Controlled Islanding Strategy Under Multi-oscillation Scenario

With the formation and development of West-East power transmission project and large-scale interconnected power grid, the traditional security and stability control of the power grid is facing unprecedented challenges. Some large disturbances may cause two-freguency or even multi-frequency oscillation. How to determine the proper splitting surfaces in a short time is a key problem for power splitting control under the situation of multi-frequency oscillation.The criteria of out-of-step separation equipment in the current domestic engineering fields are acquired based on out-of-step equivalent two-machine power system. It is worth exploring their applicability under multi-frequency oscillation. In this paper, expressions of electrical quantities are inferred based on a three-machine equivalent system. And it’s found that the out-of-step center would migrate between different transmission lines and different positions even on one transmission line. The more distinct the different modulations, the more transmission lines the out-of-step center may migrate between. Besides, the three out-of-step splitting criteria based on Ucosφ characteristic, the measured impedance locus and the apparent impedance angle can hardly adapt to the multi-frequency oscillation scenario. What’s more, it is hard for out-of-step splitting devices to cooperate with each other in terms of operation time. A lot of off-line simulations are needed to obtain the proper splitting surfaces when using the out-of-step splitting criteria, which makes it difficult to adapt to the changes of operation mode and instability mode. The controlled islanding strategy can solve this problem. And it has necessity for study.The key problem to be solved on controlled islanding is how to find out optimal islanding surfaces promptly and accurately. Based on the maximum flow-minimum cut theorem, a searching method for the optimal islanding scheme is proposed to solve the non-deterministic polynomial(NP) problem when it comes to large-scale power system. Firstly, obtain the data of rotor angle velocity from WAMS after loss of stability, and the coherency identification of generators can be implemented by using recognition method of power system coherent generators based on ant colony algorithm. According to the generators’ grouping information, the system networks with capacity are built on the basis of the maximum flow-minimum cut theorem. Secondly, the Ford-Fulkerson labeling algorithm is used to reduce searching space. Finally, based on the demand of minimal net power imbalance within the islands, the optimal islanding surface is obtained. Simulations on the New England 39-bus system and IEEE-118 bus system validate the effectiveness and rapidity of the approach proposed under two- frequency or multi-frequency oscillation scenario.