高维模型表达在概率潮流和静态安全域中的应用研究

发布时间：2018-09-17 12:29

【摘要】：近年来，，随着电力系统规模的不断扩大和电网中越来越多的不确定性因素的产生，概率潮流和静态安全域作为电力系统安全评估和调控的两个重要工具，引起了人们越来越多的关注。在概率潮流和静态安全域应用过程中寻求简单高效的方法，已成为工程研究的热点。高维模型表达（High Dimensional Model Representation，HDMR）在描述系统输出量关于多输入量之间关系方面具有独特的性能，而电网潮流状态量与网络多个节点源流注入量间正好符合HDMR的相关属性。本文在分析了HDMR模型的原理和建模过程的基础上，利用源流路径链信息找到影响关键支路潮流的负荷和发电机节点，并使用已经开发的GUI-HDMR程序，进行支路潮流的HDMR建模。通过修改输入量的阶数和正交多项式的次数对模型进行修正，用简单的多项式组合替代复杂的潮流方程。在支路开断情况下，为避免大规模的模型重建，本文结合灵敏度-补偿法对支路潮流的HDMR模型进行了修正，节省了大量时间，并用IEEE39节点算例验证了HDMR模型能够很好地拟合潮流方程，并可用于潮流状态的快速计算。本文将HDMR应用于电网潮流概率评估与调控问题：通过典型代表性的样本构建关键支路上传输的功率与电源和负荷间的HDMR关系，替换传统潮流计算方式承担潮流概率评估过程中大规模的潮流计算任务，以极大地提高关键支路潮流累积概率分布生成及其相关特征求取的效率；对关键支路潮流阻塞问题，设计了一种利用HDMR提供的全局灵敏度信息并兼顾节能减排性能指标的概率调控策略。算例表明，HDMR的应用可显著提高电网潮流概率评估的计算效率和关键支路潮流阻塞概率调控的性能。针对静态安全域计算过程中，交流潮流模型计算量大和直流潮流方程精度不高的问题，本文用HDMR模型替代静态安全域中的交流或直流潮流方程，不仅保证了计算速度，而且提高了计算精度。在已知某些发电机有功功率变化范围的情况下，提出了一种计算关注节点子集功率变化安全域的方法。算例表明，HDMR的应用可显著提高静态安全域构造的效率。
[Abstract]:In recent years, with the continuous expansion of power system scale and the emergence of more and more uncertain factors in power network, probabilistic power flow and static security region are two important tools for power system security assessment and regulation. It has attracted more and more attention. Seeking simple and efficient methods in the application of probabilistic power flow and static security domain has become a hot topic in engineering research. High-dimensional model representation (High Dimensional Model Representation,HDMR) has a unique performance in describing the relationship between system output and multiple inputs, while the power flow state and the injection of multiple nodes in the network are exactly in line with the correlation attributes of HDMR. On the basis of analyzing the principle and modeling process of HDMR model, this paper uses the information of source and stream path chain to find the load and generator node that affect the key branch power flow, and uses the developed GUI-HDMR program to model the HDMR model of branch power flow. The model is modified by modifying the order of input quantity and the degree of orthogonal polynomial, and the complex power flow equation is replaced by simple polynomial combination. In order to avoid large-scale model reconstruction, the HDMR model of branch power flow is modified with sensitivity compensation method in this paper, which saves a lot of time. The IEEE39 node example shows that the HDMR model can fit the power flow equation well and can be used to calculate the power flow state quickly. In this paper, HDMR is applied to power flow probabilistic evaluation and control problem: the HDMR relation between power transmission and power supply and load is constructed through typical representative samples. In order to improve the efficiency of generating the cumulative probability distribution of critical branch power flow and its related characteristics, the traditional power flow calculation method is used to take on the task of large-scale power flow calculation in the process of power flow probabilistic evaluation. A probabilistic control strategy based on the global sensitivity information provided by HDMR and taking into account the performance index of energy saving and emission reduction is designed. An example shows that the application of HDMR can significantly improve the efficiency of power flow probability evaluation and the performance of power flow blocking probability control in key branches. In order to solve the problem that the calculation of AC power flow model is heavy and the accuracy of DC power flow equation is not high during the calculation of static safe domain, this paper uses HDMR model to replace AC or DC power flow equation in static safe domain, which not only guarantees the calculation speed. Moreover, the accuracy of calculation is improved. Under the condition that the range of active power of some generators is known, a new method to calculate the security region of the power variation of the node subset is proposed. An example shows that the application of HDMR can significantly improve the efficiency of static security domain construction.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM73

【相似文献】