深度森林与人工神经网络在光伏出力预测的比较

Comparison of deep forest and artificial neural network in prediction of PV output

针对在光伏功率预测中传统人工神经网络预测模型需预先确定结构、对训练样本数量要求高、调参复杂等缺点,应用一种多粒度级联森林的预测模型,并分析了温度、气象因素对光伏功率的影响,以温度、湿度、风速等因素作为输入对光伏输出功率进行预测,对输入数据进行归一化处理,将预测的结果与传统人工神经网络对比,经过多种验证指标对结果评估,表明模型的预测效果和模型结构均有良好的表现。

Aiming at the shortcomings of the traditional artificial neural network prediction model in the prediction of photovoltaic power,the forecasting model of gc Forest（ multi-Grain Cascade Forest） is applied,and the temperature and meteorological factors are analyzed. The influence of photovoltaic power on the photovoltaic output power is predicted by the factors of temperature,humidity,wind speed and so on. The input data are normalized and the predicted results are compared with the traditional artificial neural network. The results are evaluated by various verification indexes,which indicats that the model＇s predictive effect and model structure are well-behaved.