Bruce Stephen

Wind Turbine Condition Assessment Through Power Curve Copula Modeling

Power curves constructed from wind speed and active power output measurements provide an established method of analyzing wind turbine performance. In this paper, it is proposed that operational data from wind turbines are used to estimate bivariate probability distribution functions representing the power curve of existing turbines so that deviations from expected behavior can be detected. Owing to the complex form of dependency between active power and wind speed, which no classical parameterized distribution can approximate, the application of empirical copulas is proposed; the statistical theory of copulas allows the distribution form of marginal distributions of wind speed and power to be expressed separately from information about the dependency between them. Copula analysis is discussed in terms of its likely usefulness in wind turbine condition monitoring, particularly in early recognition of incipient faults such as blade degradation, yaw, and pitch errors.

A Copula Model of Wind Turbine Performance

The conventional means of assessing the performance of a wind turbine is through consideration of its power curve which provides the relationship between power output and measured wind speed. In this letter, it is shown how the joint probability distribution of power and wind speed can be learned from data, rather than from examination of the implied function of the two variables. Such an approach incorporates measures of uncertainty into performance estimates, allows inter-plant performance comparison, and could be used to simulate plant operation via sampling. A preliminary model is formulated and fitted to operational data as an illustration.

Enhanced Load Profiling for Residential Network Customers

Anticipating load characteristics on low voltage circuits is an area of increased concern for Distribution Network Operators with uncertainty stemming primarily from the validity of domestic load profiles. Identifying customer behavior makeup on a LV feeder ascertains the thermal and voltage constraints imposed on the network infrastructure; modeling this highly dynamic behavior requires a means of accommodating noise incurred through variations in lifestyle and meteorological conditions. Increased penetration of distributed generation may further worsen this situation with the risk of reversed power flows on a network with no transformer automation. Smart Meter roll-out is opening up the previously obscured view of domestic electricity use by providing high resolution advance data; while in most cases this is provided historically, rather than real-time, it permits a level of detail that could not have previously been achieved. Generating a data driven profile of domestic energy use would add to the accuracy of the monitoring and configuration activities undertaken by DNOs at LV level and higher which would afford greater realism than static load profiles that are in existing use. In this paper, a linear Gaussian load profile is developed that allows stratification to a finer level of detail while preserving a deterministic representation.

Domestic Load Characterization Through Smart Meter Advance Stratification

The heterogeneity of domestic loads presents distribution network operators with operational uncertainties which may become problematic as generation capacity shrinks and network infrastructure ages. High resolution meter advances recorded by increasingly ubiquitous Smart Meters can be seen as representing base loads along with aggregations of multiple domestic appliances-in this letter, a Bayesian formulation of the finite mixture probability distribution is employed to enumerate and capture generalizations of these, from which compact representations of domestic load profiles can be formed.

Implementation of herd management systems with wireless sensor networks

The work summarises a study of the data communications requirements for agricultural livestock monitoring applications using wireless sensor networks (WSNs). Several design challenges are identified and analysed in depth based on actual global positioning system positioning data gathered from an actual herd of cattle. A wireless system including antennae diversity together with data downloads optimisation schemes utilising data collector and routers are developed and tested in a working farm environment. Two analysis metrics, connection availability and connection duration, are used to quantify the impact of cattle movement on network connectivity. The major contributions of this study stem from a definition of the communication issues in deploying animal monitoring platforms in free-ranging farm environments and the analysis and optimisation of the wireless data download performance using as the foundation knowledge gained from a series of working farm trials. Additionally, the data download protocols are designed particularly to treat animal movement. The results prove the viability of WSN-based solutions for livestock monitoring applications.

Adaptation of wireless sensor network for farming industries

In recent years, wireless sensor networks (WSN) have received considerable attention within agriculture and farming as a means to reduce operational costs and enhance animal health care. This paper examines the application of WSNs to livestock monitoring and the issues related to hardware realization. The core of this study is to overcome the aforementioned drawbacks by using alternative cheap, low power consumption sensor nodes capable of providing real-time communication at a reasonable hardware cost. In this paper, various factors i.e. radio frequency selection, channel bandwidth, etc. have been evaluated to provide a solution which can obtain real-time data from diary cattle whilst conforming to the limitations associated with WSNs implementations.

