BP神经网络结合响应面优化荷叶功能性硬糖的工艺

Process Optimization of Lotus Leaf Functional Hard Candy by BP Neural Network Combined with Response Surface

本试验以白砂糖、葡萄糖浆、木糖醇和荷叶粉为原料制作一种功能性硬糖。以感官评分为评价指标,研究葡萄糖浆、木糖醇和荷叶粉末添加量对硬糖品质的影响,在单因素试验基础上,利用响应面法优化荷叶功能性硬糖的配方,以响应面试验数据作为BP神经网络的输入值,对主要影响因素进行仿真优化。结果表明,荷叶硬糖的最优配方为葡萄糖浆15.3 g,木糖醇3.3 g,荷叶粉末2.8 g,白砂糖37 g,水20 g。在此条件下制得的荷叶硬糖糖体呈通透的墨绿色,色泽均匀,外形完整无裂痕,甜味纯正,坚硬而脆,不黏牙,具有荷叶风味。

This experiment used white sugar, glucose syrup, xylitol and lotus leaf powder as raw materials to make a functional hard candy. The sensory evaluation score was used as the evaluation index to study the effect of glucose syrup, xylitol and lotus leaf powder addition on the hard candy. On the basis of single factor test, the response surface method was used to optimize the formula of lotus leaf functional hard candy. The response surface test data was used as the input value of the BP neural network to simulate and optimize the main influencing factors. The results showed that the optimal formula of lotus leaf hard candy was glucose syrup 15.3 g, xylitol 3.3 g, lotus leaf powder 2.8 g, white granulated sugar 37 g, water 20 g. The sugar body of lotus leaf hard candy was transparent dark green, uniform color distribution, complete shape without cracks, pure sweetness, hard and brittle, non-sticky, with lotus leaf flavor.