透過您的圖書館登入
IP:3.237.178.126
  • 學位論文

基座干擾下以視覺回授為基礎之定向控制

Vision Based Orientation Control of a Pan/Tilt Platform Subjected to Base Disturbance

指導教授 : 蕭俊祥
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


本論文為pan/tilt旋轉機構的基座受到外力干擾下,以視覺回授為基礎之定向控制研究。我們應用適應性主動式振動控制與視覺感測器建立一視覺伺服系統,以期達到pan/tilt機構減振的目的。主要是利用CCD攝影機偵測目標位置,使由史都華平台提供基座干擾讓pan/tilt平台產生目標物偏離影像平面中心的偏移誤差資訊,能傳送訊號至電腦,以啟動個人電腦的控制器、輸出控制訊號給平台的伺服馬達,達到穩定鎖住目標物於影像平面中心點的定向控制目的。 在控制方面,採用適應性控制理論之Filtered-X LMS演算法及自我調節控制器與類神經網路,來驗證主動式振動控制器能消除基座振動之干擾。實驗結果顯示,應用主動振動控制,單頻訊號於單軸控制上FXLMS演算法有平均約21.56dB的減振量,STR控制方法平均約有17.41dB的減振量;雙軸同動控制上FXLMS演算法有平均約15.56dB的減振量,STR控制方法平均約有11.57dB的減振量。 複頻訊號於單軸控制上FXLMS演算法有平均約15.56 dB的減振量,STR控制方法平均約有10.00dB的減振量;雙軸同動控制上FXLMS演算法有平均約10.05dB的減振量,STR控制方法平均約有6.74dB的減振量。由此可驗證,主動式振動控制在抑制低頻的干擾上有顯著之控制效果。 最後在類神經控制方面,主要是針對基座一連串定點位移來做定位控制。經實驗顯示針對每一定點偏差,類神經控制器可在3∼5秒內完全把目標定位在影像中心。

並列摘要


This thesis focuses on the study of orientation control of a pan/tilt platform subjected to base excitation. Combination of an adaptive active vibration control and a vision system is done to form a vision servo system for suppressing vibrations of the pan/tilt platform. A CCD camera is mounted in the tilt platform for capturing an object coordinates, which can be disturbed from its center position due to base disturbances and which is then fed back to the computer. A control signal from the controller is sent to the servo motors in the pan/tilt platform hoping for constantly orienting the object in the center of the image plane. For the controller design, adaptive feedback controls based on Filtered-X LMS algorithm, and self-tuning regulator, as well as a neural network are utilized for attenuating vibration from the platform base disturbances. Experimental results reveal that for single-frequency base disturbance, about 21.56 dB attenuation and 17.41 dB attenuation can be obtained, respectively, by using the Filtered-X LMS algorithm and self-tuning regulator for independent yaw or pitch control. However, for simultaneously yaw and pitch vibration control, less vibration reductions of 15.56 dB and 11. 57 dB are observed for using Filtered-X LMS algorithm and self-tuning regulator, respectively. For dual-frequency base excitations, 15.56 dB and 10.0 dB attenuations are obtained by using the Filtered-X LMS algorithm and self-tuning regulator, respectively, for independent yaw or pitch control, whereas 10.05dB and 6.74 dB reductions are observed for simultaneously yaw and pitch vibration control. From these experimental results we can see the effectiveness of the adaptive feedback controllers for suppressing the low-frequency vibrations. Lastly, a series of fixed point disturbances is applied to the pan/tilt platform, the neural controller is shown to have the capability of re-orienting the object in the image center in about 3~5 seconds.

參考文獻


[7] 蔡博智,2002, ”影像追蹤方法應用在監控系統之研究” 碩士論文, 中原大學機械系。
[4] Papanikolopoulos, N.P., Khosla, P.K., and Kanade, T., 1993, “Visual tracking of a moving target by a camera mounted on a robot: a combination of control and vision,” IEEE Transactions on Robotics and Automation, Vol. 9, No. 1, pp. 14-35.
[6] Song, K.T. and Chien, C.C., 2002, “Visual tracking of a moving person for a home robot,”ICMT2003, Taiwan.
[11] Magana, M. and Holzapfel, F., 1998, “Fuzzy-logic control of an inverted pendulum with vision feedback,” IEEE Transactions on Education, Vol. 41, No. 2, pp.165-170.
[13] Jamro, E. and Wiatr, K., 2001, “Convolution operation implemented in FPGA structures for real-time image processing,” Proc. of the 2nd International Symposium on Image and Signal Processing and Analysis, pp. 417-422.

延伸閱讀