以大米和莲子为原料，采用响应面试验设计优化莲子黄酒的发酵条件，探索其发酵规律。在单因素试验基础上确定以发酵时间、接种量、发酵温度为影响因素，以酒精体积分数为响应值，根据Box-Behnken中心组合方法采用三因素三水平响应值试验设计优化。结果表明，莲子黄酒最佳发酵工艺条件为：发酵时间14 d、接种量1.0%、发酵温度30 ℃，酒精体积分数14.2%，与预测值14.54%基本一致。说明该模型能较好地预测莲子黄酒发酵过程中的酒精体积分数。

Response surface methodology was applied to optimize fermentation parameters for lotus rice-wine production. On the basis of single factor experiments, the fermentation time, wheat koji inoculums size, and temperature on the yield of alcohol were selected as main factors for further test. A regression model for the yield of alcohol was established using a Box-Behnken design involving the above three factors at three levels. Results showed that the optimal fermentation parameters were as follows: fermentation time 14days, inoculums size 1.0%, and temperature 30℃. Under these conditions, the alcohol yield reached14.2%, closing to predicted yield of 14.54%. This reliable model could predict the contents of alcohol in the lotus rice-wine fermentation conditions.