An Efficient State Estimation Algorithm Considering Zero Injection Constraints

Traditional state estimation methods employ large weights for zero injection pseudo-measurements, which may result in numerical instability. The normal equations with constraints will increase the order of the coefficient matrix, thus reducing the speed of the calculations. To improve computational performance, this paper presents a modified Newton method to deal with zero injection constraints. The decoupled form of the proposed technique-a modified fast decoupled state estimation method-is also given. The computational speed of the proposed modified Newton method is similar to that of conventional state estimations based on normal equations, since they use the same type of coefficient matrices. Furthermore, it offers the advantage that zero injection constraints can be strictly satisfied, and the numerical stability problem caused by large weights will no longer exist. Extensive numerical results from test systems and a real provincial system are included to verify the performance of the proposed procedure.