School of Mechanical Engineering, Shenyang University of Technology, Shenyang 110870, China.
Sensors (Basel). 2019 Nov 20;19(23):5059. doi: 10.3390/s19235059.
Great length, large weight and other factors may cause difficulty in measuring the profile accuracy of the double-headed screw rotor. To solve this problem, an on-machine measuring system based on a laser-displacement sensor (LDS) was designed and implemented in this paper by taking an LXK100 four-axis whirlwind milling machine as the carrier. To improve the measurement accuracy of the system, the generalized variable-structural-element morphological method, polynomial interpolation algorithm and ellipse fitting method were first combined to realize the rapid subpixel centroid extraction from a noise-containing spot image, thus improving the data acquisition accuracy of the LDS, and then the hybrid method was experimentally verified. Next, a wavelet threshold function with high-order differentiability and adaptive wavelet coefficient contractility was constructed based on the hyperbolic tangent function, so as to inhibit the disturbance from random errors and preserve real profile information, and this method was simulated and verified. Subsequently, a smoothing algorithm for point cloud data was proposed based on the Lagrange multiplier method to avoid the defect of the piecewise curve-fitting method, that is, function continuity and differentiability could not be satisfied at piecewise points. Finally, the profile accuracy was calculated in real time according to the data reconstruction result and the machining quality was judged. The measurement experiment of the double-headed screw rotor indicates that the proposed on-machine measuring system can complete the profile accuracy measurement for a screw pitch within 39.7 s with measurement accuracy reaching ±8 μm, and the measurement uncertainties of the major axis, minor axis and screw pitch are 0.72 μm, 0.69 μm and 1.24 μm, respectively. Therefore, the measurement accuracy and efficiency are both remarkably improved.
长度大、重量大等因素可能导致双头螺杆转子的轮廓精度测量困难。为了解决这个问题，本文以 LXK100 四轴旋风铣床为载体，设计并实现了一种基于激光位移传感器（LDS）的在位测量系统。为了提高系统的测量精度，首先结合广义变结构形态学方法、多项式插值算法和椭圆拟合方法，实现了含有噪声点图像的快速亚像素质心提取，从而提高了 LDS 的数据采集精度，并进行了实验验证。然后，基于双曲正切函数构建了一种具有高阶可微性和自适应小波系数收缩性的小波阈值函数，以抑制随机误差的干扰并保留真实轮廓信息，并进行了模拟验证。随后，提出了一种基于拉格朗日乘子法的点云数据平滑算法，避免了分段曲线拟合方法的缺陷，即在分段点处不能满足函数的连续性和可微性。最后，根据数据重构结果实时计算轮廓精度，并判断加工质量。双头螺杆转子的测量实验表明，所提出的在位测量系统可以在 39.7s 内完成一个螺距的轮廓精度测量，测量精度达到±8μm，其长轴、短轴和螺距的测量不确定度分别为 0.72μm、0.69μm 和 1.24μm。因此，测量精度和效率都得到了显著提高。
