NEURAL DYNAMIC MODELING ON EARTHQUAKE MAGNITUDE SERIES

Earthquake magnitude prediction is of vital importance for human safet y. The earthquake is a very complicated and non-linear dynamic process . It cannot be described adequately by any deterministic models. In th is paper a neural dynamic modelling for earthquake magnitude predictio n is reported. Historical records of earthquake magnitude series are u sed to construct the optimal non-linear dynamic model, and the consequ ent outcome of the earthquake behaviour is then predicted by this mode l. In turn, the latest recorded data set can be fed back to improve th e accuracy of the neural dynamic model. The modelling of experiments o f three earthquake magnitude series in China and Japan and their extra polated predictions are included in this paper. The values predicted b y extrapolation are in good agreement with the historical data.