论文部分内容阅读
分析了在小批量的产品生产方式下传统零件加工过程的统计过程控制方法的局限性；说明了基于自回归模型的简单最小二乘算法处理数据自相关性的可行性与不足之处；针对单批次小批量数据中随机噪声水平相对较高、而噪声统计特性没有充分展开的主要矛盾，借鉴自适应卡尔曼平滑算法对于观测噪声的一定抑制机制，提出将该类方法应用到小批量零件的生产加工的质量建模中来；最后，利用实验数据对以上提出的各种方法进行了比较研究，充分说明了简单最小二乘算法的可行性、自适应卡尔曼平滑算法的有效性
The limitations of the statistical process control method for the processing of traditional parts in the small-batch production mode are analyzed. The feasibility and deficiencies of the autocorrelation of the data based on the autoregressive model are illustrated. The random noise level in batch small batch data is relatively high, but the statistical characteristics of noise are not fully developed. Based on the restraining mechanism of adaptive Kalman smoothing algorithm for observation noise, this method is proposed to apply to low-volume parts Production and processing of quality modeling; Finally, the use of experimental data to compare the various methods proposed above, fully demonstrated the feasibility of the simple least squares algorithm, the effectiveness of adaptive Kalman smoothing algorithm