The invention discloses a distributed photovoltaic data virtual collection method, and the method comprises the steps: analyzing historical operation and maintenance data of a distributed
photovoltaic power station which is provided with a data collection device in a region through employing a
grey correlation theory, and obtaining a characteristic curve of
irradiance and distributed photovoltaicoutput; and then making correlation degree calculation on the regional daily
irradiance real-
time information and historical
irradiance data, selecting similar days according to the correlation degree, establishing a BP neural
network data virtual acquisition model to achieve virtual acquisition of all distributed photovoltaic data in a regional range. According to the invention, irradiance information closely related to photovoltaic output power is used to carry out virtual acquisition of photovoltaic output data; compared with an existing photovoltaic output prediction method, the invention has the advantages that the required data size is small, the
data input dimension is reduced, the
algorithm is greatly simplified, the output data of the
photovoltaic power station in a whole day can be collected systematically, the model is simplified, a
network model does not need to be trained in a time-phased mode, and the purpose of minimizing the cost is achieved while the data precision is guaranteed.