Data driven approach for active loss prediction with reactive support contingencies using ANN

Power system always demands optimum and efficient operations. For this purpose, it is very crucial to manage and reduce the system losses for both transmission and distribution network. Conventionally, load flow techniques such as Gauss-Seidel and Newton-Raphson are used to calculate system losses but these techniques are time consuming due to iterative nature. Power systems are subjected to various disturbances which in extreme condition may include tripping of reactive power support due to shunt devices such as synchronous condensers and capacitor banks. In this paper, data driven approach is proposed to estimate system active losses with consideration of reactive support contingencies. IEEE 14 bus model is used in this research article as test system.