video

sound

Persian Version

View:

1,254

Download:

0

Cites:

Information Seminar Paper

Title

PREDICTION OF SOIL ACTIVITY THROUGH ARTIFICIAL NEURAL NETWORKS AND ORDINARY LEAST SQUARE MODELS

Pages

 Start Page | End Page

Abstract

 SOIL SWELLING RATE IS ASSESSED ACCORDING TO ITS POSITIVE AND NEGATIVE EFFECTS IN CIVIL ACTIVITIES. SOIL SWELLING RATE DEPEND ITS WATER ABSORPTION RATE. THE SOIL SWELLING LOSS THE STRENGTH SOIL AND INCREASE SOIL COMPRESSIBILITY. ALTHOUGH THE SOILS OF THIS TYPE CAN BE USED TO DECREASE PERMEABILITY IN HYDRAULIC STRUCTURES SUCH AS EARTH DAMS, WATER CHANNELS AND ETC, WHIT A POSITIVE EFFECT, THEY USUALLY HAVE DEFORMATION EFFECTS ON THE STRUCTURES, PAVEMENT. SOIL SWELLING RATE IS ORDINARY OBTAINED BY CALCULATING SOIL ACTIVITY NUMBER. THE ACTIVITY NUMBER IS OBTAINED THROUGH DIVIDING THE PLASTICITY INDEX BY THE PERCENTAGE OF THE PARTICLES LESS THAN 2 MICRON. OBTAINING THE PLASTICITY INDEX IN LABORATORIES IS USUALLY SIMPLE, BUT CLAY FRACTION IS OBTAINED BY SEDIMENTATION PROCEDURES WHICH ARE PRETTY DIFFICULT, EXPENSIVE AND TIME CONSUMING. IN THIS STUDY TOGETHER WITH THE CORRELATION STUDY OF PARTICLES SMALLER THAN 75 MICRON, PLASTICITY LIMIT, LIQUID LIMIT, SHRINKAGE LIMIT AND PLASTICITY INDEX OF 214 SOIL SAMPLES ARE TESTED, AND NUMBER OF SOILS ACTIVITY IS PREDICTED THROUGH ORDINARY LEAST SQUARE (OLS) AND MULTILAYER PERCEPTRON OF ARTIFICIAL NEURON NETWORKS (ANNS) IN 5 SUGGESTED MODELS. THE RESULTS INDICATE THAT THE MULTILAYER PERCEPTRON NOT ONLY PERFORMS BETTER THAN REGRESSION'S MODELS BUT ALSO PROVIDES ACCEPTABLE AND RELIABLE OUTCOMES WITH RESPECT TO THE PREDICTED OBJECTIVES' MATERIALIZATION.

Cites

  • No record.
  • References

  • No record.
  • Related Journal Papers

  • No record.
  • Related Seminar Papers

  • No record.
  • Related Plans

  • No record.
  • Recommended Workshops






    File Not Exists.