This study used a feed-forward artificial neural network (ANN) technique in order to predict fabric permeability. The study introduced a set of input patterns on the fabric permeability during the ANN training phase. The individual weights for the interconnections between nodes were adjusted until the inputs of temperature, pressure drop and fabric type yielded the required permeability output. In this way, the ANN learned the desired input-output response behaviour. After the initial training, the ANN was tested on additional data that were not part of the training processes. The predictions of the trained network agreed very well with these new experimental data. The study indicates that ANN can be an effective tool in modeling airbag fabric behaviour. This process requires time and experimental data for training. However, once trained, only fractions of a second are needed for information assimilation and output generation. This coupled with simplicity of use and accuracy of predictions make ANN attractive for on-line applications.
Abstract