Tag Archives: control

AI & ML seminar at Stanford Feb. 24, News

Seminar at Bits & Watts (Stanford) on Machine Learning & AI for Power Systems

January 2022

I am very excited with Dr. Liang Min‘s invitation to present my Smart Grid works on power system control with machine learning and artificial intelligence in the framework of the Bits & Watts Initiative at Stanford! The seminar will take place on Feb. 24th and I will go over the use of top-down heuristically inducted binary decision trees to procure firm capacity by renewables with volatility, and on how voltage control can be modeled as a problem of classical mechanics physics. I look forward to hearing attendees’ ideas and thoughts on other machine learning and AI applications in power system optimization, planning and control.

The seminar will be in-person, so if you are faculty, student & researcher at Stanford and would like us to meet before/after the seminar, please, do not hesitate to reach out!

Inverter Interfaced Renewables Seminar at Yale Oct. 13, News

Seminar at Yale on Inverter Control for Grids Rich in Renewables

October 2021

I am extremely lucky to have recently met and exchanged ideas and research aspirations with Prof. Leandros Tassiulas, chair of Electrical Engineering (EE) at Yale.  He has honored me with an invitation to offer a seminar to the Dept. of EE and the Institute for Network Science at Yale on October 13th. I will be presenting 2 of my earlier research works on control methods to procure active power reserves from wind generators and string photovoltaic inverters. Even though these technologies typically pursue maximum use of their aerodynamic and solar potential, respectively, they must also be able to support system stability. This becomes even more critical as renewable resources slowly dominate the grid and displace conventional resources that have until recently ensured stability. I will extend my previous results into the most recent research aspirations for a grid dominated by inverter-interfaced renewables and batteries and how such aspirations may be made possible.

The seminar will be virtual, but I will make myself available to all faculty, students & researchers at Yale, who would like us to talk before/after the seminar, so, please, do not hesitate to reach out!

AI & ML seminar at RPI Sep. 29, News

Seminar at RPI on Power System Control with Machine Learning & Artificial Intelligence

August 2021

I want to thank Prof. Mona Mostafa Hella and Dr. Luigi Vanfretti, my friend and collaborator at the North American Synchrophasor Initiative (NASPI), for inviting me to offer a seminar at the Dept. of Electrical, Computer & Systems Engineering at the Rensselaer Polytechnic Institute on September 29th. I will review 2 of my works on generation control with machine learning (ML) and artificial intelligence (AI). I will start by discussing how to use top-down heuristically inducted binary decision trees of ML to actively control firm capacity by volatile resources operated (among others units) as a Virtual Power Plant. In the second part, I will present how voltage control can be modeled as a problem of classical mechanics physics; from there it can be solved as an AI implementation of the 2nd law of thermodynamics to redispatch active and reactive power generation. I plan to spark a discussion on conceiving new ML applications and AI models for power system operational control and monitoring.

The seminar will be virtual, but I will make myself available to all faculty, students & researchers of RPI, who would like us to talk before/after the seminar, so, please, do not hesitate to reach out!

News, PV-BSS seminar at OSU Nov. 10

Seminar at Ohio State University on residential PV & batteries

August 2021

I am grateful to my friend and collaborator at the IET Renewable Power Generation journal, Prof. Ramteen Sioshansi for kindly inviting me to offer a seminar at the Dept. of Integrated Systems Engineering at Ohio State University on November 10. I will talk about my earlier work at CMU on the Dept. of Energy SHINES project, on how batteries can allow residential end-customers to widely benefit from behind-the-meter photovoltaics. Reducing the costly effects of demand charges promises additional value to that of net-metering or self-consumption from photovoltaics, while the policy implications offer much food for thought on the role of utilities, cooperatives and/or microgrids.

Depending on the situation with COVID the seminar might be in-person, so follow me on Twitter & LinkedIn for updates.

IEEE Standard 2660.1 Published, News

Standard IEEE 2660.1-2020 Published

February 2021

After many efforts over about a 5-year period I am happy to report the publication of the IEEE Standard 2660.1-2020 titled “Recommended Practice for Industrial Agents.” You can find it here. We are moving fast into an IoT era where local decision making either in the form of optimization, control action or system assessment, becomes critical and widespread. The most typical software entity that performs some type of any activity at a local level and in the form of a module is the ‘agent.’ Agents have been the backbone of many approaches, paradigms and architectures in dozens of applications. However, there is little information or methodology of how an agent should be deployed for a certain purpose, how to interact with other agents or the equipment it drives and the data it collects. This standard introduces exactly that. An algorithm (in the abstract sense) that takes into account the premises of an application and ranks according to various metrics which type of programming, organizing and communication protocols would fit best the said application.

This standard represents a major step forward in opening up a wide and clearly specified path for agents to be deployed in applications of the buildings, industrial, power, energy and other sectors. Stakeholders in these fields can employ this standard to best define the value of every different agent implementation in light of each application scoped.

I have been fortunate to work with great collaborators and honored to serve as the subgroup chair for the Energy & Power systems applications.