Artificial neural network application to voltage stability considering power system configuration

In the power system area, the artificial neural network theory has been applied to various subjects. However, when applying this theory to practical power system models, the dimensional problem is often encountered. Namely, the required memory and learning time increase dramatically with the size of the power system. In this paper, we propose a neural network model which exploits power system configuration in order to reduce the required memory. The proposed neural network is applied to the calculation of a voltage stability index and estimation of the initial value for computation of the unstable power flow solutions. Numerical simulations based on a 43-bus power system model demonstrate the effectiveness of the proposed method.