www.heavensgatel2.fora.pl

[logoxgvb]

Rate limits based on statistical analysis to determine quality control


Stable at: l, the sample n is equal to 4, the requirements of a, p were 0.25%, 3.6%, can meet a, p range of control limits, k = 2.8. According to the impact of p, in determining the sample size and sample interval should take into account the efficiency of sample size, that under certain conditions the total sample, is to select a small sample of small intervals, or select a large sample of large intervals. The purpose of sampling procedures in order to detect anomalies, and thus sampling plan must consider the impact of 8. For example: a product production process,[link widoczny dla zalogowanych], when given two options, each 1 hour n: 4, n = 16 per 4 hours Author: Wang teaches East 0965. ), M, Shou County, Anhui, lecturer, is mainly engaged in military training simulation of motor transport. * 62 * for a given k: 3,: 1.5 time; n = 4 solution, p = (3-1.5 × 2) a (a 3-1.5 × 2) = 0.5; n = 16 time,[link widoczny dla zalogowanych], p = (3-1.5 × 4) a (a 3-1.5 × 4): 0.00135 Thus: n = 4 n = 16 program to achieve the accuracy of the program, the sample size is quite O. 59 ~ = 0.00135, that is, 4 × 9.5:38 a sample size, n = 16 sample size savings program. However, the loss during the run, set up units in loss of A,[link widoczny dla zalogowanych], the program found a small interval in time, can reduce waste during the loss, n = 4 savings lost during the program expectations 2.95A; n = 16 program expected during the loss of 3 .15 A. When the sample during the sampling costs and losses to the timing, you can choose the best program. Thus, the control range limits the effect of statistical control has a greater impact. Quality control in a number of practical problems, we want to control the first category in addition to the probability of wrongful convictions, often also hope that based on the requirements of the problem,[link widoczny dla zalogowanych], given significant control pre-commit the probability of the first kind of wrongful convictions, and guilty of second mistake the probability of contracting the capacity is dependent on the choice of the sample. Here, we discuss how to select the sample size makes the probability of committing Type II error control in the pre-specified limits. Definition: If c is the parameter 0 of a test problem of a test method: p (0) = Po (accepted H said the implementation of test methods characteristic function c (OC function). For the normal population mean of the test method of the OC function, described as follows: the quality inspection of the right questions Ho:: 0Hl:> 0 The OC function is) to accept a phoenix )={> .1-t +8 (8> 0 is taken constant),[link widoczny dla zalogowanych], the probability of committing Type II error of not more than a given. For example: with a large number of products, product quality indicators XN (, d2), who preferred a small, factory inspection program requirements as determined by high-quality products (≤) be a high efficient one for the buyer of a accepted. buyer requires low-quality products (≥ + can with high probability 1-p is rejected. a, p is given by the factory in consultation with the buyer, and take a sample to determine whether the product is accepted by the buyer, Q. How should arrange sampling plan, known Po = 120,8 = 20, and long experience from the factory to know: d2 = 900, and agreed that a, p are 0.05. The test problem can be expressed as: Ho: ≤ Po, Hl: ≥ Po + domain is still Susan Chan rejected: ≥ OC function is therefore:) accept {<)={- ta/4n <a) where / 4n as ≥, uo + when p () ≤ p, because p ( ) is a decreasing function, it just p (0 +) = p, namely: science, the data considered by the given n ≥ 24.35, so they chose n: 25, and when Zhu ÷: Z0.05: 1. 645, that is, when the buyer refused to 129.87 of these products, and when <129.87, the buyer to accept the quality of these products. References: [1] Zhou Mengqing. Probability random variables and random inference [M]. Beijing: Beijing Aerospace University Press, 1989. [2] Wei Zong Shu. Probability and Mathematical Statistics [M]. Beijing: People's Education Press, 1979. [3] Hou set Pi. Exploration of the theory of nonlinear evaluation [M】. Hefei: China University of Technology Press, 2001. [Received Date] 2002-07.09 [Revised] 2o03. . 03.27

相关的主题文章：


[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]