Islanding detection in distributed energy resources based on gradient boosting algorithm

Abstract

This paper proposes a novel passive‐based intelligent method for anti‐islanding. Passive methods generally suffer from the improper tuning of threshold values for measured variables and false detection when the active or reactive power mismatch is small. Conversely, intelligence‐based methods highly depend on the choice of an appropriate model, the universality of data and selected features. In addition, the risk of overfitting and underfitting for a single model due to an improper selected feature or insufficient data is always possible. This condition makes a single classification model unreliable for practical purposes. The method used in this work is a sequential ensemble of intelligence‐based models called gradient boosting. The sequential form of an ensemble is substantially designed to solve the underfitting problem. In addition, by selecting the best combination of features and ensembles, algorithm overfitting can be prevented and the attribute data space can be constructed with maximum discrimination. Furthermore, this condition yields a model with a zero non‐detection zone. As a result, a multi‐objective optimisation method called the non‐dominated sorting genetic algorithm‐III is used to tune the model. The reliability and economy of this method for distribution networks with various types of inverter‐based and synchronous distributed energy resources are demonstrated.

A computer image feature extraction algorithm based on distributed K-means algorithm

“…Depending on the number of candidates that are updated regularly, the competition strategy can be divided into two groups: Winner Taker All (WTA) and Sales Competition (SC). The concept of WTA refers to the development and maintenance of similar facilities in only one model; The SC strategy allows multiple teams to change at different levels [5][6]. Employing the SC strategy for online integration provides advantages in terms of time shifting and integration.…”

Vigil AI Assemble Leveraging Ensemble Technique for Enhanced Detection of Multi-Converter Synchronous Active Islanding Mode in Hybrid Power Distribution Systems

“…The best operational model that has been proposed will be useful for community-based, multi-party microgrids that can function in either grid-connected or island modes [37]. The goal of the project was to optimize the operation of multi-party microgrids in order to increase their efficiency, resilience, and overall performance in different operating circumstances.…”

Design of fault-tolerant control system for distributed energy resources based power network using Phasor Measurement Units

“…29 Fault control using hardware redundancies is presented by Amin and Mahmood-ul-Hasan 30 An innovative passive-based intelligent method for anti-islanding is proposed that is a subsequent collection of intelligence-based models called gradient boosting. 31 To control fault using observer-based models and create fault isolation units in case of fault occurrence, an active FTC is proposed based on fuel actuators in the fuel supply line. 32 A commonly used one-class classifier, for fault detection in the distribution systems, is proposed by Lin et al, 33 which is based on small data available on fault conditions, and on the Support Vector Data Description (SVDD) method, that only requires the normal data for its training process.…”