论文部分内容阅读
以PTT纤维碱的处理工艺提出1种编码后的线性回归法,解决正交设计L_(25)(5~6)实验中各因素的单项、交互项以及平方项对因变量的影响,还可以根据获得的线性方程,在变量的取值范围内,确定最佳的工艺条件和较优的操作范围。试验数据回归后的回代方程为:y=55.6776-15.301x_1-0.01262x_3-1.6561x_6+0.2783x_1x_6+0.005363x_3x_6-0.002801x_3x_3+0.009877x_6x_6,其中,y为减重率(%),x_1为碱浓度(mol/L),x_3为时间(min),x_6为温度(℃)将回归方程规划求解,确定较优的条件范围为:碱浓度:(1.01～1.125)mol/L,反应时间:(63～70)min,温度:81℃～90℃,预测的减重率为14.77%～16.72%。
A kind of coding linear regression method was proposed to deal with the PTT fiber alkali treatment process to solve the influence of single item, interaction term and square term on the dependent variable of each factor in the orthogonal design L_ (25) (5 ~ 6) experiment. According to the obtained linear equation, the optimal process conditions and the optimal operating range are determined within the range of the variables. The regression equation after the regression of the experimental data is: y = 55.6776-15.301x_1-0.01262x_3-1.6561x_6 + 0.2783x_1x_6 + 0.005363x_3x_6-0.002801x_3x_3 + 0.009877x_6x_6, where y is the weight loss rate (%), x_1 is the alkali concentration (mol / L), x_3 is the time (min), x_6 is the temperature (℃), the regression equation is solved. The optimum conditions are as follows: alkali concentration: 1.01 ~ 1.125 mol / L, reaction time: ~ 70) min, temperature: 81 ℃ ~ 90 ℃, the predicted weight loss rate of 14.77% ~ 16.72%.