| Droughts caused by global climate change are becoming more frequent and severe,posing a serious threat to agricultural production and food security.As one of the most important food crops in the world,rice faces severe challenges due to its sensitivity to drought.Establishing accurate and rapid rice drought detection technology is the key to achieving efficient irrigation management,improving crop yields and water resource utilization efficiency,but current drought detection technology is often divorced from the photosynthetic physiological needs of crops.Chlorophyll Fluorescence(Chl F)released during plant photosynthesis is closely related to photosynthesis and plant physiological status.It can be used as an important indicator to evaluate plant physiological status and environmental stress.index.Therefore,studying rice drought detection technology based on Chl F has important theoretical and practical significance.This study is based on the Chl F of rice and aims to establish a detection technology that can accurately and quickly identify drought conditions in rice.The main research content of this paper is as follows:(1)Research on the establishment of a rapid estimation method for the maximum photochemical quantum yield of Photosynthetic System II(PSII)based on the Least Squares Support Vector Machine(LSSVM).The ratio of variable fluorescence to maximum fluorescence(Fv/Fm)obtained from the Chl F induction curve(OJIP induction curve)under dark adaptation is usually used to reflect the maximum photochemical quantum yield of PSII and is closely related to plant physiological states such as drought.It is often used to sense drought in the literature,but it requires plant samples to dark adapt for twenty minutes or more before testing,which limits its practical application value in production and scientific experiments.In this study,it is proposed to use LSSVM to establish a rapid estimation method of Fv/Fm from the OJIP induction curve under non-dark adaptation,and at the same time reveal the hidden correlation between Chl F signals under dark adaptation and non-dark adaptation.Experiments under different drought stress durations,temperatures and other environments have verified that Fv/Fm can be estimated from the Chl F test data without dark adaptation.This technology improves the real-time performance of Fv/Fm information acquisition,making Fv/Fm practical value in rice drought detection.(2)Research on the establishment of rice drought detection method based on OJIP induction curve.Existing Chl F-based drought detection technologies in the literature often rely on several artificially selected feature point data on the OJIP induction curve,while ignoring the potentially rich physiological information present in all data on the entire OJIP induction curve,resulting in a large amount of information being discarded.In this study,a method for classifying rice drought degree using data on the entire OJIP induction curve was established based on Support Vector Machine(SVM)and verified through experiments.The results show that the classification accuracy of using the OJIP induction curve as the input data of the SVM model is higher than that of the few feature point data often used in the literature.In order to retain as much Chl F information as possible and reduce data complexity,independent component analysis(ICA)was used to reduce the dimensionality of the OJIP induction curve,and compared with various feature extraction or feature dimensionality reduction methods.The results found that the 15 new features obtained after dimensionality reduction of the induction curve by ICA have higher rice drought classification accuracy than the entire OJIP induction curve used as the SVM model input data set.The results show that automatic dimensionality reduction using ICA can make full use of the OJIP induction curve in rice drought detection.The plant physiological information carried on it improves drought detection and classification accuracy.(3)Research on Chl F analysis and rice drought detection technology based on broadband excitation.PSII is a high-order broadband system,and the traditional Chl F signal is usually based on step or pulse excitation,which is difficult to stimulate the richer dynamic characteristics of the photosynthetic system,making the Chl F signal carry more significant stress information,thus reducing the ability of Chl F to sense drought stress.sensitivity and robustness.In this study,pseudo-random binary sequence(PRBS)was used to excite the Chl F signal,and a drought stress sensing technology based on Chl F wavelet information entropy excited by the PRBS signal was established and verified through experiments.The results show that compared with the traditional Chl F excitation method,the Chl F wavelet information entropy excited by the proposed PRBS signal has better performance relative to the OJIP induction curve in distinguishing droughts such as rice.This work provides a new plant-based Physiological drought stress detection methods and Chl F analysis technology.(4)Research and development of chlorophyll fluorometer with broadband excitation function.Currently,no chlorophyll fluorometer on the market has a broadband excitation function,which limits the ability of emerging artificial intelligence algorithms to process complex signals and mine rich information.To solve this problem,we designed and developed a chlorophyll fluorometer with broadband excitation function based on PRBS.This instrument can not only measure the Chl F signal excited by PRBS,but also the Chl F signal excited by traditional step and PAM.The comparison results of Chl F information entropy under three kinds of excitation light show that the Chl F signal excited by PRBS has the highest information entropy.This instrument has reached commercial grade and has begun to enter the market.It can provide a new scientific instrument for the detection of plant physiology and environmental stress,and provide richer information data for emerging artificial intelligence algorithms. |
PDF Full Text Request