A fertilization indication method for diagnosing rice leaf failure using SPAD instrument

Abstract

The present application relates to the technical field of rice seedling cultivation, and specifically to a method for diagnosing rice leaf drop using a SPAD instrument. The method comprises: obtaining the SPAD value, nitrogen application rate, and yield in each region of various rice varieties during any growth period, and obtaining the SPAD value of each subregion; determining the optimal diagnosis period based on the correlation between the SPAD value and the yield; determining the optimal fitting model by analyzing the fitting of all nitrogen application rates and SPAD values ​​of various rice varieties during the optimal diagnosis period under different fitting methods; and analyzing the difference between the SPAD value of each region of various rice varieties during the optimal diagnosis period and the SPAD value corresponding to the maximum yield, and fertilizing the various rice varieties. The present application aims to obtain the optimal fitting model and improve the accuracy of controlling the amount of fertilizer applied to rice.

Description

Technical Field

[0001] The present application relates to the technical field of rice seedling cultivation, and in particular to a method for indicating fertilization by using a SPAD instrument to diagnose one leaf falling off in rice. Background Art

[0002] During the growth of rice, nitrogen is a key element that promotes plant growth, increases yield and improves quality. However, excessive or insufficient nitrogen fertilizer application is common in rice fields. This not only leads to rice yield and quality failing to meet expectations, but also may cause nitrogen environmental pollution. For this reason, reasonable fertilization and accurate diagnosis of rice nitrogen requirements have become one of the core tasks for achieving sustainable agricultural development. Traditional nitrogen diagnostic methods mainly rely on chemical testing and leaf color analysis. Among them, SPAD instruments, as a commonly used non-destructive testing tool, are widely used in rice fertilization management because of their ease of operation and low cost.

[0003] The existing technology analyzes the correspondence between the SPAD value of rice and the yield and nitrogen application rate, and selects a suitable fitting model to perform data fitting, thereby achieving fertilization instructions. However, the existing technology selects fitting models by using methods such as the coefficient of determination and the mean square error to evaluate the fitting accuracy of each fitting model, but does not take into account that evaluation techniques such as the coefficient of determination and the mean square error are all for evaluating the overall error. If a certain rice area is affected by soil factors and irrigation conditions, resulting in large error fluctuations in the SPAD value, it will cause a large local fitting error at that data point, which will in turn affect the overall fitting error of the fitting model, resulting in an inappropriate selection of the fitting model and reducing the accuracy of controlling the amount of fertilizer applied to rice. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a method for diagnosing a fertilization indication of a rice leaf falling using a SPAD instrument to solve the existing problems.

[0005] The present invention discloses a method for diagnosing rice leaf failure by using a SPAD instrument to indicate fertilization, which adopts the following technical solutions:

[0006] One embodiment of the present application provides a method for diagnosing a fertilization indication of a rice leaf falling using a SPAD instrument, the method comprising the following steps:

[0007] Divide the planting area of ​​various rice varieties into multiple regions, apply different amounts of nitrogen fertilizer to different regions, divide each region into multiple sub-regions, obtain the SPAD value, nitrogen application amount, and yield of each region in any growth period of various rice varieties, and obtain the SPAD value of each sub-region;

[0008] Based on the correlation between SPAD value and yield, the optimal diagnosis period was selected from all growth periods of various rice varieties. All nitrogen application rates and SPAD values ​​of various rice varieties under the optimal diagnosis period were analyzed to determine the optimal fitting model, which is as follows:

[0009] Different fitting methods were used to fit binary pairs consisting of nitrogen application rates and SPAD values ​​in all regions of various rice varieties. The distances of all binary pairs to the fitting curves obtained by any fitting method were used to form an error sequence. The error sequence was divided into multiple subsequences. The distribution of all elements in each subsequence was analyzed to determine the fluctuation value of various rice varieties under any fitting method.

[0010] Within each region of various rice varieties, the deviation of the SPAD value within each subregion from the average SPAD value of all other subregions was used as a difference index. The difference indices of all subregions were clustered to obtain a significant set for each region. Based on the distribution of all difference indices within the significant set, error elements were screened from the error sequence. The difference between each error element and the distribution of all other elements in its subsequence was compared. Combined with the fluctuation value, the deviation value of each rice variety under any fitting method was determined.

[0011] Analyze the changing trend of the fitting curves of various rice varieties under any fitting method, determine the changing index of various rice varieties under any fitting method, determine the trend value of various rice varieties under any fitting method, and determine the optimal fitting model in combination with the deviation value;

[0012] The differences between the SPAD values ​​of various rice regions during the optimal diagnosis period and the SPAD values ​​corresponding to the maximum yield were analyzed, and fertilization was performed on various rice varieties based on the optimal fitting model.

[0013] Preferably, the method for determining the fluctuation values ​​of the various rice varieties under any fitting method is:

[0014] For the error sequences of various rice varieties under any fitting method, the mean of all elements in each subsequence was calculated as the average error of each subsequence;

[0015] The variance and cumulative sum of the average errors of all subsequences in the error sequence are calculated respectively, and the product of the variance and the cumulative sum is used as the fluctuation value of various rice varieties under any fitting method.

[0016] Preferably, the process of obtaining the salient set of each region is:

[0017] The difference indexes of all sub-regions in each region of various rice varieties are clustered to obtain multiple clusters. The mean of all difference indices in each cluster is calculated and recorded as the difference mean of each cluster. Among all clusters, the cluster with the largest difference mean is taken as the significant set of each region.

[0018] Preferably, the process of filtering out all error elements from the error sequence is:

[0019] For various rice varieties under any fitting method, the difference mean of the significant sets of all regions is used as the input of the threshold segmentation algorithm, and the segmentation threshold is output. The region where the significant set corresponds to the difference mean greater than the segmentation threshold is used as the error region, and the corresponding elements of the binary groups of all error regions in the error sequence are used as error elements.

[0020] Preferably, the method for determining the deviation values ​​of the various rice varieties under any fitting method is:

[0021] For the error sequences of various rice varieties under any fitting method, the difference between each error element and the mean of all other elements in its subsequence is calculated, and the difference between all error elements is used as the error index of various rice varieties under any fitting method;

[0022] Deviation value A of the uth rice variety under fitting method i u,i The expression is: Where B u,i represents the coefficient of determination of all pairs of rice varieties u under fitting method i; C u,i represents the fluctuation value of the u-th rice under fitting method i; D u,i represents the error index of the u-th rice under fitting method i; T u represents the number of all clusters of the u-th rice variety; exp() represents an exponential function with a natural constant as the base.

[0023] Preferably, the expression of the variation index of the various rice varieties under any fitting method is:

[0024] Extract the peaks and valleys on the fitting curves of various rice varieties under any fitting method, connect adjacent peaks and valleys, calculate the slopes of the connected straight lines, and arrange the slopes of the connected straight lines between all adjacent peaks and valleys in ascending order according to the time of peak appearance to form a slope sequence of various rice varieties under any fitting method;

[0025] The variation index R of the uth rice variety under fitting method i u,i The expression is: Where, E u,i F represents the ratio of the number of all negative elements to the number of all positive elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; u,i Z represents the ratio of the number of all positive elements to the number of all negative elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; u represents the classification of the u-th rice, Z u=1 indicates that the uth rice is hybrid rice, Z u =0 means that the u-th rice variety is conventional rice.

[0026] Preferably, the expression of the trend value of the various rice varieties under any fitting method is:

[0027] Where M u,i represents the trend value of the u-th rice under fitting method i; V u,i represents the number of all elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; It represents the difference between the qth element and its adjacent previous element in the first-order difference sequence of the slope sequence of the uth rice variety under fitting method i; exp() represents the exponential function with a natural constant as the base; norm[] represents the normalization function; max() represents the maximum value function.

[0028] Preferably, the method for determining the optimal fitting model is:

[0029] Height fitting coefficient R of fitting method i i The expression is: Where, represents the mean of the trend values ​​of all hybrid rice varieties under fitting method i; represents the mean of the trend values ​​of all conventional rice varieties under fitting method i; represents the mean of the deviation values ​​of all rice varieties under fitting method i; ∈ represents a constant greater than 0;

[0030] Among all the fitting methods, the fitting method with the largest height fitting coefficient is taken as the optimal fitting model.

[0031] Preferably, the fertilizing of various rice plants comprises:

[0032] In all regions of various rice varieties within the optimal diagnostic period, the SPAD corresponding to the maximum yield is recorded as the optimal value of various rice varieties within the optimal diagnostic period. If the SPAD value of the jth region of the uth rice variety under the optimal diagnostic period is less than the optimal value of the uth rice variety, it is necessary to fertilize the uth rice variety in the jth region. The deviation between the SPAD value of the uth rice variety in the jth region and the optimal value is used as the input of the optimal fitting model, and the output is the nitrogen fertilizer content that needs to be supplemented; otherwise, it is not necessary to fertilize the uth rice variety in the jth region.

