Crop water-saving irrigation method and device, electronic equipment and storage medium

Abstract

This invention relates to the field of agricultural water-saving irrigation technology, and more particularly to a method, device, electronic device, and storage medium for water-saving irrigation of crops. The method first acquires multiple factors affecting crop growth; then, it analyzes the crop growth rate based on a crop growth data queue to obtain the crop's growth characteristics and growth stages; next, it extracts periodic features from the multiple factor data queues to obtain multiple first-factor features; finally, it inputs the growth characteristics, growth stages, and multiple factor features into a growth trend analysis model, and optimizes the irrigation strategy through the growth trend analysis model. This invention optimizes the irrigation strategy by inputting the analyzed crop growth characteristics, growth stages, and the state characteristics of the factors into a growth trend analysis model constructed using an artificial neural network, achieving the dual goals of ensuring yield and water conservation.

Description

Technical Field

[0001] This invention relates to the field of agricultural water-saving irrigation technology, and in particular to a method, apparatus, electronic device and storage medium for water-saving irrigation of crops. Background Technology

[0002] With the increasing severity of global climate change and water scarcity, water-saving irrigation technology has become an important direction for global agricultural development. This is especially true in arid and semi-arid regions, where the development and application of water-saving irrigation technology are of paramount importance. The promotion and application of water-saving irrigation technology is of great significance.

[0003] Water-saving irrigation technology has evolved from traditional irrigation methods to modern and intelligent systems. Traditional irrigation methods, such as flood irrigation, not only waste water resources but also easily lead to problems such as soil salinization. Modern water-saving irrigation technologies, such as sprinkler irrigation, drip irrigation, and micro-sprinkler irrigation, significantly improve the efficiency of irrigation water use by reducing irrigation water loss and evaporation.

[0004] However, due to a lack of professional knowledge and special skills, farmers are unable to choose appropriate water-saving irrigation methods when irrigating in different environments; they are also unable to determine the impact of irrigation and fertilization schemes on crop yields.

[0005] Therefore, it is necessary to develop and design a water-saving irrigation method for crops. Summary of the Invention

[0006] The present invention provides a method, apparatus, electronic device and storage medium for water-saving irrigation of crops, which solves the problem that existing crop irrigation strategies are difficult to balance water conservation and crop yield.

[0007] In a first aspect, embodiments of the present invention provide a water-saving irrigation method for crops, comprising:

[0008] To identify multiple factors influencing crop growth;

[0009] The growth rate of crops is analyzed based on the crop growth data queue to obtain the growth characteristics and growth stages of crops.

[0010] Periodic feature extraction is performed on multiple factor data queues to obtain multiple first factor features. The factor data queues are obtained based on influencing factors, and each factor feature corresponds to one influencing factor.

[0011] The growth characteristics, growth stages, and multiple factor features are input into a growth trend analysis model, and the irrigation strategy is optimized through the growth trend analysis model. The growth trend analysis model determines the growth trend of the crop based on the crop's growth pattern and the status of multiple factors.

[0012] In one possible implementation, the analysis of crop growth rate based on a crop growth data queue to obtain crop growth characteristics and growth stages includes:

[0013] Based on the length of the first time period, the crop growth data queue is divided into multiple first sub-queues;

[0014] For each first sub-queue, multiple basic growth models are fitted separately, and the model with the smallest deviation is selected as the candidate model from the multiple basic growth models.

[0015] Based on multiple fitting biases, the model with the smallest overall fitting bias is selected from multiple candidate models as the target model;

[0016] The target model is fitted using the crop growth data queue to obtain a crop growth characteristic model;

[0017] Based on the last crop growth data and the crop growth characteristic model, the growth stage and growth characteristics of the crop are determined, wherein the last crop growth data is the data at the end of the crop growth data queue.

[0018] In one possible implementation, the step of fitting multiple basic growth models to each first sub-queue and selecting the model with the smallest deviation from the multiple basic growth models as a candidate model includes:

[0019] For each first sub-queue, perform the following steps:

[0020] The model is selected from the plurality of basic growth models as the model to be fitted.

[0021] Generate multiple parameter arrays;

[0022] Substitute the multiple parameter arrays into the model to be fitted to obtain multiple process models;

[0023] For each process model, multiple data points are extracted from the process model based on the multiple time nodes corresponding to the first sub-queue, and these are used as the model queue.

[0024] Based on the first formula, the first sub-queue, and multiple model queues, multiple process fitting deviations are determined, where each process fitting deviation corresponds to a process model. The first formula is:

[0025]

[0026] In the formula, This represents the process fitting deviation. For the first subqueue One data point, For the model queue of the first One data point, This represents the total number of data items in the first sub-queue.

[0027] Add the fitting deviation of each process to the corresponding fitting deviation queue;

[0028] If the number of iterations has not been reached, for each parameter array, the globally optimal parameter array and the historically optimal parameter array are selected based on multiple fitting deviation queues and multiple process fitting deviations. The parameters in the arrays are adjusted based on the globally optimal parameter array and the historically optimal parameter array, and then the process jumps to the step of substituting the multiple parameter arrays into the model to be fitted to obtain multiple process models.

