Since deep learning is providing an effective way to train systems and gaining high performance, it is being transforming action in various fields. This study defines the deep learning approach for structural analysis and its predictions for exploring optimum design variables. The basic originality of this work is training dataset and prediction of design variables in the case of two types of truss structures. In order to train the neural network, various numbers of sets and classes are created for 10 bar and 25 bar trusses. In this case, implementation of this experiment involves two manners: regression and multi-classification. By these two manners, the power of deep learning for truss optimization is determined without using any structural analyzers and algorithms.
- Ian Goodfellow, Yoshua Bengio, and Aaron Courville. Deep Learning. MIT Press, November 2016.
- Vanluchene R. D. and Sun Roufei. Neural Networks in Structural Engineering.Computer-Aided Civil and Infrastructure Engineering,November 2008.
- Frank Rosenblatt. The perceptron: A probabilistic model for information storage and organization in the brain. Psychological Review,1958.
- E. Levin. Word recognition using hidden control neural architecture. In International Conference on Acoustics, Speech, and Signal Processing, pages 433–436 vol.1, April 1990.
- Hojjat Adeli. Neural Networks in Civil Engineering: 1989-2000. Computer Aided Civil and Infrastructure Engineering, March 2001.
- Seunghye Lee, Jingwan Ha, Mehriniso Zokhirova, Hyeonjoon Moon, and Jaehong Lee. Background Information of Deep Learning for Structural Engineering. Archives of Computational Methods in Engineering, January 2018.
- A. Kaveh, V. Kalatjari. Size, geometry optimization of trusses by the force method and genetic algorithm. International Journal for Numerical Methods in Engineering, October, 2003.
"Effective Approach for Truss Optimization Using Deep Learning Based on Multi – Classification,"
Acta of Turin Polytechnic University in Tashkent: Vol. 9
, Article 10.
Available at: https://uzjournals.edu.uz/actattpu/vol9/iss2/10