Scribd
Upload
12 views8 pages

ML IN POWER SYSTEM ReserachProposalPresentation

The document discusses the integration of machine learning (ML) in optimizing power systems, highlighting the limitations of traditional methods and the potential of ML to enhance decision-making and system operations. It outlines various ML applications in load modeling and frequency regulation services, emphasizing innovative methodologies that can improve economic benefits for users and communities. The research aims to expand ML utilization in power systems by incorporating physical characteristics and solving previously unexplored problems.

Uploaded by

Samuel Bhukya
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
12 views8 pages

ML IN POWER SYSTEM ReserachProposalPresentation

The document discusses the integration of machine learning (ML) in optimizing power systems, highlighting the limitations of traditional methods and the potential of ML to enhance decision-making and system operations. It outlines various ML applications in load modeling and frequency regulation services, emphasizing innovative methodologies that can improve economic benefits for users and communities. The research aims to expand ML utilization in power systems by incorporating physical characteristics and solving previously unexplored problems.

Uploaded by

Samuel Bhukya
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 8

“Power System

Optimization Using ML
application”
Name of Scholar : BHUKYA SAMUEL NAIK

PhD Admission in School : SCHOOL OF ENGINEEERING


Area of Enrollment : Ph.D Machine Learning
Introduction
• The power grid is a classical subject in electrical engineering while machine learning is a
much newer subject.
• Power generation capacity and long-distance power transmission embrace revolutionary
changes.
• The traditional power system operation and planning models have become gradually
incapable of satisfying the modern system’s requirements for stable and efficient operation.
• SCADA-integrated Energy Management System (EMS), Wide-Area Measurement Systems
(WAMS), etc.
• Machine learning gains accumulated attention from researchers in the power area in the hope
that it can assist humans on decision making and system operation.
• Most of the supervised/unsupervised ML algorithms have strong data exploration ability,
which enables them to discover the hidden information embedded in the large amount of data
from the power grid.
Literature Review
• By the increasing integration of smart appliances into the system, the power industry is
rapidly moving towards digitalization and intelligence.
• An impressive number of ML-related methodologies have been proposed in the power
system community to facilitate this digital reform.
• ML-based system security assessment strategies outperform the traditional security
assessment method and ensure the optimal planning of the power system under complex
system configurations.
• Some of these research works aim to bring economic benefits to the individual customers,
while some of them improve social welfare at the community level.
• ML algorithms can be applied to extract the features of users’ behaviors and provide a
guideline for their energy consumption.
• Typical supervised learning algorithms include artificial neural network (ANN)-based
classifier, support vector machine (SVM), adaptive network-based fuzzy inference system
(ANFIS), autoregressive integrated moving average (ARIMA), etc.
Objectives
• By incorporating the physical characteristics of the power system into the
machine learning algorithm, the range of ML utilization can be greatly
expanded.
• The ML applications proposed innovations over the existing ML algorithms
according to different needs and extend their application in new areas.
• With unique problem modeling strategies, ML algorithms successfully reach the
areas that have never been explored using ML methods.
• Verifies the long-term economics of the battery swapping station-based
frequency regulation services through the policy gradient algorithm.
• Different testing criteria can be added up to the platform to acquire the battery
swapping station’s performance boundary under different scenarios.
• The problems discussed here are for the first time being solved using ML
techniques.
Methodology
 The three ML-based load modeling works on the individual house level, community
level, and system level respectively:
A factorial hidden Markov model for energy disaggregation for a household.
It applies a stacked autoencoder in residential customer baseline load estimation.
 It implements DRL in load modeling in the transient study.
 The two studies that explore the application of deep reinforcement learning (DRL) in
V2G- based ancillary services:
 work adopts the policy gradient (PG) method in economic risk assessment for the battery-
swapping stations.
 verifies the profitability of letting battery-swapping stations (BSSs) participate in fast
frequency regulation service (FFRS).
On top of the first work, deep Q-learning (DQN) techniques are applied to the BSS for real-
time control in order to extract the maximum economic benefit from the ancillary services.
References
• [1] A. Y. S. Lam, K. Leung and V. O. K. Li, “Capacity Estimation for Vehicle-to-Grid Frequency Regulation Services
with Smart Charging Mechanism,” IEEE Trans. Smart Grid, vol. 7, no. 1, pp. 156-166, Jan. 2016.
• [2] M. R. V. Moghadam, R. Zhang, et al., “Distributed frequency control via randomized response of electric vehicles
in power grid,” IEEE Transactions on Sustainable Energy, vol. 7, no. 1, pp. 312–324, Jan. 2016.
• [3] I. Duggal and B. Venkatesh, "Short-Term Scheduling of Thermal Generators and Battery Storage With Depth of
Discharge-Based Cost Model," IEEE Trans. Power Syst., vol. 30, no. 4, pp. 2110-2118, Jul. 2015.
• [4] “Foot Traffic Report for the fuel & convenience retailing industry,” GasBuddy and Cuebig, 2018 [Online].
Avaliable: https://www.iab.com/wp-content/uploads/2018/05/GasBuddy-Foot-Traffic-Report-Q1-2018-1.pdf.
[Accessed Oct. 2019].
• [5] CaliforniaTraffic count data. [Online]. Available at: http://pems.dot.ca.gov/. [Accessed Oct. 2019].
• [6] X. Wang and G. G. Karady, "Hybrid battery charging strategy for maximizing PV customers' economic benefits," in
2016 IEEE Power and Energy Society General Meeting (PESGM), Boston, MA, 2016, pp. 1-5.
• [7] I. Grondman, L. Busoniu, G. A. D. Lopes and R. Babuska, “A Survey of Actor-Critic Reinforcement Learning:
Standard and Natural Policy Gradients,” IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications
and Reviews), vol. 42, no. 6, pp. 1291-1307, Nov. 2012.
• [8] R. S. Sutton, D. McAllester, S. Singh, Y. Mansour, “Policy Gradient Methods for Reinforcement Learning with
Function Approximation,” in Advances in neural information processing systems (NIPS) 12, Denver, CO, pp.1057-
1063, Dec. 1999.
THANK YOU

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy