Power Flow Analysis using a Multilayer Perceptron Neural Network with Interval Arithmetic in the Presence of data Uncertain

  • G. Sandhya Rani, M. Chakarcarthy, B. Mangu

Abstract

An efficient Power Flow Analysis is required for a complex Power System where the data is uncertain. This work presents an Interval Fast De-Coupled Power Flow (IFDPF) algorithm and a Multi-layer perceptron Feed forward Neural Network (MFNN) to realize power flow analysis under data uncertainty. The IFDPF algorithm is proposed to provide power flow solution of a power system, where load data and generation data is uncertain. The generator and load bus data are taken in intervals as uncertainties are present in the power system. The obtained voltage magnitudes, phase angles, active and reactive powers from IFDPF algorithm are also represented in intervals. The proposed MFNN model is fed with the generator and load bus data in intervals are as input and voltage magnitudes and phase angles which are computed from IFDPF algorithm are considered as output to train the network model to predict the voltages, phase angles at all buses for any unknown operating conditions. The proposed Interval Fast De-Coupled Power Flow (IFDPF) algorithm and developed Multilayer feed forward (MLFF) neural network is tested on IEEE-30 bus system and the results of IFDPF and MLFF are compared for an unknown network operating condition. From the result it is concluded that the developed neural network is validated for online implementation of interval power flow methods.

Published
2020-03-30
How to Cite
G. Sandhya Rani, M. Chakarcarthy, B. Mangu. (2020). Power Flow Analysis using a Multilayer Perceptron Neural Network with Interval Arithmetic in the Presence of data Uncertain. International Journal of Advanced Science and Technology, 29(3), 12625 - 12634. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/30383
Section
Articles