Analysis of Principal Component - Application of SVM Model in Prediction of Ultimate Bearing Capacity of Static Pressure Pipe Pile

Abstract:

Vertical ultimate bearing capacity of static pressure pipe pile is influenced by comprehensive factors, such as pile body, soil around pile and construction conditions., and the relationship between the impact factors and ultimate bearing capacity of a single pile is highly complexity and non-linear. This paper is based on collecting the data from static load tests in typical geological conditions of Liao-shen area, construction records, and test pile sites. And then combine analysis of principal component with SVM to analysis the prediction of the single pile’s vertical ultimate bearing capacity. This model can reduce the number of SVM input variable dimension to improve speed of training support vector effectively. At the same time it can eliminate the influence factors of multiple correlation. The results show that the proposed principal component analysis SVM model has good predictive accuracy and generalization ability, and opens up new avenue of research for analysis of static pressure pipe pile vertical bearing properties.

• Periodical: Applied Mechanics and Materials (Volumes 166-169)
• Pages: 1345-1352
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.166-169.1345
• Online since: May 2012
• Authors: Jian Bin Zhao, Chao Xu, Hui Guo
• Keywords: Analysis of Principal Component, Influence Factor, Prediction of Single Pile Ultimate Bearing Capacity, Static Pressure Pipe Pile, SVM( Support Vector Machine) Model

References

[1] Qinan Ruan. Prestressed Concrete Pipe Pile [M]. (Beijing: China building materials industry press, 2000 )(in Chinese)

[2] Daye Zheng. The Application of The Forecast Method for Full Pile's Ultimate Bearing Capacity [D] (Hefei: Hefei University of Technology, 2004), (in Chinese)

[3] Zhenye Lu. Pile foundation inspection technology [M]. (The Center of State Seismological Bureau Engineering Measure, 1993), (in Chinese)

[4] Jinmang Zai, Jinzhang Zai. High-rise building foundation analysis and design [M]. (Beijing: China Architecture &Building Press, 1993), (in Chinese)

[5] Feng Huang, Chaogang Li, Gungxin Li. Discussion of Pile bearing capacity determination method [J]. ( Tsinghua university journal, 1998, 38(l): 28-23), (in Chinese).

[6] Lee I M, Lee J H. Prediction of Pile Bearing Capacity Using Artificial Neural Networks [J]. ( Computers and Geotechnics, 1996, 18(3): 189-199.)

DOI: 10.1016/0266-352x(95)00027-8

[7] Nawari N O, Liang R, Nusairat J. Artifical intelligence techniques for the design and analysis of deep foundations [J]. (Electronic Journal of Geotechnical Engineering. 1999, 4 ) .

[8] Ziling Feng, Haitao Sun, Shujuan Wang. Forecast Vertical Ultimate Bbearing Capacity of Single Pile by Artificial Neural Network [J]. (Tongji University Journal, 1999, 27(4): 397- 401) , (in Chinese)

[9] Yongjian Liu. The Vertical Bearing Capacity of The Single Pile Prediction Method Research Based on Genetic Neural Network [J]. (Geotechnical engineering, 2004, 25(1): 60-63),(in Chinese).

[10] T. M. H. Lok, W. F. Che. Axial Capacity Prediction for Driven Piles using ANN: Model Comparison [J]. (Geotechnical Engineering for Transportation Projects, 2004: 697-704)

DOI: 10.1061/40744(154)56

[11] Yunkai Liu, Chunsheng Qiao, Shengfeng Tian. The Study on Support Vector Machine (SVM) Model and Precision Influence Factors of Slope Angle Design (Mechanics and Engineering, Vol. 24 (2) (2005), pp.328-335 (in Chinese)

[12] Yanbing Gong, Senfa Chen. Industrial Construction, Vol. 36 (2) (2006), pp.50-53 (in Chinese)

[13] Minggui Liu, Junwei Peng. Industrial Construction, Vol. 37 (2) (2007), pp.60-54 (in Chinese)

[14] Huiwen Wang: Partial Least Squares Regression and its Application (National Defence Industry Press, China 1999) (in Chinese)

[15] Huixuan Gao. Application of multivariate statistical analysis [M]. Beijing: Beijing university press, 2005(in Chinese)

[17] Naiyang Deng, Yingjie Tian. The New Way of Data Mining- Support Vector Machine (SVM)[M]. (Beijing: Science press, 2004), (in Chinese)