基于激光断面扫描卧式罐所获取的点云数据,提出了一种卧式罐倾斜比计算方法。该方法建立在卧式罐筒体为近似圆柱体的基础上,通过拟合筒体上点云数据得到轴线方程,从而得到卧式罐倾斜比计算方法。该方法首先给出了点云数据的预处理方法,通过基于曲率算法和对称性原理,将单层数据中的圆弧段与直线段进行分离;通过基于最小二乘法的直线拟合,剔除了由于附件引入的干扰点,以此获取稳定的圆柱拟合参数;然后,利用Levenberg—Marquardt非线性最小二乘法,拟舍得到圆柱体参数方程,最终换算得到卧式罐倾斜比的精确计算。实验证明,该方法运算时间短,鲁棒性强,与手工测量相比,角度偏差小于0.01°。
A new method of horizontal tank tilt ratio based on cloud data by laser scanning is presented. Horizontal tank can be supposed as approximate cylinder. The equation of the cylinder axis can be gotten by fitting the cloud data. First, pretreatment method based on the curvature algorithm and the symmetry principle was presented to separate line segments and arc segments from single layer data. The interference of attachment points had been excluded through linear least squares fitting in order to get some stable cylindrical fitting parameters. Then, Levenberg-Marquardt nonlinear least-squares method was applied for fitting to get cylinder parametric equation. Finally, horizontal tank tilt ratio was obtained by exactly calculating. Compared with manual measurement, experimental results show that this method has fast computing speed and strong robustness, and angle deviation was less than 0.01 degrees.