logoxgvb
Joined: 30 Aug 2010
Posts: 1580
Read: 0 topics
Warns: 0/5
Location: hdfttv
|
 Posted: Wed 3:14, 15 Dec 2010 Post subject: ugg boots günstig Sampling Inspection and Its App |
|
|
Sampling Inspection and Its Application
; 100% of the bad batch is still less than B greater than zero probability of being sentenced as qualified to receive grant, which received the bad batch of false positives is not as qualified as the second batch of the wrong type of error, also known as \the false is true \B the greater the greater the likelihood of a miscarriage of justice. Such errors to the user interests. Obviously, the factory hopes a smaller the better, you want to 8 as possible. Usually we set a factory and the user can receive a and B values. By limiting the interests of a way to protect the factory; with limited ways to protect the interests of the user S. Usually take a equal to 5% or 1%, taking S equal to 1O%, 5% or 1%. So we made these two types of error probability is very small, only 5% or 10%. Therefore, we are almost certain to judge by the grasp. (Continued on page 16) problems. Therefore, the definition of a new feature vector K. = [? (K), (K), ... one by one fL (K)], that it by the D (K) of the L items (except the constant), with the D (K.) Can be regarded as linear discriminant function. This new space is located in the K, the pattern class can use the linear discriminant function D (K) apart. . 4 troubleshooting steps nonlinear polynomial discriminant function diagnosis procedure is almost similar to the linear discriminant function method, as follows: (1) before the test to determine the required pattern of need to determine the type, select the characteristic parameters, the correct division of feature space, and converted into a new spatial linear separability, construct discriminant function, the coexistence of one computer. (2) O, the amount of characteristic parameters I l 【l,[link widoczny dla zalogowanych], k2, ... ... k blood, and the data for pretreatment. (3) The calculation of the various discriminant function after the test and compare the value, if D. (K) = max {Di (K) j, then the test system should be classified pattern class. 5 Conclusion As long as the various features of the measured object has the same dimensionless parameters and characteristics of non-linear time-space, you can use the proposed diagnosis method for fault diagnosis. Compared with the linear discriminant function,[link widoczny dla zalogowanych], this method can also be components, components, functional blocks of the fault detection and isolation, and examination according to successive changes in the nonlinear discriminant function to predict the trends,[link widoczny dla zalogowanych], characteristics of disadvantage is divided space and the conversion feature space requires more complex analysis and calculation, but also because of characteristics of spatial patterns can not be the complete set, it will result in the omission of certain failure. But using this method on a class of complex fault diagnosis of analog circuits to achieve certain effects. References 1 high tin Chun, Zhou Yufen. Analog Lu fault diagnosis. Beijing: People's Liberation Army Air Force Engineering Institute, 1986: Xie Ping Zongzhe a 332. Xiang-Jun. 4 Principles of pattern recognition. Xi'an: Xi'an University of Electronic Science and Technology Agency rebels out I pick ,1993:8-14c on page 36 of Volume 5 instances of Example 6 when N = 200,[link widoczny dla zalogowanych], acceptable quality level AQL = 0.4, by checking the level of Ⅱ, a sample of normal program requirements . Solution: According to GB2828-2829-87 Table 2 (sample size code) from a given volume N = 200 rows and check the level Ⅱ where the line intersection was the sample size code ... G, the national standard of table 3 (normal Check a sample program) where the line of code G with AQL = 0.4 is located at the intersection of the column was = 0, Re = 1, G in the right code was the sample size n = 32. Therefore asked for a sample of normal program (200,32,0). Example 7 when N = 2O0O, AQL = 1.5, with the inspection level I, find the second sample of normal inspection program. Solution: Same as Example 6 from the N = 2O0O and level I in GB Richard Table 2 Sample size code \・ l6 ・ [0,3,3,4], in Table 7 \0, Rl = 3; n2 = 32,[link widoczny dla zalogowanych], = 3, = 4. Reference 1 Third Department of 301 machines. total quality management foundation, 19 ~ o02 coral Northern Province, the Bureau of Standards. count Adjusting Sampling Principles and Methods , 19833 Friends of Fu Lin. Pina quality of statistical control methods. Enterprise Management Press, 19804 Liu Xun and so on. Machinery manufacturing inspection techniques. National Defense Industry Press, 19875GB2828 ~ 2829-87. batches Check Sampling procedures and sampling. cycle checkpoint count Sampling procedures and tables. Standards Press of China, 19,886 Chinese Academy of Sciences Institute of Mathematics and Statistics Group. commonly used methods of mathematical statistics. Science Press, Chinese Academy of Sciences Institute of Mathematics, Probability and Statistics 19,797 rooms. commonly used mathematical statistics. Science Press, 1979
相关的主题文章：
[link widoczny dla zalogowanych]
[link widoczny dla zalogowanych]
[link widoczny dla zalogowanych]
The post has been approved 0 times
|
|
