Abstract
As the electrical demand continuously rises up due to economic growth, the power system becomes more stressed and hard to maintain. The recent records indicate that the large-scale blackouts are usually referred from the effect of cascading failures and the precise construction of hidden failure model is necessary. The researches of hidden failure model have been focused on the impacts of several parameters and assessing possible mitigation methods. Therefore, DC hidden failure model is built to verify the system parameters such as loading level, spinning reserve and the probability of hidden failures. To improve the system with hidden failures, stochastic programming is introduced to solve generation expansions that can efficiently deal with cascading events. Benders decomposition is applied to speed up computing time of complex problem. Two types of simulation algorithm are addressed: one is that the control actions are solved before the whole hidden failure simulation and the other type is that the controls are calculated and applied between the cascading stages. Both types of procedure will be tested with several IEEE systems. Considering the shortage of DC model, the hidden failure model is extended to AC form. To guarantee the high performance of cascading model, the programs are written in python script to utilize the package function from PSS/E, a powerfully commercial program. AC hidden failure model contains the issue of voltage stability. Hence, load margin evaluation is used to measure the system margin to voltage collapse. Preventive security constraint optimal power (PSCOPF) is applied to improve the system. There are also two types of algorithms. Both of them will be simulated and compared. IEEE test systems and Taipower system will be used to verify AC hidden failure model and PSCOPF.