柔性电路板智能对位装置的研究与开发

发布时间：2018-05-15 11:38

本文选题：柔性电路板 + 机器视觉　；参考：《浙江理工大学》2017年硕士论文

【摘要】：近年来随着智能化电子产品和设备的迅速发展,柔性电路板(FPC)的应用范围越来越广阔。在柔性电路板的生产过程中,对位是其关键技术环节,本文针对自动化对位的生产需求,进行了深入的研究。开发出了基于机器视觉的智能对位装置,包括机械、运动控制和视觉系统。在开发出的装置上进行了大量的实验,分析验证了装置的精度、速度以及稳定性。全文取得的研究成果如下:(1)针对柔性电路板的特性,设计了柔性电路板智能对位系统的总体方案,包括机械系统、控制系统、视觉系统三个子系统。机械系统设计了基于伺服电机和步进电机的三坐标平台,并针对FPC的柔软轻薄的特性,设计了真空吸附装置,完成FPC的抓取和放置,减少了对FPC的机械损伤;控制系统采用了运动控制卡对电机进行控制,高效精准的完成了对FPC的一系列对位动作;视觉系统对相机、镜头、光源及其照明方式进行了详细的研究,根据技术指标和装置相关内外环境,选择出了相机、镜头、光源的型号,结合照明方式进行了装置的整体搭建。(2)对采集图像进行了预处理,分析研究各种算法的优缺点,选择使用改进的中值滤波进行平滑去噪、迭代法进行区域阈值分割,并且使用形态学操作去噪,最后使用了Canny算子进行了边缘检测。(3)在图像定位算法中,针对霍夫圆拟合不精准,单个圆拟合随机误差较大的缺点,并根据FPC定位圆的特性,在不同半径区域内,结合最小二乘法进行了多个椭圆的同时拟合定位,取其平均值,大大提高了定位圆定位的稳定性和精准性;在FPC偏移角度定位中,使用累计概率霍夫检测直线,不仅执行效率高,而且可以直接检测到端点坐标,便于计算FPC的偏移角度。(4)设计了柔性电路板智能对位的软件系统,使用Sapera LT软件库和OpenCV函数库进行图像的采集和处理;在运动控制方面使用了API函数中的定时器,对不同运动动作设置相关标志位,通过定时器扫描运动状态标志位的改变来判断执行步骤,保证了系统稳定有序的执行。大量的实验验证表明,系统对位误差不超过:±0.06mm,角度误差不超过0.0016rad,速度为3s/chip。
[Abstract]:In recent years, with the rapid development of intelligent electronic products and devices, FPC (flexible Circuit Board) is more and more widely used. Alignment is the key technology in the production of flexible circuit board. An intelligent alignment device based on machine vision is developed, including mechanical, motion control and vision systems. A large number of experiments were carried out on the developed device, and the accuracy, speed and stability of the device were analyzed and verified. According to the characteristics of flexible circuit board, the overall scheme of intelligent alignment system of flexible circuit board is designed, which includes three subsystems: mechanical system, control system and vision system. The mechanical system has designed the three-coordinate platform based on servo motor and step motor, and according to the soft and thin characteristic of FPC, the vacuum adsorption device has been designed to complete the grasping and placing of FPC and reduce the mechanical damage to FPC. The control system uses the motion control card to control the motor, and completes a series of matching actions of FPC efficiently and accurately. The visual system studies the camera, lens, light source and its lighting mode in detail. According to the technical specifications and the relative internal and external environment of the device, the models of camera, lens and light source are selected, and the whole structure of the device is built in combination with the lighting mode. (2) the preprocessing of the collected image is carried out, and the advantages and disadvantages of various algorithms are analyzed and studied. We choose to use improved median filter for smoothing denoising, iterative method for region threshold segmentation, and morphological operation for denoising. Finally, Canny operator is used for edge detection in image location algorithm. Aiming at the shortcoming of inaccurate Hoff circle fitting and large random error of single circle fitting, and according to the characteristic of FPC positioning circle, in different radius region, combined with least square method, several ellipses are fitted simultaneously, and its average value is taken. The stability and accuracy of circular positioning are greatly improved. In FPC offset angle location, the cumulative probability Hoff is used to detect straight lines, which not only performs efficiently, but also directly detects the endpoint coordinates. The software system of intelligent alignment of flexible circuit board is designed, which uses Sapera LT software library and OpenCV function library to collect and process images, and uses the timer of API function in motion control. The relative sign bits are set for different motion actions, and the execution steps are judged by the timer scanning the changes of the moving state markers, which ensures the system to execute stably and orderly. A large number of experiments show that the system has no more than 卤0.06mm in position error, less than 0.0016radin angle error, and a speed of 3 s / r.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN41;TP391.41

【相似文献】