论文部分内容阅读
风力发电发电量依赖于气象因素和环境条件的制约,存在很大的随机性和不确定性,对电网的稳定运行造成一定的影响,针对这一问题,本文提出了应用单隐含层前馈神经网络SLFN(Single-hidden Layer Feedforward Neural Network)中的极限学习机(Extreme Learning Machine,ELM)进行预测分析。首先建立风力发电的简化预测模型;然后进行网络训练;最后分别对日平均风速和日发电量进行预测分析。结果表明应用ELM减小预测误差,对实际具有一定的指导和参考价值。
Wind power generation depends on the meteorological factors and environmental conditions constraints, there is a lot of randomness and uncertainty, the steady operation of the power grid to some extent, in response to this problem, the paper proposed a single hidden layer feedforward Predictive analysis is carried out by Extreme Learning Machine (ELM) in Single-hidden Layer Feedforward Neural Network (SLFN). First, a simplified forecasting model of wind power generation is established; then network training is conducted; finally, the daily average wind speed and daily generation are predicted and analyzed respectively. The results show that the use of ELM to reduce the prediction error has practical guidance and reference value.