|
| Decentralized neural networks vibration control of tall buildings under earthquakes |
| Received:September 12, 2017 Revised:October 29, 2017 |
| View Full Text View/Add Comment Download reader |
| DOI：10.7511/jslx20170912002 |
| KeyWord:tall building seismic action neural network decentralized control |
| Author | Institution |
| 汪权 |
合肥工业大学 土木与水利工程学院, 合肥 |
| 韩强强 |
合肥工业大学 土木与水利工程学院, 合肥 |
| 王肖东 |
合肥工业大学 土木与水利工程学院, 合肥 |
| 袁加伟 |
合肥工业大学 土木与水利工程学院, 合肥 |
|
| Hits: 1201 |
| Download times: 855 |
| Abstract: |
| Vibration control of tall buildings under earthquakes based on the neural networks is studied.By combining the neural network theory and the decentralized control strategy,an algorithm named Decentralized Neural Networks Control(DNNC) is been presented for the seismic response control of tall buildings.The structural model is built by using multilayer feedforward neural network to predict the structural vibration response.The BP neural network controller is designed on the basis of Nonlinear Auto-Regressive Moving Average(NARMA) mode to study the effectiveness of the DNNC which is then compared with that of Centralized Neural Networks Control(CNNC).The performance of control strategies of a typical 20-floor seismically excited building is analyzed the performance of the control strategies.The results indicate that the DNNC strategy could simplify the structure of the neural networks and effectively mitigate structural seismic response and eliminate the time delay of the control system.It is also more reliable in ensuring the real-time response of vibration control system. |