输电线路螺栓紧固带电作业机器人的视觉搜索、识别与定位方法

发布时间：2018-12-17 06:56

【摘要】：在输电线路采用机器人进行螺栓的全自动带电紧固作业是一项非常有挑战性的工作,螺栓的自动紧固首先必须解决螺栓的自动搜索、识别与定位,由于线路环境复杂,这些工作变得十分困难,为此,提出一种新的螺栓视觉搜索识别定位方法,该方法分为两部分,基于参考物的螺栓追踪,通过设定引流线为参考物,先对引流线进行定位,然后沿着引流线方向来搜索螺栓,从而简化螺栓搜索过程,降低螺栓识别难度;基于改进Hough变换的螺栓识别算法,通过对经典Hough变换的峰值选择策略进行改进来实现螺栓的精确识别,然后利用螺栓头部圆形特征来完成螺栓中心的验证,并通过HOG和SVM技术来实现目标物体的识别分类,消除外界不相关物体对目标图像的影响,进一步提高识别精度。根据该方法,设计了机器人原理样机并进行了模拟测试和现场测试,测试结果表明,该方法能够高效地实现输电线路上螺栓的搜索、识别与定位,极大程度地提高了机器人的带电作业效率。
[Abstract]:It is a very challenging task to use robot to carry out the automatic live fastening of bolts in transmission lines. The automatic fastening of bolts must first solve the problems of automatic searching, identifying and locating of bolts, because of the complex environment of the lines. For this reason, a new method of bolt visual searching and locating is proposed. The method is divided into two parts: bolt tracking based on reference material, and the drainage line is located by setting the drainage line as reference object. Then the bolt is searched along the drainage line, which simplifies the bolt searching process and reduces the difficulty of bolt identification. The bolt recognition algorithm based on improved Hough transform is improved to realize the accurate identification of bolts by improving the peak value selection strategy of classical Hough transform, and then the bolt center is verified by using the circular feature of bolt head. The recognition and classification of target objects are realized by HOG and SVM technology, which can eliminate the influence of external objects on the target image and improve the recognition accuracy. According to this method, the prototype of the robot is designed, and the simulation test and field test are carried out. The test results show that the method can efficiently search, identify and locate the bolts on the transmission line. The working efficiency of the robot is greatly improved.
【作者单位】：长沙理工大学电气与信息工程学院;国家电网湖南省电力公司带电作业中心;
【基金】：国家自然科学基金(61473049)资助项目
【分类号】：TP242;TP391.41

【相似文献】