基于三点法的精密主轴径向回转误差分离技术研究

发布时间：2018-11-14 17:44

【摘要】：机床主轴是现代机械加工中的核心部件,主轴精度决定了工件的加工精度,因此,对主轴的研究就显得尤为重要。主轴空间振动最直接表现是圆度误差与回转误差,通过对以上两种误差的研究便能清晰把握一个机床主轴的运动状态。论文选择了三点法误差分离技术为理论基础进行研究,围绕提高圆度误差与回转误差分离精度展开论述。首先,本文对比分析了目前比较常见的误差分离方法,并最终选择了具有一定研究价值的三点法误差分离技术展开讨论。阐述了该分离方法的基本原理,分析发现权函数在三点法误差分离技术中具有重要意义,当权函数选择不合理时,可能导致谐波抑制现象的发生,从而产生错误的误差分离结果。其次,从数学分析的角度对权函数的特性进行了研究。围绕权函数,提出了改善误差分离精度的两个标准—全局性、局部性。依据以上两个标准,运用极大熵函数方将不可微优化目标函数转化为连续可微的优化函数。结合粒子群优化算法,优化得到了可预防谐波抑制现象发生的全局传感器安装角度。仿真效果表明:全局最优值很大程度上提高了误差分离精度。最后,针对现实情况下传感器噪声信号对误差分离效果影响这一问题,本文进一步提出了一套抑制传感器噪声信号的优化方法。该方法不再围绕权函数展开研究,而是将三点法误差分离技术中的权系数与权函数当作统一整体进行分析,确立了新的优化目标函数。再次结合粒子群优化算法,优化得到了噪声抑制下的传感器安装角度。理论分析与仿真结果表明,该角度下的误差分离精度仍能保持较高水平,并且该角度组合对传感器噪声信号具有良好的抑制效果。考虑到由于传感器安装角度偏差可能带来的误差分离精度下降和谐波抑制现象发生等问题,本文提出了克服此类缺点的方法,即区间优化。通过该方法的运用,有效解决了传感器安装过程中由于安装偏差所带来的误差分离精度下降问题。综上所述,本文围绕提高三点法误差分离精度展开深入研究。针对不同问题,优化得到了传感器安装角度全局最优值、噪声抑制最优值、区间最优值。最终,通过仿真分析和实验测试充分证明了理论研究的可靠性和优化结果的正确性。
[Abstract]:The spindle of machine tool is the core part in modern machining. The precision of spindle determines the machining accuracy of workpiece. Therefore, the research of spindle is particularly important. The most direct manifestation of spatial vibration of spindle is roundness error and rotation error. The motion state of a machine tool spindle can be clearly grasped by studying the above two kinds of errors. In this paper, the three-point error separation technique is chosen as the theoretical basis, and the improvement of the accuracy of roundness error and rotation error separation is discussed. Firstly, the common error separation methods are compared and discussed in this paper. Finally, the three-point error separation method with certain research value is chosen to discuss. The basic principle of the separation method is expounded. It is found that the weight function is of great significance in the three-point error separation technique. When the weight function is not reasonable, it may lead to the occurrence of harmonic suppression. This results in erroneous error separation results. Secondly, the characteristics of weight function are studied from the angle of mathematical analysis. Around the weight function, two criteria to improve the accuracy of error separation are proposed, namely, global and local. According to the above two criteria, the objective function of nondifferentiable optimization is transformed into a continuous differentiable optimization function by using the maximum entropy function. Combined with particle swarm optimization algorithm, the global sensor installation angle which can prevent harmonic suppression is obtained. Simulation results show that the global optimal value greatly improves the accuracy of error separation. Finally, in order to solve the problem that the sensor noise signal affects the error separation effect, a set of optimization methods to suppress the sensor noise signal is proposed in this paper. Instead of focusing on the weight function, the method analyzes the weight coefficient and the weight function in the three-point error separation technique as a unified whole, and establishes a new optimization objective function. Finally, the sensor installation angle under noise suppression is obtained by combining particle swarm optimization algorithm. Theoretical analysis and simulation results show that the accuracy of error separation at this angle is still high, and the combination of the angle has a good suppression effect on the sensor noise signal. Considering the problem that the error separation accuracy and harmonic suppression may be decreased due to the angle deviation of sensor installation, this paper proposes a method to overcome this kind of shortcoming, namely interval optimization. Through the application of this method, the problem of the error separation accuracy decline caused by the installation deviation during the installation of the sensor is effectively solved. To sum up, this paper focuses on improving the accuracy of three-point error separation. According to different problems, the global optimal value of sensor installation angle, the optimal value of noise suppression and the optimal value of interval are obtained. Finally, the reliability of the theoretical research and the correctness of the optimization results are fully proved by simulation analysis and experimental tests.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TG502.3

【参考文献】