Integrated planting data management method and system combined with strawberry growth and development model

Abstract

The application discloses an integrated planting data management method and system combined with a strawberry growth and development model, relates to the technical field of data processing, and comprises the following steps: obtaining K sample plant growth monitoring data set sequences and K planting environment monitoring data set sequences; obtaining K planting environment monitoring data central value sets; obtaining a planting environment monitoring data central value mean set; determining an environment deviation factor set; obtaining a sample plant growth monitoring data mean set sequence; obtaining an interactive fusion sample plant growth characteristic set; determining a strawberry growth and development evaluation factor; and performing integrated planting data management according to a planting data extraction frequency. The application solves the technical problems of low fitting degree between the strawberry planting data management and the actual growth and development of strawberries and poor management quality in the prior art, and achieves the technical effects of improving the reliability and management efficiency of planting data management.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to an integrated planting data management method and system that incorporates a strawberry growth and development model. Background Technology

[0002] To obtain high-quality strawberries, multi-dimensional data monitoring of the strawberry cultivation process is necessary, generating a large amount of cultivation data that requires management. Currently, integrated cultivation data is often extracted either in real-time or at a fixed monitoring frequency. The former requires processing a large amount of data, easily leading to data redundancy, long data processing cycles, and an inability to respond quickly to strawberry conditions. The latter tends to result in delayed analysis of changes in strawberry conditions. Existing technologies suffer from low alignment between strawberry cultivation data management and actual strawberry growth and development, resulting in poor management quality. Summary of the Invention

[0003] This application provides an integrated planting data management method and system that combines a strawberry growth and development model, which is used to address the technical problem that the existing strawberry planting data management has a low degree of consistency with the actual growth and development of strawberries, resulting in poor management quality.

[0004] In view of the above problems, this application provides an integrated planting data management method and system that combines a strawberry growth and development model.

[0005] The first aspect of this application provides an integrated planting data management method combining a strawberry growth and development model, the method comprising:

[0006] Conditional sampling is performed on strawberry plants in the target strawberry growing area to obtain K sample plants. Plant growth and planting environment are monitored on the K sample plants in a preset monitoring window to obtain a set of K sample plant growth monitoring data and a set of K planting environment monitoring data, where K is an integer greater than or equal to 1.

[0007] Traverse the K sets of planting environment monitoring data to perform intra-sequence data central value analysis, and obtain the K sets of planting environment monitoring data central values.

[0008] Using the data type of planting environment monitoring as an index, the mean values ​​of planting environment monitoring data of the same type in the central value sets of the K planting environment monitoring datasets are processed to obtain the central value mean set of planting environment monitoring datasets.

[0009] Based on the degree of deviation between the mean set of the planting environment monitoring data set and the preset planting environment data set, a set of environmental deviation factors is determined, wherein the set of environmental deviation factors includes M environmental deviation factors, and each environmental deviation factor corresponds to a planting environment monitoring data type;

[0010] perform cross-sample data mean calculation on the K sample plant growth monitoring data set sequences to obtain a sample plant growth monitoring data mean set sequence;

[0011] construct M adaptive convolutional network layers based on the M environment deviation factors, use the M adaptive convolutional network layers to perform multi-scale feature analysis on the sample plant growth monitoring data mean set sequence, and perform feature correlation interaction fusion analysis on the analysis results to obtain an interaction fusion sample plant growth feature set;

[0012] use the strawberry growth and development model to analyze the interaction fusion sample plant growth feature set, determine a strawberry growth and development evaluation factor, and determine the frequency of planting data extraction for the strawberry plants in the target strawberry growth area according to the size of the strawberry growth and development evaluation factor, and perform integrated planting data management according to the planting data extraction frequency.

[0013] In a second aspect, the present application provides an integrated planting data management system combined with a strawberry growth and development model, which comprises:

[0014] A planting environment monitoring data set sequence obtaining module is configured to conditionally sample the strawberry plants in a target strawberry growth area to obtain K sample plants, perform plant growth monitoring and planting environment monitoring on the K sample plants in a preset monitoring window, and obtain K sample plant growth monitoring data set sequences and K planting environment monitoring data set sequences, wherein K is an integer greater than or equal to 1.

[0015] A data central value set obtaining module is configured to perform sequence internal data central value analysis on the K planting environment monitoring data set sequences to obtain K planting environment monitoring data central value sets.

[0016] A planting environment monitoring data central value mean set obtaining module is configured to use planting environment monitoring data types as indexes to perform mean processing on the same type of planting environment monitoring data in the K planting environment monitoring data central value sets, respectively, and obtain a planting environment monitoring data central value mean set.

