大豆在生长过程中因病害影响其产量会急剧下降, 如果不及时判别出病害种类, 喷洒相关农药, 病害严重的大豆甚至会绝产。 及时判别病害种类进行合理施药, 阻止病害进一步发展是保证大豆安全生产的重要环节。 目前, 基于大豆植株细菌性病害的病原菌鉴定和聚合酶链式反应(polymerase chain reaction, PCR)的鉴定方法, 最短需要两天时间, 因此, 快速检测大豆病害种类的方法成为该作物, 也是建立智慧农业生产的关键环节之一。 应用拉曼光谱快速检测技术诊断大豆病害, 构建N-乙酰胞壁酸分子空间结构, 采用密度泛函理论通过利用B3LYP/6-31+(d, p)基组优化大豆细菌性病害标志物N-乙酰胞壁酸的分子结构计算其拉曼光谱, 并进行理论因子校正, 校正因子为0.985 7; 采用微区三级拉曼光谱技术探测该标志物N-乙酰胞壁酸的拉曼光谱, 采用平滑、 去基线、 截取波数范围等过程进行光谱预处理; 在理论和实验对比分析的基础上, 指认大豆测试和计算的拉曼光谱对应的特征峰, 峰值波数相差大多在0~10 cm^-1, 实验数据与理论计算结果基本一致, 判定了振动拉曼光谱的特征峰及其对应的分子结构的关系。 结果表明: 大豆细菌性病害标志物N-乙酰胞壁酸分子在200~1 650 cm^-1范围内含15个特征峰, 较强峰值和振动归属分别为229.0 cm^-1的甲基摇摆振动和764.0 cm^-1环内的摇摆呼吸振动等, 给出了键长、 键角和二面角等15个振动峰的空间结构参数, 指证了N-乙酰胞壁酸分子的特征结构。 结果也证明了可通过多种生物分子的大豆拉曼光谱测量, 筛选细菌性病害标志物N-乙酰胞壁酸分子的拉曼光谱, 能够有效识别细菌性病害。 智慧农业生产中利用拉曼光谱快速检测技术, 是农作物病害检测诊断的一种有效方法, 若结合应用机器学习方法与光谱分析识别, 以快速、 准确和便捷的方式为智慧农业的健康生产及保驾护航发挥效用, 是推进我国农业发展的重要环节。

The yield of soybean will drop dramatically due to disease during its growth. If the disease is not identified in time and no corresponding pesticides are sprayed, severely diseased soybeans can even be wiped out. It is very important to identify the disease species and apply the insecticide rationally to prevent the further development. Currently, it will take two days to make the pathogenic and polymerase chain reaction (PCR) identification of soybean bacterial diseases. Therefore, the method of quickly detecting the types of soybean diseases has become one of the key links in the intelligent agricultural production of this crop. Raman spectroscopy is used to rapidly diagnose soybean diseases. The molecular space structure of N-acetylmuramic acid is constructed, density functional theory (DFT) with B3LYP/6-31+(d,p) basis set was used to do the theoretical calculation. Through theoretically calculating the Raman spectra of soybean bacterial spot disease marker N-acetylmuramic acid, the characteristic peaks of the vibrational Raman spectra and their corresponding molecular structures of N-acetylmuramic acid are identified. The calculated Raman spectra should be corrected using the correction factor, and the correction factor is 0.985 7. In addition, the experimental Raman spectra of N-acetylmuramic acid are obtained using micro-zone three Grade Raman spectroscopy technology. The process of smoothing, baselines removal and wavenumber range interception was used to preprocess the spectra. The comparative analysis of theoretical and experimental results determines the characteristic peaks of vibrational Raman spectra and the corresponding molecular structures. The peak wavenumber difference is mostly 0~10 cm^-1. The experimental data is consistent with the theoretical calculation results. The results show that the N-acetylmuramic acid molecule, a marker of bacterial spot in soybean, contains 15 characteristic peaks in the range of 200 to 1 650 cm^-1, which can be used as a diagnostic basis. The main peak assignment at 229 and 763 cm^-1 were attributed to the methyl swing vibration and ring breathing vibration. The spatial structure parameters of 15 vibration peaks such as bond length, bond angle and dihedral angle are given to identify the structure of the N-acetylmuramic acid molecule. The results also proved that the Raman spectroscopy of soybean with a variety of biomolecules could be used to screen the Raman spectroscopy of N-acetylmuramic acid, and it could effectively identify bacterial disease. Raman spectroscopy rapid detection technology is a new method for soybean disease detection and diagnosis, which plays a part in protecting healthy products in the field of intelligent agriculture. The results should be better combine with machine learning methods in spectral analysis and identification. Exploring a fast, accurate and convenient method could obtain a lot of benefits in intelligent agriculture, which plays a vital role in promoting the development of agriculture in China.