自动线典型失控状态随机信号的模拟分析与研究

The Simulate Analysis and Research of Automatic Line Typical Incontrollable State Random Signal

使用控制图对在线零件加工质量进行检测是一种十分经济的质量控制手段。但在随机抽样检验时也不可避免地要冒两类误判的风险。为减少此类误判事件的发生,对典型在线加工失控随机信号进行多次的重复模拟分析与研究就显得十分必要。用高斯白噪声模拟失控随机信号应该说是最接近现场实际情况的。但高斯白噪声产生的随机数序列不具备可重复性,这就给分析与研究工作带来不便。运用混沌学研究的成果,选取恰当的混沌数来构造自动线典型失控随机信号,就成为进行在线质量监控动态模拟过程中防止误判的关键。

Using control charts for on-line machining quality testing is an economical means of quality control. However, random sampling tests inevitably bear the risk of two types of false positives. To reduce the incidence of such false positives, it is very necessary to do simulation analysis of on- line process control typical random signal at the repetition . That analog control with Gaussian white noise random signal is closest to the actual situation. But the Gaussian white noise sequence of random numbers do not have the repeatability, which is not convenient to analyze. Using research results of chaos and selecting the appropriately chaotic number for constructing a typically automatic line out of control random signal is the key to prevent miscarriage of justice in the process of on-line quality control dynamic simulation.