[0029] Otherwise, the minimum of the fitting deviations of the plurality of processes is added to the model deviation queue;

[0030] If the traversal of the multiple basic growth models is not completed, then proceed to the step of traversally extracting a model from the multiple basic growth models as the model to be fitted.

[0031] Otherwise, the model corresponding to the minimum value in the model deviation queue is selected as the candidate model.

[0032] In one possible implementation, the periodic feature extraction of multiple factor data queues is performed respectively to obtain multiple first factor features, including:

[0033] For each factor data queue, perform the following steps:

[0034] Obtain the basic fluctuation cycle of the influencing factors corresponding to the factor data queue, and use the frequency corresponding to the basic fluctuation cycle as the reference frequency.

[0035] Based on the reference frequency, multiple frequency amplitudes are extracted from the factor data queue, and the obtained multiple frequency amplitudes are constructed into an amplitude vector;

[0036] The combined distance from the magnitude vector to the centers of multiple vector classes is used as multiple reference distances, where the center of a vector class is the center of a vector class, which is obtained by clustering multiple historical magnitude vectors.

[0037] The vector class corresponding to the reference distance with the smallest value is taken as the target class, and the identifier of the target class is taken as the first factor feature of the factor data queue.

[0038] In one possible implementation, the extraction of multiple frequency amplitudes from the factor data queue based on the reference frequency includes:

[0039] Based on the second formula and the aforementioned benchmark frequency, multiple frequency amplitudes are extracted from the factor data queue, including:

[0040]

[0041] In the formula, For the reference frequency The frequency amplitude of the harmonic. For the first factor data queue One data point, The total number of data points in the factor data queue. It is a natural constant. The imaginary unit, As the reference frequency, This refers to the number of data points within the first data segment during the basic period. Pi;

[0042] The step of using the combined distance from the magnitude vector to the centers of multiple vector classes as multiple reference distances includes:

[0043] Calculate the cosine distances from the magnitude vectors to the centers of multiple vector classes, and use these as multiple first distances;

[0044] Calculate the Euclidean distances from the magnitude vectors to the centers of multiple vector classes, and use these distances as multiple second distances;

[0045] The distance with the largest value and the distance with the smallest value are selected from the plurality of second distances as the maximum distance and the minimum distance, respectively.

[0046] A reference distance is determined based on the third formula, the maximum distance, the minimum distance, the plurality of first distances, and the plurality of second distances, wherein the third formula is:

[0047]

[0048] In the formula, For the magnitude vector to the first Reference distance of each vector class center For the magnitude vector to the first The second distance between the center of each vector class. For the maximum distance, For the minimum distance, For the magnitude vector to the first The first distance between the center of each vector class.

[0049] In one possible implementation, inputting the growth characteristics, the growth stage, and the multiple factor features into a growth trend analysis model, and optimizing the irrigation strategy through the growth trend analysis model, includes:

[0050] Obtain multiple irrigation water volume queues, soil texture, and current soil moisture. Among them, multiple data points in the irrigation water volume queues correspond to multiple time points.

[0051] Each irrigation water volume queue, along with the soil texture, the current soil moisture, the growth characteristics, the growth stage, and the multiple factor characteristics, are input into the growth trend analysis model. The model's output is used as an estimate of crop growth rate.

[0052] Multiple crop growth rate estimates are added to multiple estimation queues, with each estimation queue corresponding to an irrigation water volume queue.

[0053] The irrigation water volume queue corresponding to the crop growth rate estimate with the largest median among the multiple crop growth rate estimates is taken as the target irrigation water volume queue.

[0054] If the number of iterations has not been reached, then for each irrigation water volume queue, adjustments are made based on the historical best irrigation water volume queue and the target irrigation water volume queue. The historical best irrigation water volume queue is the historical irrigation water volume queue corresponding to the crop growth rate estimate with the largest value in the estimated queue.

[0055] Otherwise, an irrigation strategy is determined based on the target irrigation water volume queue.

[0056] In one possible implementation, the growth trend analysis model is built upon an artificial neural network and includes:

[0057] Multiple crop growth samples were obtained. Each crop growth sample was a dataset constructed based on soil texture, irrigation strategy, soil moisture, crop growth characteristics, crop growth stage and multiple factor features. Each crop growth sample corresponded to a crop growth rate.

[0058] The multiple crop growth samples were divided into a training group and a validation group;

[0059] The crop growth samples of the training group are traversally input into the artificial neural network model. Based on the prediction error of the model, the gradient descent method is used to adjust multiple parameters of the artificial neural network model until the prediction error of the model is less than the first error threshold, thus obtaining the training model. The prediction error is determined based on the crop growth rate corresponding to the crop growth sample and the predicted crop growth rate of the model.

[0060] The crop growth samples from the validation group are traversally input into the training model to determine the overall prediction error of the training model;

[0061] If the overall prediction error is greater than the second error threshold, the number of neurons in the artificial neural network model is reduced according to the degree of deviation between the overall prediction error and the second error threshold, wherein the second error threshold is greater than the first error threshold;

[0062] Otherwise, the trained model will be used as the growth trend analysis model.

