Parallel self-organising hierarchical neural network-based fast voltage estimation

Fast voltage security monitoring and analysis have assumed importance in the present-day stressed operation of power system networks; and fast prediction of bus voltage is essential for this. An approach based on parallel self-organising hierarchical neural networks is presented to predict bus voltage in an efficient manner. Parallel self-organising hierarchical neural networks (PSHNN) are multistage networks, in which stages operate in parallel rather than in series during testing. The entropy concept has been used to identify the inputs for PSHNN. A revised back propagation algorithm is used for learning input nonlinearities, along with forward-backward training. The proposed method is used to predict bus voltage at different loading conditions and for an outage event in IEEE 30-bus and a practical 75-bus systems.