Deep Reinforcement Learning-Enabled Adaptive Forecasting-Aided State Estimation in Distribution Systems with Multi-Source Multi-Rate Data

Distribution system state estimation (DSSE) is paramount for effective state monitoring and control. However, stochastic outputs of renewables and asynchronous streaming of multi-rate measurements in practical systems largely degrade the estimation performance. This paper proposes a deep reinforcement learning (DRL)-enabled adaptive DSSE algorithm in unbalanced distribution systems, which tackles hybrid measurements with different time scales efficiently. We construct a three-step forecasting-aided state estimation framework, including DRL-based parameter identification, prediction, and state estimation, with multi-rate measurements incorporating limited synchrophasor data. Furthermore, a DRL-based adaptive parameter identification mechanism is embedded in the prediction step. As a novel attempt at utilizing DRL to enable DSSE adaptive to varying operating conditions, this method improves the prediction performance and further facilitates accurate state estimation. Case studies in two unbalanced feeders indicate that our method captures state variation with multi-source multi-rate data efficiently, outperforming the traditional methods.