Managing server energy and reducing operational cost for online service providers

The past decade has seen the energy consumption in servers and Internet Data Centers (IDCs) skyrocket. A recent survey estimated that the worldwide spending on servers and cooling have risen to above $30 billion and is likely to exceed spending on the new server hardware . The rapid rise in energy consumption has posted a serious threat to both energy resources and the environment, which makes green computing not only worthwhile but also necessary. This dissertation intends to tackle the challenges of both reducing the energy consumption of server systems and by reducing the cost for Online Service Providers (OSPs). Two distinct subsystems account for most of IDC’s power: the server system, which accounts for 56% of the total power consumption of an IDC, and the cooling and humidifcation systems, which accounts for about 30% of the total power consumption. The server system dominates the energy consumption of an IDC, and its power draw can vary drastically with data center utilization. In this dissertation, we propose three models to achieve energy effciency in web server clusters: an energy proportional model, an optimal server allocation and frequency adjustment strategy, and a constrained Markov model. The proposed models have combined Dynamic Voltage/Frequency Scaling (DV/FS) and Vary-On, Vary-off (VOVF) mechanisms that work together for more energy savings. Meanwhile, corresponding strategies are proposed to deal with the transition overheads. We further extend server energy management to the IDC’s costs management, helping the OSPs to conserve, manage their own electricity cost, and lower the carbon emissions. We have developed an optimal energy-aware load dispatching strategy that periodically maps more requests to the locations with lower electricity prices. A carbon emission limit is placed, and the volatility of the carbon offset market is also considered. Two energy effcient strategies are applied to the server system and the cooling system respectively. With the rapid development of cloud services, we also carry out research to reduce the server energy in cloud computing environments. In this work, we propose a new live virtual machine (VM) placement scheme that can effectively map VMs to Physical Machines (PMs) with substantial energy savings in a heterogeneous server cluster. A VM/PM mapping probability matrix is constructed, in which each VM request is assigned with a probability running on PMs. The VM/PM mapping probability matrix takes into account resource limitations, VM operation overheads, server reliability as well as energy effciency. The evolution of Internet Data Centers and the increasing demands of web services raise great challenges to improve the energy effciency of IDCs. We also express several potential areas for future research in each chapter.

DYNAMIC MODEL BASED STATE ESTIMATION IN A HEAVY DUTY DIESEL AFTERTREATMENT SYSTEM FOR ONBOARD DIAGNOSTICS AND CONTROLS

Estimating un-measurable states is an important component for onboard diagnostics (OBD) and control strategy development in diesel exhaust aftertreatment systems. This research focuses on the development of an Extended Kalman Filter (EKF) based state estimator for two of the main components in a diesel engine aftertreatment system: the Diesel Oxidation Catalyst (DOC) and the Selective Catalytic Reduction (SCR) catalyst. One of the key areas of interest is the performance of these estimators when the catalyzed particulate filter (CPF) is being actively regenerated. In this study, model reduction techniques were developed and used to develop reduced order models from the 1D models used to simulate the DOC and SCR. As a result of order reduction, the number of states in the estimator is reduced from 12 to 1 per element for the DOC and 12 to 2 per element for the SCR. The reduced order models were simulated on the experimental data and compared to the high fidelity model and the experimental data. The results show that the effect of eliminating the heat transfer and mass transfer coefficients are not significant on the performance of the reduced order models. This is shown by an insignificant change in the kinetic parameters between the reduced order and 1D model for simulating the experimental data. An EKF based estimator to estimate the internal states of the DOC and SCR was developed. The DOC and SCR estimators were simulated on the experimental data to show that the estimator provides improved estimation of states compared to a reduced order model. The results showed that using the temperature measurement at the DOC outlet improved the estimates of the CO , NO , NO2 and HC concentrations from the DOC. The SCR estimator was used to evaluate the effect of NH3 and NOX sensors on state estimation quality. Three sensor combinations of NOX sensor only, NH3 sensor only and both NOX and NH3 sensors were evaluated. The NOX only configuration had the worst performance, the NH3 sensor only configuration was in the middle and both the NOX and NH3 sensor combination provided the best performance.

