Cluster Computing-Based Trajectory Sensitivity Analysis Application to the WECC System

Abstract

This paper presents an application of the trajectory sensitivity method to a large realistic power system. A cluster computing platform is utilized to solve the trajectory sensitivity equations in parallel when multiple sensitivity element calculations are performed. The results show great efficiency improvement with this platform. Another important issue addressed is the error associated with the perturbation size for linear trajectory approximations. When the perturbation size varies, it is critical to ascertain the accuracy of the trajectory approximation. A technique to characterize the error and quantify it is developed. A transient stability constrained preventive control application of generation rescheduling involving trajectory sensitivity analysis, parallel computing, and accuracy evaluation is presented for a large WECC test system. This demonstration highlights the efficacy of the parallel computing capability and accuracy of the calculated preventive control.