The growth of the automobile Industry in the past 50 years is radical. The development of chassis control systems have grown drastically due to the demand for safer, faster and more comfortable vehicles. For example, the invention of Anti-lock Braking System (ABS) has resulted in saving more than a million lives since its adaptation while also allowing the vehicles to commute faster. As we move into the autonomous vehicles era, demand for additional information about tire-road interaction to improve the performance of the onboard chassis control systems, is high. This is due to the fact that the interaction between the tire and the road surface determines the stability boundary limits of the vehicles. In this research, a real-time system to classify the road surface into five major categories was developed. The five surfaces include Dry Asphalt, Wet Asphalt, Snow, and Ice and dry Concrete. tri-axial accelerometers were placed on the inner liner of the tires. An advanced signal processing technique was utilized along with a machine learning model to classify the road surfaces. The instrumented Volkswagen Jetta with intelligent tires was retrofitted with new instrumentation for collecting data and evaluating the performance of the developed real-time system. A comprehensive study on road surface classification was performed in order to determine the features of the classification algorithm. Performance of the real-time system is discussed in details and compared with offline results. / Master of Science / The automobile industry has been improving road transportation safety over the past 50 years. While we enter the autonomous vehicles era, the safety of the vehicle is of primary concern. In order to get the autonomous vehicles to production, we will have to improve the on board vehicle control systems to adapt to all surfaces. Gaining more accurate information about the tire and road interaction will help in improving the control systems. Tires have always been considered a passive element of the vehicle. However, more recently, the idea of “tire as a sensor” has surfaced and has become one of the major research thrusts in tire as well as vehicle companies. The intelligent tire research at the Center for Tire Research (CenTiRe) begun in 2010 and has been going strong. In this work, we have developed a classification algorithm to classify the road surfaces in real-time based on acceleration measured inside the tire. The information regarding the road surface would be highly beneficial for the developing new control strategies, automate service vehicles and aid surface prediction in autonomous vehicles.

This thesis investigates and develops a direct method for transient stability analysis using the energy approach [1] and the Phasor Measurement Units (PMUs). The originality of this new method results from a combination of a prediction of the post-fault trajectory based on the PMUs and the Transient Energy Function of a multimachine system. Thanks to the PMUs, the weakness of the direct methods, which is the over-simplification of the generator model, is overcome. This new method consists of fitting a curve to the data of the post-fault path provided by PMUs and identifying the controlling unstable equilibrium point (c.u.e.p.). Two second-order linear models have been estimated and evaluated from a prediction viewpoint. These are a polynomial function and an auto-regressive model. These parameters have been estimated by means of the least-squares estimator. They have been compared to the model proposed by Y. Ohura et al. [6], which has been upgraded into an iterative algorithm. The post-fault trajectory is predicted until the exit point located on the Potential Energy Boundary Surface (p.e.b.s.) is reached. In order to detect with efficiency this exit point and to find the c.u.e.p., it is proposed a combination of the so called "Ball-Drop" method [22] and an improved version of the Shadowing method. These combined procedures give accurate results when they are compared to the step-by-step method, which directly integrates the differential equations using a fourth-order Runga-Kutta method. The simulations have been carried out on a 3-machine system and on the 10-machine New-England power system. / Master of Science

As the number of processing cores contained in modern processors continues to increase, cache hierarchies are becoming more complex. This added complexity has the effect of increasing the potential cost of any cache misses on such architectures. When cache misses become more costly, minimizing them becomes even more important, particularly in terms of scalability concerns. In this thesis, we consider the problem of cache-aware real-time scheduling on multiprocessor systems. One avenue for improving real-time performance on multi-core platforms is task partitioning. Partitioning schemes statically assign tasks to cores, eliminating task migrations and reducing system overheads. Unfortunately, no current partitioning schemes explicitly consider cache effects when partitioning tasks. We develop the LWFG (Largest Working set size First, Grouping) cache-aware partitioning algorithm, which seeks to schedule tasks which share memory with one another in such a way as to minimize the total number of cache misses. LWFG minimizes cache misses by partitioning tasks that share memory onto the same core and by distributing the system's sum working set size as evenly as possible across the available cores. We evaluate the LWFG partitioning algorithm against several other commonly-used partitioning heuristics on a modern 48-core platform running ChronOS Linux. Our evaluation shows that in some cases, the LWFG partitioning algorithm increases execution efficiency by as much as 15% (measured by instructions per cycle) and decreases mean maximum tardiness by up to 60%. / Master of Science

