Abstract
Modern power systems continue to grow in size and complexity due to the high demand for electricity. Thus, dynamic security assessment (DSA) is becoming a necessary requirement in the system operation. The critical clearing time (CCT) is a key issue for DSA. Nonlinear time domain simulation (NTDS) is the most accurate method for computing CCT. Unfortunately, DSA is often confronted by the high nonlinearity of interconnected power networks. Thus, NTDS-based DSA is considered time consuming and needs heavy computational effort. In order to avoid these drawbacks, this paper deals with a new technique for online DSA of interconnected power system. Such technique is developed in two steps. Firstly, NTDS is used to compute CCTs for various loading conditions. Then, adaptive network based fuzzy inference systems (ANFIS) is used to establish the relationship between the operating conditions and the corresponding CCTs. The approach effectiveness is validated on two multimachine power systems under severe fault disturbances.