[0017] An environment deviation factor determining module is configured to determine an environment deviation factor set according to the deviation degree of the planting environment monitoring data central value mean set and a preset planting environment data set, wherein the environment deviation factor set comprises M environment deviation factors, and each environment deviation factor corresponds to a planting environment monitoring data type.

[0018] The monitoring data mean set sequence obtaining module is configured to perform cross-sample data mean calculation on the K sample plant growth monitoring data set sequences to obtain a sample plant growth monitoring data mean set sequence.

[0019] The interactive fusion sample plant growth feature set obtaining module is configured to construct M adaptive convolution network layers based on the M environment deviation factors, perform multi-scale feature analysis on the sample plant growth monitoring data mean set sequence by using the M adaptive convolution network layers, and perform feature correlation interactive fusion analysis on the analysis result to obtain an interactive fusion sample plant growth feature set.

[0020] The data management module is configured to analyze the interactive fusion sample plant growth feature set by using a strawberry growth and development model, determine a strawberry growth and development evaluation factor, determine a planting data extraction frequency for the strawberry plants in the target strawberry growth area according to the size of the strawberry growth and development evaluation factor, and perform integrated planting data management according to the planting data extraction frequency.

[0021] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0022] The application obtains K sample plants by condition sampling on the strawberry plants in the target strawberry growth area, performs plant growth monitoring and planting environment monitoring on the K sample plants in a preset monitoring window, obtains K sample plant growth monitoring data set sequences and K planting environment monitoring data set sequences, wherein K is an integer greater than or equal to 1, then iterates the K planting environment monitoring data set sequences to obtain a K planting environment monitoring data central value set through central value analysis on the data in the sequence, respectively processes the same type of planting environment monitoring data in the K planting environment monitoring data central value sets through mean value processing with the planting environment monitoring data type as an index to obtain a planting environment monitoring data central value mean value set, and then determines an environment deviation factor set according to the deviation degree of the planting environment monitoring data central value mean value set and a preset planting environment data set, wherein the environment deviation factor set includes M environment deviation factors, each environment deviation factor corresponds to a planting environment monitoring data type, then performs cross-sample data mean value calculation on the K sample plant growth monitoring data set sequences to obtain a sample plant growth monitoring data mean value set sequence, then constructs M adaptive convolution network layers based on the M environment deviation factors, performs multi-scale feature analysis on the sample plant growth monitoring data mean value set sequence by using the M adaptive convolution network layers, and performs feature correlation interactive fusion analysis on the analysis results to obtain an interactive fusion sample plant growth feature set, then analyzes the interactive fusion sample plant growth feature set by using a strawberry growth and development model to determine a strawberry growth and development evaluation factor, and determines the frequency of planting data extraction on the strawberry plants in the target strawberry growth area according to the size of the strawberry growth and development evaluation factor, and performs integrated planting data management according to the planting data extraction frequency. The technical effect of using the strawberry growth and development model to analyze the strawberry growth features with high reliability to obtain the strawberry growth state, determining the extraction frequency of reliable data extraction on the planting data according to the strawberry growth state, and improving the data management reliability and management efficiency is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0024] Figure 1 The integrated planting data management method flowchart provided by the embodiment of the present application in combination with the strawberry growth and development model;

[0025] Figure 2 The integrated planting data management system structure diagram provided by the embodiment of the present application in combination with the strawberry growth and development model.

[0026] The drawing reference is explained: a planting environment monitoring data set sequence obtaining module 11, a data set value obtaining module 12, a planting environment monitoring data set value mean set obtaining module 13, an environment deviation factor determining module 14, a monitoring data mean set sequence obtaining module 15, an interactive fusion sample plant growth feature set obtaining module 16, and a data management module 17. DETAILED DESCRIPTION

[0027] The present application provides an integrated planting data management method and system combined with a strawberry growth and development model, which is used to solve the technical problems of low fitting degree of strawberry planting data management and actual growth and development of strawberries and poor management quality in the prior art.

[0028] The technical solutions in the embodiments of the present application will be clearly and completely described 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 of ordinary skill in the art without creative labor are within the scope of protection of the present application.

[0029] It should be noted that the terms "include" and "have" are intended to cover non-exclusive inclusion, for example, a process, method, system, product or server including a series of steps or units need not be limited to only those steps or units clearly listed, but can include other steps or modules not clearly listed or inherent to these processes, methods, products or devices.