The I-81 Corridor in Virginia traverses the western part of the state, connecting Bristol in the south to Winchester in the north. A study carried out at the Virginia Tech Center for Transportation Research identified traffic safety, work zone safety and traffic control, trucking issues, and intercity traveler information needs as important issues that deserve attention on the I-81 Corridor in Virginia. Analysis of work zone accident statistics showed a need for real-time systems to enhance work zone safety. Real-time advanced warning and traffic control systems provide a means of dynamic information dissemination and advanced warning, thereby enhancing work zone safety and facilitating traffic control. The focus of this research was on the development of functional and system requirements for a real-time advanced warning and traffic control system for work zones. This task was based on the examination of work zone accidents and their causes. The functional requirements include advanced warning, surveillance, advisory, and control functions. Each of these functions consists of several sub-functions. The needs with respect to each of these functions have also been identified. System requirements such as real-time operation, credibility, portability, ease of installation, and adaptability were also identified. Evaluation criteria and potential Measures Of Effectiveness (MOEs) for the evaluation of the system were also identified. Additionally, issues related to the evaluation of the system, such as time duration for evaluation and data collection techniques were identified and examined. / Master of Science

Real-time channels are established between a source and a destination to guarantee in-time delivery of real-time messages in multi-hop networks. In this thesis, we propose two schemes to establish real-time channels for soft real-time applications whose timeliness properties are characterized using Jensen's Time Utility Functions (TUFs) that are non-increasing. The two algorithms are (1) Localized Decision for Utility accrual Channel Establishment (LocDUCE) and (2) Global Decision for Utility accrual Channel Establishment (GloDUCE). Since finding a feasible path optimizing multiple constraints is an NP-Complete problem, these schemes heuristically attempt to maximize the system-wide accrued utility. The channel establishment algorithms assume the existence of a utility-aware packet scheduling algorithm at the interfaces. The route selection is based on delay estimation performed at the source, destination, and all routers in the path, from source to destination. We simulate the algorithms, measure and compare their performance with open shortest path first (OSPF). Our simulation experiments show that for most of the cases considered LocDUCE and GloDUCE perform better than OSPF. We also implement the schemes in a proof-of-concept style routing module and measure the performance of the schemes and compare them to OSPF. Our experiments on the implementation follow the same trend as the simulation study and show that LocDUCE and GloDUCE have a distinct advantage over OSPF and accrue higher system-wide utility. These schemes also react better to variation in the loading of the links. Among the two proposed approaches, we observe that GloDUCE performs better than LocDUCE under conditions of increased downstream link loads. / Master of Science

Alcohol use disorder is common, and treatments are currently inadequate. Some of the acute effects of alcohol on the brain, such as altering the decision-making and future thinking capacities, mirror the effects of chronic alcohol use. Therefore, interventions that can address these shortcomings may be useful for reducing the negative effects of alcohol use disorder in combination with other therapies. The signature of those interventions may also be evident in the signature of large-scale, dynamic brain networks, which can show whether an intervention is effective. One such intervention is episodic future thinking, which has been shown to reduce delay discounting and orient people toward pro-social, long-term outcomes. To better understand decision making in high-risk individuals, we examined delay discounting in an adolescent population. When the decision-making faculties were challenged with difficult choices, adolescents made decisions inconsistent with their predicted preference, complemented by increased brain activity in the central executive network and salience network. Using these results and the hypothesis that the default mode network would be implicated in future thinking and intertemporal choice, we examined the neural effects of a brief behavioral intervention, episodic future thinking, that seeks to address these impairments. We showed that episodic future thinking has both acute and longer-lasting effects on consequential brain networks at rest and during delay discounting compared to a control episodic thinking condition in alcohol use disorder. Our failure to show group differences in default mode network prompted us to scrutinize it more carefully, from a position where we could measure the ability to self-regulate the network rather than its resting-state tendency. We implemented a real-time fMRI experiment to test the degree to which people along the alcohol use severity spectrum can self-regulate this network. Our results showed that default mode network suppression is impaired as alcohol use disorder severity increases. In the process, we showed that direct examination of resting-state networks with these methods will provide more information than measuring them at rest alone. We also characterized the default mode network along the real-time fMRI pipeline to show the whole-brain spatial pattern of regions associated and unassociated with the network. Our results indicate that resting-state brain networks are important markers for outcomes in alcohol use disorder and that they can be manipulated under experimental conditions, potentially to the benefit of the afflicted individual. / Doctor of Philosophy / Alcohol is the most widely-used mind-altering substance in the United States. Even though most people do not develop a problem with alcohol use, many people will at some point develop drinking patterns that classify as an alcohol use disorder. Brain damage from drinking can come from the toxicity of alcohol, but also as a result of behaviors associated with drinking too much, including injury, violence, accidents, and other health-related issues. Interventions at the behavioral level can be effective at curbing drinking patterns before damage accrues, and a better understanding of those interventions at the level of the brain may make them more effective. This work investigated the decision-making processes and the ability to think clearly about the future, two faculties that begin to become diminished in alcohol use disorder. In our first set of studies, we tested a brief behavioral intervention called episodic future thinking, which helps people orient themselves away from short-term rewards like alcohol and toward long-term goals that could happen if they stopped drinking as much. We showed that one hour-long, intensive session produced changes in the connectivity between the prefrontal cortex and the lower brain. We also generated data in a long-term experiment suggesting repeated reminders of the episodic future thinking intervention produce changes in large-scale brain networks that are disrupted in substance use disorders. In a separate set of experiments, we showed that people can gain control over one of these networks, called the default mode network, to the point of being able to control a brain-machine interface just by following simple instructions. However, we demonstrated that the degree to which someone can control this brain activity was associated with their drinking severity. In other words, the more people drank, in terms of volume and frequency, the less control they had over their own brain activity. This finding is important because many researchers have shown that activity in this brain region is related to many psychopathologies, including substance use disorders. Other researchers have been developing ways in which the ability to control this brain activity can be trained. While we did not find evidence of a training effect in a small group of healthy people (5), it may be the case that people impaired by alcohol use disorder can improve through practice or through cutting back on drinking. Ultimately, we hope that the research presented here will help to guide the development of treatments for alcohol use disorder to be more effective.

