中文摘要：

利用卟啉类化合物制成可视传感器阵列，对3类桃子（正常、机械损伤、真菌感染）进行测试．提取桃子的气味信息，将传感器阵列与桃子顶空气体发生反应前后的颜色变化作为特征值．对特征值进行主成分分析，提取前6个主成分用于最小二乘支持向量机建模．对训练集中的26个样本进行测试，模型的回判正确率达100％；对24个预测样本的识别正确率达91．7％．研究表明可视传感器阵列用于桃子的质量评定是可行的，对其他水果的评定有直接借鉴作用．

英文摘要：

A colorimetric sensor array was developed using metalloporphyrins. Odors of peaches of three different classes （normal, physical damage, fungal diseases） were detected. The RGB value changes of the sensor array before and after exposure to the headspace gas of peaches were analyzed by principal component analysis （PCA）. The former 6 principal components were selected for the model building of the least squares support vector machine （LS -SVM）. The discriminating rate of the model is 100% for 26 samples in training set, and 91.7% for 24 samples in predicting set. Experiments show that it is feasible to apply the colorimetric sensor array to evaluate the quality of the peaches. The method developed can also be used as reference for quality evaluation of other fruits.