本論文使用類神經網路與滑動模式控制針對三維天車系統設計控制器。天車系統控制目標為快速準確定位與抑制吊掛負載搖擺,所設計之類神經滑動模式控制器,藉由解耦合滑動面函數以及透過類神經網路調適滑模斜率參數,並改良順滑面函數縮短滑動模式控制的迫近模態,同時以類神經網路調整參數做到雙重調適功能,讓系統在迫近模態與滑動模態皆具有強健性能。滑動模態之等效控制力計算透過類神經網路完成,免除了複雜動態系統等效控制力推算,也避免了受控系統參數的不精確性。透過Lyapunov證明,本論文所提出之類神經滑動模式控制方法,確保了系統狀態能夠在有限時間內接觸到順滑面而在順滑平面上滑動直至進入漸進收歛之平衡點上。最後本論文以電腦模擬驗證,從模擬結果中看出本文所設計之類神經滑動模式控制器在天車系統上具有快速準確的定位功能與抑制吊掛負載搖擺之能力,而控制器所設計之特色與優點經過模擬比較,透過其結果顯示具有優異的控制能性能。
This thesis uses neural network and sliding mode control to design controller for a three dimensional overhead crane system. The control objectives are to move the load to the target as fast as possible and to avoid load swing of the crane system. Based on the principles of decoupling sliding surface function and self-tuning algorithm of the sliding slope parameters, this thesis presents a new sliding surface function combined with neural network to improve the approaching mode of the sliding mode control, and reaction time. So, the system has robust properties at the approaching mode and sliding mode. Also, the equivalent control based on neural network not only avoids computing the complex inverse dynamic control, but also performs the accurate system parameters. The stability of the proposed control scheme is also guaranteed. Finally, performance of the proposed method is demonstrated by some simulation results.