基于PSO-BP神经网络的短期光伏系统发电预测

对光伏发电影响因素进行了分析,建立了粒子群算法优化的前向神经网络光伏系统发电预测模型。该模型利用了粒子群算法来优化神经网络内部连接权值和阈值,兼具粒子群和BP神经模型的优点,具有较好的收敛速度,泛化性能与预测精度。将光伏电站发电历史数据与天气情况作为样本,运用所建立的模型进行了训练与预测。结果表明,经过粒子群优化的BP网络模型预测精度高于典型BP网络,验证了该方法的有效性。

Short-Term photovoltaic system power forecasting based on PSO-BP neural network

In order to improve the photovoltaic power forecasting accuracy,the influencing factors of photovoltaic power system’s output are analyzed and a particle swarm optimization algorithm is built for BP neural network prediction model of photovoltaic power forecasting.The particle swarm optimization algorithm is used to optimize the internal connection weights and thresholds of neural network in this model.Combining the advantages of the particle swarm optimization and BP neural model,the model achieves a better convergence speed,generalization performance and prediction accuracy.Taking photovoltaic power plant historical data and weather conditions as samples,the model completes training and prediction.The prediction results show that with the particle swarm optimization,BP neural network model prediction accuracy is higher than typical BP neural network,which verifies the effectiveness of the method.