全方位球型機器人之設計與實現:使用具強健性之模糊監督控制器

Design and implementation of omni- directional spherical robot: using robust fuzzy supervisory controller

蔡文儒(Wen-Ru Tsai)

指導教授 ： 涂世雄 ； 邱智煇

中原大學/工學院/電機工程研究所/碩士(2009年)

https://doi.org/10.6840/cycu200900328

摘要

在本論文中，我們實現了一個以球做為移動機制的機器人，此機器人與輪型（wheel-base）有著不同的驅動模式。以往的輪型機器人皆有一個共同的缺點，即轉向時無法於第一時間轉到任何一個方位，也就是須先使各輪轉動不一樣的角度，達到車身行進方向偏移，然後才能達到轉向之目的。由於無法於第一時間完成全方位轉向，如此將會使其行動敏捷程度降低。基於此考量，本作品即發展出一全方位球型機器人，其具備全方位的行動能力，同時還有低迴轉半徑及高的靈活度。並利用一般的感應裝置來完成姿態量測，降低系統的建構成本。
為完成建構全方位機器人之目標，首先我們必須設計一主動式平衡控制器，利用傳統倒單擺控制的概念，並考慮身體與球的相對角度與角速度關係當作控制變數，主動控制球的轉速與轉向。在此，我們提出了三種控制演算法，其中包含模糊監督式控制器、強健模糊控制器及強健模糊監督式控制器。傳統模糊理論之模糊規則藉由使用者經驗所設計，故對於控制非線性系統，是一簡單、好設計之控制理論。而此球型機器人存在著許多不確定之因素及雜訊，會影響系統動作，為了解決此問題，我們加上強健控制器(robust controller)及監督式控制器(supervisory controller)，並藉由李亞普諾夫定理(Lyapunov theory)推導其穩定性；此強健模糊監督式控制器有三個優點：一、運用在即時系統(on-line system)中，模糊控制器動作的響應時間會比其它傳統控制器來得更短、運算速度更快；二、具良好的強健性，經微調控制器之輸出，壓抑不確定之因素干擾或補償未調整好之模糊規則；三、可證明其穩定性。另外還加入模糊定位控制器 (Fuzzy position controller) 達到機器人定位設計。完成身體直立並且能自由轉向、前進、後退，使其成為具自主性之球型機器人的目標。最後，將會以模擬及實驗驗證本論文所實現出的機器人及提出的理論架構對於球型機器人之控制具有相當不錯的成效。

關鍵字

輪型機器人 ； 模糊控制 ； 強健控制器 ； 監督式控制器 ； 李亞普諾夫定理

並列摘要

In this thesis, we design and implementation an omni-directional spherical mobile robot control system. The mobile mechanism of spherical robot is different from the wheel-based one. Because the wheel-based robot has some constraint in mechanism, they can not move around well. The major advantage of this spherical robot is that it is capable to move for omni-directional without any constraint. It is obviously such a robot system which is high nonlinearity and is always unknown. It is difficult to establish an exact mathematical model for the design of a model-based control system. To deal with those unknown nonlinearities and external disturbances, the technique of fuzzy logic control is introduced. The fuzzy logic control is a completely different approach that does not require a precise mathematical model of the system. This control method is based on human experience to understand the behavior of the system. Thus, this kind of control design is simpler than traditional one. However, there are lots of unknown factors and disturbances to affect the system to work properly. To make it work more like we expect, we add a device on it.
However, the main drawback of Mamdani type fuzzy theory is the stability analysis. The reason is that such a control method is based on human experience. Moreover, there are many unknown and time varying terms in the robot system, such as frition, payload variation and disturbance. It is difficult to control this robot only using fuzzy theorem. In order to solve this problem, we propose a robust fuzzy supervisory controller to control the spherical robot. By the proposed control method, the stability of the system can be guaranteed. Finally, the simulation and experimental results demonstrate the good performance of the whole control system.

並列關鍵字

supervisory controller ； Lyapunov theory ； robust controller ； fuzzy control ； wheel-based robot

參考文獻

[2]. Tanaka, F., Fortenberry, B., Aisaka, K., Movellan, J.R.: “Plans for developing real-time dance interaction between QRIO and toddlers in a classroom environment”, Proc. The 4th Inter. Conf. Devel. Learn., 2005, pp. 142–147.

[3]. Gen Endo, Jun Nakanishi, Jun Morimoto, Cheng, G.:“Experimental studies of a neural oscillator for biped locomotion with QRIO” Proc. IEEE Int. Conf. Robot. Autom., 2005, pp. 596–602.

[5]. Dariush, B., Gienger, M., Bing Jian, Goerick, C., Fujimura, K.:“Whole body humanoid control from human motion descriptors” Proc. IEEE Int. Conf. Robot. Autom., 2008, pp. 2677 – 2684

[6]. Chestnutt, J., Michel, P., Kuffner, J., Kanade, T.:“Locomotion among dynamic obstacles for the honda ASIMO” IEEE/RSJ Int. Intel. Robot. Syst., 2007, pp. 2572 – 2573

[10]. Cole, J.B., Grimes, D.B., Rao, R.P.N.:“Learning full-body motions from monocular vision: dynamic imitation in a humanoid robot” IEEE/RSJ Int. Intel. Robot. Syst., 2007, pp. 240 – 246