基于CNN-WSN与SHO-KELM的电子鼻食品质量检测方法

An Electronic Nose Food Quality Detection Method Based on CNN-WSN and SHO-KELM

摘要: 食品质量的检测对于人体健康与工业生产具有重要意义，但是当下的常见检测手段难以实现快速、准确、无损的检测需求。因此在这项工作中，提出了一种基于CNN-WSN与SHO-KELM相结合的电子鼻食品质量检测方法。首先基于卷积神经网络(CNN)与小波散射网络(WSN)得到了能够有效表征食品质量原始信息的CNN-WSN融合特征。然后利用海马优化算法(SHO)对核极限学习机(KELM)模型的核参数与正则化系数进行优化，解决了关键参数选择困难的问题。为了验证提出方法的有效性，最后自主搭建了一套电子鼻系统并对牛奶样本进行了采集与测试。实验结果证实了该方法具有良好的食品质量检测效果。

Abstract: The detection of food quality is of great importance for human health and industrial production, but the current common detection methods are difficult to achieve the fast, accurate and non-destructive detection needs. Therefore, in this work, an electronic nose food quality detection method based on the combination of CNN-WSN and SHO-KELM is proposed. Firstly, CNN-WSN fusion features that can effectively characterize the original information of food quality are obtained based on convolutional neural network (CNN) and wavelet scattering network (WSN). Then the kernel parameters and regularization coefficients of the kernel-extreme learning machine (KELM) model are optimized using the hippocampus optimization algorithm (SHO), which solves the problem of the difficulty in selecting key parameters. In order to verify the validity of the proposed method, finally, an electronic nose system was built independently and milk samples were collected and tested. The experimental results confirm that the proposed method has good food quality detection effect.