As China Challenges The U.S. in AI, Big Data And Machine Learning Are Reshaping The Energy Industry

#artificialintelligence

Machine learning, Big Data, and automation are revolutionizing global industry – and the energy sector is no exception. Innovation is driving technological progress, boosting economic efficiency, creating smarter business operations, and leading to more resilient infrastructure. It's why businesses and governments around the world are making advanced technology – including artificial intelligence – a top economic and national security priority. Energy companies are implementation big data and AI in versatile ways – and the sector is growing. The market for AI software in the oil and gas industry is expected to reach a whopping $2.85 billion by 2022.


Real-time Anomaly Detection and Classification in Streaming PMU Data

arXiv.org Machine Learning

--Ensuring secure and reliable operations of the power grid is a primary concern of system operators. Phasor measurement units (PMUs) are rapidly being deployed in the grid to provide fast-sampled operational data that should enable quicker decision-making. This work presents a general interpretable framework for analyzing real-time PMU data, and thus enabling grid operators to understand the current state and to identify anomalies on the fly. Applying statistical learning tools on the streaming data, we first learn an effective dynamical model to describe the current behavior of the system. Next, we use the probabilistic predictions of our learned model to define in a principled way an efficient anomaly detection tool. Finally, the last module of our framework produces on-the-fly classification of the detected anomalies into common occurrence classes using features that grid operators are familiar with. We demonstrate the efficacy of our interpretable approach through extensive numerical experiments on real PMU data collected from a transmission operator in the USA. Traditional supervisory control and data acquisition (SCADA) systems provide information regarding the system state at the order of seconds to the operator. However, such fidelity, considered appropriate in prior decades, is not sufficient to observe or predict disturbances at faster timescales that are increasingly being observed in today's stochastic grid [1]. To provide more rapid measurement data, phasor measurement units (PMUs) have gained widespread deployment. PMUs [2] are time-synchronized by GPS timestamps and collect measurements of system states (Eg.


How Big Data Changes the Economics of Renewable Energy

#artificialintelligence

Jason Bordoff (@JasonBordoff), a former energy adviser to President Obama, is a professor of professional practice in international and public affairs and founding director of the Center on Global Energy Policy at Columbia University. Thanks to the exponential growth of information commonly known as "big data" and our increasingly sophisticated methods to analyze it, the machine learning revolution is increasingly disrupting old industries from advertising and transportation to fashion and the legal profession. The energy industry is no exception. Big data analytics are optimizing oil-field production and estimating oil storage levels via satellite images and remote sensing methods. But perhaps nowhere in the energy sector is the impact of big data more revolutionary than in the operations of the electricity system, where it will play an increasingly pivotal role integrating more and more renewables into the power mix.


On the Prior Sensitivity of Thompson Sampling

arXiv.org Machine Learning

The empirically successful Thompson Sampling algorithm for stochastic bandits has drawn much interest in understanding its theoretical properties. One important benefit of the algorithm is that it allows domain knowledge to be conveniently encoded as a prior distribution to balance exploration and exploitation more effectively. While it is generally believed that the algorithm's regret is low (high) when the prior is good (bad), little is known about the exact dependence. In this paper, we fully characterize the algorithm's worst-case dependence of regret on the choice of prior, focusing on a special yet representative case. These results also provide insights into the general sensitivity of the algorithm to the choice of priors. In particular, with $p$ being the prior probability mass of the true reward-generating model, we prove $O(\sqrt{T/p})$ and $O(\sqrt{(1-p)T})$ regret upper bounds for the bad- and good-prior cases, respectively, as well as \emph{matching} lower bounds. Our proofs rely on the discovery of a fundamental property of Thompson Sampling and make heavy use of martingale theory, both of which appear novel in the literature, to the best of our knowledge.


Fueling utility innovation through analytics

#artificialintelligence

Utilities around the world are making big investments in advanced analytics. Getting the full value, however, requires rethinking their strategy, culture, and organization. Advanced analytics can deliver enormous value for utilities and drive organizations to new frontiers of efficiency-- but only with the right approach. There's little to be gained from just bolting on a software solution. The real value comes from embedding data analytics as a core capability in the organization and using it to detect pain points, design solutions, and enable decision making.