[0030] Embodiment one, as shown in the present application provides an integrated planting data management method combined with a strawberry growth and development model, wherein the method comprises: Figure 1

[0031] Step S100: Conditionally sampling the strawberry plants in the target strawberry growth area to obtain K sample plants, performing plant growth monitoring and planting environment monitoring on the K sample plants in a preset monitoring window to obtain K sample plant growth monitoring data set sequences and K planting environment monitoring data set sequences, wherein K is an integer greater than or equal to 1;

[0032] Further, conditionally sampling the plants in the target strawberry growth area to obtain K sample plants, the step S100 of the present application embodiment further comprises:

[0033] Obtaining a preset sampling number, randomly sampling the plants in the target strawberry growth area based on the preset sampling number to obtain K random sample plants, wherein the preset sampling number is K;

[0034] ​The K random sample plants are enumerated two by two to obtain a plurality of plant enumeration combinations, and the distribution position similarity of the plurality of plant enumeration combinations is calculated using cosine similarity to obtain a plurality of position combination similarities.

[0035] It is judged whether the plurality of position combination similarities satisfy a preset sampling condition, and if so, the K random sample plants are authenticated by sampling as K sample plants, wherein the preset sampling condition is that the number of combinations in the plurality of position combination similarities that are greater than or equal to a preset position combination similarity is less than a preset number threshold.

[0036] In one possible embodiment, the number K of random sampling is set by a person skilled in the art according to the plant scale and monitoring requirements of the target strawberry growing area, wherein K is an integer greater than or equal to 1. The target strawberry growing area is any area that needs integrated planting data management. The K sample plants meet the requirements of the preset sampling condition, that is, the plants authenticated by sampling. The preset monitoring window is a time period for monitoring the growth conditions and planting environment conditions of sample plants, which can be 3 days, 5 days, etc., and can be set by a person skilled in the art according to actual conditions.

[0037] In one embodiment, plant growth monitoring indicators (including plant height, leaf area, stem diameter, flower bud differentiation state, fruit weight, etc.) and planting environment monitoring indicators (including temperature, humidity, light intensity, soil moisture, soil pH value, etc.) are obtained, and then the K sample plants are respectively subjected to plant growth monitoring and planting environment monitoring using a sensor array (an integrated sensor for monitoring the growth state and environment state of sample plants, such as a sensor, a camera, etc.) in the preset monitoring window, and the monitoring data is recorded in time sequence to form K sample plant growth monitoring data set sequences and K planting environment monitoring data set sequences. The K sample plant growth monitoring data set sequences reflect the changes in the plant growth state of sample plants in the preset monitoring window, and the K planting environment monitoring data set sequences reflect the changes in the planting environment state of sample plants in the preset monitoring window.

[0038] By obtaining the K sample plant growth monitoring data set sequences and K planting environment monitoring data set sequences, the goal of comprehensively monitoring the growth conditions and planting environment of the K sample plants is achieved, and corresponding data set sequences are generated, achieving the technical effect of providing high-quality data support for subsequent feature analysis and planting data management.

[0039] For example, assuming there are 500 strawberry plants in the target area, and a preset sampling size K = 10, 10 plants are randomly selected from these 500 plants as the initial sample. Pairwise combinations of these 10 plants are enumerated, resulting in 45 combinations. The similarity of the distribution locations of each combination is calculated using the cosine similarity formula, and the number of combinations with a similarity higher than a preset threshold (e.g., 0.8) is counted. If the number of highly similar combinations is less than the preset threshold (e.g., 15), the sample passes authentication and enters data monitoring; otherwise, resampling is performed.

[0040] By using conditional sampling, representative plant samples are obtained, avoiding overly concentrated sample distributions that could hinder reliable data for subsequent monitoring and analysis. This achieves the technical goal of improving the quality of plant sample selection and providing a foundation for subsequent analysis.

[0041] Step S200: Traverse the K sets of planting environment monitoring data to perform intra-sequence data central value analysis to obtain the K sets of planting environment monitoring data central values;

[0042] Furthermore, by traversing the K planting environment monitoring data sets and performing intra-sequence data central value analysis to obtain the K planting environment monitoring data central value sets, step S200 of this application embodiment also includes:

[0043] Extract a first planting environment monitoring data set sequence from the K planting environment monitoring data set sequences, wherein the first planting environment monitoring data set sequence includes multiple first planting environment monitoring data sub-sequences of multiple planting environment monitoring data types;

[0044] The mean of the data is calculated by traversing the multiple first planting environment monitoring data subsequences to determine multiple first data means. The multiple first data means are used as the starting points for multiple central value analysis. Iterative analysis is performed on the multiple first planting environment monitoring data subsequences according to a preset iteration step size to determine multiple central values ​​of the first planting environment monitoring data. The multiple central values ​​of the first planting environment monitoring data are then summarized to obtain the set of central values ​​of the first planting environment monitoring data.