[0033] This application has at least the following beneficial effects:

[0034] This application constructs a fluctuation value by analyzing the fluctuation of the fitting effect between the nitrogen application amount and the SPAD value of rice under different fitting methods, which can preliminarily evaluate the fitting effect of the fitting method for large data; further, by analyzing the difference in SPAD values ​​between different sub-regions in the same region, a deviation value is constructed, which eliminates the influence of environmental factors on the fitting effect, improves the accuracy of the fitting result, and helps to more clearly judge the fitting effect of different fitting methods; further, by analyzing the changes in SPAD values ​​of different types of rice with nitrogen application during the optimal diagnosis period, a trend consistency coefficient is constructed, which reflects the real change trend of the data and helps to select the optimal fitting model for different rice varieties, so as to control the fertilizer application amount according to the optimal fitting model, thereby improving the accuracy of rice fertilizer application control. This application evaluates the fitting effect of different fitting methods from multiple aspects such as overall, local, and data trends, selects the optimal fitting model that is most suitable for different rice varieties in the optimal diagnosis period, and improves the accuracy of rice fertilizer application control. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0036] Figure 1 A flowchart of a method for diagnosing a fertilization indication of a rice leaf falling using a SPAD instrument according to an embodiment of the present application is provided;

[0037] Figure 2 A flowchart of the steps for obtaining the optimal fit model provided in one embodiment of the present application. DETAILED DESCRIPTION

[0038] To further illustrate the technical means and effectiveness of this application to achieve the intended invention objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of a method for diagnosing rice leaf failure using a SPAD instrument, as proposed in this application. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0039] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0040] The following describes in detail a specific scheme of a method for diagnosing one-leaf-falling rice fertilizer application using a SPAD instrument provided by the present application with reference to the accompanying drawings.

[0041] One embodiment of the present application provides a method for indicating fertilization using a SPAD instrument to diagnose rice with one leaf falling. Specifically, the following method for indicating fertilization using a SPAD instrument to diagnose rice with one leaf falling is provided. Figure 1 , the method comprises the following steps:

[0042] S1: Divide the planting area of ​​various rice varieties into multiple regions, apply different amounts of nitrogen fertilizer to different regions, divide each region into multiple sub-regions, obtain the SPAD value, nitrogen application amount and yield of various rice varieties in each region at any growth period, and obtain the SPAD value of each sub-region.

[0043] The entire growth process of rice includes tillering stage, jointing stage, panicle differentiation stage, booting stage and heading stage. This example takes the mainstream hybrid rice "Shenyou 26", "Shenyou 28" and conventional rice "Nanjing 46" and "Chongshang 2022" as the target, and uses 0kg / hm 2 、105kg / hm 2 、135kg / hm 2 、150kg / hm 2 , 225kg / hm 2 、300kg / hm 2 、337.5kg / hm 2 、375kg / hm 2 Eight nitrogen application rate treatments were applied to four rice varieties.

[0044] In this embodiment, the planting area of ​​various rice varieties is evenly divided into 8 areas, different amounts of nitrogen fertilizer are applied to different areas, and each area is divided into 3 sub-areas. Six rice plants are selected in each sub-area, and a SPAD instrument is used to measure the SPAD value of the fallen leaf of these six rice plants at any growth stage during their growth process. The measurement position is 1 / 2 from the base of the leaf, and the leaf veins and leaf edges are avoided during the measurement. The average of the SPAD values ​​of the six rice plants is used as the SPAD value of the sub-area where the six rice plants are located, and the average of the SPAD values ​​of all sub-areas is used as the SPAD value of each area. At the same time, the nitrogen application amount and yield of various rice varieties in each area at any growth stage are obtained.

[0045] S2: Based on the correlation between SPAD values ​​and yield, the optimal diagnostic period was selected from all growth periods of various rice varieties. The optimal fitting model was determined by analyzing the fitting of all nitrogen application rates of various rice varieties under the optimal diagnostic period and SPAD values ​​under different fitting methods.

[0046] After obtaining relevant rice data, we can build a fitting model to analyze the corresponding relationships between various data. We can then use the fitting model and existing data to predict rice yields and provide guidance on rice fertilization. Therefore, it is crucial to select a model with high fitting accuracy.

[0047] S201: Using different fitting methods, fit the binary groups consisting of nitrogen application rate and SPAD value in all regions of various rice varieties to obtain fitting curves under different fitting methods, and form an error sequence based on the distances from all binary groups to the corresponding fitting points on the fitting curve under any fitting method. The error sequence is divided into multiple subsequences, and the fluctuation value of various rice varieties under any fitting method is determined by analyzing the average distribution and discreteness of all elements in each subsequence.

[0048] (1) Obtain the optimal diagnosis period for various rice varieties.

