2831 ニューラルネットワークを用いた屋外石油タンクの擬似AE源位置標定(S37-2 非破壊評価とモニタリング(2),S37 非破壊評価とモニタリング)

概要

論文の詳細を見る
The development of the technique to monitor the acceleration of corrosion in outdoor oil tank is desired to prevent environmental pollution accident. Present study utilized the acoustic emission technique to monitor the corrosion. Artificial AE signals were generated by breaking pencil leads on 73 locations of the bottom of the real tank (300kl) to investigate an accuracy of predicted location. AE source location was carried out using a neural network. Arrival time intervals of each sensor were used for input units of the NN. Two methods, visual observation and threshold technique for normalized AE wave, were examined to detect arrival time. The result suggested that accurate location is possible for both methods.
社団法人日本機械学会の論文
2006-09-15