JUDGING METHOD OF LIQUEFACTION CAUSED BY EARTHQUAKE WITH NEURAL NETWORK

摘要

PURPOSE: To improve the judging accuracy of a liquefaction phenomenon at the time of an earthquake and serve as a disaster prevention countermeasure against liquefaction by introducing unknown factors. ;CONSTITUTION: A judging device constituted of a processing means 1 and a neural network 2 is used, and the actual values in liquefaction phenomena by earthquakes occurring in the past, e.g. geological features, terrains, sizes of earthquake movements, and degrees of observed liquefaction, are given to the neural network 2 as teacher data for learning. The geological feature, terrain, and size of earthquake movement at the point to be judged on liquefaction are given to the neural network 2 as input data, and the index indicating the degree of liquefaction is obtained as output data. The obtained index is compared with the threshold value set in advance in a processing circuit 4 to judge liquefaction.;COPYRIGHT: (C)1996,JPO