神经网络优化非水相脂肪酶催化合成抗坏血酸油酸酯的条件研究

Optimization of Lipase Catalyzed Synthesis of Ascorbyl Oleic Acid Ester by Neural Network

抗坏血酸油酸酯具有强抗氧化作用.为了获得脂肪酶催化合成抗坏血酸油酸酯的最适条件,主要研究了反应温度、脂肪酶量、油酸量对抗坏血酸油酸酯合成效果的影响.采用中心组合设计和动量梯度下降神经网络对反应条件网络进行训练仿真,并利用训练好的网络对催化酯化工艺条件进行预测.研究结果表明:经过训练的网络可以很好的模拟反应条件,得到了脂肪酶催化反应的最佳工艺参数.当抗坏血酸0.8g时,反应温度56℃,油酸量0.95g,固定化脂肪酶量0.74g,添加分子筛条件下,抗坏血酸油酸酯的转化率为46.5%.该方法为抗坏血酸酯化催化效果的预测提供了一条可行的途径.

Ascorbyl oleic acid ester is a good kind of antioxidant. The effects of temperature, oleic acid amount and enzyme amount on conversion rate were carried out for obtaining optimal parameters. The simulation of lipase catalyzed synthesis of ascorbyl oleic acid ester based on central composite experiment and back-propagation algorithm with momentous factor was carried out to train the neural network followed by prediction of synthesis reaction. The results showed that lipase reaction was well simulated and optimal technological conditions of lipase reaction could be obtained by the trained neural network. When the ascorbic acid of 0.8g, reaction temperature of 56℃, oleic acid of 0.95g and lipase amount of 0.74g in the presence of molecular sieve, conversion rate of ascorbyl oleic acid ester reached 46.5%. It provided a feasible approach for the prediction of synthesis of ascorbyl fatty acid ester.