Machine Learning for the New York City Power Grid

Abstract

Abstract-Power companies can benefit from the use of knowledge discovery methods and statistical machine learning for preventive maintenance. We introduce a general process for transforming historical electrical grid data into models that aim to predict the risk of failures for components and systems. These models can be used directly by power companies to assist with prioritization of maintenance and repair work. Specialized versions of this process are used to produce 1) feeder failure rankings, 2) cable, joint, terminator and transformer rankings, 3) feeder MTBF (Mean Time Between Failure) estimates and 4) manhole events vulnerability rankings. The process in its most general form can handle diverse, noisy, sources that are historical (static), semi-real-time, or real-time, incorporates state-of-the-art machine learning algorithms for prioritization (supervised ranking or MTBF), and includes an evaluation of results via cross-validation and blind test. Above and beyond the ranked lists and MTBF estimates are business management interfaces that allow the prediction capability to be integrated directly into corporate planning and decision support; such interfaces rely on several important properties of our general modeling approach: that machine learning features are meaningful to domain experts, that the processing of data is transparent, and that prediction results are accurate enough to support sound decision making. We discuss the challenges in working with historical electrical grid data that were not designed for predictive purposes. The "rawness" of these data contrasts with the accuracy of the statistical models that can be obtained from the process; these models are sufficiently accurate to assist in maintaining New York City's electrical grid.Index Terms-applications of machine learning, electrical grid, smart grid, knowledge discovery, supervised ranking, computational sustainability, reliability !

Interactions between learning and decision making

“…Machine learning models have started to be used for proactive maintenance in NYC, where supervised ranking algorithms are used to rank the manholes in order of predicted susceptibility to failure (fires, explosions, smoke) so that the most vulnerable manholes can be prioritized [Rudin et al, 2010[Rudin et al, , 2012a. The machine learning algorithms make reasonably accurate predictions of manhole vulnerability; however, they do not (nor would they, using any other prediction-only technique) take the cost of repairs into account when making the ranked lists.…”

“…In New York City, there are several separate new preemptive maintenance programs, including the targeted inspection program for electrical service structures (manholes), programs that perform extensive repairs that were placed on a waiting list after the manhole was inspected, and the vented cover replacement program, where each manhole is replaced with a vented cover that allows gases to escape, mitigating the possibility and effects of serious events including fires and explosions. Con Edison, the power company in NYC, has the ability to use machine learning models in Manhattan, Brooklyn and the Bronx for scheduling of manhole inspection and repair work [Rudin et al, 2010[Rudin et al, , 2012b. This project was the motivation for the development of the ML&TRP and we use data from the NYC power grid for our experiments.…”

Fuzzy clustering and prediction of electricity demand based on household characteristics

“…Computational techniques in the fields of statistical and machine learning are starting to be extensively used in order to extract knowledge from the data generated by the grid [2,3]. This paper proposes a methodology for: (1) the segmentation of residential electricity consumers, and (2) the prediction of electricity demand profiles based on household characteristics.…”

AdaBoost$^{+}$: An Ensemble Learning Approach for Estimating Weather-Related Outages in Distribution Systems

“…Ensemble learning is a technique that embodies one of the main directions of current machine learning research. Although in this paper the ensemble's constituent units are referred to as neural networks, an ensemble could equally well comprise of other learning models, such as support vector machines [13,14], kernel-based models [15], radial basis function networks [16,17], decision trees [14,18], fuzzy logic [19] or ARMA models [21] in addition to neural networks [16,17,[19][20][21]. Random forests [22] are another good method for classification and regression.…”