Subsystem boundary values methods for power system transient stability distributed simulation

Abstract

This paper studies the efficiency of power system transient stability distributed simulation and the methods to enhance the computation performance. With parallel in space scheme, power system is split into several subsystems. The bottlenecks in distributed computation are the low speed network communication and the mass times of boundary values exchange for subsystem time domain step solution. For that the iterative boundary values convergence property is the key factor to affect the efficiency, three boundary values methods (BVM) is presented to improve the subsystem solution convergence property. With the BVM, the necessary exchanging times of boundary values are reduced, the communication times for one time domain step is shorten and the distributed computation efficiency is improved.