论文部分内容阅读
A feedforword neural network of multi-layer topologies for systems with hysteretic nonlinearity is constructed based on Bouce Wen differential model. It not only reflects the hysteresis force characteristics of the Bouce Wen model, but also determines its corresponding parameters. The simulation results show that restoring forceedisplacement curve hysteresis loop is very close to the real curve. The model trained can accurately predict the time response of system. The model is checked under the noise level. The result shows that the model has higher modeling precision, good generalization capability and a certain anti-interference ability.
A not for only reflecting the hysteresis force characteristics of the Bouce Wen model, but also determining its corresponding parameters. The simulation results show that restoring The model trained can precisely predict the time response of the system. The model shows the model under the noise level. The result shows that the model has higher modeling precision, good generalization capability and a certain anti-interference ability.