A Deep Learning Method for Power Fluctuation Identification from Frequency Fluctuations

Abstract

Since power fluctuations caused by renewable energy sources and flexible loads may lead to instability problem even blackout of power systems, power balance plays a significant role in improving the power quality and stability of power systems. However, it is very difficult to obtain the values of power mismatch in both transmission and distribution systems, especially when there is a high penetration level of renewable energies. Thus, the long short-term memory recurrent neural network is proposed in this paper to estimate power fluctuations from frequency fluctuations. With the help of the identified power fluctuations, the system frequency can be better maintained by automatic generation control by the new accurate reference. The proposed algorithm was tested on an established model of Singapore power system under comparison with various classical methods including the state-of-the-arts in the field of artificial neural networks. The simulation results clearly demonstrate the necessity for power fluctuation identification, and the effectiveness of the proposed method.