This thesis presents the design and evaluation of an object-oriented (OO) operating system kernel for real-time embedded systems based on dataflow architecture. Dataflow is a software architecture that is well suited to applications that involve signal flows and value transformations. Typically, these systems comprise numerous processes with heavy inter-process communications. The dataflow style has been adopted for the control software for PEBB (Power Electronic Building Block) systems by the Center for Power Electronic Systems (CPES), Virginia Tech., which is involved in a research effort to modularize and standardize power electronic components. The goal of our research is to design and implement an efficient object-oriented kernel for the PEBB system and compare its performance vis-à-vis that of a non-OO kernel. It presents strategies for efficient OO design and a discussion of how OO performance issues can be ameliorated. We conclude the thesis with an evaluation of the advantages gained by using the OO paradigm both from the standpoint of the classically cited advantages of OO programming and other crucial aspects. / Master of Science

This research presents the development and proof of concept of a TeleGeoMonitoring (TGM) system for spatially monitoring and analyzing, in real-time, data derived from vehicle-mounted sensors. In response to the concern for vibration related injuries experienced by equipment operators in surface mining and construction operations, the prototype TGM system focuses on spatially monitoring vehicle vibration in real-time. The TGM vibration system consists of 3 components: (1) Data Acquisition Component, (2) Data Transfer Component, and (3) Data Analysis Component. A GPS receiver, laptop PC, data acquisition hardware, triaxial accelerometer, and client software make up the Data Acquisition Component. The Data Transfer Component consists of a wireless data network and a data server. The Data Analysis Component provides tools to the end user for spatially monitoring and analyzing vehicle vibration data in real-time via the web or GIS workstations. Functionality of the prototype TGM system was successfully demonstrated in both lab and field tests. The TGM vibration system presented in this research demonstrates the potential for TGM systems as a tool for research and management projects, which aim to spatially monitor and analyze data derived from mobile sensors in real-time. / Master of Science

This paper describes the implementation of a scanning signal detector and automatic modulation classification system. The classification technique is a completely blind method, with no prior knowledge of the signal's center frequency, bandwidth, or symbol rate. An energy detector forms the initial approximations of the signal parameters. The energy detector used in the wideband sweep is reused to obtain fine estimates of the center frequency and bandwidth of the signal. The subsequent steps reduce the effect of frequency offset and sample timing error, resulting in a constellation of the modulation of interest. The cumulant of the constellation is compared to a set of known ideal cumulant values, forming the classification estimate. The algorithm uses two platforms that together provide high speed parallel processing and flexible run-time operation. High-rate spectral scanning using an energy detector is run in parallel with a variable down sampling path; both are highly pipelined structures, which allows for high data throughput. A pair of processing cores is used to record spectral usage and signal characteristics as well as perform the actual classification. The resulting classification system can accurately identify modulations below 5 dB of signal-to-noise ratio (SNR) for some cases of the phase shift keying family of modulations but requires a much higher SNR to accurately classify higher-order modulations. These estimates tend toward classifying all signals as binary phase shift keying because of limits of the noise power estimation part of the cumulant normalization process. Other effects due to frequency offset and synchronization timing are discussed. / Master of Science

This thesis describes the design and implementation of real-time image-processing tasks on a custom computing platform called Splash-2. The tasks that have been implemented are image histogram generation, median filtering using 3X3 neighborhoods, and morphological dilation and erosion. These problems are computationally intensive, involving large amounts of data. The problems are especially difficult when the images need to be processed at real-time rates, typically 30 frames per second. Splash-2 is a reconfigurable FFPGA-based attached processor featuring several programmable processing elements and programmable communication paths. Although not designed specifically for image-processing applications, it possesses architectural properties that make it well suited for high speed computations and data transfer rates that are characteristic of this class of problems. This thesis discusses the design process which has been used to map the tasks to Splash-2, and presents results which demonstrate the effectiveness of custom computing platforms for high performance image processing. / Master of Science