• Finhasali I. Navsariwala
• Dr. N. K. Solanki

Artificial neural network is the tool which can solve many structural engineering related problems with less effort. Analysis of truss involves calculating axial forces in each member of the truss. Member forces varies as per loading and analysing entire truss takes time even in software. Hence an attempt is made in this work to connect the analysis of truss with Neural Networking to get the results without using design software. In the present work, fifteen number of trusses of varying span, height, number of panels, number of members and point loads are used to generate data for training, simulating and testing the network created from MATLAB’s ANN module. Another aim of the study is to decide ideal pair of transfer function which are to be specified before developing a network. At the end of simulation, it was found that the results obtained by simulation of network are same as that calculated by software with very minimal errors. Moreover, it was determined that the combination of tansig-purelin as transfer functions between input, hidden and output nodes validates to be the best pair as the root mean square error for this pair of transfer function is nearly zero.

Artificial Neural Network, MATLAB, Transfer Function, truss, member forces.

"Analysis Of 2-D Howe Type Truss Using MATLAB’s Artificial Neural Network Module", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 6, page no.993-997, June 2019, Available :http://www.jetir.org/papers/JETIR1907595.pdf

"Analysis Of 2-D Howe Type Truss Using MATLAB’s Artificial Neural Network Module", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 6, page no. pp993-997, June 2019, Available at : http://www.jetir.org/papers/JETIR1907595.pdf