[0049] Since each rice variety has multiple growth stages, and the correspondence between SPAD values ​​and rice yield in different growth stages is not consistent, it is necessary to select the growth stage with the strongest correlation between SPAD value and yield from multiple cultivation periods and build a model based on the relevant data of rice in this growth stage to ensure the reliability of the prediction model. The specific process of determining the optimal diagnostic period is as follows:

[0050] The yield and SPAD value of each region of various rice varieties at any growth stage are combined into two-dimensional data points, and the two-dimensional data points of all regions are used as inputs of a quadratic function regression equation, and the determination coefficient is output as the determination coefficient of various rice varieties at any growth stage;

[0051] Among all growth periods of various rice varieties, the growth period corresponding to the maximum determination coefficient is taken as the optimal diagnosis period for each rice variety.

[0052] Among them, the method of using the quadratic function regression equation to obtain the determination coefficient is a well-known technology, and its specific principle and process will not be repeated here.

[0053] (2) Analyze the fitting accuracy of various fitting methods.

[0054] Different fitting methods were used to fit all binary pairs consisting of nitrogen application rates and SPAD values ​​for various rice varieties to obtain fitting curves using different fitting methods, where the abscissa represents the nitrogen application rate and the ordinate represents the SPAD value. In this embodiment, the different fitting methods include at least: a quadratic polynomial fitting method, a logarithmic fitting method, a power function fitting method, an exponential function fitting method, a least squares method, and a linear regression model.

[0055] Furthermore, for various rice varieties under any fitting method, the distances from all the binary groups to the fitting curves under any fitting method are used to form an error sequence, and the error sequence is evenly divided into multiple subsequences, and the mean of all elements in each subsequence is calculated as the average error of each subsequence; if the mean of all elements in the subsequence is larger, that is, the larger the average error, it indicates that the fitting effect of the current fitting method in the corresponding area under the corresponding nitrogen application rate is poor; conversely, if the mean of all elements in the subsequence is smaller, that is, the smaller the average error, it indicates that the fitting effect of the current fitting method in the corresponding area under the corresponding nitrogen application rate is better.

[0056] Furthermore, the variance and cumulative sum of the average errors of all subsequences in the error sequence are calculated respectively, and the product of the variance and the cumulative sum is used as the fluctuation value of various rice varieties under any fitting method.

[0057] According to the fluctuation values ​​of various rice varieties under any fitting method, it can be understood that if the variance of the average error of all subsequences is larger, it means that the effect fluctuation of the current fitting method under different nitrogen application rates is larger, indicating that the overall fitting effect is poor, and the larger the cumulative sum of the average errors of all subsequences is, the worse the fitting effect of the current fitting method between SPAD value and nitrogen application rate under different nitrogen application rates is, that is, the larger the variance, the larger the cumulative sum, and the larger the fluctuation value obtained, indicating that the fitting effect of the current fitting method between SPAD value and nitrogen application rate under different nitrogen application conditions is larger, and the overall fitting effect is poor;

[0058] On the contrary, if the variance of the average error of all subsequences is smaller, it means that the fitting effect of the current fitting method under different nitrogen application rates is more stable, indicating that the overall fitting effect is better, and the smaller the cumulative sum of the average errors of all subsequences is, it means that the fitting effect between SPAD value and nitrogen application rate under different nitrogen application rates using the current fitting method is better, that is, the smaller the variance, the smaller the cumulative sum, and the smaller the obtained fluctuation value, indicating that the fitting effect of the current fitting method for SPAD value and nitrogen application rate under different nitrogen application conditions is more stable, and the overall fitting effect is better.

[0059] S202: Within each region of various rice varieties, the deviation between the SPAD value within each sub-region and the average SPAD value of all other sub-regions is used as a difference index, and the difference indices of all sub-regions are clustered to obtain a significant set for each region. Based on the average distribution of all difference indices within the significant set of each region, all error elements are screened out from the error sequence, and the difference between each error element and the average distribution of all other elements in its sub-sequence is compared. Combined with the fluctuation value, the deviation value of various rice varieties under any fitting method is determined.

[0060] Factors such as soil environment, irrigation conditions, and light intensity can all cause variations in rice SPAD, necessitating further analysis. If nitrogen application rates are consistent across rice fields, the SPAD values ​​detected should also be relatively consistent. Therefore, the greater the difference in SPAD values ​​between a particular region and those receiving the same nitrogen application rate, the greater the degree to which environmental factors have affected that region's SPAD value, and the greater the original error in the SPAD value for that region.

