Mathematical Methods for Data-driven Power Systems

PI: Juan Miguel Morales González

Acronym: PowerMath

Abstract: The primary objective of PowerMath is to develop a disruptive methodology that facilitates the smart use of data for power systems operations. The research team will challenge this classical scheme for decision-making and develop a new one whereby decisions are directly inferred from data, thus averting the need for estimating a prediction model. That is, under the proposed new paradigm, data is not used to model the processes driving the decision-making, but to estimate the decisions themselves.
The procedures and techniques that the proposed framework will produce will be the result of bridging two different disciplines, namely, those of statistical learning and optimization-based decision-making. The advantages of the proposed methodology will be demonstrated on the solution to paradigmatic problems arising in the operation of power systems with a large penetration of renewables and distributed energy sources. 
More particularly, the proposed methodology will enable us to reformulate procedures for power system operations such as economic dispatch, unit commitment, microgrid and demand-side management, etc. in a way such that they can make the most of the available data to make efficient operating decisions. This will allow reaping all the benefits brought by the ICT-based modernization of the power sector while ensuring a resilient, reliable, and secure operation of the electrical infrastructure.