基于量子遗传算法的圆度误差测量研究

Detecting Roundness Error Based on Quantum Genetic Algorithm

提出采用量子遗传算法,以提高圆度测量精度。首先用最小二乘法拟合获得建模数据中圆度图像的圆心坐标和半径;再通过圆度计算剔除不符合要求的圆度;然后用量子遗传算法进行多进制编码,量子旋转门非固定步长调整更新;最后给出圆度误差测量流程。实验仿真显示该算法获得了精确的测量数值,与三坐标测量机测量结果误差相差小于0.005 8 mm,半径相对误差小于0.19%,测量最大误差均在0.01%以内,同时最大误差波动比较平稳,测量不确定度比其它方法值较低。

In order to improve the roundness measurement accuracy, quantum genetic algorithm was proposed. Firstly, the center coordinates and the radius of the roundness image of the modeling data were established using the least square fitting. Secondly, unacceptable roundness data were rejected in the roundness calculation. Thirdly, multi-bit coding was completed using quantum genetic algorithm, and quantum revolving gate non-fixed step size adjustment and updating were implemented. Finally, the procedure for measuring the roundness error was presented. Simulation results indicated that the proposed algorithm produces accurate measurements, and when compared with the three-coordinate measuring machine, the error is within 0.0058mm, and the relative error of radius is within -0.19%, the maximum measurement errors all being within 0.01%.Further, the maximum error fluctuation is stable, and relative to other methods, the measurement uncertainty is lower.