CNN-based cyber-attack detection research for photovoltaic new energy stations

With the rapid development of new power system technologies, the intelligentization, automation, and digital transformation of photovoltaic new energy stations are gradually being realized. However, with the expansion of network scale and the increase in data traffic, the power Internet of Things is facing serious cybersecurity risks. Security issues such as malicious attacks, data breaches, and equipment tampering pose a great threat to the stability and reliability of the power system, making effective detection of network traffic crucial for ensuring the security of the power Internet of Things. This paper explores the network traffic detection methods for photovoltaic new energy stations, researching and designing a convolutional neural network-based network attack detection for photovoltaic new energy stations, by monitoring and analyzing the network traffic generated in the power Internet of Things, using feature extraction and anomaly detection technologies to identify potential malicious traffic. Experimental results show that the proposed method can effectively identify various types of attacks and has high accuracy and robustness.