基于响应面和人工神经网络-遗传算法优化鲜切萝卜光动力杀菌工艺

Optimization of photodynamic pasteurization process of fresh-cutradish based on response surface methodology and artificialneural network-genetic algorithm

[目的]本文旨在寻找有效建模方法以预测基于姜黄素介导的光动力技术(photodynamic technology, PDT)对鲜切萝卜的杀菌效果,优化其杀菌工艺。[方法]以白萝卜为材料,通过单因素试验及中心组合试验设计,探讨姜黄素浓度、光照时间、光照强度以及孵育时间对鲜切萝卜光动力杀菌的影响。在此基础上分别选用响应面法(response surface methodology, RSM)和人工神经网络-遗传算法(artificial neural network-genetic algorithm, ANN-GA)构建光动力杀菌模型。[结果]单因素试验及中心组合试验结果表明,影响鲜切萝卜光动力杀菌的主要因素从大到小依次为光照强度、光照时间、姜黄素浓度、孵育时间。通过验证试验及模型参数分析,发现基于RSM和ANN-GA方法构建的模型均能对杀菌效果进行较为准确预测,且后者预测能力优于前者。通过ANN-GA法确定鲜切萝卜的最佳杀菌工艺为:姜黄素浓度30μmol·L^(-1),光照强度100μmol·m^(-2)·s^(-1),光照时间24.50 min,孵育时间14.75 min,此时可获得较好的品质保留(维生素C、总酚含量分别为25.14、183.99 mg·100 g^(-1))。[结论]ANN-GA法可以全面、有效、准确地预测基于姜黄素介导的光动力技术对鲜切萝卜的杀菌效果,可为提高光动力技术在食品杀菌领域的应用提供理论基础。

[Objectives]This study aimed to explore an effective modeling method to predict the pasteurization effect of curcumin-mediated photodynamic technology(PDT)on fresh-cut radish and optimize its pasteurization process.[Methods]The effects of curcumin concentration,light intensity,illumination time and incubation time on the PDT pasteurization process of fresh-cut white radishes were investigated through single factor analysis and central composite designs.On this basis,response surface methodology(RSM)and artificial neural network-genetic algorithm(ANN-GA)were employed to develop predictive models for the PDT pasteurization process respectively.[Results]The results of the single factor analysis and central composite designs showed that the main influencing factors of fresh-cut radishes PDT pasteurization were in the order of light intensity,illumination time,curcumin concentration,and incubation time.The results of validation experiments and analysis of the statistical parameters of models indicated that models developed either by RSM or ANN-GA could well predict PDT pasteurization of fresh-cut radishes,and the predicting ability of the latter model was better than that of the former model.The optimal PDT pasteurization parameters for fresh-cut radishes determined by the ANN-GA method were curcumin with a concentration of 30μmol·L^(-1),light intensity of 100μmol·m^(-2)·s^(-1),illumination time of 24.50 min and the incubation time of 14.75 min.Products pasteurized under this condition could retain their quality well,with vitamin C content of 25.14 mg·100 g-1 and total phenols content of 183.99 mg·100 g-1.[Conclusions]The ANN-GA method can thoroughly,effectively,and accurately predict curcumin-mediated PDT pasteurization effect on fresh-cut radishes,therefore can provide a theoretical foundation for improving the application of PDT in food pasteurization.