【摘要】 提出一种根据苹果气味对苹果进行无损检测的新方法,介绍一套适合苹果气味检测的电子鼻系统,阐述该系统识别苹果气味的过程 对超市所购的好坏苹果各50个进行了检测,在获得传感器阵列数据的基础上,从每个传感器曲线中提取了5个特征参数,该参数作为模式识别的输入向量 用主成分分析法和遗传神经网络对所测的样本进行分析,主成分分析的结果是能较好地区分好坏苹果,遗传神经网络对训练集的回判正确率和对测试集的测试正确率分别为100%和96 4% 试验证明该方法和电子鼻装置对苹果质量进行评定是有效的,同时也可以用于其他水果的检测更多还原

【Abstract】 A new method to classify apples by their odor is given. An electronic nose equipment to classify apples is introduced. Fifty good apples and fifty bad apples bought from the super-market have been examined. Five feature parameters are developed from each data curves of sensor arrays and are taken as input vectors. Principal component analysis (PCA) and genetic algorithm neural network (GA-NN) are used to analyze the feature parameters. Good separation is obtained using principal component analysis. The recognition probability of the GA-NN to the learning samples and the testing samples are 100% and 96.4% respectively. The new method can be applied to other fruit classification.更多还原