PREDICTION OF PORE PRESSURE CHANGE IN ROCK-FILL DAMS DURING ITS FIRST IMPOUNDING BY USING REVISED ANN MODEL

Abstract

The safety condition of a rock-fill dam has been evaluated by comparing the data measured apparatuses with results getting from numerical analyses. Recently numerical analyses have been highly developed. But, calculated results have not always fitted accurately the real movement. The object of this report is to develop a new back propagation rule for studying the parameters in ANN model. It is found that this new technique can reduce the time for estimating parameters. Furthermore the estimated values by the technique could give the better approximation of measured data. It was concluded that this new method could be aptly applied to the evaluation and prediction of pore pressure change in real rock-fill dams.