An Active Power Predictive Control Method Considering the Scheduling of Photovoltaic Cluster Field Stations

In order to reduce the output power fluctuations and randomness of photovoltaic clusters, ensure the power demand of users and the safe operation of the power system, a method of active power prediction and control considering the output power sorting of internal stations in the photovoltaic cluster is proposed. In order to improve the accuracy of the scheduling curve prediction of the photovoltaic cluster, the SSA initialization rule is improved using the sine chaotic mapping, and the SSA position update rule is improved. Then, the improved SSA is used to optimize the hyperparameters of the LSTM network. In order to improve the accuracy of scheduling tracking, the actual output of the station, the typical meteorological day curve, and the scheduling prediction curve are used to construct output conformity, cosine similarity, scheduling follow-up degree and other effectiveness indicators for station control, and the internal stations of the cluster are sorted and controlled, so that they can adaptively select the output mode and ultimately optimize the cluster output. Simulation results have verified that the proposed method can reasonably allocate the output of each station, making the cluster output smoother.