Transient stability contingency screening and ranking based on data mining

Contingencies can impact security of power system operations. Severity of contingencies varies with pre-fault conditions. Therefore, it is of interest to use a database of analyzed pre-fault scenarios in order to recognize scenarios similar to the current, thus eliminating the need for continual and extensive on-line CSR, limiting this process only to faults identified as critical in similar cases from a database. Smart grid applications have recently been developed in order to perform supervisory functions, with the crucial goal of reducing the risks associated with insecure operating conditions. CSR belongs to this type of applications. This paper presents a novel approach for carrying out CSR, based on data mining. Considering operational statistics related to loading condition changes, a database was built and employed for system state clustering. Data mining techniques are used for real-time similar state(s) recognition. The proposal is tested on the IEEE 39-bus system. Results demonstrate feasibility of the proposal in on-line CSR.