[0063] In a second aspect, embodiments of the present invention provide a crop water-saving irrigation device for implementing the crop water-saving irrigation method as described in the first aspect or any possible implementation thereof, the crop water-saving irrigation device comprising:

[0064] The factor acquisition module is used to acquire multiple factors that affect crop growth;

[0065] The crop growth characteristic analysis module is used to analyze the growth rate of crops based on the crop growth data queue to obtain the growth characteristics and growth stages of crops.

[0066] The factor analysis module is used to extract periodic features from multiple factor data queues to obtain multiple first factor features. The factor data queues are obtained based on influencing factors, and each factor feature corresponds to one influencing factor.

[0067] as well as,

[0068] An irrigation strategy determination module is used to input the growth characteristics, growth stage, and multiple factor features into a growth trend analysis model, and optimize the irrigation strategy through the growth trend analysis model. The growth trend analysis model determines the growth trend of the crop based on the crop's growth pattern and the status of multiple factors.

[0069] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement the steps of the method as described in the first aspect or any possible implementation of the first aspect.

[0070] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in the first aspect or any possible implementation thereof.

[0071] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0072] This invention discloses a water-saving irrigation method for crops. First, it identifies multiple factors influencing crop growth. Then, it analyzes the crop growth rate based on a crop growth data queue to obtain the crop's growth characteristics and growth stages. Next, it performs periodic feature extraction on the multiple factor data queues to obtain multiple first-factor features. Each factor feature corresponds to one influencing factor. Finally, it inputs the growth characteristics, growth stages, and multiple factor features into a growth trend analysis model to optimize the irrigation strategy. The growth trend analysis model determines the crop's growth trend based on the crop's growth patterns and the status of multiple factors. This invention optimizes the irrigation strategy by inputting the analyzed crop growth characteristics, growth stages, and factor status characteristics into a growth trend analysis model constructed using an artificial neural network, achieving the dual goals of ensuring yield and water conservation. Attached Figure Description

[0073] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0074] Figure 1 This is a flowchart of the crop water-saving irrigation method provided in the embodiments of the present invention;

[0075] Figure 2 This is a schematic diagram illustrating the process of constructing a crop growth characteristic model according to an embodiment of the present invention.

[0076] Figure 3 This is a functional block diagram of the crop water-saving irrigation device provided in the embodiments of the present invention;

[0077] Figure 4 This is a functional block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0078] In the following description, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limitation, so as to provide a thorough understanding of embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0079] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0080] The embodiments of the present invention will be described in detail below. This example is implemented based on the technical solution of the present invention, and provides detailed implementation methods and specific operation processes. However, the protection scope of the present invention is not limited to the following embodiments.

[0081] Figure 1 A flowchart of a crop water-saving irrigation method provided for an embodiment of the present invention.

[0082] like Figure 1 As shown, a flowchart illustrating the implementation of the crop water-saving irrigation method provided by an embodiment of the present invention is presented, and is described in detail below:

[0083] In step 101, multiple factors affecting crop growth are identified.

[0084] In step 102, the growth rate of the crop is analyzed based on the crop growth data queue to obtain the crop growth characteristics and growth stages.

[0085] In some implementations, the step of analyzing the crop growth rate based on a crop growth data queue to obtain the crop's growth characteristics and growth stage includes:

[0086] Based on the length of the first time period, the crop growth data queue is divided into multiple first sub-queues;

[0087] For each first sub-queue, multiple basic growth models are fitted separately, and the model with the smallest deviation is selected as the candidate model from the multiple basic growth models.

[0088] Based on multiple fitting biases, the model with the smallest overall fitting bias is selected from multiple candidate models as the target model;

[0089] The target model is fitted using the crop growth data queue to obtain a crop growth characteristic model;

[0090] Based on the last crop growth data and the crop growth characteristic model, the growth stage and growth characteristics of the crop are determined, wherein the last crop growth data is the data at the end of the crop growth data queue.

[0091] In some implementations, the step of fitting multiple basic growth models to each first sub-cohort and selecting the model with the smallest deviation from the multiple basic growth models as a candidate model includes:

[0092] For each first sub-queue, perform the following steps:

[0093] The model is selected from the plurality of basic growth models as the model to be fitted.

[0094] Generate multiple parameter arrays;

[0095] Substitute the multiple parameter arrays into the model to be fitted to obtain multiple process models;

[0096] For each process model, multiple data points are extracted from the process model based on the multiple time nodes corresponding to the first sub-queue, and these are used as the model queue.

[0097] Based on the first formula, the first sub-queue, and multiple model queues, multiple process fitting deviations are determined, where each process fitting deviation corresponds to a process model. The first formula is:

[0098]

[0099] In the formula, This represents the process fitting deviation. For the first subqueue One data point, For the model queue of the first One data point, This represents the total number of data items in the first sub-queue.

