NEURAL-NETWORK FOR EARTHQUAKE SELECTION IN STRUCTURAL TIME HISTORY ANALYSIS

A neural network is employed to select earthquake waves in a time hist ory approach for structural dynamics. The neural network is a preferab le alternative to an expert system because knowledge can easily be ren ewed. It involves a back propagation model having three layers (one in put, one hidden and one output layer) and is used to avoid inappropria te earthquake input prior to practical numerical computations. Knowled ge to categorize the earthquake waves is acquired through network trai ning with earthquake response spectra and structural responses. The tr ained network is tested by categorizing the responses of three types o f unknown structures caused by 50 previously recorded earthquakes. Com parisons are made with analogous data from the traditional site domina nt period method. Results demonstrate that, unlike the latter method, a neural network is generally more successful as the number of trainin g patterns increases.