應用滑模控制策略於自動導引自行車之平衡控制

Abstract

本論文旨在發展自動導引自行車系統，並結合陀螺儀平衡器的穩定性與龍頭控制的機動性，設計滑動模式控制器 (Sliding-mode control, SMC) 與模糊滑動模式控制器 (Fuzzy sliding-mode control, FSMC) 於自動導引自行車之平衡控制。本研究系統包含四個主要部分：自行車平台、陀螺儀平衡器、自行車姿態量測與主控制系統。自行車系統先接收姿態量測資料，並將自行車之傾斜角與角速度回授輸入至自行車上控制器中進行計算，給予龍頭、陀螺儀平衡器與自行車驅動馬達輸入命令，藉由此三個控制變數來達到自行車的穩定平衡。 本研究利用Lagrange方程式推導陀螺儀平衡器之動力學模型結合自行車動力模型，並利用PID控制、滑動模式控制與模糊滑動模式控制等理論來設計控制器，該平衡器可產生陀螺儀效應，藉此產生抵抗自行車傾倒之反力矩。滑動模式控制的設計對於干擾與模型誤差，具有高度的強健性。模糊滑動控制具有提升系統強健性與控制器設計簡單的優點，滑動平面增加了受控系統的強健性，而模糊邏輯利用經驗法則可大幅降低設計的複雜度，因此兩種方法都極適用於自行車此種不穩定系統維持穩定。 在模擬中利用上述三種控制器來模擬自行車在具有初始傾斜角、突起地磚與坑洞干擾等三種情況的系統表現。模擬結果可發現三種控制器都可使自行車達成穩定，而SMC與FSMC對於外界干擾具有較佳系統表現。實驗中利用零速度實驗測試自行車平衡穩定性，衝擊實驗利用2公升水瓶落下撞擊自行車，給予突然的外界干擾測試系統強健性。零速度實驗結果中，PID可維持自行車平衡12秒，SMC可維持自行車平衡20秒，而FSMC則可使自行車持續維持穩定而不傾倒，證明FSMC控制器相較其他兩者的優越性。衝擊實驗結果中，FSMC控制器在自行車系統受到劇烈撞擊後仍可維持自行車平衡，證明本研究提出之FSMC控制器有較佳的抗干擾強健性與系統穩定性。未來規劃為基於龍頭控制與陀螺儀平衡器的權重調整，實現自行車系統的路徑規劃。

This research aims to develop a self-riding bicycle that achieves stability and maneuverability through gyroscopic balancer and steering control. And, we incorporate sliding mode control (SMC) and fuzzy sliding mode control (FSMC) into the controller to enhance the stability of the bicycle. The bicycle system contains four parts: bicycle platform, gyroscopic balancer, control system, and initial measurement unit (IMU) for acquiring posture information. When the self-riding bicycle system receives data from IMU, the micro-controller would generate appropriate command for steering mechanism, gyroscopic balancer, and driving motor of the bicycle based on lean angle and rate of lean angle of the bicycle system. The dynamic model of the self-riding bicycle system with the gyroscopic balancer and bicycle is derived by Lagrange equation, and then it is employed in the design of PID control, sliding-mode control and fuzzy sliding-mode control. Moreover, the gyroscopic balancer can produce a torque by gyroscopic precession to overcome the leaning of the bicycle. The sliding-mode control strategy has high robustness in the situation with unknown uncertainties due to model error or disturbances. The fuzzy sliding-mode control has even more advantages on improving the robustness of control system and making design easier. The sliding surface design improves the robustness of control system, and the fuzzy logic design which is constructed by human experience decreases the complexity of controller design. Therefore, these two methods would be appropriate for stabilizing a statically unstable system like bicycles. The simulations are carried out for the comparison of PID controller, SMC controller, and FSMC controller with initial lean angle, uneven brick disturbance or pothole disturbance on the bicycle. The simulation results show that three controllers can stabilize the bicycle, and SMC controller and FSMC controller have the better performance under external disturbance. In field test, the bicycle with zero velocity and impact disturbance is tested for evaluating the stability and robustness. 2-liter bottle is falling to give the bicycle sudden external disturbances. The experiment results show that the PID controller can stabilize the bicycle up to 12 seconds, the SMC controller up to 20 seconds, and the FSMC controller can sustain the bicycle without falling down. Therefore, the FSMC controller is better than the other two. The FSMC controller is also proved to be more robust under the impact. In the future work, the path tracking of the bicycle system will be implemented by further weighting adjustment between steering control and gyroscopic balancer.