Self-learning load characteristic models for smart appliances

It is generally accepted that if dynamic electricity pricing tariffs were to be introduced, their effectiveness in controlling domestic loads will be curtailed if consumers were relied on to respond in their own interests. The complexities of relating behavior to load to price are so burdensome that at least some degree of automation would be required to take advantage of pricing signals. However, a major issue with home automation is fitting in with the lifestyles of individual consumers. Truly smart appliances that can learn the details of their routine operation may be several years away from widespread adoption making integrated home energy management systems unfeasible. Similarly, usage patterns of these same appliances may be substantially different from household to household. The contribution of this paper is the proposal and demonstration of a set of probabilistic models that act in a framework to reduce appliance usage data into contextual knowledge that accounts for variability in patterns in usage. Using sub-metered load data from various domestic wet appliances, the proposed technique is demonstrated learning the appliance operating likelihood surfaces from no prior knowledge.

Compositional Modeling of Partial Discharge Pulse Spectral Characteristics

Partial discharge (PD) monitoring is an established method for insulation health monitoring in high-voltage plant. A number of different approaches to PD defect diagnosis have been developed to extract defect-specific information from PD pulse data in both the time and frequency domains. Frequency-based PD pulse analysis has previously been demonstrated to offer a low-power approach to PD defect identification, where a mixture of passive and active analog electronics can be used to generate diagnostic features in a low-power device suited to wireless sensor network operation. This paper examines approaches to implementing diagnostic methods for frequency-based PD pulse diagnosis targeted at compositional frequency spectrum features in a computationally efficient manner. Dirichlet and Gaussian distributions are used to demonstrate the complex probabilistic form of fault class decision surfaces, which motivates the proposed application of the log ratio transform to frequency composition data. The results demonstrate that PD defects can be differentiated using these frequency-based methods and that employing the log ratio transform to the compositional frequency content data yields increases in classification accuracy without necessarily resorting to more complex classifiers.

Practical Applications of Data Mining in Plant Monitoring and Diagnostics

Using available expert knowledge in conjunction with a structured process of data mining, characteristics observed in condition monitoring data (which represent modes of plant operation) may be understood, explained and quantified. Knowledge and understanding of satisfactory and unsatisfactory plant condition can be gained and made explicit from the analysis of data observations and subsequently used to form the basis of condition assessment and diagnostic rules/models implemented in decision support systems supporting plant maintenance. This paper proposes a data mining method for the analysis of condition monitoring data, and demonstrates this method in its discovery of useful knowledge from trip coil data captured from a population of in-service distribution circuit breakers and empirical UHF data captured from laboratory experiments simulating partial discharge defects typically found in HV transformers. This discovered knowledge then forms the basis of two separate decision support systems for the condition assessment/defect clarification of these respective plant items.

The Use of Hidden Markov Models for Anomaly Detection in Nuclear Core Condition Monitoring

Unplanned outages can be especially costly for generation companies operating nuclear facilities. Early detection of deviations from expected performance through condition monitoring can allow a more proactive and managed approach to dealing with ageing plant. This paper proposes an anomaly detection framework incorporating the use of the Hidden Markov Model (HMM) to support the analysis of nuclear reactor core condition monitoring data. Fuel Grab Load Trace (FGLT) data gathered within the UK during routine refueling operations has been seen to provide information relating to the condition of the graphite bricks that comprise the core. Although manual analysis of this data is time consuming and requires considerable expertise, this paper demonstrates how techniques such as the HMM can provide analysis support by providing a benchmark model of expected behavior against which future refueling events may be compared. The presence of anomalous behavior in candidate traces is inferred through the underlying statistical foundation of the HMM which gives an observation likelihood averaged along the length of the input sequence. Using this likelihood measure, the engineer can be alerted to anomalous behaviour, indicating data which might require further detailed examination. It is proposed that this data analysis technique is used in conjunction with other intelligent analysis techniques currently employed to analyse FGLT to provide a greater confidence measure in detecting anomalous behaviour from FGLT data.