A multilayer feed forward neural network (MFFNN) is suggested in this research to analyse nonlinear functions for online power flow assessment under data uncertainty. To create the training data for the proposed MFFNN model, the Interval fast decoupled power flow method (IFDPF) is used. The IFDPF method provides a power flow solution for a PS with unknown load and generation data. The IFDPF technique produces voltage magnitudes, phase angles, active and reactive powers, which are used to train the network model to forecast voltages and phase angles at all buses under any unknown operating conditions. The proposed Multilayer feed forward (MFFNN) neural network is tested on IEEE-30, IEEE 57, and IEEE 118 bus systems, and the results of MFFNN and IFDPF under an unknown network operating condition are compared. From the results it is concluded that the constructed neural network has been validated for online implementation of interval power flow methods.

Interval arithmetic, Interval Fast Decoupled Power Flow method, Multilayer feed forward neural network.