Photovoltaic Pseudo-Measurement Generation Based on Power Forecasting in Active Distribution Systems

With the large-scale integration of photovoltaic (PV) generation, state estimation in Active Distribution Systems (ADS) faces challenges due to insufficient real-time measurements. This paper proposes a comprehensive solution that combines PV power forecasting with Gaussian pseudo-measurement generation. First, an advanced spatio-temporal graph convolutional network (ST-GCN) algorithm is employed to achieve high-precision short-term forecasting of PV power, ensuring accurate predictions across various time periods and weather conditions. Then, based on the forecasted results and the topology of the distribution system, Gaussian-distributed pseudo-measurement data is generated to supplement areas with insufficient measurements, enhancing the completeness and reliability of state estimation. Experimental results demonstrate that this method achieves robust estimation performance across different scenarios, significantly improving the situational awareness of ADS and providing critical support for the intelligent dispatch and optimized operation of the distribution network.