There is, I think, something ethereal about i —the square root of minus one. I remember first hearing about it at school. It seemed an odd beast at that time—an intruder hovering on the edge of reality.

Usually familiarity dulls this sense of the bizarre, but in the case of i it was the reverse: over the years the sense of its surreal nature intensified. It seemed that it was impossible to write mathematics that described the real world in …

I and i

Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Data Mining - Concepts and Techniques.

Data Mining Practical Machine Learning Tools and Techniques

High level of delivered power quality (PQ) is becoming one of the key performance indicators for both contemporary and future power networks. The increased proliferation of converter connected generation and load in power networks, increased sensitivity to network disturbances of some of these new types of devices and requirements for more flexible operation of power networks led to the revision of some of PQ standards and introduction of modified or in some cases new requirements for PQ compliance. Although almost all PQ phenomena, with exception of voltage transients, are well defined and appropriate thresholds for individual phenomena are set in international standards, there is no standardised nor commonly accepted way to describe and evaluate the overall PQ performance at buses. This study presents an analytic hierarchy process (AHP) inspired methodology for assessing the overall PQ performance at a bus based on several different PQ phenomena considered simultaneously. Compound bus PQ index is defined using AHP to present the overall PQ performance at the bus with respect to voltage sag, harmonics and voltage unbalance. The application of the methodology is illustrated on a 295 bus generic distribution network.

/pdf/compound-index-for-power-quality-evaluation-and-benchmarking-3hdqewnlns.pdf

Compound index for power quality evaluation and benchmarking

This paper is concerned with the analysis of voltage sags characteristics and propagation in radial networks following the faults in the system. Both symmetrical and asymmetrical faults are simulated at the different voltage levels and different locations in the network. Realistic pre-fault voltages and voltage phase-angles obtained from load flow study are used in all calculations. The influence of the size of the embedded generation, dynamic loads, load structure and composition on voltage sags characteristics is also investigated.

Characteristics of voltage sags in radial networks with dynamic loads and embedded generators

This paper provides a detailed overview of the techniques and applications related to distribution system state estimation (DSSE), together with the classification of various types of state estimation. The paper also provides the state-of-art techniques applied in DSSE including forecasted-aided state estimation, close-loop DSSE methods, the application of computation intelligence in DSSE and the use of smart meters and phasor measurement units in load estimation. As more and more active elements and functionalities will be integrated in future distribution network, e.g. demand-side management, the increased penetration of decentralized generation and dynamically controlled storage/devices, DSSE will be one of the critical functionalities for secure operation of future distribution networks. As a pathway to cost-efficient DSSE for future distribution networks, this paper addresses the characteristics of future distribution grids that will affect DSSE, and also discusses the techniques and data that can be used to better understand/model the network behaviors and load profiles in order to ultimately improve the accuracy of DSSE.

/pdf/pathway-to-cost-efficient-state-estimation-of-future-4i6uw6lx4b.pdf

Pathway to cost-efficient state estimation of future distribution networks

Accurate prediction of load composition at bulk supply points can significantly improve power system planning, electricity market analysis and demand side management. This paper discusses an artificial neural network (ANN) based approach to forecasting load composition at the bulk supply bus based on RMS measurement of voltage, real and reactive power and local forecasted weather. Probabilistic distributions and confidence levels of the prediction under different prediction error intervals have been derived and analysed. It is demonstrated that the approach yields prediction of load composition with errors typically less than 10%.

Accuracy of ANN based methodology for load composition forecasting at bulk supply buses

With the increase in the scale of demand side management (DSM) activities, the load characteristic of power system at the load reduction and recovery times will experience significant changes due to DSM actions and hence affect the dynamic performance of power systems. This paper proposes a methodology to assess the effect of DSM actions on power system network performance — small signal stability of transmission network. A case study of shifting of different type of load from one hour to the other is tested and investigated to what extent the impact of these DSM activities can have on grid performance. A generic 295 bus distribution network and the IEEE standard 9-bus transmission network are used to demonstrate this effect. In the proposed methodology, different types of load sectors (domestic, commercial, industrial) with corresponding mix of load categories (small and large induction motors, constant power, constant impedance and constant current load) in each of them are modelled in DigSILENT/PowerFactory based on typical structure and composition of demand in distribution network. The paper demonstrates and quantifies the impact of the DSM actions on transmission network performance and establish the extent to which DSM can contribute, or otherwise, to power system stability. It will pave the way to development of optimal DSM that will ensure that not only total demand is managed but also the way how it should be managed in order not to endanger network performance in case of disturbances.

Jovica V. Milanovic

Papers

Compound index for power quality evaluation and benchmarking

Characteristics of voltage sags in radial networks with dynamic loads and embedded generators

Pathway to cost-efficient state estimation of future distribution networks

Accuracy of ANN based methodology for load composition forecasting at bulk supply buses

Assessment of the impact of demand side management on power system small signal stability