| Owing broad arable land and amounts of peasant population, China is a country with a long agricultural civilization. Cultivation of crops is an important means of survival and economic source for farmers. For some annual periodicity seeding planting crops, seed quality and cultivar performance directly affect the crops’ every important link from growth to harvest, thereby further affect food production and farmers’ income. Therefore, crop varieties and seed quality problem are the hot issue in farmers. Today, there are a wide variety of crop seed in the market, in which seed quality varies greatly and more fake seeds are found in news reports from time to time. Therefore, it has important practical significance to achieve fast, convenient, accurate and reliable identification of seeds for market management.Nonlinear oscillating chemical fingerprint analysis method is a method of generating a characteristic spectrum utilizing B-Z oscillating reaction system with biological sample complex redox component interactions. It can indirectly reflects the overall chemical characteristics of a biological sample not fully grasping the complex biological samples component, in order to achieve the purpose of the analysis of samples identified.The second chapter studied the application of nonlinear chemical fingerprints in soybean cultivar identification, and discussed several factors of B-Z oscillating system. The results showed that the chemical fingerprint parameters of different varieties of soybean seed quite were different,which both can be accurately isolated and identificated by the cluster analysis(CA) and principal component analysis(PCA); and the case of soybean seeds under different batches of the same species, of which spectrum feature parameter difference is small(maximum relative standard deviation of 3.3%), were successfully classified as a class in the cluster analysis and principal component analysis. Therefore, nonlinear chemical fingerprint technology is a new method of analysis to effectively identify soybean varieties.The third chapter identified the hybrid rice seed, conventional rice seed and hybrid rice seed offspring utilizing chemical oscillation fingerprint, and discussed the reaction temperature, the additive amount of sample, and spectra reproducibility. The result showed that, the chemical oscillation fingerprints of different varieties of rice seeds had obvious characteristic, and the fingerprint of offspring hybrid rice seeds and original seed existed big difference. The oscillation spectrum characteristic parameters exhibited good stability under 45 OC, and the identification of the vast majority of rice varieties can be completed within 15 minutes. Nonlinear Chemical Oscillation fingerprinting provided a new solving idea for identifying rice varieties.The fourth chapter analyzed and identified 17 common varieties of maize utilizing nonlinear oscillating chemical fingerprint, and discussed the differences between different production batches of maize seed. The results showed that the nonlinear chemical fingerprint integrated pattern recognition method could extract and exhibit good overall chemical characteristics of each maize seed. Tree cluster analysis diagram and principal component analysis diagram translated the large amounts of data directly into visual information easy to observe, which provided a simple and effective method of analysis. The study enriched the scope of application of nonlinear chemical fingerprint and established a rapid and effective identification of nonlinear chemical methods. |
