rac99111's starred repositories
PPGN-Physics-Preserved-Graph-Networks
The increasing number of variable renewable energy (solar and wind power) causes power grids to have more abnormal conditions or faults. Faults may further trigger power blackouts or wildfires without timely monitoring and control strategy. Machine learning is a promising technology to accelerate the automation and intelligence of power grid monitoring systems. Unfortunately, the black-box machine learning methods are weak to the realistic challenges in power grids: low observation, insufficient labels, and stochastic environments. To overcome the vulnerability of black-box machine learning, we preserve the physics of power grids through graph networks to efficiently and accurately locate the faults even with limited observability and low label rates. We first calculate the graph embedding of power grid infrastructure by establishing a reduced graph network with the observed nodes, then efficiently locate the fault on the node level using the low-dimensional graph embedding. To augment the location accuracy at low label rates, we build another graph network representing the physical similarity of labeled and unlabeled data samples. Importantly, we provide the physical interpretations of the benefits of the graph design through a random walk equivalence. We conduct comprehensive numerical experiments in the IEEE 123-node. Our proposed method shows superior performance than three baseline classifiers for different fault types, label rates, and robustness to out-of-distribution (OOD) data. Additionally, we extend the proposed method to the IEEE 37-node benchmark system and validate the effectiveness of the proposed training strategy.
PowerFDNet
PowerFDNet annd FDIA datasets
attacks_on_complex_networks
In this work, we examine the resilience of two complex network types (Erdos Renyi, and Power-Law/Scale-free) to potential delivered attacks and random errors.
GAN-energy-modeling
Energy Modeling via standard GANs.
TimevariantGAN-AEMO-dataset
Synthetic Energy Data Generation using Time Variant Generative Adversarial Network
PINNpapers
Must-read Papers on Physics-Informed Neural Networks.
Physics-Informed-Neural-Networks
Investigating PINNs
PINNs-TF2.0
TensorFlow 2.0 implementation of Maziar Raissi's Physics Informed Neural Networks (PINNs).
easy_decison_tree
credit default analysis using decision trees and a couple of other advanced algorithms
Open-source-power-dataset
We present PSML, a first-of-its-kind open-access multi-scale time-series dataset, to aid in the development of data-driven machine learning (ML) based approaches towards reliable operation of future electric grids.
PowerSystems
Free (standard conform) library that is intended to model electrical power systems at different levels of detail both in transient and steady-state mode.
PowerSimulationsDynamics.jl
Julia package to run Dynamic Power System simulations. Part of the Scalable Integrated Infrastructure Planning Initiative at the National Renewable Energy Lab.
Predicting-Smart-Grid-Stability---Deep-Learning-Analytics-ANN-
The structure of the DSGC system is based on a 4-node star architecture which includes one power source and three consumption nodes. The inputs/features are the total power balance, the reaction time of the consumers to the price changes, and the energy price elasticity. We predict the stability of smart grids based on the Artificial Neural Network (ANN) models and compute accuracies based on different structures and hyperparameters.
irep_2022_closing_the_loop
Implementation of the work "Closing the Loop: A Framework for Trustworthy Machine Learning in Power Systems" submitted to IREP 2022
Machine_Learning-Power_System_Fault_Detection
Developing multiple ML Classifiers using SVM, PCA, Decision Tree, Random Forest, GMM & MLP Models for detecting Faults in Power Systems and comparing their accuracies & performances.
Fault-Detection-in-Power-Microgrid
This project presents the concept of fault detection and location in a Power Microgrid making use of the machine learning concepts like Artificial Neural Network. The electronic equipment used in microgrids is in essential need of more secure protection against short circuit faults. Due to the high current at the time of fault occurrence, the whole system might be de-energized which would have a severely negative impact on the entire system. A fault occurs when two or more conductors come in contact with each other or ground. Ground faults are considered as one of the main problems in power systems and account for more than 80% of all faults. An effective method to detect, isolate, and protect the power microgrid system against the effects of short circuit faults is extremely important. In this project we worked on a highly effective new method to protect the microgrid system using an Artificial Neural Network (ANN) that will detect and find the location of the fault before it affects other parts of the system. It would, therefore, be more dependable for microgrid protection. This protection network is distributed all along the power microgrid system protecting the entire microgrid network and is connected to the other protective devices in the system. This project focuses on detecting faults and identifying the location of the faults on electric power transmission lines in the power microgrid network.