With advances in technology and the increasing use of power electronic components in the design of household and industrial equipment, harmonic distortion has become one of the major power quality problems in power systems. Identifying the harmonic sources and quantifying the contributions of these harmonic sources provides utility companies with the information they require to effectively mitigate harmonics in the system. This thesis proposes the use of observers for harmonic estimation and harmonic source identification. An iterative observer algorithm is designed for performing harmonic estimation in measured voltage or current signals taken from a power distribution system. The algorithm is based on previous observer designs for estimating the power system states at the fundamental frequency. Harmonic estimation is only carried out when the total harmonic distortion (THD) exceeds a specified threshold. In addition, estimation can be performed on multiple measurements simultaneously. Simulations are carried out on an IEEE distribution test feeder. A number of scenarios such as changes in harmonic injections with time, variations in fundamental frequency and measurement noise are simulated to verify the validity and robustness of the proposed iterative observer algorithm. Furthermore, an observer-based algorithm is proposed for identifying the harmonic sources in power distribution systems. The observer is developed to estimate the system states for a combination of suspicious nodes and the estimation error is analysed to verify the existence of harmonic sources in the specified node combinations. This method is applied to the identification of both single and multiple harmonic sources. The response of the observer-based algorithm to time varying load parameters and variations in harmonic injections with time is investigated and the results show that the proposed harmonic source identification algorithm is able to adapt to these changes. In addition, the presence of time delay in power distribution system measurements is taken into consideration when identifying harmonic sources. An observer is designed to estimate the system states for the case of a single time delay as well as multiple delays in the measurements. This observer is then incorporated into the observer-based harmonic source identification algorithm to identify harmonic sources in the presence of delayed measurements. Simulation results show that irrespective of the time delay in the measurements, the algorithm accurately identifies the harmonic sources in the power distribution system.

No

No

Smart inverters have been considered the primary fast solution for voltage regulation in power distribution systems. Optimizing the coordination between inverters can be computationally challenging. Reactive power control using fixed local rules have been shown to be subpar. Here, nonlinear inverter control rules are proposed by leveraging machine learning tools. The designed control rules can be expressed by a set of coefficients. These control rules can be nonlinear functions of both remote and local inputs. The proposed control rules are designed to jointly minimize the voltage deviation across buses. By using the support vector machines, control rules with sparse representations are obtained which decrease the communication between the operator and the inverters. The designed control rules are tested under different grid conditions and compared with other reactive power control schemes. The results show promising performance. / With advent of renewable energies into the power systems, innovative and automatic monitoring and control techniques are required. More specifically, voltage regulation for distribution grids with solar generation is a can be a challenging task. Moreover, due to frequency and intensity of the voltage changes, traditional utility-owned voltage regulation equipment are not useful in long term. On the other hand, smart inverters installed with solar panels can be used for regulating the voltage. Smart inverters can be programmed to inject or absorb reactive power which directly influences the voltage. Utility can monitor, control and sync the inverters across the grid to maintain the voltage within the desired limits. Machine learning and optimization techniques can be applied for automation of voltage regulation in smart grids using the smart inverters installed with solar panels. In this work, voltage regulation is addressed by reactive power control.

Doctor of Philosophy / Department of Electrical and Computer Engineering / Anil Pahwa, Sanjoy Das / Occurrence of outages in overhead distribution systems is a significant factor in determining distribution system reliability. Analysis of animal-related outages has practical value since animals cause a large number of outages in overhead distribution systems. This dissertation presents several different methods to investigate the impact of weather and time of the year on the animal-related outage rate. The animal-related outages from year 1998 to year 2007 for different cities in Kansas are provided by Westar Energy. From examinations of the historical data, two factors which influence the animal-related outages, the month type and the number of fair weather days are taken as inputs along with historical outage data for prediction models. Poisson regression model, neural network model, wavelet based neural network model and Bayesian model combined with Monte Carlo simulations are applied to the weekly data of different cites. Even though Poisson regression models, Bayesian models and neural network models are able to recognize the changing pattern of outage rates under different weather conditions, they are limited in their ability to follow the high peaks in the time series of weekly animal-related outages. The introduction of wavelet transform techniques overcomes this problem. Simulation results indicate that the wavelet based neural network models are able to capture the pattern of fast fluctuations in the weekly outages of different cities in Kansas of various sizes. A hyperpermutation method inspired by artificial immune system algorithm is used to solve the overtraining problem in the application of neural networks. Finally, Monte Carlo simulations based on conditional probability tables from Bayesian models are used to find out the confidence intervals of the predictions. We aggregate the weekly data and carry out the analysis on a monthly and yearly basis too. Simulation results indicate that the models are able to capture the pattern as at least 90% of the observed values are within the upper limits of 95% confidence in the predictions for weekly, monthly and yearly animal-related outages of different cities in Kansas. The results obtained from Monte Carlo simulations are compared with the wavelet based neural network model to indentify years with more than expected level of outages.

No description available.