[0100] Add the fitting deviation of each process to the corresponding fitting deviation queue;

[0101] If the number of iterations has not been reached, for each parameter array, the globally optimal parameter array and the historically optimal parameter array are selected based on multiple fitting deviation queues and multiple process fitting deviations. The parameters in the arrays are adjusted based on the globally optimal parameter array and the historically optimal parameter array, and then the process jumps to the step of substituting the multiple parameter arrays into the model to be fitted to obtain multiple process models.

[0102] Otherwise, the minimum of the fitting deviations of the plurality of processes is added to the model deviation queue;

[0103] If the traversal of the multiple basic growth models is not completed, then proceed to the step of traversally extracting a model from the multiple basic growth models as the model to be fitted.

[0104] Otherwise, the model corresponding to the minimum value in the model deviation queue is selected as the candidate model.

[0105] For example, the embodiments of the present invention are intended to provide a method for determining crop irrigation strategies based on the changing patterns of crop growth characteristics, growth stages, and external influencing factors.

[0106] Regarding growth characteristics and stages, this invention selects a suitable growth curve from several typical growth curves to determine the crop's growth characteristics and current growth stage. Regarding the changing patterns of external influencing factors, this invention performs fluctuation analysis on factor monitoring data from a previous period and categorizes them into factor states. For example, influencing factors include: temperature, light intensity, wind speed, and air humidity. Taking temperature as an example, fluctuation analysis is performed on monitoring data from the previous week to determine the current temperature condition characteristics. In determining the crop irrigation strategy, the growth characteristics, current growth stage, and external factor state characteristics obtained from the aforementioned steps are fed together with the irrigation strategy into a crop growth rate prediction model constructed based on an artificial neural network to optimize the irrigation strategy and thus determine a better irrigation strategy.

[0107] To achieve the above objectives, such as Figure 2 As shown, specifically, in determining the growth characteristics and current growth stage of crops, this invention divides the previously observed crop growth data (in some scenarios, growth data determined by crop height, diameter, plant coverage area, leaf area, or number of leaves) into multiple first sub-cohorts 202 by arranging them sequentially. For each first sub-cohort 202, a basic growth model 203 is fitted. The model 205 that best fits the first sub-cohort 202 is selected from multiple fitting deviations 204. Finally, the model with the best overall fit is selected from the fitted models of multiple first sub-cohorts as the crop growth characteristic model.

[0108] Regarding basic models, a common one is the linear model, which takes the following form:

[0109]

[0110] In the formula, For crop growth data, As the initial value, The growth rate constant is denoted by . For time.

[0111] There is a Logistic model, in the following form:

[0112]

[0113] In the formula, For crop growth data, Let be the initial crop growth constant. Let be the intrinsic growth rate constant. This represents the maximum crop growth.

[0114] There is an exponential growth model, in the following form:

[0115]

[0116] In the formula, For crop growth data, This represents the initial crop growth. It is the exponential growth rate constant. For time.

[0117] For the basic model described above, when fitting the model using the first sub-queue, the parameters in the model will be set (e.g., in the formula). , , (Parameters), then use the time nodes corresponding to the data in the first sub-queue as model input, and sort the obtained outputs according to the time nodes. The fitting bias is then calculated using the first formula:

[0118]

[0119] In the formula, This represents the process fitting deviation. For the first subqueue One data point, For the model queue of the first One data point, This represents the total number of data items in the first sub-queue.

[0120] The fitting deviation is used as the basis for modifying the model parameters to adjust the model parameters.

[0121] In some methods, multiple models are generated simultaneously for the same model. The fitting deviation of multiple models is obtained through the first formula. The parameters are then adjusted using the model with the smallest fitting deviation and the historical best model with the smallest deviation in each iteration. The specific adjustment method can be the particle swarm optimization algorithm or the gray wolf algorithm.

[0122] After several iterations, the first sub-queue selects the model with the smallest fitting deviation from the aforementioned basic models. Then, from multiple first sub-queues, the model with the smallest overall fitting deviation is selected; this is the crop growth model. Once the growth model is determined, the model type and parameters related to growth rate (e.g., ...) can be defined. , As a characteristic of crop growth, linear models do not have obvious growth stages. For the Logistic model, the ratio of the current crop growth to the maximum crop growth is used as its growth stage. For the exponential growth model, the ratio of its derivative to the exponential growth rate constant is used as its growth stage.

[0123] In step 103, periodic feature extraction is performed on multiple factor data queues to obtain multiple first factor features. The factor data queues are obtained based on influencing factors, and each factor feature corresponds to one influencing factor.

[0124] In some implementations, the step of periodically extracting features from multiple factor data queues to obtain multiple first factor features includes:

[0125] For each factor data queue, perform the following steps:

[0126] Obtain the basic fluctuation cycle of the influencing factors corresponding to the factor data queue, and use the frequency corresponding to the basic fluctuation cycle as the reference frequency.

[0127] Based on the reference frequency, multiple frequency amplitudes are extracted from the factor data queue, and the obtained multiple frequency amplitudes are constructed into an amplitude vector;

[0128] The combined distance from the magnitude vector to the centers of multiple vector classes is used as multiple reference distances, where the center of a vector class is the center of a vector class, which is obtained by clustering multiple historical magnitude vectors.

[0129] The vector class corresponding to the reference distance with the smallest value is taken as the target class, and the identifier of the target class is taken as the first factor feature of the factor data queue.

[0130] In some implementations, the extraction of multiple frequency amplitudes from the factor data queue based on the reference frequency includes:

[0131] Based on the second formula and the aforementioned benchmark frequency, multiple frequency amplitudes are extracted from the factor data queue, including:

[0132]

[0133] In the formula, For the reference frequency The frequency amplitude of the harmonic. For the first factor data queue One data point, The total number of data points in the factor data queue. It is a natural constant. The imaginary unit, As the reference frequency, This refers to the number of data points within the first data segment during the basic period. Pi;

[0134] The step of using the combined distance from the magnitude vector to the centers of multiple vector classes as multiple reference distances includes:

[0135] Calculate the cosine distances from the magnitude vectors to the centers of multiple vector classes, and use these as multiple first distances;

[0136] Calculate the Euclidean distances from the magnitude vectors to the centers of multiple vector classes, and use these distances as multiple second distances;

[0137] The distance with the largest value and the distance with the smallest value are selected from the plurality of second distances as the maximum distance and the minimum distance, respectively.

[0138] A reference distance is determined based on the third formula, the maximum distance, the minimum distance, the plurality of first distances, and the plurality of second distances, wherein the third formula is:

[0139]

[0140] In the formula, For the magnitude vector to the first Reference distance of each vector class center For the magnitude vector to the first The second distance between the center of each vector class. For the maximum distance, For the minimum distance, For the magnitude vector to the first The first distance between the center of each vector class.

[0141] For example, in determining the factor characteristics of each influencing factor, the present invention first defines a basic fluctuation period, for example, the fluctuation period of temperature is one day, and uses the frequency corresponding to this period as the reference frequency. Then, using the second formula, the amplitudes of multiple harmonics of the reference frequency are extracted:

[0142]

[0143] In the formula, For the reference frequency The frequency amplitude of the harmonic. For the first factor data queue One data point, The total number of data points in the factor data queue. It is a natural constant. The imaginary unit, As the reference frequency, This refers to the number of data points within the first data segment during the basic period. Pi is the mathematical constant of a circle.

[0144] The amplitude values ​​are arranged according to their harmonics to obtain an amplitude vector. This amplitude vector is then compared with multiple class center vectors to calculate a comprehensive distance. The class containing the class center with the smallest comprehensive distance is selected as its class, and its class identifier is used as its factor feature. The comprehensive distance is determined based on Euclidean distance and cosine distance. Specifically, it is obtained by normalizing the maximum and minimum values ​​of the Euclidean distance and subtracting them from the cosine distance. The formula is expressed as:

[0145]

[0146] In the formula, For the magnitude vector to the first Reference distance of each vector class center For the magnitude vector to the first The second distance between the center of each vector class. For the maximum distance, For the minimum distance, For the magnitude vector to the first The first distance between the center of each vector class.

[0147] As for the process of obtaining the class center, it is obtained by extracting the frequency amplitude of the historical factor data queue through the aforementioned process, constructing historical amplitude vectors, and obtaining vector classes by clustering multiple historical amplitude vectors.

[0148] In step 104, the growth characteristics, growth stages, and multiple factor features are input into the growth trend analysis model. The irrigation strategy is optimized through the growth trend analysis model, wherein the growth trend analysis model determines the growth trend of the crop based on the crop's growth pattern and the status of multiple factors.

[0149] In some implementations, inputting the growth characteristics, the growth stage, and the multiple factor features into a growth trend analysis model, and optimizing the irrigation strategy through the growth trend analysis model, includes:

[0150] Obtain multiple irrigation water volume queues, soil texture, and current soil moisture. Among them, multiple data points in the irrigation water volume queues correspond to multiple time points.

[0151] Each irrigation water volume queue, along with the soil texture, the current soil moisture, the growth characteristics, the growth stage, and the multiple factor characteristics, are input into the growth trend analysis model. The model's output is used as an estimate of crop growth rate.

[0152] Multiple crop growth rate estimates are added to multiple estimation queues, with each estimation queue corresponding to an irrigation water volume queue.

[0153] The irrigation water volume queue corresponding to the crop growth rate estimate with the largest median among the multiple crop growth rate estimates is taken as the target irrigation water volume queue.

[0154] If the number of iterations has not been reached, then for each irrigation water volume queue, adjustments are made based on the historical best irrigation water volume queue and the target irrigation water volume queue. The historical best irrigation water volume queue is the historical irrigation water volume queue corresponding to the crop growth rate estimate with the largest value in the estimated queue.

[0155] Otherwise, an irrigation strategy is determined based on the target irrigation water volume queue.

[0156] In some implementations, the growth trend analysis model is built based on an artificial neural network, including:

[0157] Multiple crop growth samples were obtained. Each crop growth sample was a dataset constructed based on soil texture, irrigation strategy, soil moisture, crop growth characteristics, crop growth stage and multiple factor features. Each crop growth sample corresponded to a crop growth rate.

[0158] The multiple crop growth samples were divided into a training group and a validation group;

[0159] The crop growth samples of the training group are traversally input into the artificial neural network model. Based on the prediction error of the model, the gradient descent method is used to adjust multiple parameters of the artificial neural network model until the prediction error of the model is less than the first error threshold, thus obtaining the training model. The prediction error is determined based on the crop growth rate corresponding to the crop growth sample and the predicted crop growth rate of the model.

[0160] The crop growth samples from the validation group are traversally input into the training model to determine the overall prediction error of the training model;

[0161] If the overall prediction error is greater than the second error threshold, the number of neurons in the artificial neural network model is reduced according to the degree of deviation between the overall prediction error and the second error threshold, wherein the second error threshold is greater than the first error threshold;

[0162] Otherwise, the trained model will be used as the growth trend analysis model.

[0163] For example, in optimizing irrigation strategies, this invention initializes multiple irrigation water volume queues. Each irrigation water volume queue, along with soil texture, current soil moisture, growth characteristics and growth stage obtained in the preceding steps, and multiple factor characteristics, is input into a growth trend analysis model. The model's output is used as an estimate of crop growth rate. From the multiple growth rate estimates, the estimate with the largest value is selected, and this estimated irrigation water volume queue is taken as the globally optimal irrigation water volume queue. Each irrigation water volume queue is adjusted using its historically optimal irrigation water volume queue and the globally optimal irrigation water volume queue. After adjustment, the above process of inputting into the model is repeated. After repeating this process a preset number of times, the final globally optimal irrigation water volume queue is the target irrigation strategy.

[0164] Regarding the growth trend analysis model, this invention uses crop growth samples to adjust and train an artificial neural network model. Specifically, the crop growth samples are datasets constructed based on soil texture, irrigation strategy, soil moisture, crop growth characteristics, crop growth stage, and multiple other factors. Each crop growth sample corresponds to a crop growth rate. Multiple crop growth samples are divided into a training group and a validation group. The training group is used to train the artificial neural network, enabling it to accurately fit the relationship between the crop growth samples and the crop growth rate. The validation group is used to verify whether the artificial neural network still performs well on data outside the training group.

[0165] During training, the data from the training group is traversed and input into the artificial neural network model. The prediction error is determined based on the crop growth rate corresponding to the crop growth sample and the model output. The model parameters are then adjusted using gradient descent based on the error. After adjustment, the process returns to the traversal input step. This process is repeated until the prediction error reaches an acceptable level.

[0166] For the validation process, crop growth samples from the validation group are input into the trained model. The overall prediction error of the model is determined based on the model's output and the corresponding crop growth rate in the validation group. If the overall prediction error is large, the number of neurons is reduced, usually the number of neurons in the hidden layers. After reduction, the training process is repeated until the overall prediction error is acceptable. This artificial neural network model is then the growth trend analysis model.

[0167] The present invention provides an implementation method for water-saving irrigation of crops. First, it identifies multiple factors influencing crop growth. Then, it analyzes the crop growth rate based on a crop growth data queue to obtain the crop's growth characteristics and growth stages. Next, it performs periodic feature extraction on the multiple factor data queues to obtain multiple first-factor features. Each factor feature corresponds to one influencing factor. Finally, it inputs the growth characteristics, growth stages, and multiple factor features into a growth trend analysis model. The irrigation strategy is optimized through this model, which determines the crop's growth trend based on the crop's growth patterns and the status of multiple factors. This invention optimizes irrigation strategies by inputting the analyzed crop growth characteristics, growth stages, and factor status characteristics into a growth trend analysis model constructed using an artificial neural network, achieving the dual goals of ensuring yield and water conservation.

[0168] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0169] The following are embodiments of the apparatus of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0170] Figure 3 This is a functional block diagram of the crop water-saving irrigation device provided in the embodiments of the present invention, with reference to... Figure 3 The crop water-saving irrigation device includes: a factor acquisition module 301, a crop growth characteristic analysis module 302, a factor analysis module 303, and an irrigation strategy determination module 304, wherein:

[0171] Factor acquisition module 301 is used to acquire multiple influencing factors affecting crop growth;

[0172] The crop growth characteristic analysis module 302 is used to analyze the growth rate of crops based on the crop growth data queue to obtain the growth characteristics and growth stages of crops.

[0173] The factor analysis module 303 is used to extract periodic features from multiple factor data queues to obtain multiple first factor features. The factor data queues are obtained based on influencing factors, and each factor feature corresponds to an influencing factor.

[0174] The irrigation strategy determination module 304 is used to input the growth characteristics, the growth stage, and the multiple factor characteristics into the growth trend analysis model, and optimize the irrigation strategy through the growth trend analysis model. The growth trend analysis model determines the growth trend of the crop based on the growth law of the crop and the status of multiple factors.

[0175] Figure 4 This is a functional block diagram of the electronic device provided in an embodiment of the present invention. For example... Figure 4 As shown, the electronic device 4 of this embodiment includes a processor 400 and a memory 401, wherein the memory 401 stores a computer program 402 that can run on the processor 400. When the processor 400 executes the computer program 402, it implements the steps of the various crop water-saving irrigation methods and embodiments described above, for example... Figure 1 Steps 101 to 104 are shown.

