提高风电场风速预测能力的软测量方法

Soft Measurement Method for Wind Speed Prediction in Wind Farm

为了降低风电场风速预测的误差,基于软测量技术,从改进算法、提高预测的空间连续性及时间连续性入手,并以支持向量机为基本方法进行分析。提出了改进算法(小波分析—支持向量机),以气象数据为参考,增强了空间连续性;以风机级联并且缩短采样周期为手段,增强了时间连续性。通过现场数据进行仿真计算,验证了上述方法可以提高风电场风速的预测能力。

For decreasing the error of wind speed prediction in wind farm, improvements from the aspects of arithmetic, space and time has been proposed based on the soft sensor. Support Vector Machine as the basic prediction method is used to analyze. MRA-SVM testifies the improvement of algorithm can increase the performance targets. The modified model about meteorological information achieves the space continuity and the cascading of turbines profits the time continuity. Based on the actual measured data, the calculated result shows that it is feasible to improve the wind ＇speed prediction from the aspects of algorithm, space and time.