ABSTRACT Accurate identification of distributed photovoltaic (PV) generation at the distribution transformer area level is essential for improving grid operation and renewable energy management. However, in practical scenarios, individual behind‐the‐metre PV systems are typically not equipped with dedicated monitoring devices, and only aggregated transformer‐side total power data are available. To address this challenge, this paper develops a distributed PV power identification method based on a multiscale convolutional neural network (CNN). Solar irradiance estimation is incorporated to maintain physical consistency with radiation variation, whereas the multiscale CNN structure captures temporal variation characteristics of PV generation. By establishing a nonlinear mapping relationship between reference PV station generation and aggregated transformer‐side total power data, the proposed method enables separation of PV generation from load demand without requiring individual PV measurements. Case studies demonstrate that the proposed method achieves superior identification accuracy compared with benchmark models. Ablation and sensitivity analysis further validate the robustness and practical applicability of the framework under limited sensing conditions.
Liu et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: