Diversified rice planting method in composite environment based on big data analysis

Abstract

The invention discloses a diversified rice planting method in a compound environment based on big data analysis, and particularly relates to the technical field of agricultural information processing. The method comprises the following steps: collecting multi-dimensional monitoring data in a rice planting area, carrying out nonlinear correlation analysis, and extracting interaction characteristics among environment variables; carrying out multi-scale reconstruction analysis in combination with spatial scale difference, and identifying a main influence area of an environment variable; mapping analysis is carried out based on historical rice growth state data and environment variable interaction characteristics, and the comprehensive influence on rice growth is evaluated; analyzing the action time sequence of different environment variables in the rice growth cycle, and extracting the influence weight change rule of the environment variables in each growth stage; according to the method, environmental variable interaction characteristics and influence weight rules are fused, the growth potentials of different rice varieties are evaluated, the planting varieties and planting density of rice are optimized, and precise planting under the multi-variety collaborative adaptation and compound ecological condition of rice is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of agricultural information processing, more particularly, the present application relates to a diversified rice planting method in a complex environment based on big data analysis. BACKGROUND

[0002] In the existing agricultural planting management system, the environment monitoring and regulation of the rice planting area are mostly centered on typical climate factors, and static factor thresholds, linear weighted models or empirical classification rules are used for variety selection and planting configuration.

[0003] However, under the background of complex ecological environment, the growth of rice is affected by the coupling of multiple meteorological, soil and biological factors. These factors have high heterogeneity in time and space, and their interaction relationship presents nonlinear, feedback and stage characteristics. The existing analysis method lacks the dynamic analysis capability of the interaction between multiple variables and its evolution law, and it is difficult to accurately identify the key environmental response mechanism, resulting in uneven planting density layout and mislocation of variety adaptation.

[0004] In order to solve the above problems, a technical scheme is provided. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide a diversified rice planting method and system in a complex environment based on big data analysis to solve the problems raised in the background art.

[0006] To achieve the above object, the present application provides the following technical scheme:

[0007] A diversified rice planting method in a complex environment based on big data analysis, comprising the following steps:

[0008] S1: Obtain multi-dimensional monitoring data in the rice planting area, and perform nonlinear correlation analysis to extract the interaction characteristics between different environmental variables;

[0009] S2: Perform multi-scale reconstruction analysis on the interaction characteristics between different environmental variables based on spatial scale difference, and identify the main influence area of different environmental variables;

[0010] S3: According to the main influence area of different environmental variables, perform mapping analysis on historical rice growth state data, and evaluate the influence of the interaction characteristics between different environmental variables on rice growth;

[0011] S4: Analyze the action time sequence of different environmental variables in the rice growth cycle, and extract the influence weight change law of different environmental variables in different growth stages;

[0012] S5: based on the evaluation results of the influence of the interaction characteristics between different environmental variables on the growth of rice and the influence weight change law of different environmental variables at different growth stages, the growth potential of different rice varieties in the rice planting area is evaluated;

[0013] S6: according to the growth potential evaluation results, the planting variety and planting density of rice in the rice planting area are adjusted.

[0014] In a preferred embodiment, S1, specifically:

[0015] Collecting multi-dimensional monitoring data in the rice planting area; the multi-dimensional monitoring data includes meteorological environment monitoring data, soil environment monitoring data and rice growth state monitoring data;

[0016] The multi-dimensional monitoring data is subjected to outlier rejection, missing value interpolation and standardization processing;

[0017] Based on the multi-dimensional monitoring data after standardization processing, the non-linear correlation analysis is carried out by using mutual information analysis method, and the non-linear correlation strength between different environmental variables is calculated;

[0018] Based on the non-linear correlation strength between different environmental variables, the interaction characteristics between different environmental variables are extracted.

[0019] In a preferred embodiment, S2, specifically:

[0020] The rice planting area is divided into multiple spatial scale levels according to the spatial scale difference;

[0021] For each spatial scale level, the interaction characteristics between environmental variables are subjected to scale reconstruction analysis by using wavelet multi-scale analysis method, and the environmental variable interaction characteristic reconstruction data of each spatial scale level is obtained;

[0022] According to the environmental variable interaction characteristic reconstruction data, the spatial hot spot analysis method is used to identify and mark the main influence area of different environmental variables in each spatial scale level.

[0023] In a preferred embodiment, S3, specifically:

[0024] According to the main influence area of different environmental variables in each spatial scale level, the historical rice growth state data in the rice planting area is obtained;

[0025] For each main influence area in the spatial scale level, the spatial mapping correlation coefficient between the historical rice growth state data and different environmental variables is calculated respectively;

[0026] Based on the spatial mapping correlation coefficient, the nonlinear regression method is used to evaluate the comprehensive influence degree of the interaction characteristics between different environmental variables on the growth of rice, and the evaluation results of the influence of the interaction characteristics between different environmental variables on the growth of rice are obtained.

[0027] In a preferred embodiment, S4, specifically:

[0028] Based on the growth stages of rice in the whole growth cycle, the growth stages of rice include jointing and booting stage, heading and flowering stage and grain filling and ripening stage.

[0029] In each growth stage, the values of each environmental variable in the standardized multi-dimensional monitoring data are extracted, and the time series modeling of the value changes of the environmental variables in different growth stages is performed.

[0030] Based on the time series modeling, the influence weight values of each environmental variable in different growth stages are calculated, and the influence weight change rule of each environmental variable in different growth stages of rice is obtained.

[0031] In a preferred embodiment, S5, specifically:

[0032] The evaluation results of the influence of the interaction characteristics between different environmental variables on the growth of rice are fused with the influence weight change rule of each environmental variable in different growth stages of rice.

[0033] Based on the fusion result, a matching matrix of rice varieties and environmental response is constructed.

[0034] According to the matching matrix of rice varieties and environmental response, the growth potential of each rice variety in the rice planting area is evaluated, and the adaptability score of each rice variety under the current complex environment is output.

[0035] In a preferred embodiment, S6, specifically:

[0036] A predetermined adaptability score threshold is compared with the adaptability score, and the rice varieties with adaptability score higher than the adaptability score threshold are screened.

[0037] The planting range of each screened rice variety in the rice planting area is determined, and the planting density of each rice variety in the planting range is determined combined with the standardized multi-dimensional monitoring data.

[0038] The technical effects and advantages of the diversified rice planting method under complex environment based on big data analysis are as follows:

[0039] By nonlinear correlation analysis on multi-dimensional monitoring data, the interaction characteristics among environmental factors such as weather, soil and crop state can be deeply mined, and the recognition ability of complex coupled disturbance can be significantly improved; by multi-scale reconstruction analysis, the main influence area of different environmental variables is identified, and spatial support is provided for differentiated management in different regions; the main influence area is mapped to the historical rice growth state data, the quantitative evaluation of the influence of environmental interaction effect on rice growth is realized, and the data foundation for accurate adjustment of the target area is laid; combined with time series analysis in the growth cycle, the weight change of different environmental variables in the growth cycle of rice is clarified, which provides basis for accurate regulation; the spatial and time series evaluation results are fused, the growth potential of each rice variety under the current complex environment is evaluated, and reliable basis is provided for multi-variety collaborative planting; according to the growth potential evaluation result, the variety configuration and planting density are dynamically optimized, the efficient use of resources and the steady increase of yield are realized, and the accuracy and response speed of planting decision are significantly improved, which meets the planting demand under complex ecological conditions. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 A multi-element rice planting method under a complex environment based on big data analysis is given. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0042] EMBODIMENT

[0043] Figure 1 A multi-element rice planting method under a complex environment based on big data analysis is given, which includes the following steps:

[0044] S1: Obtain multi-dimensional monitoring data in the rice planting area, and perform nonlinear correlation analysis to extract the interaction characteristics between different environmental variables;

[0045] S2: Based on the spatial scale difference, the multi-scale reconstruction analysis is performed on the interaction characteristics between different environmental variables to identify the main influence area of different environmental variables;

[0046] S3: According to the main influence area of different environmental variables, the historical rice growth state data is mapped and analyzed to evaluate the influence of the interaction characteristics between different environmental variables on rice growth;

[0047] S4: Analyze the timing of the effects of different environmental variables on the growth cycle of rice, and extract the influence weight change law of different environmental variables at different growth stages;

[0048] S5: Based on the evaluation results of the influence of the interaction characteristics between different environmental variables on the growth of rice and the influence weight change law of different environmental variables at different growth stages, evaluate the growth potential of different rice varieties in the rice planting area;

[0049] S6: According to the growth potential evaluation results, adjust the planting varieties and planting density of rice in the rice planting area.

[0050] S1: Obtain multi-dimensional monitoring data in the rice planting area, and perform nonlinear correlation analysis to extract the interaction characteristics between different environmental variables, including:

[0051] Collect multi-dimensional monitoring data in the rice planting area; the multi-dimensional monitoring data includes meteorological environmental monitoring data, soil environmental monitoring data and rice growth state monitoring data;

[0052] Specifically, a plurality of environmental monitoring collection devices are arranged in the rice planting area, including meteorological environmental monitoring devices, soil environmental monitoring devices and rice growth state monitoring devices. Among them, the meteorological environmental monitoring device includes an air temperature sensor, an air humidity sensor, a light intensity sensor and a wind speed sensor, the soil environmental monitoring device includes a soil moisture content sensor, a soil organic matter content detector and a soil pH sensor, and the rice growth state monitoring device includes a rice plant height measuring instrument, a leaf area index measuring device and a chlorophyll content measuring instrument. Each device is provided with a plurality of monitoring nodes to ensure that the collected data covers the entire rice planting area.

[0053] The meteorological environmental monitoring device records the data of air temperature, air humidity, light intensity and wind speed continuously with a cycle of minutes; the soil environmental monitoring device records the data of soil moisture content, soil organic matter content and soil pH with a frequency of once every hour; the rice growth state monitoring device records the data of rice plant height, leaf area index and chlorophyll content with a frequency of once every day. The collection of all the above data is called multi-dimensional monitoring data.

[0054] The multi-dimensional monitoring data is subjected to outlier rejection, missing value interpolation and standardization processing;

[0055] Specifically, the outlier rejection is processed by the boxplot method, that is, the data distribution of each environmental variable and crop growth variable is calculated by the boxplot statistical method, and the upper and lower boundaries of each variable data are calculated. The data points outside the boundary range are removed as outliers. The missing value interpolation adopts the method of average interpolation based on adjacent spatial data, that is, when the monitoring data is missing, the average value of the same type of variable data of multiple monitoring sites closest to the missing data in space is used to interpolate the missing data to ensure the continuity and accuracy of the monitoring data. The standardization processing adopts the extreme value normalization processing method, that is, the data of each variable is normalized to between zero and one by the difference between the maximum value and the minimum value, so as to eliminate the dimension effect of the data.

[0056] Based on the multi-dimensional monitoring data after standardization processing, the mutual information analysis method is used for nonlinear correlation analysis to calculate the nonlinear correlation strength between different environmental variables.

[0057] Specifically, the mutual information analysis method is used to calculate the nonlinear correlation strength between each two different environmental variables. The calculation process of mutual information is as follows: the data entropy of each environmental variable is calculated, which is the sum of the probability distribution of the environmental variable and the logarithmic product of its probability; the joint entropy of the data pair composed of each two different environmental variables is calculated, that is, the sum of the probability of the simultaneous occurrence of the two environmental variables and the logarithmic product of its probability; the mutual information value: the data entropy of the two environmental variables is added respectively, and then the joint entropy of the data pair of the two environmental variables is subtracted, that is, the mutual information value between the two environmental variables is obtained. The greater the mutual information value, the higher the nonlinear correlation strength between the two environmental variables.

[0058] Based on the nonlinear correlation strength between different environmental variables, the interaction features between different environmental variables are extracted.

[0059] Specifically, according to the preset mutual information threshold, different environmental variable pairs with significant correlation are screened out. The environmental variable pair with significant correlation refers to the environmental variable combination with the mutual information value higher than the mutual information threshold.

[0060] Interaction features are extracted from the environmental variable pairs with significant correlation. The interaction features are the nonlinear correlation strength of the environmental variable combination with significant correlation, the coupling relationship mode between them and the coupling trend changing with time.

[0061] The coupling relationship mode between each other refers to that when two environmental variables present an upward trend or a downward trend at the same time, the environmental variable pair is defined as a positive coupling mode; when one environmental variable presents an upward trend and the other presents a downward trend, the environmental variable pair is defined as a negative coupling mode; otherwise, the environmental variable pair is defined as a non-monotonic coupling mode. For example, air humidity and leaf area index present a positive coupling mode when they both rise or both fall at the same time.

[0062] The coupling trend over time refers to the dynamic change characteristics of the coupling relationship over time. For example, the regression curve slope between air temperature and soil moisture content is small at the jointing and booting stage, and the regression curve slope significantly increases at the heading and flowering stage, so it can be determined that the coupling relationship between air temperature and soil moisture content becomes more significant from the jointing and booting stage to the heading and flowering stage, and has dynamic trend characteristics with the change of growth stage.

[0063] S2: Based on the spatial scale difference, the interaction characteristics between different environmental variables are analyzed by multi-scale reconstruction, and the main influence area of different environmental variables is identified, including:

[0064] According to the spatial scale difference, the rice planting area is divided into multiple spatial scale levels;

[0065] Specifically, the rice planting area is divided into three spatial scale levels of micro scale, meso scale and macro scale. The micro scale spatial level is defined as a local area in the range of several meters to tens of meters, the meso scale spatial level is defined as a plot level area in the range of several hundred meters, and the macro scale spatial level is defined as an area in the range of kilometers.

[0066] For each spatial scale level, the wavelet multi-scale analysis method is used to analyze the scale reconstruction of the interaction characteristics between environmental variables, and the environmental variable interaction characteristic reconstruction data of each spatial scale level is obtained;

[0067] Specifically, the wavelet multi-scale analysis method is to decompose the interaction characteristics between different environmental variables into different scales by wavelet function, and obtain the variable interaction characteristic reconstruction data exhibited at different spatial scales. Specifically, the spatial position data corresponding to different spatial scale levels is extracted from the interaction characteristics between different environmental variables and used as input data; the Morlet wavelet of continuous wavelet transform is selected as the wavelet basis function; the interaction characteristics between different environmental variables of each spatial scale level are respectively subjected to continuous wavelet transform by the Morlet wavelet, so as to obtain the scale components of the interaction characteristics between different environmental variables at different scales; the scale components of the same spatial scale level are recombined by using the reconstruction method of the scale components, so as to form the environmental variable interaction characteristic reconstruction data of the corresponding spatial scale level.

[0068] The environmental variable interaction feature reconstruction data represents the distribution characteristics of the interaction features between different environmental variables changing with the spatial position. For example, at the micro-scale spatial scale level, the environmental variable interaction feature can be dominated by the interaction between the local soil moisture of the field and the air humidity, while at the macro-scale spatial scale level, the overall spatial distribution pattern between the soil organic matter content and the regional climate variables can be represented.

[0069] According to the environmental variable interaction feature reconstruction data, a spatial hot spot analysis method is used to identify and mark the main influence area of different environmental variables at each spatial scale level.

[0070] Specifically, the implementation process of the spatial hot spot analysis method is as follows: spatial autocorrelation calculation is performed on the environmental variable interaction feature reconstruction data, and the spatial autocorrelation calculation method is the calculation of the local spatial autocorrelation coefficient. Specifically, the spatial grid in each spatial scale level is taken as the unit, the spatial neighborhood structure of each grid cell is defined, and the eight adjacent grids around it are selected as the analysis neighborhood. The environmental variable interaction feature reconstruction data of each grid cell is standardized, the product of the standardized values between the current cell and the neighborhood cells is calculated, and the total variance of all cells is normalized, and finally the local spatial autocorrelation coefficient is obtained, which is used to identify the spatial hot spot area. According to the calculation result of the local spatial autocorrelation, it is judged whether the local spatial autocorrelation coefficient exceeds the significance threshold value, and the significance threshold value is determined according to the Monte Carlo simulation calculation. The spatial position exceeding the significance threshold value is defined as the spatial hot spot area, and the spatial position not exceeding the significance threshold value is defined as the non-hot spot area.

[0071] In the spatial hot spot area of each spatial scale level, the main influence area of each environmental variable is identified and marked. For example, in the micro-scale spatial level, the interaction feature of soil moisture content and air humidity is identified as a hot spot area; while in the macro-scale spatial scale level, the interaction feature of light intensity and wind speed is identified as a hot spot area.

[0072] S3: According to the main influence area of different environmental variables, the historical rice growth state data is mapped and analyzed to evaluate the influence of the interaction features between different environmental variables on the growth of rice, including:

[0073] According to the main influence area of different environmental variables at each spatial scale level, the historical rice growth state data in the rice planting area is obtained.

[0074] Specifically, according to the different environmental variables identified and marked in the micro-scale spatial scale hierarchy, the meso-scale spatial scale hierarchy and the macro-scale spatial scale hierarchy, the corresponding historical rice growth state data, including the plant height, leaf area index and chlorophyll content of the rice plant, are collected. The rice growth state data is the observation data of the corresponding spatial region in the past complete rice planting cycle, and the observation frequency is regular observation in each rice growth cycle, for example, once a week, to ensure that the observation data time series is continuous and complete.

[0075] For each main influence region in each spatial scale hierarchy, the spatial mapping correlation coefficient between the historical rice growth state data and different environmental variables is calculated respectively;

[0076] Specifically, based on the main influence region of each spatial scale hierarchy, the spatial mapping correlation coefficient between the historical rice growth state data and each environmental variable is calculated respectively to quantify the spatial influence degree of the environmental variable on the rice growth. The spatial mapping correlation coefficient calculation method is: for each main influence region of each environmental variable in each spatial scale hierarchy, the environmental variable interaction feature reconstruction data is used as the spatial variable, and the historical rice growth state data is used as the target variable, to calculate the correlation degree between the two variables at the spatial position respectively; the spatial bivariate correlation analysis method is used, that is, the spatial mapping correlation coefficient is obtained by calculating the degree of change of the environmental variable and the rice growth state variable at the same spatial position. The calculation is first to calculate the standardized values of the environmental variable and the rice growth state variable in space respectively, then to calculate the average value of the product of the two sets of spatial standardized values, and finally the average value is obtained. The spatial mapping correlation coefficient close to one indicates that the environmental variable and the rice growth state data present a high spatial correlation.

[0077] For example, in the micro-scale spatial scale hierarchy, the spatial mapping correlation coefficient between the soil moisture content and the rice leaf area index is calculated; in the meso-scale spatial scale hierarchy, the spatial mapping correlation coefficient between the air humidity and the rice plant height is calculated; in the macro-scale spatial scale hierarchy, the spatial mapping correlation coefficient between the light intensity and the rice chlorophyll content is calculated. The spatial mapping correlation coefficient of each spatial scale hierarchy is calculated separately to avoid interference between scales.

[0078] Based on the spatial mapping correlation coefficient, a nonlinear regression method is used to evaluate the comprehensive influence degree of the interaction features between different environmental variables on the rice growth, and the evaluation result of the influence of the interaction features between different environmental variables on the rice growth is obtained;

[0079] Specifically, the interaction characteristics between the spatial mapping correlation coefficient and different environmental variables are taken as independent variables, and the rice growth state monitoring data are taken as dependent variables, and a nonlinear regression model is selected for fitting. First, a set of candidate nonlinear regression models is determined, including exponential function type nonlinear model, logarithmic function type nonlinear model and power function type nonlinear model; the parameter optimization of the candidate model is performed by using cross-validation method, the cross-validation method is to divide the historical observation data into multiple subsets, one part is used for model fitting, and the other part is used for model precision test, the fitting and verification are alternately performed, and the fitting degree of the model is evaluated; the parameter configuration scheme is determined from the model with the highest fitting degree, and the model parameter determination method is least square method, that is, the residual sum of squares between the predicted value of the nonlinear regression model and the actual historical rice growth state data is minimized by adjusting the parameters.

[0080] After completing the nonlinear regression model fitting, the evaluation results of the influence degree of the interaction between the environmental variables on the rice growth state are obtained through the model parameters and the nonlinear regression function, and the evaluation results are embodied as the proportion or trend of the change of the rice plant height, leaf area index and chlorophyll content data caused by the change of the environmental variable interaction value.

[0081] For example, the nonlinear regression analysis may find that the interaction of soil water content and air humidity is increased by ten percent, which will cause the average increase of rice leaf area index by five percent; and in another spatial scale level, the interaction of light intensity and soil organic matter content is increased by ten percent, which will cause the increase of rice chlorophyll content by three percent; the evaluation results of the influence of the interaction between different environmental variables on the rice growth.

[0082] S4: analyze the time sequence of the effects of different environmental variables on the rice growth cycle, and extract the change rule of the influence weight of different environmental variables in different growth stages, including:

[0083] The growth stages of rice are divided based on the complete growth cycle of rice, and the growth stages of rice include jointing and booting stage, heading and flowering stage and grain filling and ripening stage.

[0084] Specifically, according to the growth law of rice crops in actual production and the practical experience of agricultural planting management, the complete growth cycle of rice is reasonably divided into three typical growth stages of jointing and booting stage, heading and flowering stage and grain filling and ripening stage. The jointing and booting stage starts from the beginning of the stem of rice, to the end of the formation of the rice ear; the heading and flowering stage starts from the ear of rice out of the leaf sheath and begins to pull out, to the end of the complete flowering of the ear; the grain filling and ripening stage starts from the completion of the ear flowering, and the grain filling and development of rice starts, and ends until the completion of the grain filling before maturity.

[0085] At each growth stage, the values of each environmental variable in the standardized multi-dimensional monitoring data are extracted respectively, and the time series modeling of the value changes of the environmental variables in different growth stages is performed;

[0086] Specifically, for the jointing and booting stage, the heading and flowering stage, and the grain filling and fruiting stage, the values of all environmental variables in each stage are independently extracted from the multi-dimensional monitoring data after outlier removal, missing value interpolation and standardization, including meteorological environmental monitoring data, soil environmental monitoring data and rice growth state monitoring data. Among them, the meteorological environmental monitoring data are the values of air temperature, air humidity, light intensity and wind speed, the soil environmental monitoring data are the values of soil water content, soil organic matter content and soil acid-base degree, and the rice growth state monitoring data are the values of rice plant height, leaf area index and chlorophyll content. The extraction method is: according to the start and end dates corresponding to each stage, the standardized data are extracted day by day to form continuous value sequences of environmental variables in each stage, and data sets for each stage are constructed.

[0087] The data sets for the jointing and booting stage, the heading and flowering stage, and the grain filling and fruiting stage are respectively subjected to time series modeling. The time series modeling analysis method is the autoregressive moving average model. The modeling method of the autoregressive moving average model is: the values of the environmental variables in each growth stage are regarded as time series, the autocorrelation relationship between the current variable value and the variable values at multiple previous time points and the moving average relationship between the variable value and the random disturbance items at multiple previous time points are calculated according to the position of the variable value on the time axis; then, the optimal parameters of the autoregressive term order and the moving average term order in the model are determined by the autocorrelation function and the partial autocorrelation function; the optimal value of the model parameters is determined by the maximum likelihood estimation method; finally, the accuracy of the model is evaluated by residual analysis and goodness-of-fit test method to ensure that the model meets the actual situation of rice growth.

[0088] For example, for the air temperature environmental variable in the jointing and booting stage, the modeling process of the autoregressive moving average model is: the value sequence of the air temperature in the jointing and booting stage is calculated, the correlation degree of the current air temperature value and the air temperature values of the past continuous days is analyzed, and the most suitable order and model parameters are determined by maximum likelihood estimation, so as to obtain the dynamic model of the air temperature variable changing with the time of the jointing and booting stage.

[0089] Based on the time series modeling, the influence weight values of each environmental variable in different growth stages are calculated to obtain the influence weight change law of each environmental variable in different growth stages of rice;

[0090] Specifically, after the time series modeling is completed, the influence weight values of each environmental variable in the jointing and booting stage, the heading and flowering stage, and the grain filling and fruiting stage are calculated. The calculation method of the influence weight value is as follows: first, the covariance between the fluctuation amplitude of the environmental variable value and the fluctuation amplitude of the rice growth state variable value is calculated through the time series model of each environmental variable; then the covariance is divided by the sum of the covariances of all environmental variables, thereby obtaining the standardized influence weight value of the environmental variable in the corresponding growth stage. The influence weight value of each environmental variable represents the contribution size of the environmental variable to the change of the rice growth state.

[0091] The influence weight values of each environmental variable obtained in the jointing and booting stage, the heading and flowering stage, and the grain filling and fruiting stage are arranged in time sequence to form the weight change sequence of each environmental variable in different growth stages, which shows the dynamic evolution law of the influence mode of the environmental variable on the rice growth stage.

[0092] For example, through the above analysis, the air temperature environmental variable may have a higher influence weight on rice growth in the jointing and booting stage, and the weight gradually decreases in the heading and flowering stage and the grain filling and fruiting stage; while the light intensity environmental variable has a lower influence weight in the jointing and booting stage, and gradually increases in the grain filling and fruiting stage.

[0093] The influence weight change law can clearly express the change of the importance of different environmental variables in different growth stages of rice, and provide decision basis for the growth potential evaluation of rice varieties and the formulation of planting management scheme.

[0094] S5: Based on the evaluation results of the influence of the interaction characteristics between different environmental variables on rice growth and the influence weight change law of different environmental variables in different growth stages, the growth potential of different rice varieties in the rice planting area is evaluated, including:

[0095] The evaluation results of the influence of the interaction characteristics between different environmental variables on rice growth are fused with the influence weight change law of each environmental variable in different growth stages of rice;

[0096] Specifically, according to the evaluation results of the influence of the interaction characteristics between different environmental variables on rice growth and the change rule of the influence weight of each environmental variable in the jointing and booting stage, the heading and flowering stage and the grain filling and ripening stage, data fusion operation is implemented. The fusion method is a weighted data fusion method, and the implementation process of the weighted data fusion method is as follows: for each environmental variable, the influence weight value of the environmental variable in each growth stage is taken as a weight coefficient, and the influence weight value of the environmental variable in each growth stage is multiplied by the evaluation value of the interaction of the environmental variable and other variables on the growth of rice to obtain the weighted interaction influence value of the environmental variable; and then the weighted interaction influence values of all environmental variables are summed up, and the summed value is the fusion result. The weighted data fusion method can reasonably consider the spatial characteristics of the interaction between environmental variables and the dynamic change characteristics of each environmental variable in different periods, so as to accurately reflect the influence mode of the comprehensive action of environmental variables on rice growth.

[0097] For example, taking air temperature and air humidity as two environmental variables, in the jointing and booting stage, firstly, the influence weight value of air temperature on rice growth is multiplied by the influence evaluation value of the interaction between air temperature and air humidity; then the influence weight value of air humidity on rice growth is multiplied by the influence evaluation value of the interaction between air humidity and air temperature; and then all the weighted values obtained above are summed up to form the fusion result of the interaction of air temperature and air humidity environmental variables in the jointing and booting stage. The data fusion process in the heading and flowering stage and the grain filling and ripening stage is the same as the above method, and fusion calculation is performed respectively.

[0098] Based on the fusion result, a matching matrix of rice varieties and environmental response is constructed;

[0099] Specifically, different rice varieties to be evaluated in the rice planting area are taken as the rows of the matching matrix; and the comprehensive influence environmental scenarios formed by the combination of the jointing and booting stage, the heading and flowering stage and the grain filling and ripening stage and each environmental variable are taken as the columns of the matching matrix; and the adaptability of each rice variety in each comprehensive influence environmental scenario is scored one by one. The adaptability score is determined by the similarity between the historical planting performance data of the rice variety and the fusion result, that is, the cosine value between the historical growth state performance vector of each rice variety and the fusion result vector is calculated by the cosine similarity algorithm, and the larger the cosine value is, the higher the adaptability of the rice variety to the corresponding environmental scenario is. Finally, all the adaptability score values formed by the above method constitute a complete matching matrix of rice varieties and environmental response.

[0100] The calculation method of the cosine similarity algorithm is: first, the growth performance data of each rice variety under historical environmental conditions, such as plant height, leaf area index and chlorophyll content, are taken as the historical growth state performance vector; then the current environmental variable fusion result is taken as the current environmental feature vector; the inner product of the historical performance vector and the current environmental feature vector, as well as the length of the historical performance vector and the length of the current environmental feature vector, are calculated respectively; finally, the cosine value is determined by dividing the inner product of the vectors by the product of the lengths of the two vectors. The calculation result of the cosine value ranges between negative one and positive one, and the closer the value is to positive one, the higher the matching degree of the rice variety and the environmental condition, and the higher the adaptability score.

[0101] According to the matching matrix of the rice variety and the environmental response, the growth potential of a plurality of rice varieties in the rice planting area is evaluated, and the adaptability score of each rice variety under the current compound environment is output;

[0102] Specifically, the adaptability score in the matching matrix is used to evaluate the growth potential of a plurality of rice varieties in the rice planting area. The growth potential evaluation method is: for each rice variety, the adaptability scores of the jointing and squaring stage, the heading and flowering stage, and the grain filling stage are arithmetically averaged, that is, the adaptability scores of the three stages are added and then divided by the number of scores, to obtain the adaptability score of each rice variety under the current compound environment. The higher the adaptability score, the stronger the growth potential of the rice variety under the compound environment.

[0103] The current compound environment refers to a multifunctional agricultural ecological system mainly composed of rice fields, which has a highly intertwined compound characteristic in terms of spatial structure, resource utilization and ecological elements. It is usually distributed in semi-mountainous or hilly areas with large terrain undulations and various types of cultivated land. In the same cultivated unit, rice and other dry grain crops such as sorghum are planted in three dimensions, and fish, shrimp, turtle and other aquaculture activities are carried out in combination with the water layer conditions of the rice field.

[0104] S6: According to the growth potential evaluation result, the planting variety and planting density of rice in the rice planting area are adjusted, including:

[0105] A preset adaptability score threshold is compared with the adaptability score to screen rice varieties with adaptability scores higher than the adaptability score threshold.

[0106] Specifically, the adaptability score threshold for variety screening is set. The adaptability score threshold is set by a dynamic setting method based on statistical quantile: first, the comprehensive adaptability scores of all rice varieties are sorted in descending order; then, according to the set target retention ratio (for example, thirty percent), the value of the adaptability score at the quantile point is determined as the adaptability score threshold.

[0107] After setting the adaptive score threshold, the comprehensive adaptive scores of all rice varieties are screened, and the rice varieties with the comprehensive adaptive scores higher than the adaptive score threshold are retained as the target planting variety set in the current complex environment.

[0108] The planting range of each screened rice variety is determined in the rice planting area, and the planting density of each rice variety in the planting range is determined based on the standardized multi-dimensional monitoring data.

[0109] Specifically, based on the target planting variety set, spatial positioning and planting range demarcation are performed in the rice planting area: the standardized multi-dimensional monitoring data are input into the matching matrix of each rice variety and the environment to determine the matching degree of each spatial grid with each target rice variety under the current environment. Each spatial grid only retains the target rice variety with the highest matching degree, and the spatial grid is demarcated as the recommended planting range of the corresponding rice variety.

[0110] For example, in a complex ecological planting area with high air humidity, stable soil water content, and weak light intensity, input the multi-dimensional monitoring data, and through comparison with the matching matrix of multiple screened rice varieties, it is found that the matching degree of a certain rice variety in the complex ecological planting area is higher than that of other rice varieties. Therefore, the complex ecological planting area is demarcated as the recommended planting range of the corresponding rice variety.

[0111] After determining the planting range of each rice variety, the specific planting density and management mode are determined in the planting range. The determination of the planting density is based on the combination of the growth characteristics of the rice variety and the local resource and environmental carrying capacity: the combination state of the air temperature, air humidity, light intensity, and soil water content in the planting area is taken as the input feature; the best density interval is consulted by combining the growth habit database of the variety, and the recommended planting density of the target plot is determined by interval mapping. For example, the suitable density of a certain variety under low light and high humidity conditions is thirty plants per square meter, and the suitable density under high light and low humidity conditions is adjusted to twenty-five plants per square meter.

[0112] The above formulas are dimensionless numerical calculations, and the formulas are obtained by software simulation of a large amount of data to obtain a formula closest to the real situation. The preset parameters and threshold values in the formula are set by a person skilled in the art according to the actual situation.

[0113] The above-described embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented by software, the above-described embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are wholly or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center through a wired (for example, infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. containing one or more available medium collections. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state disk.

[0114] Those of ordinary skill in the art can realize that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0115] Those of ordinary skill in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, device, and module can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0116] In several embodiments provided in the present application, it should be understood that the disclosed system, device, and method can be implemented in other ways. For example, the above-described device embodiments are only schematic, for example, the division of the modules is only a logical function division, and actual implementation can have another division manner, for example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed ones can be indirect coupling or communication connection through some interfaces, devices, or modules, which can be electrical, mechanical, or other forms.

[0117] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment.

[0118] In addition, the functional modules in each embodiment of the present application can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0119] If the functions are realized in the form of software function modules and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application can be embodied in the form of software products, and the computer software products are stored in a storage medium, including a plurality of instructions for making a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, and various program code storage media.

[0120] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0121] Finally: the above is only the preferred embodiment of the present application, and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application, should be included in the protection scope of the present application.

Claims

1. A diversified rice planting method in a complex environment based on big data analysis, characterized in that: The steps include: S1: Obtain multidimensional monitoring data within the rice-growing area and perform nonlinear correlation analysis to extract the interaction characteristics between different environmental variables; S2: Conduct multi-scale reconstruction analysis of the interaction characteristics between different environmental variables based on spatial scale differences to identify the main impact areas of different environmental variables; S3: Mapping analysis of historical rice growth status data based on the main impact areas of different environmental variables to evaluate the impact of the interaction characteristics between different environmental variables on rice growth; S4: Analyze the temporal effects of different environmental variables during the rice growth cycle and extract the changing patterns of the influence weights of different environmental variables at different growth stages; S5: Evaluate the growth potential of different rice varieties within a rice-growing area based on the evaluation results of the interaction characteristics of different environmental variables on rice growth and the changing patterns of the influence weights of different environmental variables at different growth stages; S6: Adjust the rice varieties and planting density within the rice-growing area based on the growth potential assessment results.

2. The diversified rice planting method under a complex environment based on big data analysis according to claim 1, characterized in that: S1, specifically: Collect multi-dimensional monitoring data within the rice-growing area; the multi-dimensional monitoring data includes meteorological environment monitoring data, soil environment monitoring data, and rice growth status monitoring data; Eliminate outliers, interpolate missing values ​​and standardize multidimensional monitoring data; Based on the standardized multidimensional monitoring data, the mutual information analysis method was used to conduct nonlinear correlation analysis and calculate the nonlinear correlation strength between different environmental variables. Based on the nonlinear correlation strength between different environmental variables, the interaction characteristics between different environmental variables are extracted.

3. The diversified rice planting method under a complex environment based on big data analysis according to claim 2, characterized in that: S2, specifically: The rice-growing area is divided into multiple spatial scale levels according to the spatial scale differences; For each spatial scale level, the wavelet multi-scale analysis method is used to reconstruct the scale of the interaction characteristics between environmental variables, and the reconstructed data of the interaction characteristics of environmental variables at each spatial scale level are obtained; The data were reconstructed based on the interaction characteristics of environmental variables, and the spatial hotspot analysis method was used to identify and mark the main impact areas of different environmental variables at each spatial scale level.

4. The diversified rice planting method under a complex environment based on big data analysis according to claim 3, characterized in that: S3, specifically: Obtain historical rice growth status data within the rice-growing area based on the main impact areas of different environmental variables at each spatial scale level; For the main impact areas at each spatial scale level, the spatial mapping correlation coefficients between historical rice growth status data and different environmental variables were calculated; Based on the spatial mapping correlation coefficient, the nonlinear regression method was used to evaluate the comprehensive impact of the interaction characteristics between different environmental variables on rice growth, and the evaluation results of the impact of the interaction characteristics between different environmental variables on rice growth were obtained.

5. The diversified rice planting method under a complex environment based on big data analysis according to claim 4 is characterized in that: S4, specifically: The rice growth stages are divided based on the complete growth cycle of rice, which includes the jointing and booting stage, heading and flowering stage, and grain filling and fruiting stage. At each growth stage, the values ​​of each environmental variable in the standardized multidimensional monitoring data were extracted, and time series modeling was performed on the changes in the values ​​of environmental variables at different growth stages. Based on time series modeling, the impact weight values ​​of various environmental variables in different growth stages were calculated, and the changing patterns of the impact weights of various environmental variables in different growth stages of rice were obtained.

6. The diversified rice planting method under a complex environment based on big data analysis according to claim 5, characterized in that: S5, specifically: The evaluation results of the interaction characteristics between different environmental variables on rice growth are integrated with the change rules of the influence weight of each environmental variable in different rice growth stages; Based on the fusion results, a matching matrix between rice varieties and environmental responses was constructed; Based on the matching matrix between rice varieties and environmental responses, the growth potential of multiple rice varieties in the rice-growing area is evaluated separately, and the adaptability score of each rice variety in the current complex environment is output.

7. The diversified rice planting method under a complex environment based on big data analysis according to claim 6, characterized in that: S6, specifically: Presetting an adaptability score threshold, comparing the adaptability score with the adaptability score threshold, and screening rice varieties with adaptability scores higher than the adaptability score threshold; Determine the planting range of each screened rice variety within the rice-growing area, and determine the planting density of each rice variety within the planting range based on the standardized multi-dimensional monitoring data.

Images