[0045] Intra-sequence data central value analysis is performed on the K sets of planting environment monitoring data to determine the central value set of the K sets of planting environment monitoring data.

[0046] Furthermore, step S200 in this embodiment of the application also includes:

[0047] The multiple lumped value analysis starting points are iterated in the multiple first planting environment monitoring data subsequences according to the preset iteration step size to obtain multiple iterative data;

[0048] determining whether the concentration density of the plurality of iteration data is greater than or equal to the concentration density of the plurality of concentration value analysis starting points, if yes, updating the plurality of iteration data to the plurality of concentration value analysis starting points, and performing iteration in the plurality of first planting environment monitoring data subsequences according to the preset iteration step length respectively;

[0049] until a preset iteration number is met, taking the plurality of iteration data obtained by the last iteration as the plurality of first planting environment monitoring data central values.

[0050] In one possible embodiment, the data general distribution situation analysis in the sequence is performed on the K planting environment monitoring data set sequences respectively, and the planting environment monitoring data central values capable of representing the general situation of each planting environment monitoring data in the preset monitoring window are determined, so that the K planting environment monitoring data central value set is obtained. The technical effect of performing data concentration analysis on the K planting environment monitoring data central value set and obtaining representative planting environment monitoring data central values is achieved, and reliable data is provided for subsequent analysis on the basis of reducing the data analysis amount.

[0051] In one embodiment, each planting environment monitoring data set sequence contains a plurality of planting environment monitoring data types (such as temperature, humidity, light intensity, etc.). A first planting environment monitoring data set sequence is extracted from the K planting environment monitoring data set sequences, and the first planting environment monitoring data set sequence is analyzed to determine the corresponding first planting environment monitoring data central value set.

[0052] Optionally, the first planting environment monitoring data set sequence includes a plurality of first planting environment monitoring data subsequences of a plurality of planting environment monitoring data types. Each first planting environment monitoring data subsequence corresponds to one planting environment monitoring data type.

[0053] Optionally, the plurality of first data means are determined by traversing the plurality of first planting environment monitoring data subsequences to perform data mean calculation. The plurality of first data means reflect the average level of the monitoring data corresponding to each planting environment monitoring data type in the case of including accidental data.

[0054] Further, the plurality of first data means are taken as the plurality of concentration value analysis starting points, and iteration analysis is performed in the plurality of first planting environment monitoring data subsequences according to the preset iteration step length respectively, to obtain a plurality of iteration data. The preset iteration step length is the data difference value of two adjacent iterations preset by a person skilled in the art.

[0055] Optionally, a plurality of iteration neighborhoods of the plurality of iteration data are constructed with the plurality of iteration data as the center and the preset iteration step length as the radius, wherein each iteration neighborhood includes a plurality of first planting environment monitoring data with a difference value of the iteration data being the preset iteration step length, and then the amount of data contained in the plurality of iteration neighborhoods is counted, and the counting result is compared with the preset iteration step length to obtain the concentration density of the plurality of iteration data. The concentration density of the plurality of iteration data reflects the intensive degree of the first planting environment monitoring data gathered around the plurality of iteration data. Based on the same principle as obtaining the concentration density of the plurality of iteration data, the concentration density of the plurality of concentration value analysis starting points is obtained.

[0056] Optionally, it is judged whether the concentration density of the plurality of iteration data is greater than or equal to the concentration density of the plurality of concentration value analysis starting points. If yes, it indicates that the intensive degree of the data distributed around the plurality of iteration data is higher than that of the plurality of concentration value analysis starting points. At this time, the plurality of iteration data is updated as the plurality of concentration value analysis starting points, and iteration is performed in the plurality of first planting environment monitoring data subsequences according to the preset iteration step length, until a preset iteration number is met. The plurality of first planting environment monitoring data corresponding to the plurality of iteration data obtained in the last iteration is taken as the plurality of first planting environment monitoring data central values.

[0057] If not, it indicates that the intensive degree of the data distributed around the plurality of iteration data is not higher than that of the plurality of concentration value analysis starting points. At this time, iteration is stopped, and the plurality of concentration value analysis starting points are taken as the plurality of first planting environment monitoring data central values. Then, the plurality of first planting environment monitoring data central values are summarized to obtain the first planting environment monitoring data central value set.

[0058] Optionally, based on the same principle as obtaining the first planting environment monitoring data central value set, intra-sequence data central value analysis is performed on the K planting environment monitoring data sequence sets to determine the K planting environment monitoring data central value sets. This achieves the technical effect of obtaining high-representative planting environment monitoring data central values, ensuring the accuracy and reliability of the environment data, and providing high-quality input data for subsequent model analysis.

