基于强对流天气判别的风功率爬坡预报方法研究  

Research on Wind Power Ramping Prediction Based on the Strong Convective Weather Discriminant Method

作　　者：熊一[1] 査晓明[1] 秦亮[1] 陈正洪 欧阳庭辉 夏添[1] 

机构地区：[1]武汉大学电气工程学院,湖北省武汉市430072 [2]湖北省公共气象服务中心,湖北省武汉市430074

出　　处：《中国电机工程学报》2016年第10期2690-2698,共9页Proceedings of the CSEE

基　　金：国家重点基础研究发展计划项目(973计划)(2012CB215101);国家自然科学基金项目(51177113)~~

摘　　要：该文讨论一种基于强对流天气判别的风功率爬坡预报模型。首先选出指定区域中可以表征历史强对流天气的动力学和热力学特征的预报因子,再采用费希尔判别法将历史大风型强对流天气进行归纳分析,以得到预报因子的加权系数,进而确定判别强对流天气的预报方程。根据数值天气预报的数据分析得出强对流天气的预报结果,引入模板参数法将强对流天气参数库进行爬坡气象类型识别,并修正了风速预报数据,从而得到更准确的预报结果。结合风电场实际运行状况、电力系统的调度方式,以及区域电网的热备用启动速度和承受能力确定风功率爬坡定义。由此引入启发式分割算法对强对流天气预报结果进行突变检测,可得到风功率爬坡场景的定性预报结果,最终形成基于强对流天气判别的风功率爬坡预测方法。A forecasting model of wind power ramping based on convective weather discrimination was discussed.Firstly the kinetic and thermodynamic characteristics of history strong convective weather can be characterized by the predictors in the elected region.Then Fisher discriminant method was used in order to obtain the weighted predictor coefficients,and to determine the discriminant forecasting formulation.The parameter template method can be introduced to identify the type of ramping weather in strong convective weather library,according to the data analysis of numerical weather prediction（NWP）.Meanwhile,the original forecasting wind speed data was modified,which obtains more accurate wind speed.Combined with the actual operating condition of wind farms,power systems scheduling and regional distributed arrangement lead to determine the wind power ramping definition.Thus the introduction of catastrophe detection Bernaola Galvan algorithm（BGA） on severe convective weather forecasting results in getting determined wind power prediction ramping scenario.Eventually the ramping wind power forecasting method based on the discrimination of convective weather is formed.

关 键 词：风功率预测 强对流天气 爬坡定义 数值天气预报 费希尔判别法 参数模板法 启发式分割算法 

分 类 号：TM73[电气工程—电力系统及自动化]