A study on the decomposition and reconstruction of flow field of large-span flexible photovoltaic array based on dynamical mode decomposition

The flow field of a large-span flexible photovoltaic (PV) array exhibits high-dimensional dynamic characteristics with significant inter-row interference, and it is difficult to directly capture the full dynamic distribution characteristics of the flow under the inter-row interference through the physical wind tunnel test, but the data-driven reduced-order algorithm provides a new idea to realize the accurate reconstruction and prediction of its complex flow field. A 40-meter-span flexible photovoltaic array demonstration project by the State Power is analyzed using large eddy simulation to study the distribution of average and fluctuating wind pressure on the array, and the driving mechanism and vorticity evolution law of flow field are revealed. The flow field of a photovoltaic array is decomposed using the dynamical mode decomposition (DMD), and a reduced-order model of flow field evolution based on the I-mode criterion is established for reconstruction and error analysis within the snapshot number. The results show that the interference effect of the photovoltaic array is most significant in the second row on the windward side; the main modes capture key flow features and characterize the internal mechanism of flow field evolution, and the results of the flow field reconstruction by the reduced-order model of flow field evolution within the snapshot number can describe and fit the basic characteristics of the original flow field. The conclusions are significant for identifying the basic physical mechanisms and dynamic evolution of the flow field, providing a reference for optimizing and predicting the wind load design of large-span flexible photovoltaic arrays.