[0061] Therefore, by analyzing the differences in SPAD values ​​of various rice varieties in different sub-regions within the same region, the influence of environmental factors on the fitting method was eliminated. The specific process is as follows:

[0062] In each region of various rice varieties, the deviation of the SPAD value in each subregion from the average SPAD value of all other subregions is used as a difference index. The difference indices of all subregions in each region are clustered to obtain multiple clusters. The mean of all difference indices in each cluster is calculated and recorded as the difference mean of each cluster. Among all clusters, the cluster with the largest difference mean is taken as the significant set of each region; among them, if the difference index is larger, the difference in SPAD value of rice between different subregions in the region with the same nitrogen application amount is greater; conversely, if the difference index is smaller, the difference in SPAD value of rice between different subregions in the region with the same nitrogen application amount is smaller.

[0063] It should be noted that there are many commonly used clustering algorithms. In this embodiment, the k-means clustering algorithm is used to cluster the difference index. In actual application, as other implementation methods, the implementer may also adopt other clustering methods such as the DPC density peak clustering algorithm. This embodiment does not impose any special restrictions on the selection of clustering methods.

[0064] The k-means clustering algorithm is a well-known technology, and its clustering principle will not be described in detail.

[0065] Furthermore, for various rice varieties under any fitting method, the difference mean of the significant sets of all regions is used as the input of the threshold segmentation algorithm, and the segmentation threshold is output. The region where the significant set corresponds to the difference mean greater than the segmentation threshold is used as the error region, and the corresponding elements of the binary groups of all error regions in the error sequence are used as error elements.

[0066] For the error sequences of various rice varieties under any fitting method, the difference between each error element and the mean of all other elements in its subsequence is calculated, and the difference between all error elements is used as the error index of various rice varieties under any fitting method;

[0067] Furthermore, based on the error index, the fluctuation value, and the coefficient of determination of various rice varieties under any fitting method, the deviation values ​​of various rice varieties under any fitting method are determined, specifically:

[0068] Deviation value A of the uth rice variety under fitting method i u,i The expression is: Where B u,i represents the coefficient of determination of all pairs of rice varieties u under fitting method i; C u,i represents the fluctuation value of the u-th rice under fitting method i; D u,i represents the error index of the u-th rice under fitting method i; T u represents the number of all clusters of the u-th rice variety; exp() represents an exponential function with a natural constant as the base.

[0069] The process of obtaining the coefficient of determination according to the fitting method is a well-known technique, and the specific process will not be described in detail.

[0070] According to the deviation values ​​of various rice varieties under any fitting method, it can be understood that if T u When T is 1, that is, the number of clusters is 1, it means that the SPAD values ​​of the sub-regions under the same nitrogen application rate are relatively consistent, and there is no significant error in the original data. At this time, the deviation value can be determined by the determination coefficient and the fluctuation value. The smaller the fluctuation value and the larger the determination coefficient, the smaller the deviation value obtained, indicating that under the same nitrogen application rate, there is no significant difference in the SPAD value of rice in the sub-region. If T u When it is greater than 1, that is, when the number of clusters is greater than 1, it means that the SPAD value of rice in the sub-region is affected by the environment and an initial error occurs. Therefore, the deviation value is determined by the determination coefficient, fluctuation value and error index. At this time, the larger the fluctuation value, the smaller the determination coefficient, and the larger the error index, the larger the deviation value obtained, which reflects that under the same nitrogen application amount, there is an error in the initial SPAD value of the sub-region. Therefore, it is necessary to further evaluate the fitting effect of the rice area with error.

[0071] S203: Divide the fitting curves of various rice varieties under any fitting method into multiple curve segments, analyze the change trend of each curve segment, determine the change index of various rice varieties under any fitting method, determine the trend value of various rice varieties under any fitting method, and determine the optimal fitting model in combination with the deviation value.

[0072] By analyzing the data change characteristics between the SPAD value of rice and the nitrogen application rate during the optimal diagnosis period, it can be seen that for hybrid rice at the booting stage, with the increase of nitrogen application rate, the SPAD value of the top leaf of hybrid rice shows a trend of first increasing, then tending to be stable, and finally gradually decreasing; while for conventional rice at the panicle differentiation stage, with the increase of nitrogen application rate, the SPAD value of the top leaf of conventional rice shows an increasing trend.

[0073] Therefore, by analyzing the fitting between SPAD values ​​and nitrogen application rates of different rice varieties during their respective optimal diagnostic periods, the height fitting coefficient of each fitting method was determined, specifically:

[0074] Extract the peaks and valleys on the fitting curves of various rice varieties under any fitting method, connect adjacent peaks and valleys, calculate the slopes of the connected straight lines, and arrange the slopes of the connected straight lines between all adjacent peaks and valleys in ascending order according to the time of peak appearance to form a slope sequence of various rice varieties under any fitting method;

[0075] It should be noted that there are many commonly used peak and trough extraction algorithms. In this embodiment, an automatic multi-scale peak search algorithm is used to obtain the peaks and troughs on the fitting curve. In actual application, as other implementation methods, implementers may also adopt other methods. This embodiment does not impose any special restrictions on the selection of peak and trough extraction algorithms.

