CPPS optimal defense strategy game method for uncertain attacks

• Main Authors: HUANG GANG, YANG QIANG, YAN BINGJING
• Format: Patent
• Language: Chinese; English
• Published: 23.01.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; PHYSICS; TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHICCOMMUNICATION

The invention discloses a CPPS optimal defense strategy game method for uncertain attacks, and belongs to the field of smart grid security. According to the method, expected benefits in sample data under different action spaces are calculated based on the game theory, instant reward indexes in a deep learning algorithm are evaluated according to the expected benefits, learning is carried out through a deep Q network (DQN) algorithm, and an optimal defense strategy is output. According to the method, not only can a more comprehensive defense strategy be provided for an uncertain attack method, but also the challenge of making a decision for an attack mode (namely, mixed multi-node attack) containing multiple modes and targets of a system is solved. In consideration of rapid iteration of attack and defense strategies, the method can be updated online, the model can be continuously trained under the condition that trusted samples exist, and the detection precision and the detection range of the model are improve