Online Estimation of Power Shortage in Power Systems Driven by Local Frequency Measurement Data

With the growing penetration of renewable generation, the inertia and frequency support ability of the, renewable-dominated power svstem are continuously reduced, which leads to frequency collapsewhen the system is disturbed. Therefore, it is of great significance to promptly and accurately evaluate the power shortage after a major disturbance to fill the power shortage quickly. This paper proposes an online estimation approach of power shortage in power system based on deep convolution and long short term memory composite neural network driven by local frequency measurement data. First, since using the synchronous measurements to obtain the frequency of center of inertia (COI) cannot adapt to the rapidity of online estimation, this paper employs the local frequency measurements to estimate the COI frequency, avoiding the delay effect caused by the complex communication. Then, it designs a deep composite neural network to mine the correlation information between massive frequency data and power shortage. Finally, it tests the simulations on a 39-bus system to verify the effectiveness of the proposed approach. The results demonstrate that the proposed approach is effective and fast.