[0076] The automatic multi-scale peak search algorithm is a well-known technology, and its specific principle will not be described in detail.

[0077] Furthermore, based on the changing trends of all elements in the slope sequence, the change index is determined. The specific process is as follows:

[0078] The variation index R of the uth rice variety under fitting method i u,i The expression is: Where, E u,i F represents the ratio of the number of all negative elements to the number of all positive elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; u,i Z represents the ratio of the number of all positive elements to the number of all negative elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; u represents the classification of the u-th rice, Z u =1 indicates that the uth rice is hybrid rice, Z u =0 means that the u-th rice variety is conventional rice.

[0079] According to the change index of various rice varieties under any fitting method, it can be understood that the optimal diagnostic period for hybrid rice is the booting stage. During the optimal diagnostic period of hybrid rice, with the increase of nitrogen application rate, the SPAD value of the top leaf of hybrid rice shows a trend of first increasing, then stabilizing, and finally gradually decreasing. Therefore, when the fitting results are good, the number of negative elements in the first-order difference sequence of the slope sequence is greater than the number of positive elements. If the number of negative elements is less than the number of positive elements, it means that the current fitting method has a poor effect on the fitting of SPAD value and nitrogen application rate. For conventional rice in the panicle differentiation stage, with the increase of nitrogen application rate, the SPAD value of the top leaf of conventional rice shows an increasing trend. Therefore, when the fitting results are good, the number of positive elements in the first-order difference sequence of the slope sequence is greater than the negative elements. If the number of positive elements is less than the positive elements, it means that the current fitting method has a poor effect on the fitting of SPAD value and nitrogen application rate.

[0080] According to the change index of various rice varieties under any fitting method, the trend value of various rice varieties under any fitting method is further determined. The specific process is as follows:

[0081] The trend value M of the u-th rice variety under fitting method i u,i The expression is:

[0082] Where V u,i represents the number of all elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; It represents the difference between the qth element and its adjacent previous element in the first-order difference sequence of the slope sequence of the uth rice variety under fitting method i; exp() represents the exponential function with a natural constant as the base; norm[] represents the normalization function; max() represents the maximum value function.

[0083] According to the trend values ​​of various rice varieties under any fitting method, it can be understood that if the change index under the current fitting method is larger, and the difference between adjacent elements in the first-order difference sequence of the slope sequence is smaller, and the number of elements in the first-order difference sequence of the slope sequence is smaller, then the trend value is larger, indicating that the current fitting method has a better fitting effect between the SPAD value and the nitrogen application rate; conversely, if the change index under the current fitting method is smaller, and the difference between adjacent elements in the first-order difference sequence of the slope sequence is larger, and the number of elements in the first-order difference sequence of the slope sequence is larger, then the trend value is smaller, indicating that the current fitting method has a worse fitting effect between the SPAD value and the nitrogen application rate.

[0084] Furthermore, the trend value and deviation value are combined to determine the height fitting coefficient of various fitting methods, specifically:

[0085] Height fitting coefficient R of fitting method i iThe expression is: Where, represents the mean of the trend values ​​of all hybrid rice varieties under fitting method i; represents the mean of the trend values ​​of all conventional rice varieties under fitting method i; represents the mean of the deviation values ​​of all rice varieties under fitting method i; ∈ represents a preset constant greater than 0, which is used to prevent the denominator from being 0. In this embodiment, the value of ∈ is 0.01. Under the premise of ensuring that the denominator is not 0 and does not excessively affect the calculation results, the implementer can also set it according to the specific situation. This embodiment does not impose any special restrictions.

[0086] According to the height fitting coefficients of various fitting methods, it can be understood that if the trend value is larger and the deviation value is smaller when the current fitting method is used, the obtained height fitting coefficient is larger, which means that the current fitting method has a better fitting effect on the SPAD value and nitrogen application rate of rice; conversely, if the trend value is smaller and the deviation value is larger when the current fitting method is used, the obtained height fitting coefficient is smaller, which means that the current fitting method has a worse fitting effect on the SPAD value and nitrogen application rate of rice.

[0087] Finally, among all the fitting methods, the fitting method with the largest height fitting coefficient is selected as the optimal fitting model.

[0088] Preferably, the flowchart of the steps for obtaining the optimal fitting model provided in this embodiment is as follows: Figure 2 shown.

[0089] S3: By analyzing the differences between the SPAD values ​​of various rice regions during the optimal diagnosis period and the SPAD values ​​corresponding to the maximum yield, and combining the best fitting model, fertilization is carried out on various rice varieties.

