The application of photovoltaic power prediction technology

Power load forecasting is an important part of the power system planning, and accurate load forecasting can provide necessary basis data for the dispatcher, which is extremely important in the planning and operation of power system. Neural network can approximate any nonlinear mapping with arbitrary precision, and its distributed information storage and processing structure have a certain fault tolerant. So neural network is suitable for complex system modeling, and it can be used as the main method to predict photovoltaic power operation state variables. This paper uses the neural network algorithm to establish photovoltaic power system´s load forecast model, and except the generate historical data, meteorological forecast information is added to the algorithm, then the model is trained and tested, the high precision prediction results can show the effectiveness of the algorithm.