摘要
风电场功率预测对电力系统稳定运行起着决定性作用。首先对传统BP神经网络进行改进,以某一风电场获取的2月1日-10日的天气预报(NWP)数据和功率数据作为改进后BP神经网络的训练数据,对神经网络进行训练;其次以2月11号3小时的数值天气预报数据作为改进后BP神经网络的输入数据,对未来3小时的输出功率进行预测。预测过程和结果显示,改进后的BP神经网络在满足低预测误差的同时,能够提高BP神经网络的稳定性和收敛速度。
Wind power prediction of wind farm is plays a decisive role in stable electric power system operation. Firstly to improve the traditional BP neural network, to use the numerical weather prediction data and power data, from feb. 01 to 10, as the training data of the improved BP neural network; secondly to use 3 hours data about numerical weather prediction of feb. 11 as inputting data, to predict The next 3 hours of output power. Process and result of prediction displays, the improved BP neural network ensures lower prediction error, at the same time, it can improves the stability of BP neural networks and the rate of convergence.
出处
《软件导刊》
2013年第4期31-33,共3页
Software Guide
关键词
风电场功率预测
天气预报(NWP)
BP神经网络
预测误差
Wind Power Prediction of Wind Farm
Numerical Weather Prediction
BP Neural Network
Prediction Error