Water distribution systems are large and complex structures. Hence, their construction, management and improvements are time consuming and expensive. But nearly all the optimisation methods, whether aimed at design or operation, suffer from the need for simulation models necessary to evaluate the performance of solutions to the problem. These simulation models, however, are increasing in size and complexity, and especially for operational control purposes, where there is a need to regularly update the control strategy to account for the fluctuations in demands, the combination of a hydraulic simulation model and optimisation is likely to be computationally excessive for all but the simplest of networks. The work presented in this thesis has been motivated by the need for reduced, whilst at the same time appropriately accurate, models to replicate the complex and nonlinear nature of water distribution systems in order to optimise their operation. This thesis attempts to establish the ground rules to form an underpinning basis for the formulation and subsequent evaluation of such models. Part I of this thesis introduces some of the modelling, simulation and optimisation problems currently faced by water industry. A case study is given to emphasise one particular subject, namely reduction of water distribution system models. A systematic research resulted in development of a new methodology which encapsulate not only the system mass balance but also the system energy distribution within the model reduction process. The methodology incorporates the energy audits concepts into the model reduction algorithm allowing the preservation of the original model energy distribution by imposing new pressure constraints in the reduced model. The appropriateness of the new methodology is illustrated on the theoretical and industrial case studies. Outcomes from these studies demonstrate that the new extension to the model reduction technique can simplify the inherent complexity of water networks while preserving the completeness of original information. An underlying premise which forms a common thread running through the thesis, linking Parts I and II, is in recognition of the need for the more efficient paradigm to model and simulate water networks; effectively accounting for the discontinuous behaviour exhibited by water network components. Motivated largely by the potential of contemplating a new paradigm to water distribution system modelling and simulation, a further major research area, which forms the basis of Part II, leads to a study of the discrete event specification formalism and quantised state systems to formulate a framework within which water distribution systems can be modelled and simulated. In contrast to the classic time-slicing simulators, depending on the numerical integration algorithms, the quantisation of system states would allow accounting for the discontinuities exhibited by control elements in a more efficient manner, and thereby, offer a significant increase in speed of the simulation of water network models. The proposed approach is evaluated on a number of case studies and compared with results obtained from the Epanet2 simulator and OpenModelica. Although the current state-of-art of the simulation tools utilising the quantised state systems do not allow to fully exploit their potential, the results from comparison demonstrate that, if the second or third order quantised-based integrations are used, the quantised state systems approach can outperform the conventional water network simulation methods in terms of simulation accuracy and run-time.

To date, closed form optimal solutions for stocking levels in arborescent multiechelon inventory systems have not been obtained. These problems exhibit the joint difficulties of requiring an allocation policy as well as a stocking policy, and the multidimensional nature of their state space makes dynamic programming formulations impractical. In this dissertation, we introduce procedures that approximate multiechelon networks with sets of single installation problems. We first use this technique to solve for base-stock levels in a distribution network with asymmetric retailers. Second, we use this technique to analyze delayed differentiation production processes and provide guidance as to when the strategy is most warranted. Third, we modify the technique to account for inventory that exhibits perishability and solve for stocking policies for distribution systems when the inventory has a fixed shelf life.

No description available.

To protect public health, detection methods have been developed to monitor drinking water for pathogens. The goal of this dissertation is to evaluate and utilize novel methods that enhances detection and further reduces the risk of waterborne pathogens. The study in Appendix A developed a method to monitor the microbial quality of treated drinking water at the tap utilizing point-of-use (POU) filter. Tap water supplies were monitored in vending machines throughout Southern Arizona using solid block carbon (SBC) filters as a monitoring tool. Out of 48 SBC filters 54.2% were positive for at least one organism. The number of filters positive for total coliforms, E. coli, Enterococci, and enterovirus was 13, 5, 19, and 3, respectively, corresponding to 27.1%, 10.4%, 39.6%, and 6.3% of the total filters. These results suggest that the SBC filter can be used to monitor large volumes of treated drinking water and detect the incidence of indicators and pathogens. The study in Appendix B evaluated the fate of infectious prions in multiple water sources quantitatively utilizing a method that only detects infectious prions. A reduction of PrPˢᶜ was observed at 25°C and 37°C ranging between 0.41-log₁₀ and 1.4-log₁₀ after 1 week. After 8 weeks at 25°C and 37°C, inactivation ranged between 1.65-log₁₀ and 2.15-log₁₀. A maximum rate of inactivation in water occurred at 50°C, ranging from 2.0-log₁₀ and 2.51-log₁₀ after one week. The results from all types of water suggest that dissolved organic matter and temperature influence PrPˢᶜ infectivity. The study in Appendix C evaluated real-time sensors for monitoring microbial contaminants. Most sensor parameters evaluated exhibited an increase in sensor response to an increase in E. coli concentrations. Responses to E. coli concentrations at or below 10³ cfu/mL were very low due to near background levels, and responses to concentrations above 10⁶ cfu/mL exceeded threshold levels for sensors that use light scattering methods due to saturation in the flow cell. The data produced effectively shows that the sensors evaluated could be used to monitor microbial intrusion events in water distribution systems.