Sukanya Saxena and K.K.Pathak. / International Journal of New Technologies in Science and Engineering Vol. 2, Issue.2, Aug 2015, ISSN 2349-0780 53 Available online @ www.ijntse.com Application of Artificial Neural Networks for Fully Stressed Design of Warren Trusses Sukanya Saxena K.K Pathak* National Institute of Technical Teachers' Training and Research, Bhopal (MP) India *Corresponding Author: kkpathak@nitttrbpl.ac.in Abstract: In this study, artificial neural network has been employed for fully stressed design of Warren trusses. Fully stressed design (FSD) of Warren truss for 24 different loading conditions are carried out using STAAD.Pro software. Using these data of FSD, an artificial neural network has been trained. The successfully trained network is further used for FSD of four new cases which are validated by STAAD.Pro software. It is observed that ANN offers a powerful tool for FSD of Warren trusses. Application of ANN will be highly useful in absence of costly analysis and design software. Keywords: Artificial Neural Networks, Fully Stressed Design, Warren truss, STAAD.Pro I. INTRODUCTION In the present scenario artificial neural networks has proved to be a very powerful tool for structural design. It is used to interpolate or extrapolate from a set of design information. Artificial neural networks are computational models inspired by biological neural networks, and are used to approximate functions that are generally unknown. Particularly, they are inspired by the behavior of neurons and the electrical signals they convey between input (such as from the eyes or nerve endings in the hand), processing, and output from the brain (such as reacting to light, touch, or heat). Neural networks can be hardware- (neurons are represented by physical components) or software-based (computer models), and can use a variety of topologies and learning algorithms. Neural networks have been used for structural analysis and design of trusses. Mehrjoo et.al.[1] dealt with recent developments in artificial neural networks (ANNs) like structural identification of large truss bridge structures where in situ measured data are expected to be imprecise and often incomplete. The natural frequencies and mode shapes were used as input parameters to the neural network for damage identification. Kang and Yoon [2] developed a two-layer neural network (single-layer perceptron) for truss design. Reza Kamyab Moghadas et.al. [3] dealt with determination of the optimal design and maximum deflection of double layer grids spending low computational cost using neural networks. Vinay Agrawal et.al [4] presented study of conceptual design of communication towers using Artificial Neural Network approach to prove its reliability in structural engineering. Alam and Berke [5] worked on neural network models for material response into nonlinear elastic truss analysis. Laszlo Berke et.al.[6] worked on application of artificial neural networks to capture structural design. Hassan Aghabaratia and Mohsen Tehranizadeh [7] worked on the application of three main artificial neural networks (ANNs) in damage detection of steel bridges. Cheng et.al. [8] dealt with the application of artificial neural networks (ANN) to the response prediction of geometrically nonlinear truss structures. Kaveh Kumarci and Afsaneh Banitalebi Dehkordi [9] considered the training or learning algorithms in telecommunication towers based on the artificial neural networks to calculate accurately their natural frequency in different supporting conditions. Seong Beom Kim et.al.[10] dealt with the applicability of the prototype of the Neural Network base model for Optimum Structural Design (NNOSD). The artificial neural network is used as an alternative design model of the warren truss since it can handle uncertainty through the probability method. Hajela and Berke [11] carried out optimum design of trusses using neural networks. The input data consisted of length and height of trusses while output data consisted of optimized bar areas and total weight of truss. Lin Niu [12] considered neural network on structural parameter identification and damage detection approach using displacement measurement time series is proposed, and the performance of the approach is validated