Control algorithms for large scale adaptive optics

In this dissertation, the problem of creating effective large scale Adaptive Optics (AO) systems control algorithms for the new generation of giant optical telescopes is addressed. The effectiveness of AO control algorithms is evaluated in several respects, such as computational complexity, compensation error rejection and robustness, i.e. reasonable insensitivity to the system imperfections. The results of this research are summarized as follows: 1. Robustness study of Sparse Minimum Variance Pseudo Open Loop Controller (POLC) for multi-conjugate adaptive optics (MCAO). The AO system model that accounts for various system errors has been developed and applied to check the stability and performance of the POLC algorithm, which is one of the most promising approaches for the future AO systems control. It has been shown through numerous simulations that, despite the initial assumption that the exact system knowledge is necessary for the POLC algorithm to work, it is highly robust against various system errors. 2. Predictive Kalman Filter (KF) and Minimum Variance (MV) control algorithms for MCAO. The limiting performance of the non-dynamic Minimum Variance and dynamic KF-based phase estimation algorithms for MCAO has been evaluated by doing Monte-Carlo simulations. The validity of simple near-Markov autoregressive phase dynamics model has been tested and its adequate ability to predict the turbulence phase has been demonstrated both for single- and multiconjugate AO. It has also been shown that there is no performance improvement gained from the use of the more complicated KF approach in comparison to the much simpler MV algorithm in the case of MCAO. 3. Sparse predictive Minimum Variance control algorithm for MCAO. The temporal prediction stage has been added to the non-dynamic MV control algorithm in such a way that no additional computational burden is introduced. It has been confirmed through simulations that the use of phase prediction makes it possible to significantly reduce the system sampling rate and thus overall computational complexity while both maintaining the system stable and effectively compensating for the measurement and control latencies.

A comparison of ground cover and frequency estimation methods for post-harvest soil monitoring

The amount and type of ground cover is an important characteristic to measure when collecting soil disturbance monitoring data after a timber harvest. Estimates of ground cover and bare soil can be used for tracking changes in invasive species, plant growth and regeneration, woody debris loadings, and the risk of surface water runoff and soil erosion. A new method of assessing ground cover and soil disturbance was recently published by the U.S. Forest Service, the Forest Soil Disturbance Monitoring Protocol (FSDMP). This protocol uses the frequency of cover types in small circular (15cm) plots to compare ground surface in pre- and post-harvest condition. While both frequency and percent cover are common methods of describing vegetation, frequency has rarely been used to measure ground surface cover. In this study, three methods for assessing ground cover percent (step-point, 15cm dia. circular and 1x5m visual plot estimates) were compared to the FSDMP frequency method. Results show that the FSDMP method provides significantly higher estimates of ground surface condition for most soil cover types, except coarse wood. The three cover methods had similar estimates for most cover values. The FSDMP method also produced the highest value when bare soil estimates were used to model erosion risk. In a person-hour analysis, estimating ground cover percent in 15cm dia. plots required the least sampling time, and provided standard errors similar to the other cover estimates even at low sampling intensities (n=18). If ground cover estimates are desired in soil monitoring, then a small plot size (15cm dia. circle), or a step-point method can provide a more accurate estimate in less time than the current FSDMP method.

Estimation of the flammability zone boundaries with thermodynamic and empirical equations

The flammability zone boundaries are very important properties to prevent explosions in the process industries. Within the boundaries, a flame or explosion can occur so it is important to understand these boundaries to prevent fires and explosions. Very little work has been reported in the literature to model the flammability zone boundaries. Two boundaries are defined and studied: the upper flammability zone boundary and the lower flammability zone boundary. Three methods are presented to predict the upper and lower flammability zone boundaries: The linear model The extended linear model, and An empirical model The linear model is a thermodynamic model that uses the upper flammability limit (UFL) and lower flammability limit (LFL) to calculate two adiabatic flame temperatures. When the proper assumptions are applied, the linear model can be reduced to the well-known equation yLOC = zyLFL for estimation of the limiting oxygen concentration. The extended linear model attempts to account for the changes in the reactions along the UFL boundary. Finally, the empirical method fits the boundaries with linear equations between the UFL or LFL and the intercept with the oxygen axis. xx Comparison of the models to experimental data of the flammability zone shows that the best model for estimating the flammability zone boundaries is the empirical method. It is shown that is fits the limiting oxygen concentration (LOC), upper oxygen limit (UOL), and the lower oxygen limit (LOL) quite well. The regression coefficient values for the fits to the LOC, UOL, and LOL are 0.672, 0.968, and 0.959, respectively. This is better than the fit of the "zyLFL" method for the LOC in which the regression coefficient’s value is 0.416.

