【作者】 彭谦；

【导师】 杨以涵； 姜彤；

【作者基本信息】 华北电力大学（北京） ， 电力系统及其自动化， 2009， 博士

【摘要】 潮流计算、状态估计计算是电力系统最基础的电气计算。潮流计算的基本方法有:高斯法、牛顿法和快速解耦法等。目前牛顿法、快速解耦法应用较多。但牛顿法计算效率低;快速解耦法计算收敛受网络参数影响较大。因此,研究一种计算速度快、收敛性好的潮流算法是有意义的。潮流算法是状态估计算法的基础,潮流算法的发展会促进状态估计算法的发展。本文以高斯法为基础提出了改进型快速Ybus潮流算法、三相快速Ybus潮流算法和配电网高斯状态估计算法。文中首先将高斯潮流算法与牛顿潮流算法和快速解耦潮流算法进行对比。对比结果表明,高斯潮流算法计算效率高、收敛性好。但是,传统高斯潮流算法对网络中PV节点的处理方式影响计算收敛,需要在此方面进行更深入的研究。提出了改进型快速Ybus潮流算法。建立网络高斯潮流方程后,消去其中的PQ节点,得到PV节点方程,并应用固定因子表的迭代方法求解。得到PV节点电压迭代解后,将结果代入高斯潮流方程求解PQ节点电压迭代解,实现计算可靠收敛。然后研究算法的收敛原理,给出算法适用范围。不对称潮流分析需要应用三相潮流算法,将改进型快速Ybus潮流算法扩展到三相,提出三相快速Ybus潮流算法。建立网络高斯潮流方程后,消去其中的PQ节点,得到PV节点方程并应用改进型快速Ybus潮流算法求解,将计算结果代入高斯潮流方程求解PQ节点电压迭代解,实现计算可靠收敛。提出了配电网高斯状态估计算法。将各节点注入功率方程、节点电压量测方程变形,根据最小二乘法,构造配电网高斯状态估计方程。算例结果表明该方法收敛性能差,在理论分析的基础上,提出一种改进算法,提高算法的收敛可靠性。然后将配电网高斯状态估计算法及其改进算法应用于三相不对称网络计算。以鞍山配电网为背景,对本文方法进行了应用。增加变压器零序阻抗解决配电变压器节点导纳矩阵奇异问题,变换节点导纳矩阵应用实测线电压计算,应用相序变换消除节点注入电流零序分量实现计算可靠收敛。最后,针对鞍山配电网量测数据存在部分节点数据冗余、部分节点数据缺失情况,结合配电网特点,提出了一种少量测点潮流计算方法。更多还原

【Abstract】 Load flow algorithm and state estimation algorithm are the basicest computation algorithm for electric power system.The primary algorithm of load flow is:Gauss load flow algorithm,Newton load flow algorithm and fast decoupled load flow algorithm,etc.Newton load flow algorithm and fast decoupled load flow algorithm are the most popular algorithm used now.But Newton load flow algorithm calculates ineffectively.And network parameter influences the convergence of fast decoupled load flow.So studying a high speed and reliable convergence load flow algorithm is significance.Load flow algorithm is the foundation of state estimation algorithm,and its progress can promote the performance of state estimation algorithm.Fast Ybus gauss load flow algorithm、three phase fast Ybus gauss load flow algorithm and Gauss state estimation algorithm for distribution network is studied based on Gauss algorithm in this paper.Firstly,Gauss load flow algorithm、Newton load flow algorithm and fast decoupled load flow algorithm are compared and the result shows that Gauss load flow algorithm has the advantage of efficient calculation and good convergence.But when there are PV nodes in the network,traditional Gauss load flow algorithm’s convergence is bad and study in this field is requirement.Modification fast Ybus gauss load flow algorithm is studied.Network equations are created firstly and PV nodes equations can be gotten by eliminating PQ nodes.State elements factor table is founded to get the result of PV nodes.Calculate PV equations and put the result into the Gauss equations to get the result of PQ nodes.Secondly, convergence principle and application scope of the algorithm is studied.Asymmetric load flow analysis needs three phase load flow algorithm.Fast Ybus gauss load flow algorithm is used for three phase load flow analysis and three phase fast Ybus gauss load flow algorithm is studied.Network equations are created firstly and PV nodes equations can be gotten by eliminating PQ nodes.Fast Ybus gauss load flow algorithm is used to get the result of PV nodes.And PQ nodes’ result can be resolved based on the PV nodes’ result.State estimation algorithm for distribution network is studied.Modificat equations of injection power and voltage,and use Least-squares procedure to get state estimation algorithm for distribution network.The result shows the algorithm’s convergence is bad.A modification algorithm is studied based on theory analysis to make calculation convergence reliablely.At last,the algorithm is used to compute asymmetric network.Fast Ybus gauss load flow algorithm and Gauss state estimation algorithm for distribution network are used in Anshan network.Singular problem of distribution transformer admittance matrix is solved by zero-sequence resistance attachment. Amend admittance matrix to make it suitable for line-voltage measuration. Zero-sequence component effects constriction.So its must be erased when computing.In the end of the paper,a load flow algorithm for anshan real network that some nodes have redundancy measuration and some do not have measuration is studied based on the feature of distribution network.更多还原