论文部分内容阅读
在保留法方程法和正交变换法优点的基础上，提出了一种直角坐标形式的混合法状态估计方法。通过对节点编号和量测处理顺序的编号进行优化，减少吉文斯变换次数，提高状态估计计算速度；同时，在对不良数据检测与辨识时采用递推法，避免重新做状态估计和存储残差灵敏度矩阵，以及对因子更新利用稀疏矢量技术的部分重新因子化法，从而避免了不必要的计算；另外，利用判据分析避开关键量测量，保证了系统的可辨识性。对ＩＥＥＥ６节点系统进行测试，证明所提出的算法是有效的。
Based on the advantages of the method of law equation and orthogonal transform, a method of state estimation in the form of rectangular coordinates is proposed. By optimizing the number of nodes and the number of measurement processing sequence, the number of times of Guernsey transformation is reduced and the speed of state estimation calculation is improved. At the same time, recursive method is adopted to detect and identify bad data to avoid re-state estimation and storage residual Sensitivity matrix, and partial re-factorization using sparse vector techniques for factor updating, thereby avoiding unnecessary calculations. In addition, the use of criterion analysis avoids the key quantity measurement and ensures the system’s identifiability. The IEEE 6-bus system is tested to prove that the proposed algorithm is effective.