Dynamic state estimation for distribution networks with photovoltaic power forecasting based on long short-term memory

With the large-scale integration of renewable energy, the accuracy and dependability of the dynamic state estimation for modern distribution networks might be compromised by the uncertainty and fluctuation of renewable sources. To address these challenges, the paper proposed a new dynamic state estimation method based on Long Short-Term Memory instead of traditional Kalman Filter method. The proposed method exhibits promising accuracy with less time-cost by mitigating the negative influences of uncertainty and fluctuation brought by PV, all the while requiring limited measurements. In the paper, an improved photovoltaic power forecasting method was firstly introduced. The distribution network model considering PV forecasting effect was established by through the application of LSTM. Then, a dynamic state estimation method was developed based on established distribution network model. To prove the effectiveness of the method, the real time simulations based on RTDS platform and comparisons with traditional method were conducted.