双孢蘑菇质地柔嫩、 营养丰富, 具有很好的降血压、 降血脂、 消炎护肝等多种保健价值, 其新鲜度是反映内外部品质的重要指标之一。 目前双孢蘑菇新鲜度鉴别大多依据其外观品质变化(褐变), 缺乏精准的量化评价指标与方法, 因此提出了以贮藏天数为新鲜度检测的量化指标, 并利用近红外光谱技术对双孢蘑菇新鲜度进行检测分析。 依据存储天数不同, 将双孢蘑菇样本分为1~5组, 每组40个样本, 依次采集每组双孢蘑菇的近红外光谱数据。 针对采集的原始光谱数据, 首先选用卷积平滑滤波(SG)与多元散射校正(MSC)消除原始光谱噪声、 基线平移以及光散射的影响, 并选取399.81~999.81 nm的光谱波段作为数据处理范围; 然后分别使用主成分分析(PCA)和连续投影算法(SPA)进行光谱降维和特征波长选择, 继而建立极限学习机(ELM)分类模型; 同时考虑到ELM模型中初始值对分类准确率影响较大, 分别选用粒子群优化算法(PSO)、 海鸥优化算法(SOA)对ELM中初始权值及阈值进行寻优, 形成PSO-ELM, SOA-ELM优化组合分类模型; 最后分别将全光谱、 提取主成分以及所选的特征波长{556.87, 445.51, 481.15, 885.10, 802.25, 720.90, 861.34, 909.79, 924.44, 873.17 nm}输入到分类模型中, 建立不同输入、 不同分类模型的双孢菇新鲜度检测模型。 最终试验结果表明, 当ELM为分类模型, 以全光谱、 主成分以及特征波长为输入时的预测精度分别为75%, 95%, 88%; 以SPA优选特征波长作为输入的PSO-ELM、 SOA-ELM分类模型训练集精度为96.25%, 93.25%, 预测集精度为92.5%, 94%。 可知, SPA波长选择算法可以有效降低光谱信息中存在的冗余信息, 加快建模效率, 同时海鸥优化算法能较好的优化ELM分类模型的初始参数, 分类精度较ELM模型提高了6.8%, 同时不产生过拟合现象。 因此, 利用光谱特征可以快速、 准确无损的识别双孢蘑菇的新鲜度, 研究结果为便携式双孢蘑菇新鲜度快速无损检测设备的开发提供了理论依据。

Agaricus bisporus is fragile and nutritious, which helps lower blood pressure, lowering blood lipids, reducing inflammation and protecting the liver. The freshness is one of the most important indicators to reflect the internal and external quality of Agaricus bisporus. At present, the freshness identification of Agaricus bisporus is mostly based on appearance quality (browning), and there is a lack of an accurate quantitative evaluation method. Therefore, in this research, a quantitative index for freshness detection was proposed based on storage days, which was used to analyze the freshness of Ag aricus bisporus with VIS-NIR spectroscopy technology. According to the different storage days, the samples of Agaricus bisporus were divided into 1 to 5 groups, each with 40 samples, and the near-infrared spectral data of each group was collected in turn using a fiber optic spectrometer. For the collected raw spectral data, firstly, the SG and MSC transform methods were selected to correct and eliminate the effects of spectral noise, baseline shift and light scattering. Moreover, the spectral band sranging from 399.81 to 999.81 nm were selected as the data processing range simultaneously.Then the method of principal components analysis (PCA) and successive projections algorithm (SPA) were respectively used to reducethe spectral dimensionalities and select the characteristic wavelengths. And the Extreme Learning Machine (ELM) classifier was established based on the spectral features. Since the initial parameters have a greater impact on the classification accuracy of the ELM model, the Particle Swarm Optimization (PSO) and Seagull Optimization Algorithm (SOA) was used to optimize the initial values of weight and threshold for ELM classifier to establish PSO-ELM and SOA-ELM classifiers. Finally, the full spectrum, the extracted principal components and the selected characteristic wavelengths {556.87, 445.51, 481.15, 885.10, 802.25, 720.90, 861.34, 909.79, 924.44, 873.17} nm were input into the classification model to establish the freshness detection model of Pleurotus ostreatus with different inputs and different classification models. The final test results show that when the ELM is the classification model, the prediction accuracy with full spectrum, principal component and characteristic wavelength as input is 75%, 95% and 88% respectively; the training set accuracy of PSO-ELM and SOA-ELM classification model with SPA preferred characteristic wavelength as input is 96.25%, 93.25%, and the accuracy of prediction set is 92.5%, 94%. It can be seen that the method of SPA was effective to reduce the redundant information of VIS-NIR spectra and accelerate the modeling. At the same time, the SOAwas better to optimize the initial parameters of the ELM classifier and significantly improve the classification accuracy, and the classification accuracy is 6.8% higher than that of the ELM model. Therefore, the freshness of Agaricus bisporus can be identified quickly and accurately by using spectral features. The research results provide a theoretical basis for the development of portable equipment for rapid non-destructive testing of the freshness of Agaricus bisporus.