基于GWO-BP模型的短期风力发电预测

Short Term Wind Power Generation Prediction Based on GWO-BP Model

近年来,风力发电装机规模在全球范围内迅速增长。风力发电功率取决于风速及其所携带的能量,而风速是一个不断变化的随机变量。为了使风电场高效有序运行,减少因波动性、随机性对电网的冲击,降低弃风限电率,准确的短期风功率预测是必不可少的。基于灰狼优化算法(Grey Wolf Optimizer,GWO)-反向传播(Back Propagation,BP)神经网络模型预测风电场风功率,结果表明,它对短期风力发电的预测准确可靠。应用的GWO-BP模型所得结果与标准BP模型和遗传算法优化的BP模型进行比较,证明此模型预测精度更高。

In recent years,the installed scale of wind power generation has grown rapidly worldwide.The power of wind power generation depends on wind speed and the energy it carries,and wind speed is a constantly changing random variable.In order to make the wind farm operate efficiently and orderly,reduce the impact on the power grid due to volatility and randomness,and reduce the rate of wind curtailment,accurate short-term wind power prediction is essential.In this paper,wind speed and power generation are predicted based on the Grey Wolf Optimizer(GWO)-Back Propagation(BP)neural network model.The results of this model show that it is accurate and reliable for predicting wind power of wind farm.The results obtained from the applied GWO-BP model were compared with standard BP model and BP model optimized by genetic algorithms,proving that this model has higher prediction accuracy.