Abstract: An online complex tool path smooth compression algorithm was proposed based on interpolation curve fitting method. The original data points were preprocessed off-line according to the selected strategy of the dominant points， and then followed by an online interpolation curve fitting of the dominant points and non-dominant points-- error detection， and a B-spline curve was generated to meet the accuracy requirements. The dominant points were selected according to the curvature threshold， the curvature maximum value， the curve inflection point and the error maximum points after Bezier curve fitting for the dominant points. A length equalization strategy was proposed to extract the length mutation points as the new dominant points to ensure the accuracy of the local Bezier curves fitting with C2 continuity. After B-spline curve interpolation fitting of the dominant points， a contour error following method was applied to detect the errors of non-dominant points， which was more faster and effective than Newton iteration method. The simulation results show that the proposed algorithm herein may smooth and compress the complex tool paths， and the accuracy of the error detection method is satisfied.

Key words: dominant point, interpolation curve fitting, Bezier curve, contour error following method

摘要： 提出了一种基于插值拟合的在线复杂刀具轨迹平滑压缩算法，该算法依据主导点的选取策略，对原始数据点进行离线预处理，然后进行主导点的在线插值拟合以及非主导点的误差检测，进而生成一条满足拟合精度要求的B样条曲线。主导点依据离散数据点的曲率阈值、曲率极大值、曲线拐点，以及分段Bezier曲线逼近拟合后的误差最大值点进行选取。在具有C2连续性的分段Bezier曲线逼近拟合前，需要利用长度均分策略，提取长度突变点作为新增的主导点，以保证拟合的准确性。对主导点进行B样条插值拟合后，利用轮廓误差跟随法对非主导点到拟合曲线的误差进行检测。该方法与牛顿迭代法相比，其计算速度更快且能提高算法效率。仿真结果表明，提出的算法可对复杂刀具轨迹进行平滑压缩，且误差检测的精度能够满足要求。

关键词: 主导点, 插值拟合, Bezier曲线, 轮廓误差跟随法