接觸與非接觸式障礙物規避之人機協作任務之研究

Human-Robot Collaborative Task Execution with Contact and Non-Contact Obstacle Avoidance Control

郭孟勳(Mong-Hsun Kuo)

指導教授 ： 羅仁權

國立臺灣大學/電機資訊學院/電機工程學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201703335

摘要

在此論文提出一個具備接觸性與非接觸性的避障系統應用於機器人手臂，透過此系統不但能達到人機協同的目標還能確保作業員與機器的安全性。近年來製造業將人類與機械手臂的互相合作導入工廠生產線中欲達成智慧工廠的目標，因此人機協同成為重要的研究方向。然而在生產線上機器人與操作員共用工作空間，因此兩者間的碰撞意外必然會發生，這樣的碰撞不但會傷害操作員的身體也會造成精密儀器的損毀，如何保障兩者的安全性將會是一大考驗。
欲解決此問題，有兩種方法，一種為接觸式避障而另一種則為非接觸式避障。接觸式避障透過對於控制器的電流監測與分析去判定是否有外力矩作用；而非接觸式避障則是機械手臂具備感知能力並能夠在操作員進入其運動軌跡時進行回避的程序，因此在本文內使用具有感測深度資訊的感測器Kinect用於偵測障礙物。然而只使用一台感測器會因為感測死角的緣故造成碰撞的可能性，因此在此論文的研究中將應用兩台感測器達到全方位零死角的感測。從感測器來的資訊會經由向量軌跡產生法推導障礙物對於機械手臂每個關節的排斥向量與吸引向量，將此兩向量的總和決定是否需要啟動避障的功能。若是避障功能啟動，就會由兩向量總合計算機械手臂新的運動軌跡並操控手臂進行障礙物歸避動作。然而單純的向量軌跡演算法無法及時控制正在高速度和高加速度動作的手臂，因此減速度模型將會再感測器偵測到障礙物時預先啟動以降低手臂之線性速度。最後實驗結果是使用國立臺灣大學智慧機器人及自動化國際研究中心(NTU-iCeiRA)設計製作的16公斤級六軸工業型手臂。

關鍵字

避障系統 ； 感測器 ； 人機協同 ； 工廠自動化 ； 六軸工業型手臂

並列摘要

In this paper, an industrial robot manipulator integrated repulsive and attractive vector generator in order to achieve non-contact collision avoidance is proposed. In recent years, human-robot collaboration (HRC) becomes a huge research topic because more and more manufacturers want to make industry automation by depending on collaboration between mankind and robotics in the production line. In HRC system, the working environment is shared by both machines and operators. As a result, the collision between man and robot might happen accidentally due to some reasons such as the fatigue of mankind and the manipulator‘s suddenly broken. These accidents will not only injury operators but also make great damages to high-priced and precise equipment. Therefore, the highest priority is to prevent the collision happened in order to guarantee the safety of both sides.
To solve this problem, there are two different types of methods. The first one is contact obstacle avoidance while the other one is non-contact collision avoidance. In contact algorithm, the current of the manipulator’s controller is monitored and analyzed for recognizing whether there is an extra torque applied to the robot arm. Non-contact collision avoidance is that manipulator is equipped with the skill which can sidestep operators when human step into its moving trajectory. Kinect, a RGB-D sensor, which can provide not only color data but also three dimensions position information is an appropriate choice for obstacle detection. Nevertheless, multiple cameras providing comprehensive sights for robot arm to find the objects are used in the working space because there must be some blind angles when applying just single Kinect. After the position information is transformed to reference coordinate, the vector trajectory generator (VTG) system will calculate the repulsive vector and attracting vector of each joint to the obstacle. If the combination value of two vectors surpasses the expected value, the new moving trajectory will be computed in order to avoid the collision in a short time. However, the collision accident might happen if the robot’s velocity or acceleration is too swift no matter how rapid the collision avoidance algorithm responds. Therefore, to minimize the probability of collision, the deceleration model is implemented. By using background variation method, VTG can deaccelerate the speed of manipulator when any external obstacle is detected by dual cameras immediately.
In the experiment scenario, an industrial manipulator which has 16-kilogram payload is developed by our laboratory, NTU-iCeiRA. The task was to pick processing article up and place it on the platform. The linear speed of robot motion declined to original’s 50 percent while any external obstacles were detected by perception system. Also, collision avoidance algorithm was implemented by VTG when any object approached robot arm. The experiment result shows that VTG has the real time and extensive reaction to avoid collision happening by producing new reaction vector from the calculation of repulsive vector and attracting vector.

並列關鍵字

obstacle avoidance system ； sensor ； human-robot collaboration ； factory automation ； 6 DOF industrial robot manipulator

參考文獻

[1] Khatib, Oussama. "Real-time obstacle avoidance for manipulators and mobile robots." The international journal of robotics research 5.1 (1986): 90-98.

[2] S. G. Khan, G. Herrmann, T. Pipe, C. Melhuish, and A. Spiers, "Safe adaptive compliance control of a humanoid robotic arm with anti-windup compensation and posture control," in International Journal of Social Robotics, vol. 2, pp. 305-319, 2010.

[5] A. De Luca, F. Flacco, A. Bicchi, and R. Schiavi, "Nonlinear decoupled motion-stiffness control and collision detection/reaction for the VSA-II variable stiff-ness device," in IEEE/RSJ International Conference on Intelligent Robots and Systems. (IROS), St. Louis, MO, USA, Oct. 11-15, 2009, pp. 5487-5494.

[7] F. Flacco, A. De Luca, and O. Khatib, "Motion control of redundant robots under joint constraints: Saturation in the null space," in IEEE International Conference on Robotics and Automation (ICRA), St. Paul, MN, USA, May 14-18, 2012, pp. 285-292.

[8] P. K. Allen and P. Michelman, “Acquisition and Interpretation of 3-D Sensor Data from Touch,” in IEEE Transactions on Robotics and Automation, vol. 6, no. 4, pp. 99404, Aug. 1990.