Development of a physically-based method for delineation of hydrologically homogeneous regions and flood quantile estimation in ungauged basins via the index flood method

Regional flood frequency techniques are commonly used to estimate flood quantiles when flood data is unavailable or the record length at an individual gauging station is insufficient for reliable analyses. These methods compensate for limited or unavailable data by pooling data from nearby gauged sites. This requires the delineation of hydrologically homogeneous regions in which the flood regime is sufficiently similar to allow the spatial transfer of information. It is generally accepted that hydrologic similarity results from similar physiographic characteristics, and thus these characteristics can be used to delineate regions and classify ungauged sites. However, as currently practiced, the delineation is highly subjective and dependent on the similarity measures and classification techniques employed. A standardized procedure for delineation of hydrologically homogeneous regions is presented herein. Key aspects are a new statistical metric to identify physically discordant sites, and the identification of an appropriate set of physically based measures of extreme hydrological similarity. A combination of multivariate statistical techniques applied to multiple flood statistics and basin characteristics for gauging stations in the Southeastern U.S. revealed that basin slope, elevation, and soil drainage largely determine the extreme hydrological behavior of a watershed. Use of these characteristics as similarity measures in the standardized approach for region delineation yields regions which are more homogeneous and more efficient for quantile estimation at ungauged sites than those delineated using alternative physically-based procedures typically employed in practice. The proposed methods and key physical characteristics are also shown to be efficient for region delineation and quantile development in alternative areas composed of watersheds with statistically different physical composition. In addition, the use of aggregated values of key watershed characteristics was found to be sufficient for the regionalization of flood data; the added time and computational effort required to derive spatially distributed watershed variables does not increase the accuracy of quantile estimators for ungauged sites. This dissertation also presents a methodology by which flood quantile estimates in Haiti can be derived using relationships developed for data rich regions of the U.S. As currently practiced, regional flood frequency techniques can only be applied within the predefined area used for model development. However, results presented herein demonstrate that the regional flood distribution can successfully be extrapolated to areas of similar physical composition located beyond the extent of that used for model development provided differences in precipitation are accounted for and the site in question can be appropriately classified within a delineated region.

Stochastic knock detection model for spark ignited engines

This report presents the development of a Stochastic Knock Detection (SKD) method for combustion knock detection in a spark-ignition engine using a model based design approach. Knock Signal Simulator (KSS) was developed as the plant model for the engine. The KSS as the plant model for the engine generates cycle-to-cycle accelerometer knock intensities following a stochastic approach with intensities that are generated using a Monte Carlo method from a lognormal distribution whose parameters have been predetermined from engine tests and dependent upon spark-timing, engine speed and load. The lognormal distribution has been shown to be a good approximation to the distribution of measured knock intensities over a range of engine conditions and spark-timings for multiple engines in previous studies. The SKD method is implemented in Knock Detection Module (KDM) which processes the knock intensities generated by KSS with a stochastic distribution estimation algorithm and outputs estimates of high and low knock intensity levels which characterize knock and reference level respectively. These estimates are then used to determine a knock factor which provides quantitative measure of knock level and can be used as a feedback signal to control engine knock. The knock factor is analyzed and compared with a traditional knock detection method to detect engine knock under various engine operating conditions. To verify the effectiveness of the SKD method, a knock controller was also developed and tested in a model-in-loop (MIL) system. The objective of the knock controller is to allow the engine to operate as close as possible to its border-line spark-timing without significant engine knock. The controller parameters were tuned to minimize the cycle-to-cycle variation in spark timing and the settling time of the controller in responding to step increase in spark advance resulting in the onset of engine knock. The simulation results showed that the combined system can be used adequately to model engine knock and evaluated knock control strategies for a wide range of engine operating conditions.

