Research on new energy power plant network traffic anomaly detection method based on EMD

Abstract

Overview

As Photovoltaic (PV) systems connect into the grid and depend on digital technology, risks develop from obsolete components, insufficient security measures, and insecure access points.

Objectives

Improve the safety and dependability of the main data communication network by conducting research on Proactive Coordinated Fault-Tolerant Federation (PCFT) security structure, building an early warning simulation to handle power data interactions network abnormalities, studying algorithms and technologies for abnormal traffic control, and effectively managing abnormal network traffic like DDoS, network scanning, along with surge traffic; Improve the capability of the SLA hierarchical service, enhance the service quality of the core services executed through the backbone network, as well as strengthen the security capability of the access system for the new energy power plant’s communication network by combing and analyzing the traffic of the main network of the current information communication network.

Methodology

This research propose Network Quality Assessment (NQA) traffic management algorithms to prevent illegal access and data breaches, this involves strong security measures such as encryption, firewalls, and encrypted communication methods. To maximize the efficiency of solar energy systems and allow for prompt maintenance, our suggested framework provides a practical and dependable method for detecting anomalies in PV cells in real-time. The incorporation of these state-of-the-art convolutional methods into the CNN-GRU model enhances detection capabilities and opens up new avenues for exploration in the realm of anomaly detection based on deep learning.

Results

The grid deployment of large-scale PV power facilities relies heavily on dependable communication networks. The efficiency of PV power plants and their ability to meet application requirements are both impacted by the communication infrastructures that are responsible for real-time monitoring. Presenting simulations of the communication networks of the PV power system, this research sought to evaluate possible futures of PV power plants. We test our model on a massive PV cell dataset and show that it outperforms state-of-the-art approaches in terms of resilience, speed, and accuracy of identification.