基于 PCA-BP算法的鸡精风味质量模型研究

摘要

鸡精调味料风味质量评估主要依赖于仪器分析和人工感官评价，在生产过程中无法进行实时品质检测。以6种不同添加成分的鸡精调味料风味感官数据为研究对象，提出了主成分分析和改进的BP神经网络算法相结合的算法，并且建立了鸡精风味质量控制模型。该模型能够对鸡精进行快速、准确的分类，同时能够检测出不满足预定风味要求的产品，为提高鸡精风味质量提供一种简单、快速的检测方法。%The evaluation of chicken essence flavor quality depends on instrumental analysis and artificial sensory assessment,which could not be used in real time detection of quality in producing process.In this paper,six kinds of chicken essence flavor sensory data are researched,and an algorithm combining the principal components analysis and improved BP algorithm is presented.A chicken essence flavor quality control model is implemented,which can classify chicken essence rapidly and accurately,and can detect the product not meeting the intended flavor,which has provided a simple and rapid detection method for improving the flavor quality of chicken essence.