兩輪式移動機器人，亦稱為差速驅動移動機器人，在機器人領域中日益受到關注並 已被廣泛應用於工業倉儲、環境巡檢及服務型機器人等領域。然而，受限於現實環境中 的各種不確定項與外部擾動等，差速驅動移動機器人如何克服上述實務問題並可以精確 地抵達目標位置仍是當前極具挑戰性的課題之一。 本論文著重在差速驅動移動機器人的軌跡控制器設計，首先將打滑現象與系統不確 定項納入機器人的動力學模型中。接著，在控制策略上採用高強健性的超螺旋滑模控制 器來解緩外部干擾的影響，同時結合徑向基函數神經網路來即時估測系統的不確定項。 此外，本研究也設計出一套基於超螺旋滑模策略的打滑即時補償機制，用以確保機器人 的輪胎在打滑時能繼續維持其軌跡的精確追蹤，並同時透過嚴謹的數學推導來證明系統 之控制穩定性。最後，將本論文所提出之先進控制器與傳統滑模控制、積分滑模控制以 及標準超螺旋滑模控制進行一系列的軌跡追蹤實驗比較。結果證實，所提出之控制器在 軌跡追蹤精度均優於上述三種控制器，同時也展現出優異的打滑補償能力。;Two-wheeled mobile robots (TWMRs)， also known as differential drive mobile robots， have gained increasing attention in the robotics field and have been widely applied in areas such as industrial warehousing， environmental inspection， and service robotics. However， due to various uncertainties and external disturbances present in real-world environments， enabling differential drive mobile robots to overcome these practical challenges and accurately reach target positions remains a highly challenging issue. This thesis focuses on the design of a trajectory controller for differential drive mobile robots. First， wheel slip and system uncertainties are incorporated into the robot’s dynamic model. Then， a highly robust super-twisting sliding mode controller is adopted to mitigate the effects of external disturbances. At the same time， a radial basis function neural network is integrated to estimate system uncertainties in real time. Furthermore， this study develops a real- time slip compensation mechanism based on the super-twisting sliding mode strategy to ensure that the robot can maintain precision trajectory tracking even under wheel slip conditions. The stability of the control system is also rigorously proven through mathematical analysis. Finally， the proposed advanced controller is compared with conventional sliding mode control， integral sliding mode control， and standard super-twisting sliding mode control through a series of trajectory tracking experiments. The results demonstrate that the proposed controller achieves superior tracking accuracy compared to the other three methods， while also exhibiting excellent slip compensation capability.
