研究方法与论文写作 第六章 响应面法试验设计（回归设计）.ppt

* * The second step, we need to choose a suitable uniform design table. The uniform designs are tabulated, and each table has a notation like this: (the word in the middle appears), U stands for uniform design, n stands for the number of experiments, q stands for the number of levels, and s stands for the maximum number of factors. * * * First step, the experimenter carefully chooses factors and experimental domain, and determines suitable number of levels for each factor. In this example, the experimenters chose three factors: (table shows) x-one, the ratio of raw materials, x-two the amount of pyridine, x-three, the time length of reaction, and for each factor, they chose 6 levels as in the table. * For example, (table 2.3 appears) U-seven, seven-to-the-fourth, represents a uniform design table of 7 experiments, four factors and each factor consists of seven levels. Or for example, (table 2.4 appears) U-nine, nine-to-the-fourth, represents a uniform design table of nine experiments, four factors and each factor consists of nine levels. * The uniform design software will advise the user how to choose the suitable columns. For example, for U-seven, seven-to-the-fourth, we have a relative discrepancy table like this. (The first table appears) Using the previous example, (table 2.3 appears) U-seven, if we have only three factors in our experiments, we can look up the discrepancy table, (the blue box appears) and it suggested using column one, two and three of the uniform design table.(the red box appears) and therefore we have the design like this: (the arrow moves from left to right and the blue table appears). * After choosing the uniform design table, the next step is to make an experimental plan with this table. (the table appears) First, put x-one, x-two and x-three into column one, two and three. (the x’s appears) Replace one to seven of the first column by the seven levels of factor x-one, the ratio of raw materials (the first column changed)