提出了一种运用BP 神经网络来进行微生物特征分类的方法,仿真试验结果表明,该算法具有分类速度快、误分类率的特点;不仅加速了微生物的识别进程,而且为建立基于知识的分类技术提供了有力的技术支持和实现的可能。 关键词:神经网络;BP 算法;微生物分类 Abstract: An approach of Stored-grain Microbe Classification Based on BP Neutral Networks is proposed. The simulation research results indicated that this arithmetic has the capability of low misdiagnosing rate and rapid diagnosis. Not only accelerate the identify process of Microbe, moreover the powerful technical support is provided and the realization possibility for the establishment based on the knowledge diagnosis technology are presented. Keywords: neural network; BP arithmetic; Stored-grain microbe classification