[CS12-47] Modeling of hysteresis loops by neural network and its application to nonlinear dynamic analysis
Keywords:neural network, earthquake response analysis
In general, nonlinear hysteresis loops of new materials are modeled by selecting the most suitable model that captures the features of new material from relationship between the load and the displacemment obtained by the experiment and adjusting parameters. However, sometimes the model become complicated or not be able to reproduce the actual beavior. On the other hand, a neural network has a great capacity for learning and can be used for dynamic response analysis. In this research, we consider a method of performing a dynamic analysis with not modeling by using hysteresisloops based on conventional formulation but a learned neural network.
Abstract password authentication.
Password is required to view the abstract. Please enter a password to authenticate.