一种基于模糊准则的发酵过程残糖质量浓度的预测方法

Predictive Algorithm for Residual Sugar Concentration in Fermentation Process Based on Fuzzy Rule

摘要:

发酵过程残糖质量浓度的在线检测通常具有较大滞后，极大影响了残糖质量浓度的控制。发酵工艺人员根据菌体生长耗糖的惯性特征，设计了一种残糖质量浓度的经验估算算法，但是在残糖质量浓度变化较大时，估算效果不理想。作者首先分析工艺人员的经验估算算法的缺点，结合反馈校正思想，提出了经验估算改进（即有补偿）算法，接着针对工艺人员的残糖质量浓度估算模型的不准确性，结合模糊智能技术，进行模糊模拟，分别研究模糊预测无补偿、模糊预测有补偿情况下，残糖质量浓度的预测情况，最后通过实验验证了模糊残糖质量浓度估算模型以及补偿算法的准确性和有效性。

Abstract:

In the fermentation process，it is difficult to implement a good online control of residual sugar concentration with delayed measurements. Due to the inherit characteristics of biomass growth，the field operator can achieve an effective estimation on the residual sugar concentration according to operating experience or knowledge. However，it is ineffective in the large fluctuation of the residual sugar concentration. Based on the operating experience or knowledge，this paper introduces a feedback correction idea to compensate the experience estimation algorithm. Moreover，considering the inaccuracy model of the residual sugar concentration，this paper proposes a fuzzy predictive estimation algorithm，a fuzzy predictive estimation algorithm with compensation combining fuzzy intelligent technology. Afterwards，the residual sugar concentration is forecast by these two predictive estimation algorithms. The accuracy and effectiveness of the fuzzy model and compensation algorithm are verified by a real experiment.