计算机辅助提高内圆磨削质量

Computer-Aided Improvement on Internal Grinding Quality

本文为了利用计算机实时控制和预报轴承内环加工误差,针对轴承内环磨削加工的精度和速度要求,建立了基于Marple算法的AR(n)模型,以及离散勒让德多项式模型.依此研制的微机动态统计控制系统用于现场试验,显著减少了轴承内环加工的返修品和废品.

Based on an analysis of the grinding quality of bearing inner rings, a mathematical model for workpiece machining errors is developed. With this model, the machining errors of the following workpiece can be forecasted so that the grinding size undulation caused mainly by the inaccuracy and non-stability of the active measuring instrument and feed system, by the tolerance dispersiveness of the preceding process and by uneven wheel wear can be well controlled. Focus of attention is on the Time Series AR(n) model and the Discrete Legendre Polynomial L(m,n) model on their forecasting accuracy, modelling speed and real-time control possibility. Theoretical analysis and experimental results show that the Time Series AR (n) model founded by Marple algorithm and the Discrete Legendre Polynomial L(m,n) model are suitable for controlling the internal grinding process on real time. A Dynamic Statistical Control System for grinding inner rings of types 203, 205, and 206 bearing on a type MZ208 automatic internal grinder is designed. On-site experiments show that the control system can reduce the process rejection rate from 2%~3% to 0.5% and the repair rate from 3%~5% to1%.