New energy photovoltaic power generation power prediction method

The invention discloses a new energy photovoltaic power generation power prediction method, and relates to the technical field of power system photovoltaic power generation power prediction. According to the photovoltaic power prediction method based on the variational mode decomposition and the TPA-TCN, a training sample is generated through an information processing module according to a historical data sample, the training sample is subjected to continuous iterative training through a TPA-TCN prediction model until the precision reaches an expected value, and the model is reserved. According to the new energy photovoltaic power generation power prediction method, instability and volatility of a time sequence are reduced through VMD decomposition, and similar weathers are classified through a similar day method, so that a neural network can learn potential laws more easily. The method is based on a time pattern attention mechanism, can fully extract the implicit relationship among the multi-factor time sequ