|
Spatial Interpolation of Unconfined Compressive Strength for Soft Bangkok Clay Via Random Technique and Modified Inverse Distance Weight Method |
|---|---|
| รหัสดีโอไอ | |
| Creator | Thongchai Phothong |
| Title | Spatial Interpolation of Unconfined Compressive Strength for Soft Bangkok Clay Via Random Technique and Modified Inverse Distance Weight Method |
| Contributor | Puttipol Dumrongchai, Sayan Sirimontree, Boonsap Witchayangkoon |
| Publisher | TUENGR Group |
| Publication Year | 2560 |
| Journal Title | International Transaction Journal of Engineering, Management, & Applied Sciences & Technologies |
| Journal Vol. | 8 |
| Journal No. | 1 |
| Page no. | 13-22 |
| Keyword | Soft Clay, Spatial Interpolation, Inverse Distance Weight, Random Technique, Anisotropy Angle, Geographic Modification |
| URL Website | http://tuengr.com |
| Website title | International Transaction Journal of Engineering, Management, & Applied Sciences & Technologies |
| ISSN | 2228-9860 |
| Abstract | The basic of the inverses distance weight method has been improved by many researches. Not only power parameter but also sensitivity, anisotropy ratio, anisotropy angle and searching radius are incorporated to the model. In addition the cross validation processes is introduced to filter the best parameters for the observed data. To further develop, this study proposes a simple technique and a small modification of the IDW (modified IDW (mIDW)) function but increase ability to improve the estimation results. First, the random technique can be added in cross validation searching step to receive the possibly best parameters to better represent the natural phenomena of the interested areas. Second, with mIDW a coefficient of anisotropy angle parameter is modified to respond the anisotropy effect. The modified IDW is added a coefficient call geographic modification, to reflect the anisotropy angle parameter. This study uses 164 bore holes with 23 layers data for Bangkok to test the model. This study evaluates four models: IDW, Tomczak IDW (Tomczak, 1998), mIDW, and Cubic Spline. With the random technique the observed data patterns that are feed to the searching parameters step reveal the best parameters to imitate the natural phenomena of the area. This study finds mIDW model is better than IDW, Tomczak IDW, and Cubic Spline. The root means square errors of the study case decreases numerously from the worst random to best random case by 11 %. |
