- 學位論文
輻狀基底函數與模糊滑動模式之主動結構控制
Radial Basis Function combination Fuzzy Sliding Mode to Active Structural Control
摘要
本文主旨發展類神經網路與模糊滑動模式（Fuzzy sliding mode control, FSMC）在結構控制之結合應用，以輻狀基底函數學習速度快速以及對外在環境能快速適應之優點，達到最佳的控制效果。首先推導受控系統之運動方程式，作為動態反應之依據，並利用神經網路學習能力以及FSMC來做控制，研究主動式調質阻尼器應用於樓房結構系統之控制效果，並進行數值模擬分析，強制振動對兩種樓房結構系統進行驗證，分別探討被動控制與模糊滑動模式控制及類神經網路控制在單自由度樓房結構系統與多自由度樓房結構系統之可行性，並與未控制前比較，進一步探討控制結構系統之行為。 經過模擬結果證實，本文使用之類神經網路控制可以有效應用於主動式樓房結構控制系統，樓房結構系統在單自由度與多自由度受不同外擾力皆可降低樓房結構動態反應，達到良好的控制效益。
並列摘要
The purpose of this research is to develop a methodology combing neural network and fuzzy sliding mode to control structural, the principal prat of radial basis function is learning fast and qickly adapt to environment. First the state equation of the control system was derived to determine the dynamic responses. Then, control is carried out by taking advantages of the learning ability of neural network and fuzzy sliding mode control, forced vibration analysis are conducted to assess the effectiveness of the proposed algo-rithm.We studied the performance of FSMC and neural network on struc-ture,and compared with the systems without controlling. The simulation results show that the neural network control method of active control for building structures can decrease structural dynamic re-sponses. The results indicate that the control effectiveness is greatly influ-enced by neural network control in single or multi-degree of freedom system.