The architecture of dynamic security assessment processing system (DSAPS) is proposed to address online dynamic security assessment (DSA) with focus of the dissertation on low-probability, high-consequence events. DSAPS upgrades current online DSA functions and adds new functions to fit into the modern power grid.

Trajectory sensitivity analysis is introduced and its applications in power system are reviewed. An index is presented to assess transient voltage dips quantitatively using trajectory sensitivities. Then the framework of anticipatory computing system (ACS) for cascading defense is presented as an important function of DSAPS. ACS addresses various security problems and the uncertainties in cascading outages. Corrective control design is automated to mitigate the system stress in cascading progressions. The corrective controls introduced in the dissertation include corrective security constrained optimal power flow, a two-stage load control for severe under-frequency conditions, and transient stability constrained optimal power flow for cascading outages.

With state-of-the-art computing facilities to perform high-speed extended-term time-domain simulation and optimization for large-scale systems, DSAPS/ACS efficiently addresses online DSA for low-probability, high-consequence events, which are not addressed by today's industrial practice. Human interference is reduced in the computationally burdensome analysis.