Theoretical estimation of optical absorption and photoluminescence in nanostructured silicon; an approach to improve efficiency in nanostructured solar cells

Renewable energy is growing in demand, and thus the the manufacture of solar cells and photovoltaic arrays has advanced dramatically in recent years. This is proved by the fact that the photovoltaic production has doubled every 2 years, increasing by an average of 48% each year since 2002. Covering the general overview of solar cell working, and its model, this thesis will start with the three generations of photovoltaic solar cell technology, and move to the motivation of dedicating research to nanostructured solar cell. For the current generation solar cells, among several factors, like photon capture, photon reflection, carrier generation by photons, carrier transport and collection, the efficiency also depends on the absorption of photons. The absorption coefficient,α, and its dependence on the wavelength, λ, is of major concern to improve the efficiency. Nano-silicon structures (quantum wells and quantum dots) have a unique advantage compared to bulk and thin film crystalline silicon that multiple direct and indirect band gaps can be realized by appropriate size control of the quantum wells. This enables multiple wavelength photons of the solar spectrum to be absorbed efficiently. There is limited research on the calculation of absorption coefficient in nano structures of silicon. We present a theoretical approach to calculate the absorption coefficient using quantum mechanical calculations on the interaction of photons with the electrons of the valence band. One model is that the oscillator strength of the direct optical transitions is enhanced by the quantumconfinement effect in Si nanocrystallites. These kinds of quantum wells can be realized in practice in porous silicon. The absorption coefficient shows a peak of 64638.2 cm-1 at = 343 nm at photon energy of ξ = 3.49 eV ( = 355.532 nm). I have shown that a large value of absorption coefficient α comparable to that of bulk silicon is possible in silicon QDs because of carrier confinement. Our results have shown that we can enhance the absorption coefficient by an order of 10, and at the same time a nearly constant absorption coefficient curve over the visible spectrum. The validity of plots is verified by the correlation with experimental photoluminescence plots. A very generic comparison for the efficiency of p-i-n junction solar cell is given for a cell incorporating QDs and sans QDs. The design and fabrication technique is discussed in brief. I have shown that by using QDs in the intrinsic region of a cell, we can improve the efficiency by a factor of 1.865 times. Thus for a solar cell of efficiency of 26% for first generation solar cell, we can improve the efficiency to nearly 48.5% on using QDs.

Parameter estimation for transformer modeling

Large Power transformers, an aging and vulnerable part of our energy infrastructure, are at choke points in the grid and are key to reliability and security. Damage or destruction due to vandalism, misoperation, or other unexpected events is of great concern, given replacement costs upward of $2M and lead time of 12 months. Transient overvoltages can cause great damage and there is much interest in improving computer simulation models to correctly predict and avoid the consequences. EMTP (the Electromagnetic Transients Program) has been developed for computer simulation of power system transients. Component models for most equipment have been developed and benchmarked. Power transformers would appear to be simple. However, due to their nonlinear and frequency-dependent behaviors, they can be one of the most complex system components to model. It is imperative that the applied models be appropriate for the range of frequencies and excitation levels that the system experiences. Thus, transformer modeling is not a mature field and newer improved models must be made available. In this work, improved topologically-correct duality-based models are developed for three-phase autotransformers having five-legged, three-legged, and shell-form cores. The main problem in the implementation of detailed models is the lack of complete and reliable data, as no international standard suggests how to measure and calculate parameters. Therefore, parameter estimation methods are developed here to determine the parameters of a given model in cases where available information is incomplete. The transformer nameplate data is required and relative physical dimensions of the core are estimated. The models include a separate representation of each segment of the core, including hysteresis of the core, λ-i saturation characteristic, capacitive effects, and frequency dependency of winding resistance and core loss. Steady-state excitation, and de-energization and re-energization transients are simulated and compared with an earlier-developed BCTRAN-based model. Black start energization cases are also simulated as a means of model evaluation and compared with actual event records. The simulated results using the model developed here are reasonable and more correct than those of the BCTRAN-based model. Simulation accuracy is dependent on the accuracy of the equipment model and its parameters. This work is significant in that it advances existing parameter estimation methods in cases where the available data and measurements are incomplete. The accuracy of EMTP simulation for power systems including three-phase autotransformers is thus enhanced. Theoretical results obtained from this work provide a sound foundation for development of transformer parameter estimation methods using engineering optimization. In addition, it should be possible to refine which information and measurement data are necessary for complete duality-based transformer models. To further refine and develop the models and transformer parameter estimation methods developed here, iterative full-scale laboratory tests using high-voltage and high-power three-phase transformer would be helpful.

