人工神经网络和响应面法优化薏苡仁酒发酵条件

采用Box-Behnken试验设计对薏苡仁酒的发酵条件进行优化,并对Box-Behnken（BB）试验结果分别进行响应面法（RSM）和人工神经网络（ANN）分析。结果表明,RSM、ANN优化发酵条件分别为薏苡仁∶糯米为1∶2（g∶g）、酵母A1接种量为4.7%、温度为31.7 ℃、初始pH为3.0;薏苡仁∶糯米为1∶1.9（g∶g）、酵母A1接种量为4.2%、温度为28.1 ℃、初始pH为3.0,ANN、RSM分别在其最优条件下的实际值和预测值都基本一致。ANN、RSM拟合模型的相关系数（R）、决定系数（R2）、均方误差（MSE）、均方根误差（RMSE）、平均绝对误差（MAE）、平均绝对百分比误差（MAPE）分别为0.994 5、0.988 9、0.011 7、0.108 4、0.072 2、0.486 3%和0.983 6、0.967 5、0.028 9、0.170 1、0.143 7、0.985 7%。ANN具有更高拟合能力和准确性,拟合效果更好,更适合应用于薏苡仁酒发酵条件优化。

Optimization of fermentation conditions of coix seed wine by artificial neural network and response surface method

The fermentation conditions of coix seed wine were optimized by Box-Behnken(BB)experiments,and the results were analyzed by response surface methodology(RSM)and artificial neural network(ANN).The results showed that the optimal fermentation conditions optimized by RSM were coix seed:glutinous rice 1∶2(g∶g),yeast A1 inoculum 4.7%,temperature 31.7℃,and initial pH 3.0.The optimal fermentation conditions optimized by ANN were coix seed:glutinous rice 1∶1.9(g∶g),yeast A1 inoculum 4.2%,temperature 28.1℃,and initial pH 3.0,respectively.At the same time,the actual and predicted values of ANN and RSM under their optimal conditions were almost the same.The correlation coefficient R,coefficient of determination(R2),mean-square error(MSE),root mean square error(RMSE),mean absolute error(MAE),mean absolute percentage error(MAPE)correlation coefficients of the artificial neural network fitting model and the response surface methodology were 0.9945,0.9889,0.0117,0.1084,0.0722,0.4863%and 0.9836,0.9675,0.0289,0.1701,0.1437,0.9857%,respectively.Overall,the artificial neural network had higher fitting ability and accuracy than the response surface methodology and generated better fitting effect.Furthermore,the artificial neural network was more suitable for the optimization of fermentation condition of coix seed wine.