CAJ | 학술논문

针对四自由度串并混联拟人空间机械臂本体，构建了基于运动控制卡模式的机器人软件和硬件系统。采用基于位置的控制结构，利用所求得的机械臂逆运动学方程，设计了由图像反馈和末端运动组成的视觉伺服控制系统。提出了一种基于SURF特征的静态目标识别算法，算法首先对目标图像提取SURF特征，并利用欧氏距离实现模板图像与目标图像特征点匹配，然后计算已匹配特征点的质心来获得目标的位置信息，最终实现了稳定的伺服定位。实验结果证明了系统的可靠性，并且证明了四自由度串并混联机械臂本体在实际运动的精确性和稳定性。
침대사자유도천병혼련의인공간궤계비본체，구건료기우운동공제잡모식적궤기인연건화경건계통。채용기우위치적공제결구，이용소구득적궤계비역운동학방정，설계료유도상반궤화말단운동조성적시각사복공제계통。제출료일충기우SURF특정적정태목표식별산법，산법수선대목표도상제취SURF특정，병이용구씨거리실현모판도상여목표도상특정점필배，연후계산이필배특정점적질심래획득목표적위치신식，최종실현료은정적사복정위。실험결과증명료계통적가고성，병차증명료사자유도천병혼련궤계비본체재실제운동적정학성화은정성。
For the four degrees of freedom serial-parallel personification space robot arm, a new system which is based on the robot motion control card is proposed. By using the location of control structure and IKP inferential of robot arm, the system which consists of image feedback and terminal movement is designed. With the SURF, a static target identification algorithm is conducted. Firstly, the SURF features of template image and target image are extracted and matched by the Euclid distance method. Then the coordinates of object are gotten by calculating controid of features matched, and a stable servo positioning system is performed finally. The experimental results prove that the system is effective, and the movement is accurate and stable.