they present to users. All inorder to make the information more relevant and engaging for the end consumer. Howdoes this personalization affect journalists who increasingly search the web as a partof their journalistic research? In this essay we have looked at the effects thatpersonalization has on the journalists of tomorrow by conducting a survey amongstudents of journalism. We have also done a study of literature and theories to try anddetermine what consequences personalization will have on the internet in the future.We make an in depth study of the search engine Google as this is one of the mainsources of information for journalists and we have tried to tie this to theories of filter-bubbles and gatekeeping. We find that the question is in need of further studies toreally determine the threat that we face but conclude that information onpersonalization should be far more evident when it occurs.

Many problems in tracking control have been identified over the years, such as the availability of systems states, the presence of noise and system uncertainties, and speed of response, just to name a few. This thesis is concerned with developing novel integrated control and estimation algorithms to overcome some of these problems in order to achieve an efficient tracking performance. Since there are some significant advantages associated with Sliding Mode Control (SMC) or Variable Structure Control (VSC), (fast regulation rate and robustness to uncertainties), this research reviews and extends new filtering concepts for state estimation, referred to as the Variable Structure Filter (VSF)and Smooth Variable Structure Filter (SVSF). These are based on the philosophy of Sliding Mode Control.<p>The VSF filter is designed to estimate some of the states of a plant when noise and uncertainties are presented. This is accomplished by refining an estimate of the states in an iterative fashion using two filter gains, one based on a noiseless system with no uncertainties and the second gain which reflects these uncertainties. The VSF is combined seamlessly with the Sliding Mode Controller to produce an integrated controller called a Sliding Mode Controller and Filter (SMCF). This new controller is shown to be a robust and effective integrated control strategy for linear systems. For nonlinear systems, a novel integrated control strategy called the Smooth Sliding Mode Controller and Filter (SSMCF), fuses the SMC and SVSF in a particular form to address nonlinearities. The gain term in the SVSF is redefined to form a new algorithm called the SVSF with revised gain in order to obtain a better estimation performance. Its performance is compared to that of the Extended Kalman Filter (EKF) when applied to a particular nonlinear plant.<p>The SMCF and SSMCF are applied to the experimental prototype of a precision positioning hydraulic system called an ElectroHydraulic Actuator (EHA) system. The EHA system is known to display nonlinear characteristics but can approximate linear behavior under certain operating conditions, making it ideal to test the robustness of the proposed controllers.<p>The main conclusion drawn in this research was that the SMCF and SSMCF as developed and implemented, do exhibit robust and high performance state estimation and trajectory tracking control given modeling uncertainties and noise. The controllers were applied to a prototype EHA which demonstrated the use of the controllers in a real world application. It was also concluded that the application of the concepts of VSC for the controller can alleviate a challenging mechanical problem caused by a slip-stick characteristic in friction. Another conclusion is that the revised form of the SVSF could obtain robust and fast state estimation for nonlinear systems.<p>The original contributions of the research include: i) proposing the SMCF and SSMCF, ii) applying the Sliding Mode Controller to suppress cross-over oscillations caused by the slip-stick characteristics in friction which often occur in mechanical systems, iii) the first application of the SVSF for state estimation and iv) a comparative study of the SVSF and Extended Kalman Filter (EKF) to the EHA demonstrating the superiority of the SVSF for state estimation performance under both steady-state and transient conditions for the application considered.<p>The dissertation is written in a paper format unlike the traditional Ph.D thesis manuscript. The content of the thesis discourse is based on five manuscripts which are appended at the end of the thesis. Fundamental principles and concepts associated with SMC, VSF, SVSF and the fused controllers are introduced. For each paper, the objectives, approaches, typical results, conclusions and major contributions are presented. Major conclusions are summarized and original contributions reiterated.

