概率统计及SPC基础.ppt

* * * * * * * * * Discuss the difference between the Sample and Population statistics. We will discuss the importance of these later. Example: Avg. Height of Class Vs. USA avg ht. Measuring class is easy and can get a good measure of avg height. Measuring everyone in USA is tough. Want to predict the average height based on a significant sample size. The question is how confident you are in your estimate based on the sample size, measurement system, etc. * Example: House prices People looking at houses and find 4 houses in the area priced as follows: $115K, $80K, $100K, $175K. Avg and Median pretty close together ($112.5K). Add 1 more house at $12.8M. Mean is greatly shifted upwards while median stays at $115K. If you have even # of values, median is the average of the center 2 values. Median is less affected by shift in values, mean is greatly affected by shift in values. Median is a very poor measure of Central Tendency. * 11 Traditional range based sample standard deviation estimation table: n d2 . 2 1.128 3 1.693 4 2.059 5 2.326 6 2.534 7 2.704 8 2.847 9 2.970 10 3.078 ___________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ * * * * * * * * * * * * * * * * * * * * * * * * * * 計算Ca,Cp,Cpk公式 規格中心m LSL + 3 ? - 3 ? 製程寬度6 ? 規格寬度T USL Su SL Cp: Capability of Precision精確度: 實際中心 - X m X 當僅有下限時:Cp = ( -SL)/(3σ) 對雙邊規格: Cp = T/(6σ) 當僅有上限時: Cp = (Su- )/(3σ) X X 等級 Cp值 A Cp≧1.33 B 1.00≦ Cp1.33 C 0.67≦Cp1.00 D Cp0.67 分級標準如下: 主值 計算Ca,Cp,Cpk公式 Cpk: 指制程能力參數, 是Cp和Ca的綜合. ? 對雙邊規格: Cpk=(1-│Ca│)*Cp= Min[(Su- )/(3σ), ( -SL)/(3σ)] 對單邊規格, 可以認為T為∞, 則 Ca= ( -μ)/ (T/2)= 0 Cpk= (1-│Ca│)*Cp= Cp 等級 Cpk值 評价 A Cpk≧1.33 理想 B 1.00≦Cpk1.33 正常 C Cpk1.0 不足 分級標準如下: X X X SPC介紹 SP