Research and analysis on the preparation of C4 olefin by ethanol coupling method based on regression algorithm with computer technology

Abstract

Computer technology has been used to study the different combination of catalyst and temperature on the ethanol conversion (one-way conversion rate of 100% per unit time ethanol (alcohol intake - ethanol residue)/ethanol intake) and C4 olefin selectivity (the proportion of all products of C4 olefin), it is necessary to decompose the catalyst composition into several indicators. The effects of several indexes on ethanol conversion and C4 olefins selectivity were analyzed by computer technology. Canonical correlation analysis was used to study them, and the coefficient before each component represented the degree of importance. The results showed that the higher the catalyst combination temperature, the higher the weight of hydroxyapatite, the higher the ethanol conversion and C4 olefins selectivity, and vice versa. When the C4 olefin yield (its value is the selectivity of C4 olefin to ethanol conversion) is as high as possible, the multiple linear stepwise regression function of catalyst parameters and temperature on C4 olefin yield is obtained. MATLAB software was used to calculate the function formula. The absolute value of coefficients determines the importance of each index to C4 olefin yield. Combined with chemical reaction, when the mass ratio is 200:200 and the temperature is 400℃, the yield of C4 olefins can be maximized, and the yield of C4 olefins is 44.728%. Through the establishment of decision tree regression model by computer technology, it can be intuitively concluded that when the temperature is lower than 350℃, the best experimental condition is the C4 olefins yield of 17.264%.