[0176] For example, the computer program 402 may be divided into one or more modules / units, which are stored in the memory 401 and executed by the processor 400 to complete the present invention.

[0177] The electronic device 4 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. The electronic device 4 may include, but is not limited to, a processor 400 and a memory 401. Those skilled in the art will understand that... Figure 4 This is merely an example of electronic device 4 and does not constitute a limitation on electronic device 4. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 4 may also include input / output devices, network access devices, buses, etc.

[0178] The processor 400 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0179] The memory 401 can be an internal storage unit of the electronic device 4, such as a hard disk or memory. The memory 401 can also be an external storage device of the electronic device 4, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 401 can include both internal and external storage units of the electronic device 4. The memory 401 is used to store the computer program 402 and other programs and data required by the electronic device 4. The memory 401 can also be used to temporarily store data that has been output or will be output.

[0180] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the aforementioned method embodiments, and will not be repeated here.

[0181] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0182] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these 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 implementations should not be considered beyond the scope of this invention.

[0183] In the embodiments provided by this invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0184] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0185] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0186] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above-described embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various methods and apparatus embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0187] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method of water conservation irrigation of crops, characterized in that, The method comprises the following steps: obtaining a plurality of influencing factors affecting crop growth; analyzing the growth speed of the crop according to the crop growth data queue to obtain the growth characteristics and the growth stage of the crop; extracting periodic characteristics of the plurality of factor data queues respectively to obtain a plurality of first factor characteristics, wherein the factor data queue is obtained according to the influencing factor, and each factor characteristic corresponds to an influencing factor; inputting the growth characteristics, the growth stage and the plurality of factor characteristics into a growth trend analysis model to optimize the irrigation strategy through the growth trend analysis model, wherein the growth trend analysis model determines the growth trend of the crop according to the growth law of the crop and the trend of a plurality of factors; wherein the step of extracting periodic characteristics of the plurality of factor data queues respectively to obtain a plurality of first factor characteristics comprises: for each factor data queue, the following steps are performed respectively: obtaining the basic fluctuation period of the influencing factor corresponding to the factor data queue, and taking the frequency corresponding to the basic fluctuation period as the reference frequency; extracting a plurality of frequency amplitudes of the factor data queue according to the reference frequency, and constructing the obtained plurality of frequency amplitudes into an amplitude vector; taking the comprehensive distance of the amplitude vector to a plurality of vector class centers as a plurality of reference distances, wherein the vector class center is the center of the vector class, and the vector class is obtained by clustering a plurality of historical amplitude vectors; taking the vector class corresponding to the reference distance with the smallest value as the target class, and taking the identification of the target class as the first factor characteristic of the factor data queue; the step of extracting a plurality of frequency amplitudes of the factor data queue according to the reference frequency comprises: extracting a plurality of frequency amplitudes of the factor data queue according to the second formula and the reference frequency, comprising: wherein, a frequency amplitude of a a frequency amplitude of a a first data in the factor data queue, a total number of data in the factor data queue, a natural constant, an imaginary unit, a reference frequency, a number of data in a first data segment within a basic period duration, a number of data in a first data segment within a basic period duration, a circular constant; the step of taking the comprehensive distance of the amplitude vector to a plurality of vector class centers as a plurality of reference distances comprises: calculating the cosine distance of the amplitude vector to a plurality of vector class centers as a plurality of first distances; calculating the Euclidean distance of the amplitude vector to a plurality of vector class centers as a plurality of second distances; selecting the distance with the maximum value and the distance with the minimum value from the plurality of second distances as the maximum distance and the minimum distance; determining the reference distance according to the third formula, the maximum distance, the minimum distance, the plurality of first distances and the plurality of second distances, wherein the third formula is: wherein is a reference distance of the magnitude vector to the first vector class center, is a second distance of the magnitude vector to the first vector class center, is a maximum distance, is a minimum distance, is a first distance of the magnitude vector to the first vector class center.

2. The crop water saving irrigation method according to claim 1, characterized in that, the step of analyzing the growth speed of the crop according to the crop growth data queue to obtain the growth characteristics and the growth stage of the crop comprises: dividing the crop growth data queue into a plurality of first sub-queues according to the first time period length; for each first sub-queue, fitting a plurality of growth basic models respectively, and selecting a model with the minimum deviation from the plurality of growth basic models as a candidate model; selecting a model with the minimum comprehensive fitting deviation from the plurality of candidate models as a target model according to a plurality of fitting deviations; fitting the target model using the crop growth data queue to obtain a crop growth characteristic model; According to the last crop growth data and the crop growth characteristic model, a growth stage and a growth characteristic of the crop are determined, wherein the last crop growth data is data at the end of the queue of the crop growth data.

