Enhanced dynamic state estimation of regional new energy power system under different abnormal scenarios

Abstract

The high proportion of renewable energy sources in the power grid increases the failure probability of the system, which becomes a new challenge for the safe and stable operation of the regional power grid. To ensure stable control of the power network with substantial renewable energy integration, this article proposes a new method that combines the long short-term memory (LSTM) neural networks and adaptive cubature Kalman filter (ACKF) to improve the prediction accuracy of mutation data inherited from the renewable generation. Four abnormal scenarios, including low voltage ride-through (LVRT), high voltage ride-through (HVRT), continuous fault ride-through and bad data injection of the regional power grid are investigated through extensive case studies. The proposed method is implemented on the IEEE 30-node system for performance verification. The simulation results demonstrate that the proposed method has considerably higher robustness than the traditional Kalman filter algorithm and can effectively improve the overall state estimation accuracy of the renewable energy power system under different scenarios.