Online Distribution System Topology Monitoring With Limited Smart Meter Communication

Accurate topology monitoring is a prerequisite for advanced control and optimization of distribution power systems. However, online topology monitoring usually results in high communication facility investments and operating costs. This research proposes an online system topology monitoring method to reduce communication capacity and network traffic by adjusting the smart meter communication interval. Measurements from some critical buses are acquired with short communication intervals. In contrast, other non-critical buses can still maintain much longer communication intervals. We design a three-step angle-free optimization algorithm framework to track system topology and detect branch switching events online. First, we perform load forecasting to obtain active and reactive power pseudo measurements at non-critical buses. Power flow calculations are used to generate voltage pseudo measurements. Then, we perform the weighted least square method and an improved extended Kalman filter to calibrate both pseudo measurements and critical bus measurements. Finally, a topology tracking algorithm based on voltage estimation errors of critical buses is proposed to estimate and correct the current topology. In addition, we propose a critical bus selection method based on historical data statistics. The numerical results on IEEE 33-bus case show that our framework achieves a high online topology monitoring accuracy
$F1$
score of 89.91% with only 10 critical buses. Further tests on a practical 108-bus system in China also demonstrate the effectiveness of our framework.