DEVELOPMENT OF A CONCEPTUAL MODEL AND PARAMETERIZATION OF A MATHEMATICAL MODEL FOR SEDIMENT PHOSPHORUS DIAGENESIS, SORPTION AND RELEASE FROM ONONDAGA LAKE SYRACUSE, NEW YORK

Eutrophication is a persistent problem in many fresh water lakes. Delay in lake recovery following reductions in external loading of phosphorus, the limiting nutrient in fresh water ecosystems, is often observed. Models have been created to assist with lake remediation efforts, however, the application of management tools to sediment diagenesis is often neglected due to conceptual and mathematical complexity. SED2K (Chapra et al. 2012) is proposed as a "middle way", offering engineering rigor while being accessible to users. An objective of this research is to further support the development and application SED2K for sediment phosphorus diagenesis and release to the water column of Onondaga Lake. Application of SED2K has been made to eutrophic Lake Alice in Minnesota. The more homogenous sediment characteristics of Lake Alice, compared with the industrially polluted sediment layers of Onondaga Lake, allowed for an invariant rate coefficient to be applied to describe first order decay kinetics of phosphorus. When a similar approach was attempted on Onondaga Lake an invariant rate coefficient failed to simulate the sediment phosphorus profile. Therefore, labile P was accounted for by progressive preservation after burial and a rate coefficient which gradual decreased with depth was applied. In this study, profile sediment samples were chemically extracted into five operationally-defined fractions: CaCO3-P, Fe/Al-P, Biogenic-P, Ca Mineral-P and Residual-P. Chemical fractionation data, from this study, showed that preservation is not the only mechanism by which phosphorus may be maintained in a non-reactive state in the profile. Sorption has been shown to contribute substantially to P burial within the profile. A new kinetic approach involving partitioning of P into process based fractions is applied here. Results from this approach indicate that labile P (Ca Mineral and Organic P) is contributing to internal P loading to Onondaga Lake, through diagenesis and diffusion to the water column, while the sorbed P fraction (Fe/Al-P and CaCO3-P) is remaining consistent. Sediment profile concentrations of labile and total phosphorus at time of deposition were also modeled and compared with current labile and total phosphorus, to quantify the extent to which remaining phosphorus which will continue to contribute to internal P loading and influence the trophic status of Onondaga Lake. Results presented here also allowed for estimation of the depth of the active sediment layer and the attendant response time as well as the sediment burden of labile P and associated efflux.

Direction finding in the presence of a more realistic environment model

Direction-of-arrival (DOA) estimation is susceptible to errors introduced by the presence of real-ground and resonant size scatterers in the vicinity of the antenna array. To compensate for these errors pre-calibration and auto-calibration techniques are presented. The effects of real-ground constituent parameters on the mutual coupling (MC) of wire type antenna arrays for DOA estimation are investigated. This is accomplished by pre-calibration of the antenna array over the real-ground using the finite element method (FEM). The mutual impedance matrix is pre-estimated and used to remove the perturbations in the received terminal voltage. The unperturbed terminal voltage is incorporated in MUSIC algorithm to estimate DOAs. First, MC of quarter wave monopole antenna arrays is investigated. This work augments an existing MC compensation technique for ground-based antennas and proposes reduction in MC for antennas over finite ground as compared to the perfect ground. A factor of 4 decrease in both the real and imaginary parts of the MC is observed when considering a poor ground versus a perfectly conducting one for quarter wave monopoles in the receiving mode. A simulated result to show the compensation of errors direction of arrival (DOA) estimation with actual realization of the environment is also presented. Secondly, investigations for the effects on received MC of λ/2 dipole arrays placed near real-earth are carried out. As a rule of thumb, estimation of mutual coupling can be divided in two regions of antenna height that is very near ground 0