【作者】 王睿；

【导师】 李建华； 班剑锋；

【作者基本信息】 兰州理工大学 ， 机械工程（专业学位）， 2021， 硕士

【摘要】 铝及合金棒材是有色金属铸造生产的重要产品,可通过二次加工来制作各类型材,被广泛运用于各大工业领域。铝及合金棒材生产信息的标识是生产管理信息化与智能化的重要支撑,激光标刻作为兴起的直接部件标识技术,以其非接触、零耗材、低功耗、低成本等优势,已经成为很多行业内首选的物料标识技术。实现基于激光标刻的铝及合金棒材生产信息标刻应用难,主要问题包括:棒材堆放空间位置不固定,端面空间定位难;棒材端面不均匀、高反光,激光标刻效果难以保障。针对以上问题,本文完成了以下研究工作:(1)对工业生产环境下的棒材标识进行分析,明确了棒材激光标识的相关难点,确定了开发专用激光自动标刻机器人的必要性和可行性。(2)提出了基于六轴臂式工业机器人的驱动平台,在遵循机械设计原则的前提下对标刻平台进行整体和分块设计,对相关设备进行参数计算和选型,并对关键零部件进行必要的应力和模态分析。(3)提出了基于视觉伺服的空间定位与找正方法,对实际工作环境中影响标刻定位精度的主要因素进行试验及分析,设计了基于卷积神经网络的标刻面特征识别的定位算法,并通过找正算法实现了激光的自动聚焦。(4)在各工艺参数下对铝合金棒材端面进行单因素标刻实验和正交试验,建立多元非线性回归方程。利用差分进化法进行模型参数拟合,得到激光控制参数与标刻图像灰度值之间相关性的方程。同时,通过手眼校准和相机标定,减少空间定位与找正误差。(5)设计并开发激光标刻控制平台软件,完成了激光标刻平台试验样机的实验,对实验结果进行分析和展示,结果显示试验样机达到了预期的目标。更多还原

【Abstract】 Aluminum and alloy bars are important products produced by non-ferrous metal casting.Various types of materials can be produced through secondary processing and are widely used in various industrial fields.The identification of aluminum and alloy bar production information is an important support for the informatization and intelligence of production management.As a direct component identification technology that has emerged,laser marking has achieved non-contact,zero consumables,low power consumption,and low cost advantages.It has become the preferred material identification technology in many industries.It is difficult to realize the application of laser marking-based aluminum and alloy bar production information marking.The main problems include: the position of the bar stacking space is not fixed,and it is difficult to locate the end face space;the end face of the bar is uneven and highly reflective,and the laser marking effect is difficult to guarantee.In response to the above problems,this article has completed the following research work:(1)Analyzing the bar marking in the industrial production environment,clarifying the related difficulties of bar laser marking,and determining the necessity and feasibility of developing a dedicated laser automatic marking robot.(2)A drive platform based on a six-axis arm industrial robot is proposed.Under the premise of following the principles of mechanical design,the marking platform is designed as a whole and in blocks,parameter calculation and selection of related equipment,and key components Perform the necessary stress and modal analysis.(3)A spatial positioning and alignment method based on visual servoing is proposed,and the main factors that affect the accuracy of marking positioning in the actual working environment are tested and analyzed,and a positioning algorithm based on convolutional neural network for facet feature recognition is designed.,And realize the automatic focusing of the laser through the alignment algorithm.(4)Carry out single factor marking experiment and orthogonal experiment on the end face of aluminum alloy bar under various process parameters,and establish multiple nonlinear regression equation.The differential evolution method is used to fit the model parameters,and the correlation equation between the laser control parameters and the gray value of the marking image is obtained.At the same time,through hand-eye calibration and camera calibration,the spatial positioning and alignment errors are reduced.(5)Designed and developed the laser marking control platform software,completed the experiment of the laser marking platform test prototype,analyzed and displayed the experimental results,and the results showed that the test prototype reached the expected goal.更多还原