Indian Journal of Science and Technology
DOI: 10.17485/ijst/2014/v7i6.1
Year: 2014, Volume: 7, Issue: 6, Pages: 729–733
Original Article
N. Jayaramappa*, A. Krishna, B. P. Annpurna and T. Kiran
Faculty of Engineering-Civil, UVCE, JBC, Bangalore University, Bangalore-560056, India; jakasauj@gmail.com
The main objective of this paper is to predict the base shear for a three dimensional four storey single bay RC frame, subjected to lateral load, using Artificial Neural Network. Analytical Pushover analysis is carried out using ETAB software. Pushover curve is plotted for Base shear versus roof displacement, to the prototype model tested at Central Power Research Institute (CPRI) Bangalore. The analytical results obtained from ETAB analysis are compared with the results predicted from Artificial Neural Network (ANN). Displacement and plastic hinges are selected as input vectors to ANN. Back propagation algorithm has been used for training. A learning rate constant of 0.85, error tolerance 0.001, and 5500 cycles are used for training ANN. Delta rule is used for adjusting the weights. It is found that a four-layer network (5-6-6-1) which consists of 5 input neurons, two hidden layers of 6 neurons each and one output neuron efficiently converges almost close to analytical results with an error range of 0.17% to 8.25%.
Keywords: ANN, Back Propagation, Base Shear, Delta Rule, Pushover
Subscribe now for latest articles and news.