基于PCA与BP混合神经网络算法的电子鼻系统

Electronic nose system based on PCA and BP hybrid neural network algorithm

为了对食物品质进行非接触式评价,基于6种费加罗金属氧化物气体传感器阵列,通过由数据采集模块和微处理器模块组成的硬件设计方案,设计并研制了可对被测食物进行实时、无损检测的电子鼻系统。在软件设计方案上,该系统采用主成分分析(PCA)和反向传播(BP)混合神经网络模式,通过LabVIEW对气体"指纹信息"数据库进行分析。实验结果表明:该设计的电子鼻系统可以很好地区分不同种类的食醋,并提供了一种对食醋品质评价的便利方法。

For non contact food quality evaluation,through hardware design scheme consists of data acquisition module and microprocessor module,an electrical noise system is designed and developed based on six kinds of Figaro metal-oxide gas sensor array for real time food non-destructive examination （NDE）.As for the software design scheme,the principal component analysis（PCA） and back propagation（BP） are utilized and used to analyze gas＂ fingerprint information＂ database by LabVIEW.The experimental results show that the designed electronic nose system can distinguish different types of vinegar accurately,and provide a convenient method for vinegar quality evaluation.