Solar Generation Estimation in Distribution Network Using Limited Measurements

Today, high renewable energy penetration in distribution networks (DNs), such as rooftop photovoltaic (PV) generations increases the complexity of DNs, where new control and monitoring methodology is needed. State Estimation (SE) methods are commonly used in DNs to improve the visibility of the low voltage (LV) feeders. However, having PV in these areas violates the assumptions of the white noise, the basis for SE methods, since it imposes a correlation between different measurements. In this work, the proposed weighted least square method is deployed to disaggregate the solar fraction from the measured total load using one reference PV in the area. The proposed methodology is validated on five different LV areas, using the Australian distribution network real dataset with a resolution of 30 minutes. According to the results, the proposed method can improve the PV estimation efficiency in each LV area compared with the ordinary method.