Crack Damage Identification of Reinforced Concrete Simply Supported Beam Based on BP Neural Network

Abstract:

Considering the good forecasting capability of BP neural network, a new crack damage identification method for reinforced concrete simply supported beam is proposed in this paper. After simulating the crack damage of a reinforced concrete simply supported beam, the natural frequency of the beam is chosen as the input parameters of the BP neural network. The data before and after damage of the simply supported beam are put into the trained neural network to judge the structural damage. The results demonstrate that the approach has a better application prospects in structural damage identification.

• Periodical: Advanced Materials Research (Volumes 468-471)
• Pages: 738-741
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.468-471.738
• Online since: February 2012
• Authors: Xiao Ming Yang, Fu Li
• Keywords: BP Neural Network (BPNN), Damage Identification, Natural Frequency, Reinforced Concrete Simply Supported Beam.

References

[1] Xiang Li and Qing Wang.Modern Physics(2001).(In Chinese)

[2] Yigui Xu ,Shi Tie Lin and Yang Shu Zi. Journal of Vibration Engineering(1997).(In Chinese)

[3] Bosheng Wang and Jiang Dinghao.Engineering Design(1999). (In Chinese)

[4] Jingzhi Wu and Yigang Zhang.World Earthquake Engineering(2005). (In Chinese)

[5] Xianguo Ye.Journal of Civil Engineering(2008). (In Chinese)

[6] Suwen Cheng and Guoqiang Li .Vibration, detection and Diagnosis(2001). (In Chinese)