3. The crop water saving irrigation method according to claim 2, characterized in that, The fitting of the plurality of growth basic models is performed respectively for each first subqueue, and a model with the minimum deviation is selected from the plurality of growth basic models as a candidate model, including: For each first subqueue, the following steps are performed respectively: A model is taken from the plurality of growth basic models as a model to be fitted; A plurality of parameter arrays are generated; The plurality of parameter arrays are respectively substituted into the model to be fitted to obtain a plurality of process models; For each process model, a plurality of data are extracted from the process model according to a plurality of time nodes corresponding to the first subqueue as a model queue; A plurality of process fitting deviations are determined according to a first formula, the first subqueue and the plurality of model queues, wherein each process fitting deviation corresponds to a process model, and the first formula is: In the formula, This represents the process fitting deviation. For the first subqueue One data point, For the model queue of the first One data point, This represents the total number of data items in the first sub-queue. Each process fitting deviation is respectively added to a corresponding fitting deviation queue; If the number of iterations is not reached, for each parameter array, a global optimal parameter array and a historical optimal parameter array are selected according to the plurality of fitting deviation queues and the plurality of process fitting deviations, parameters in the array are adjusted according to the global optimal parameter array and the historical optimal parameter array, and the step of substituting the plurality of parameter arrays into the model to be fitted to obtain a plurality of process models is jumped to; Otherwise, the minimum value in the plurality of process fitting deviations is added to a model deviation queue; If the plurality of growth basic models are not traversed, the step of taking a model from the plurality of growth basic models as a model to be fitted is jumped to; Otherwise, a model corresponding to the minimum value in the model deviation queue is taken as a candidate model.

4. The method of water saving irrigation of crops according to any one of claims 1-3, characterized in that, The growth characteristic, the growth stage and the plurality of factor characteristics are input into a growth trend analysis model, and an irrigation strategy is optimized by the growth trend analysis model, including: A plurality of irrigation water quantity queues, soil texture and current soil moisture conditions are obtained, wherein a plurality of data of the irrigation water quantity queue correspond to a plurality of time nodes; Each irrigation water quantity queue is input into the growth trend analysis model together with the soil texture, the current soil moisture condition, the growth characteristic, the growth stage and the plurality of factor characteristics as input data, and an output of the model is taken as a crop growth speed estimation; A plurality of crop growth speed estimations are added to a plurality of estimation queues respectively, wherein each estimation queue corresponds to an irrigation water quantity queue; An irrigation water quantity queue corresponding to a crop growth speed estimation with the maximum value among the plurality of crop growth speed estimations is taken as a target irrigation water quantity queue; If the number of iterations is not reached, for each irrigation water quantity queue, a historical optimal irrigation water quantity queue is adjusted according to the target irrigation water quantity queue, wherein the historical optimal irrigation water quantity queue is a historical irrigation water quantity queue corresponding to a crop growth speed estimation with the maximum value in the estimation queue; Otherwise, determine an irrigation strategy according to the target irrigation water amount queue.

5. The crop water saving irrigation method according to claim 4, characterized in that, The growth trend analysis model is constructed based on an artificial neural network and includes: Obtain a plurality of crop growth samples, wherein each crop growth sample is a dataset constructed based on soil texture, irrigation strategy, soil moisture, crop growth characteristics, crop growth stage, and a plurality of factor characteristics, and each crop growth sample corresponds to a crop growth speed; Divide the plurality of crop growth samples into a training group and a verification group; Iteratively input the crop growth samples of the training group into an artificial neural network model, adjust a plurality of parameters of the artificial neural network model using gradient descent method according to a prediction error of the model, until the prediction error of the model is less than a first error threshold, and obtain a training model, wherein the prediction error is determined according to the crop growth speed corresponding to the crop growth sample and the predicted crop growth speed of the model; Iteratively input the crop growth samples of the verification group into the training model, and determine a comprehensive prediction error of the training model; If the comprehensive prediction error is greater than a second error threshold, reduce the number of neurons in the artificial neural network model according to the deviation degree of the comprehensive prediction error and the second error threshold, wherein the second error threshold is greater than the first error threshold; Otherwise, take the training model as the growth trend analysis model.

6. A water saving irrigation device for crops, characterized in that, The crop water-saving irrigation device for implementing the crop water-saving irrigation method according to any one of claims 1-5 includes: a factor obtaining module configured to obtain a plurality of influence factors affecting crop growth; a crop growth characteristic analysis module configured to analyze the growth speed of the crop according to the crop growth amount data queue, and obtain the growth characteristics and growth stage of the crop; a factor analysis module configured to perform periodic feature extraction on the plurality of factor data queues respectively, and obtain a plurality of first factor characteristics, wherein the factor data queue is obtained according to the influence factors, and each factor characteristic corresponds to an influence factor; and an irrigation strategy determination module configured to input the growth characteristics, the growth stage, and the plurality of factor characteristics into a growth trend analysis model, and optimize the irrigation strategy through the growth trend analysis model, wherein the growth trend analysis model determines the growth trend of the crop according to the growth law of the crop and the state of a plurality of factors.

7. An electronic device comprising a memory and a processor, said memory having stored therein a computer program operable on said processor, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1-5.

8. A computer-readable storage medium storing a computer program, the computer-readable storage medium comprising: The computer program is executed by the processor to implement the steps of the method according to any one of claims 1-5.

Images