Comprehensive Evaluation of Photovoltaic Power Prediction Error Based on TOPSIS Considering the Influence of Cloud Motion

The comprehensive evaluation of photovoltaic power prediction errors is the basis of improving the prediction model, enhancing the prediction accuracy and providing guidance for the rational application of prediction results, which is of great significance for the safe and economic operation of the new power system containing large-scale renewable energy. Currently, the photovoltaic power prediction mainly adopts mean value evaluation indexes of the 15-minute level, which cannot reflect the influence of the sudden fluctuations on low-inertia power systems and the adaptability of prediction systems at the minute level prediction situation. Therefore, considering the influence of cloud motion on photovoltaic power generation and the corresponding characteristics, this paper proposes a comprehensive evaluation index system including random indexes and volatility indexes. Then, a comprehensive evaluation method of photovoltaic power prediction errors based on technique for order preference by similarity to ideal solution (TOPSIS) is proposed. Finally, the prediction performance of different prediction models is evaluated comprehensively, which verifies the feasibility of the proposed evaluation method. This evaluation method could provide some theoretical support for the selection of prediction model and performance evaluation of prediction system in a photovoltaic station.