[0090] In all rice regions during the optimal diagnosis period, the SPAD corresponding to the maximum yield is recorded as the optimal value of each rice region during the optimal diagnosis period. The SPAD value of each rice region during the optimal diagnosis period is compared with the optimal value to determine whether fertilization is needed in the corresponding region. Specifically,

[0091] If the SPAD value of the jth region of the uth variety of rice during the optimal diagnosis period is less than the preset critical value, it means that the uth variety of rice in the jth region is nitrogen deficient and needs to be supplemented with a large amount of nitrogen fertilizer; if the SPAD value of the jth region of the uth variety of rice during the optimal diagnosis period is greater than or equal to the critical value and less than the optimal value, it means that the nitrogen of the uth variety of rice in the jth region is relatively sufficient and a small amount of nitrogen fertilizer can be supplemented; if the SPAD value of the jth region of the uth variety of rice during the optimal diagnosis period is greater than or equal to the optimal value, it means that the nitrogen of the uth variety of rice in the i-th region is excessive and no nitrogen fertilizer is needed.

[0092] For rice areas that need additional nitrogen fertilizer, the deviation between the SPAD value of the corresponding rice in the area and the optimal value is used as the input of the optimal fit model, and the output is the nitrogen fertilizer content that needs to be supplemented.

[0093] It should be noted that the value of the preset critical value is set artificially, where the value of the preset critical value is smaller than the optimal value. In this embodiment, the value of the preset critical value is 0.9 times the optimal value. In actual application, as other implementation methods, the implementer can also combine it with a value within 0.9-0.95 and set it according to the specific situation. This embodiment does not impose any special restrictions.

[0094] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0095] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0096] The above-described embodiments are only used to illustrate the technical solutions of the present application, and not to limit them. Modifications to the technical solutions described in the aforementioned embodiments, or equivalent replacements of some of the technical features therein, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A method for diagnosing rice leaf failure by using a SPAD instrument, characterized in that: The method comprises the following steps: Divide the planting area of ​​various rice varieties into multiple regions, apply different amounts of nitrogen fertilizer to different regions, divide each region into multiple sub-regions, obtain the SPAD value, nitrogen application amount, and yield of each region in any growth period of various rice varieties, and obtain the SPAD value of each sub-region; Based on the correlation between SPAD value and yield, the optimal diagnosis period was selected from all growth periods of various rice varieties. All nitrogen application rates and SPAD values ​​of various rice varieties under the optimal diagnosis period were analyzed to determine the optimal fitting model, which is as follows: Different fitting methods were used to fit binary pairs consisting of nitrogen application rates and SPAD values ​​in all regions of various rice varieties. The distances of all binary pairs to the fitting curves obtained by any fitting method were used to form an error sequence. The error sequence was divided into multiple subsequences. The distribution of all elements in each subsequence was analyzed to determine the fluctuation value of various rice varieties under any fitting method. Within each region of various rice varieties, the deviation of the SPAD value within each subregion from the average SPAD value of all other subregions was used as a difference index. The difference indices of all subregions were clustered to obtain a significant set for each region. Based on the distribution of all difference indices within the significant set, error elements were screened from the error sequence. The difference between each error element and the distribution of all other elements in its subsequence was compared. Combined with the fluctuation value, the deviation value of each rice variety under any fitting method was determined. Analyze the changing trend of the fitting curves of various rice varieties under any fitting method, determine the changing index of various rice varieties under any fitting method, determine the trend value of various rice varieties under any fitting method, and determine the optimal fitting model in combination with the deviation value; The differences between the SPAD values ​​of various rice regions during the optimal diagnosis period and the SPAD values ​​corresponding to the maximum yield were analyzed, and fertilization was performed on various rice varieties based on the optimal fitting model.

2. The method for diagnosing a rice leaf failure by using a SPAD instrument as claimed in claim 1, wherein: The optimal diagnosis period is selected from all growth periods of various rice varieties, including: The yield and SPAD value of each region of various rice varieties at any growth stage are combined into two-dimensional data points, and the two-dimensional data points of all regions are used as inputs of a quadratic function regression equation, and the determination coefficient is output as the determination coefficient of various rice varieties at any growth stage; Among all growth periods of various rice varieties, the growth period corresponding to the maximum determination coefficient is taken as the optimal diagnosis period for each rice variety.

3. The method for diagnosing rice leaf failure by using a SPAD instrument as claimed in claim 1, wherein: The method for determining the fluctuation value of various rice varieties under any fitting method is as follows: For the error sequences of various rice varieties under any fitting method, the mean of all elements in each subsequence was calculated as the average error of each subsequence; The variance and cumulative sum of the average errors of all subsequences in the error sequence are calculated respectively, and the product of the variance and the cumulative sum is used as the fluctuation value of various rice varieties under any fitting method.

