False Data Injection Detection using H Infinity Filter in an Automatic Generation Controlled Two Area Power System

Abstract

Latest advancements in power systems lead towards the concept of a smart grid which largely depends on the communication infrastructures. These improvements offer better monitoring, control, and protection for effective power flow management in the electrical network. These dependencies of information hamper the system's operation in the existence of any cyber-attacks. The effect of these attacks must be mitigated by compensating for these attacks so that normality can be restored. This work tests an H infinity-based filtering technique to detect the attack signal like step and sine attacks. The test is performed in a two-area power system connected through a tie-line for power exchange. Automatic generation control (AGC) tries to maintain the frequency to a nominal value by providing the essential area control error (ACE) signals to the area controllers. The proposed filter works better than the Kalman fflter in the presence of non-Gaussian noises, also with high noise magnitudes. Further validation is camed out in a test environment consisting of a real-time digtal simulator (RTDS) and Raspberry Pi 4 board connected through an Ethernet]ink. The communication structure incorporated server-client-based TCP/IP socket protocol to transfer test environment information.