论文部分内容阅读
通用生成函数法(universal generating function,UGF)具有计算简单,运算快等优点,但其传统的应用中忽略了时序性,评估结果粗糙。为此,提出基于改进UGF法的随机生产模拟算法。首先,采用主成分分析法(principal componentanalysis,PCA)和分层聚类法相结合的方法获得典型风电出力时序样本,然后采用均摊法建立风电出力UGF模型。在建立UGF模型时,通过选择合适的公因子,将状态值转化为公因子的整数倍,以有效抑制状态数指数增长,减缓了状态数爆炸问题。其次,对聚类后的时序风电出力模式进行UGF建模,得到时序UGF模型,该模型既可以体现风电出力的波动性,也可以体现风电出力的不确定性,克服了传统UGF法因缺乏时序性所致的风电特征代表性差的问题。最后,采用修正的IEEE-RTS79算例验证了所提算法的有效性。
The universal generating function (UGF) has the advantages of simple calculation and fast calculation, but its traditional application ignores the time series and the evaluation results are rough. Therefore, a stochastic production simulation algorithm based on improved UGF method is proposed. First, a typical sample of wind power output was obtained by a combination of principal componentanalysis (PCA) and hierarchical clustering. Then an UGF model of wind power output was established by using an equal distribution method. When establishing the UGF model, by selecting the appropriate common factor, the state value is converted into an integer multiple of the common factor to effectively suppress the exponential growth of the state number, which alleviates the state explosion problem. Secondly, UGF modeling is performed on the clustered timing wind power output model to obtain the time series UGF model. This model not only reflects the fluctuation of wind power output but also reflects the uncertainty of wind power output, overcomes the lack of timing of the traditional UGF law Due to the poor representativeness of the wind power characteristics caused by the problem. Finally, a modified IEEE-RTS79 example is used to verify the effectiveness of the proposed algorithm.