Abstract Since Van der Meer proposed new empirical formulas to calculate the stability number of rock armor based on his own experimental data in 1987, the data have also been used for the development of artificial neural network (ANN) models. However, the ANN models are seldom used because they are not easy to verify in spite of high accuracy. In this study, an accurate easy-to-use ANN-based model is developed. The stability number is calculated by ensemble-averaging the outputs of 500 ANN models which were developed with different training data. The accuracy of the model is markedly improved compared with previous empirical formulas or ANN models. A spreadsheet calculator is also provided so that it can be easily used by engineers without a deep knowledge of ANN. It calculates the stability number by using the pre-determined weights and biases of the 500 ANN models. The confidence intervals of several confidence levels are also calculated by the standard deviation or the quantiles of the 500 model outputs. A similar model is developed for Tetrapod as well.
Spreadsheet Calculators for Stability Number of Armor Units Based on Artificial Neural Network Models
KSCE Journal of Civil Engineering ; 23 , 12 ; 4961-4971
2019-11-12
11 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
HENRY – Bundesanstalt für Wasserbau (BAW) | 1974
|Airplane stability programs for pocket calculators
NTRS | 1979
|Stability of Armor Units at Concave Section of Man-made Island
British Library Conference Proceedings | 1997
|CORE-LOC: Optimized Concrete Armor Units
Online Contents | 1995
|