针对香蕉品质检测的专用电子鼻系统

Special Electronic Nose System for Banana Quality Detection

研究开发了一套针对香蕉品质检测的专用电子鼻系统。针对香蕉品质变化过程中释放的特征气体选择了特定的气敏传感器阵列,利用传感器检测腔的环境控制优化模块来稳定气体传感器的工作环境,提高系统检测精度和稳定性。同时通过传感器阵列及气腔的小型化设计,实现系统的便携化,极大地提高了系统的实用性。系统采用数据预处理和BP神经网络的组合识别算法,对香蕉的不同品质状态进行了分类识别。实验结果表明,系统的识别率达到了100%。

A special electronic nose system was developed for the detection of banana quality.According to the gas released in the process of banana corruption,we selected specialized gas sensors to form the sensor array.In order to stabilize the work situation and increase the measurement precision and stability,we designed the temperature control subsystem and humidity control subsystem to control the circumstance of sensors chamber.Meanwhile we minimized the sensors chamber to realize the portability and advance the practicability.Finally the system succeeded in the classification of various qualities of bananas by combining data preprocessing with BP neural network.The optimal experiments present that the classification success rate is 100%.