An on-line generation dispatching algorithm to improve small signal stability

To prevent or suppress the low-frequency oscillation in real time, the modal sensitivity with respect to operating parameters is calculated. It is applied to generate the effective control strategy that can improve the system damping. Initially, Fuzzy ISODATA clustering method is used to obtain the coherent generation group, from which candidate generators for rescheduling are formed and further refined by using participation factors. With the efficient parallel computation of eigen-sensitivity, the relationship between modes of interest and operating parameters can be obtained and used to provide system operators with real-time system operation strategies to improve the system damping. The proposed control strategy satisfies many constraints, such as power output limits and real time operational constraints. The feasibility and applicability of the proposed method has been tested using an actual large scale power system of China with more than 20000 nodes. The time taken to solve this large system is just over 2 minute with parallel computation. The proposed method has been implemented in the on-line dynamic security assessment and early warning system of State Grid Corporation of China (SGCC).