4. The method for diagnosing a rice leaf failure by using a SPAD instrument as claimed in claim 1, wherein: The acquisition process of the salient set of each region is as follows: The difference indexes of all sub-regions in each region of various rice varieties are clustered to obtain multiple clusters. The mean of all difference indices in each cluster is calculated and recorded as the difference mean of each cluster. Among all clusters, the cluster with the largest difference mean is taken as the significant set of each region.

5. The method for diagnosing rice leaf failure by using a SPAD instrument as claimed in claim 1, wherein: The process of filtering out all error elements from the error sequence is as follows: For various rice varieties under any fitting method, the difference mean of the significant sets of all regions is used as the input of the threshold segmentation algorithm, and the segmentation threshold is output. The region where the significant set corresponds to the difference mean greater than the segmentation threshold is used as the error region, and the corresponding elements of the binary groups of all error regions in the error sequence are used as error elements.

6. The method for diagnosing a rice leaf failure by using a SPAD instrument as claimed in claim 4, wherein: The method for determining the deviation values ​​of the various rice varieties under any fitting method is as follows: For the error sequences of various rice varieties under any fitting method, the difference between each error element and the mean of all other elements in its subsequence is calculated, and the difference between all error elements is used as the error index of various rice varieties under any fitting method; Deviation value A of the uth rice variety under fitting method i u,i The expression is: Where B u,i represents the coefficient of determination of all pairs of rice varieties u under fitting method i; C u,i represents the fluctuation value of the u-th rice under fitting method i; D u,i represents the error index of the u-th rice under fitting method i; T u represents the number of all clusters of the u-th rice variety; exp() represents an exponential function with a natural constant as the base.

7. The method for diagnosing rice leaf failure by using a SPAD instrument as claimed in claim 1, wherein: The expression of the variation index of various rice varieties under any fitting method is: Extract the peaks and valleys on the fitting curves of various rice varieties under any fitting method, connect adjacent peaks and valleys, calculate the slopes of the connected straight lines, and arrange the slopes of the connected straight lines between all adjacent peaks and valleys in ascending order according to the time of peak appearance to form a slope sequence of various rice varieties under any fitting method; The variation index R of the uth rice variety under fitting method i u,i The expression is: Where, E u,i F represents the ratio of the number of all negative elements to the number of all positive elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; u,i Z represents the ratio of the number of all positive elements to the number of all negative elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; u represents the classification of the u-th rice, Z u =1 indicates that the uth rice is hybrid rice, Z u =0 means that the u-th rice variety is conventional rice.

8. The method for diagnosing rice leaf failure by using a SPAD instrument as claimed in claim 7, wherein: The expression of the trend value of various rice varieties under any fitting method is: Where M u,i represents the trend value of the u-th rice under fitting method i; V u,i represents the number of all elements in the first-order difference sequence of the slope sequence of the u-th rice under fitting method i; It represents the difference between the qth element and its adjacent previous element in the first-order difference sequence of the slope sequence of the uth rice variety under fitting method i; exp() represents the exponential function with a natural constant as the base; norm[] represents the normalization function; max() represents the maximum value function.

9. The method for diagnosing rice leaf failure by using a SPAD instrument as claimed in claim 7, wherein: The method for determining the optimal fitting model is: Height fitting coefficient R of fitting method i i The expression is: Where, represents the mean of the trend values ​​of all hybrid rice varieties under fitting method i; represents the mean of the trend values ​​of all conventional rice varieties under fitting method i; represents the mean of the deviation values ​​of all rice varieties under fitting method i; ∈ represents a constant greater than 0; Among all the fitting methods, the fitting method with the largest height fitting coefficient is taken as the optimal fitting model.

10. The method for diagnosing rice leaf failure by using a SPAD instrument as claimed in claim 1, wherein: The fertilizing of various rice plants comprises: In all regions of various rice varieties within the optimal diagnostic period, the SPAD corresponding to the maximum yield is recorded as the optimal value of various rice varieties within the optimal diagnostic period. If the SPAD value of the jth region of the uth rice variety under the optimal diagnostic period is less than the optimal value of the uth rice variety, it is necessary to fertilize the uth rice variety in the jth region. The deviation between the SPAD value of the uth rice variety in the jth region and the optimal value is used as the input of the optimal fitting model, and the output is the nitrogen fertilizer content that needs to be supplemented; otherwise, it is not necessary to fertilize the uth rice variety in the jth region.

Images