| The modal fuzzy algorithm based on BP neural network modeling for structural vibration control was presented in this paper,for the structural seismic response mitigation in civil engineering.At first,the structural analysis model was established from BP(back-forward propagation)network training with the computed structural seismic responses.Then the modal fuzzy control rules were formulated,which for the structural state variables-the time domain modal coordinates to decline the system governing grade.These efforts will enable the number of logic reasoning to be within an acceptable level during the implantation of modal fuzzy control.As a verification example,the structural model was established from BP network with EL-Centro wave being the network input and with the seismic response data being the network output.The fuzzy control rules then were formed in order for the object of minimizing the whole system mechanical energy.Six earthquake waves from different site records were input into the modal,with their PGA all being 4m/s2.The seismic dynamic responses(inter storey displacements,storey displacements,storey velocities and storey accelerations)under several control force limitation was compared with those from non-control scheme and from AMD active optimistic control strategy.The main conclusions are as follows:1)Only the mode fuzzy control is adopted for the first mode of the structure,and the results are compared with the interlayer displacement and acceleration in the uncontrolled state,which can achieve satisfactory damping results.2)When AMD's optimal control amplitude is adopted as the control domain of each floor,there is a gap between the damping effect of the modal fuzzy control and that of the control,but the fuzzy control is easy to implement and has good robustness.3)Increasing the domain of control force,the damping efficiency gradually approximates the optimal AMD control,and better damping effect can be obtained.Figure 37;Table 9;Reference 57... |
PDF Full Text Request