A model-based estimation process is implemented in simulation of a water transmission line for the purpose of leak detection. The objective of this thesis is aimed at determining, through simulation results, the effectiveness of the Extended Kalman Filter for leak detection. Water distribution systems often contain large amounts of unknown losses. The range in magnitude of losses varies from 10 to over 50 percent of the total volume of water pumped. The result is a loss of product, including water and the chemicals used to treat it, environmental damage, demand shortfalls, increased energy usage and unneeded pump capacity expansions. It is clear that more control efforts need to be implemented on these systems to reduce losses and increase energy efficiencies. The problems of demand shortfalls, resulting from lost product, are worsened by the limited availability of water resources and a growing population and economy. The repair of leakage zones as they occur is not a simple problem since the vast majority of leaks, not considered to be major faults, go undetected. The leak detection process described in the work of this thesis is model based. A transient model of a transmission line is developed using the Method of Characteristics. This method provides the most accurate results of all finite-difference solutions to the two partial differential equations of continuity and momentum that describe pipe flow. Simulations are run with leakage within the system and small transients are added as random perturbations in the upstream reservoir head. The head measurements at the two pipe extremes are used as inputs into the filter estimation process. The Extended Kalman Filter is used for state estimation of leakage within the transmission line. The filter model places two artificial leakage states within the system. The estimates of these fictitious leakage states are then used to locate the actual position and magnitude of leakage within the transmission line. This method is capable of locating one leak within the line. The results of the Extended Kalman Filter (EKF) process show that it is capable of locating the position and magnitude of small leaks within the line. It was concluded that the EKF could be used for leak detection, but field tests need to be done to better quantify the ability of these methods. It is recommended that a multiple filtering method be implemented that may be able to locate the occurrence of multiple leakage.

A model based algorithm of pipeline flow is developed and tested to determine if the model is capable of detecting a leak in a pipeline. The overall objective of this research is to determine the feasibility of applying the Extended Kalman Filter and a new technique defined as the Extended Boundary Approach to the detection of leakages in a physical water distribution system. <p>The demands on the water supply system increase as the human population grows and expands throughout the world. Water conservation is required to ensure an adequate supply of water remains for future generations. One way to conserve this water is by reducing the leakages in underground water distribution systems. Currently between 10 to 50 percent of the pumped water is lost due to unrecognized leakages. This results in a huge revenue loss of water, chemicals and energy that is required for transporting the water. The detection of underground leakages is a very complex problem because many leakages are small and go unnoticed by todays leak detection technology. <p>A model based leak detection technique is developed and tested in this thesis. The Method of Characteristics is used to develop a model of a single pipeline. This method is extensively used and provides the most accurate results of the two partial differential equations of continuity and momentum that describe pipe flow. The Extended Kalman Filter is used to estimate two fictitious leakages at known locations along the pipeline. In order to ensure the model is observable four pressure measurements are needed at equally spaced nodes along the pipeline. With the development of the Extended Boundary Approach only the upstream and downstream pressure measurements are required, however; the upstream and downstream flow measurements are also required. Using the information from the two fictitious leaks the actual leak location and magnitude can be determined. This method is only capable of detecting one leak in a single pipeline. <p>The results of the developed model show that the approach is capable of theoretically determining the leak location and magnitude in a pipeline. However, at this time, the feasibility of implementing the proposed leak detection method is limited by the required level of accuracy of the sensors which is beyond that found in todays technology. It was also found that the EKF used primarily steady state information to predict the leakage. It is recommended that further research explore alternate models which might better enhance the EKF approach using transient information from the pipeline. This may allow implementation on a real pipeline.

Firewalls are an integral part of network security. They are pervasive throughout networks and can be found in mobile phones, workstations, servers, switches, routers, and standalone network devices. Their primary responsibility is to track and discard unauthorized network traffic, and may be implemented using costly special purpose hardware to flexible inexpensive software running on commodity hardware. The most basic action of a firewall is to match packets against a set of rules in an Access Control List (ACL) to determine whether they should be allowed or denied access to a network or resource. By design, traditional firewalls must sequentially search through the ACL table, leading to increasing latencies as the number of entries in the table increase. This is particularly true for software firewalls implemented in commodity server hardware. Reducing latency in software firewalls may enable them to replace hardware firewalls in certain applications. In this thesis, we propose a software firewall architecture which removes the sequential ACL lookup from the critical path and thus decreases the latency per packet in the common case. To accomplish this we implement a Bloom filter-based, stochastic pre-classification stage, enabling the bifurcation of the predicted good and predicted bad packet code paths, greatly improving performance. Our proposed architecture improves firewall performance 67% to 92% under anonymized trace based workloads from CAIDA servers. While our approach has the possibility of incorrectly classifying a small subset of bad packets as good, we show that these holes are neither predictable nor permanent, leading to a vanishingly small probability of firewall penetration.