Pile capacity determination using neural networks
การหาค่ารับน้ำหนักบรรทุกของเสาเข็มโดยโครงข่ายสมองกล

Saravut Jaritngam, D. of Civil Eng., F. of Eng., PSU.
Somchai Chuchom, D. of Industrial Eng., F. of Eng., PSU.
Rittisak Jaritngam, D. of Physics, Faculty of Sci., Thaksin U., Songkhla
Corresponding e-mail : jsaravut@ratree.psu.ac.th

Presented : The 1st Seminar on Collaboration in Computer Communications and IC Design Technological in Thailand, 25-26 December 2000, Prince of Songkla University, Hat Yai, Songkla, Thailand : 21-26
Key words : pile capacity, neural networks

Normally, engineers can apply various methods to determine pile capacity. For the construction purpose, engineers must test the pile after driving by Static load test. Because of the significant ad-vances in computer technology, the estimation of pile capacity can be achieved by dynamic load test. Dynamic measurements of force and velocity can be used to quantify pile capacity, generally the load test instrumentation system consists of strain transducers and accelerometers. For each hammer blow, the pile driving analyzer converts the strain and acceleration signals into force and velocity data vesus time (Stress Wave). These data sets are used for computing pile capacity using CAPWAP (Case Pile Analysis Program). The operation of CAPWAP requires trained and experienced person. Even though, the process can be very time-consuming.

This paper demonstrates the use of back-propagation neural networks which is a part of Expert System to solve determining pile capacity. The outcome can get a decision support system by appli-cation of Information Technology. This approach will be a new method to solve civil engineering problems. Finally, it can be shown the relationship among a number of the interacting factors.

---

BACK