基于卡尔曼滤波的焊缝跟踪技术研究  

Study on seam tracking technique based on the Kalman Filtering

作　　者：陈建辉[1] 高向东[1] 刘明涛[1] 丁度坤[1] 赵传敏[1] 

机构地区：[1]广东工业大学机电工程学院,广东广州510006

出　　处：《焊接技术》2008年第1期11-12,共2页Welding Technology

基　　金：国家自然科学基金资助项目(60375012);广东省自然科学基金资助项目(07001764;6021444)

摘　　要：运用计算机视觉传感方法获取熔池图像,并提取熔池边缘,确定熔池形心。以熔池形心为描述焊缝偏差的特征矢量,建立熔池形心状态方程和测量方程。应用卡尔曼滤波将熔池形心预测值和测量值进行优化,抑制过程噪声和测量噪声,得到最小均方差条件下的熔池形心最优估计值,实现弧焊过程中焊缝的精确跟踪。The Kalman Filtering was used to improve seam tracking accuracy.Computer visual sensor and digital image processing methods were applied to acquire the weld pool image,and then the edge of weld pool was attracted to determine the centroid.A state equation and a measurement equation based on the welding pool centroid which is considered as the measurement eigenvector of the weld position error were constructed.The advantages of Kalman Filtering on optimizing the estimate and measurement,restraining the process and measurement noise were taken to get the optimum weld pool centroid under the least mean square error.Experiments prove that this method has good performance in improving the seam tracking accuracy.

关 键 词：图像处理 卡尔曼滤波 焊缝跟踪 

分 类 号：TG409[金属学及工艺—焊接]