摘要
风电出力特性与负荷需求的形状(负荷需求具有双峰性的典型形状)不一样,难以直观得到风电出力特性曲线的大概轮廓。因此,提出改进的自适应模糊聚类算法,采用表征多时空尺度的风电波动性、同时率评价指标对风电出力数据归一化处理,分段聚合降低维度,计算自适应函数α(c),确定各时段最佳聚类数c,对风电出力类型分类;引入变异离散度系数βi,剔除风电出力畸变数据;提出分区加权中位值法,辨识并提取风电出力特性概率区间带。针对新疆区域电网2015年1月份的风电出力数据,仿真计算分析风电出力特性曲线及概率区间带,验证了所提方法的有效性和稳定性。
The shape of wind power output characteristic is different to load demand(load demand has a typical shape of bimodal),so it is difficult to obtain the approximate outline of wind power output characteristic curve intuitively.In order to solve this problem,the wind power volatility index and simultaneous evaluation index characterizing multiple time-space scales are firstly used to normalize wind power output data and then the dimension of wind power output information is reduced by piecewise aggregation,and secondly,an improved adaptive fuzzy clustering algorithm is proposed to calculate the adaptive function α(c) for determining the optimal clustering number c,and then the type of wind power output is classified.The distortion data of wind power output is eliminated by introducing variation dispersion coefficient βi.The probability interval band of wind power output characteristic is identified and extracted using proposed zoning weighted median method.According to the wind power output data of Xinjiang regional power grid in January of 2015,the characteristic curve and probability interval band of wind power output characteristic are calculated to verify the effectiveness and stability of proposed method.
作者
蔺红
徐邦恩
LIN Hong;XU Bang'en(College of Electrical Engineering,Xinjiang University,Urumqi 830047,Xinjiang,China)
出处
《水力发电》
北大核心
2018年第12期95-99,共5页
Water Power
基金
国家自然科学基金项目(51667019)
新疆维吾尔自治区自然科学基金项目(2017D01C029)
关键词
分段聚合近似
模糊聚类
分区加权中位值法
风电出力特性
piecewise aggregate approximation
fuzzy clustering
zoning weighted median method
wind power output characteristic
