呙倩 于宝成 徐文霞

摘要：为解决工业生产线灰尘清理以及产线过度磨损的问题，本文研发设计了一种基于图像检索算法的产线智能吸尘机器人。首先设计了产线吸尘机器人的机械结构，其次，为提升其吸尘效率，基于摄像头获取的实时传送带监控图像，采用小波变换和GLCM算法检索图像纹理特征，使用根据距离度量法计算图像的相似度，相似度越低，吸尘口风速越大，然后，控制直流电机的转速调整吸尘口风速。经实际产线环境测试实验，吸尘机器人的底座能与产线无缝配合，沿传送带自动吸尘。此外，将图像处理技术应用于产线上，把吸尘效率作为吸尘的量化评价指标，产线吸尘机器人在传送带上吸尘可以达到93%以上的吸尘效率。

关键词：产线吸尘机器人;图像处理;灰尘检测;灰度共生矩阵;小波变换

Abstract： In order to solve the problems of dust cleaning and excessive wear of industrial production line， an intelligent dust cleaning robot based on image retrieval algorithm is developed and designed in this paper. Firstly， the mechanical structure of the production line vacuum cleaning robot is designed. Secondly， in order to improve its dust collection efficiency， the wavelet transform and GLCM algorithm are used to retrieve the image texture features based on the real-time monitoring image of the conveyor belt obtained by the camera. The similarity of the image is calculated by the distance measurement method. The lower the similarity is， the greater the wind speed of the dust suction port is. Then， the speed of the DC motor is controlled to adjust the dust suction port Wind speed. According to the actual production line environment test， the base of the vacuum cleaning robot can cooperate with the production line seamlessly and automatically dust along the conveyor belt. In addition， the image processing technology is applied to the production line， and the dust collection efficiency is taken as the quantitative evaluation index of dust collection. The dust collection efficiency of the production line vacuum cleaning robot on the conveyor belt can reach more than 93%.

Key words： production line vacuum robot; image processing; dust detection; gray level co-occurrence matrix; wavelet transform

世界已经进入了工业4.0时代，主要目标是实现无人化工厂以及工业设备智能化。实际工厂中，工业上的互感器检测线工作一段时间后，上面会产生大量的灰尘，从而影响继续生产，而灰尘积累过多会损伤产线，传送带损伤情况如图1所示。为了正常地进行生产，需要清理传送带上的灰尘。通常依靠人力来清洁，不仅难度比较大，时间长，没有实现自动化吸尘，还会造成产线严重磨损。研发产线智能吸尘机器人，将会节省大量的人力物力，能够在线吸尘，实现智能化生产，不会造成产线损伤，同时，根据传送带灰尘和碎屑含量来控制机器人的吸尘力度，这样能提高产线智能机器人的效率，节约能源，并提高经济效益。……