ABSTRACT Many studies have investigated the performance simulation of bifacial PV power stations, but existing models generally suffer from performance overestimation. Furthermore, some models often fail to converge when calculating rear‐side irradiance under sparse layout conditions. This work presents an improved discretized analytical integration strategy for optical calculations. Unlike available research relying on short‐term data, this work uses full‐year, minute‐level data from the Taizhou demonstration base to validate the model. The results show that the model possesses high calculation accuracy and robustness, with a power R 2 of 0.9929, an MAE of less than 5%, and an annual power generation deviation of less than 2%. An intelligent optimization framework aimed at maximizing net present value (NPV) is established. Coupled analysis of the installation parameters facilitates a global optimum for the power station design. Test cases indicate that, compared with traditional empirical schemes, this optimization strategy increases NPV by 4.1% and 3.1% in different scenarios, providing practical guidelines for the refined design and cost‐efficiency improvement of bifacial PV power stations.
Wang et al. (Sun,) studied this question.