Robust linear state estimation using multi-level power system models with different partitions

摘要

The work in this paper is motivated by the computational challenges in monitoring very large scale interconnected power transmission systems. The main idea is to exploit the robustness property of the recently developed PMU-based linear estimator while eliminating hierarchical computational complexity. It is recognized that state estimation can be implemented for observable islands in an efficient and robust manner as long as the measurement redundancy remains above a certain threshold. When using distributed computational schemes, local redundancy is inadvertently reduced at the area boundaries and additional effort is needed to compensate for this deficiency. In this paper a new approach is proposed where multiple copies of a multi-area estimator run in parallel, making sure that every bus appears as an “internal area” bus in at least one of these copies. The concept and its envisioned implementation will be explained using the IEEE 30-bus test system as an example.