[0059] Step S300: Taking the planting environment monitoring data type as the index, the same type of planting environment monitoring data in the K planting environment monitoring data central value sets is subjected to mean value processing respectively to obtain a planting environment monitoring data central value mean value set;

[0060] Step S400: determining an environment deviation factor set according to the deviation degree of the set of central values of the planting environment monitoring data set and the preset planting environment data set, wherein the environment deviation factor set includes M environment deviation factors, and each environment deviation factor corresponds to a type of planting environment monitoring data;

[0061] Step S500: performing cross-sample data mean value calculation on the K sample plant growth monitoring data set sequences to obtain a sample plant growth monitoring data mean value set sequence;

[0062] In one possible embodiment, the K planting environment monitoring data set of central values is classified according to the types of planting environment monitoring data (such as temperature, humidity, light intensity, soil humidity, etc.) to generate a plurality of same-type planting environment monitoring data set of central values. Each same-type planting environment monitoring data set of central values corresponds to a type of planting environment monitoring data. The plurality of same-type planting environment monitoring data set of central values is subjected to mean value calculation within the sub-set to obtain a plurality of same-type planting environment monitoring data set of central values, which are collected to obtain the set of central values of the planting environment monitoring data. The set of central values of the planting environment monitoring data is obtained by fusing the environment monitoring data of a plurality of sample plants, and reflects the general situation of the planting environment of the strawberry plants in the target strawberry growth area.

[0063] Optionally, the difference values of the corresponding data in the set of central values of the planting environment monitoring data and the preset planting environment data set are calculated respectively, and the calculation results are compared with the corresponding data in the preset planting environment data set to obtain an environment deviation factor set, wherein the environment deviation factor set includes M environment deviation factors, and each environment deviation factor corresponds to a type of planting environment monitoring data. Each environment deviation factor reflects the deviation of the type of planting environment monitoring data in the actual planting process.

[0064] Taking the types of plant growth monitoring data as indexes, the K sample plant growth monitoring data set sequences are subjected to cross-sample data mean value calculation to determine the general data situation of each type of plant growth monitoring data changing with time within a preset window, and a sample plant growth monitoring data mean value set sequence is obtained. This achieves the technical effect of providing reliable data support for subsequent analysis.

[0065] Step S600: constructing M adaptive convolution network layers based on the M environment deviation factors, performing multi-scale feature analysis on the sample plant growth monitoring data mean value set sequence by using the M adaptive convolution network layers, and performing feature correlation interaction fusion analysis on the analysis results to obtain an interaction fusion sample plant growth feature set;

[0066] Further, based on the M environment deviation factors, M adaptive convolutional network layers are constructed, multi-scale feature analysis is performed on the sample plant growth monitoring data mean set sequence by using the M adaptive convolutional network layers, and feature correlation interactive fusion analysis is performed on the analysis result to obtain an interactive fusion sample plant growth feature set. The step S600 of the embodiment of the present application further includes:

[0067] The M adaptive convolutional network layers perform data feature extraction on the sample plant growth monitoring data mean set sequence according to M receptive fields to obtain M sample plant growth feature sets, wherein the receptive field is the extraction width of the adaptive convolutional network layer when performing feature extraction on the sample plant growth monitoring data mean set sequence.

[0068] The M sample plant growth feature sets are mapped and sorted according to the order from large to small of the M receptive fields to obtain a sample plant growth feature set sequence.

[0069] Feature correlation interactive fusion analysis is performed on the sample plant growth feature sets in the sample plant growth feature set sequence in turn to obtain an interactive fusion sample plant growth feature set.

[0070] Further, feature correlation interactive fusion analysis is performed on the sample plant growth feature sets in the sample plant growth feature set sequence in turn to obtain an interactive fusion sample plant growth feature set. The step S600 of the embodiment of the present application further includes:

[0071] Feature correlation interactive fusion analysis is performed on the first sample plant growth feature set and the second sample plant growth feature set of the sample plant growth feature set sequence to obtain a first correlation interactive fusion sample plant growth feature set.

[0072] The third sample plant growth feature set of the sample plant growth feature set sequence is extracted, and feature correlation interactive fusion analysis is performed on the third sample plant growth feature set and the first correlation interactive fusion sample plant growth feature set to determine a second correlation interactive fusion sample plant growth feature set.

[0073] Based on the second correlation interactive fusion sample plant growth feature set, feature correlation interactive fusion analysis is continuously performed on the sample plant growth feature set sequence until the end of the sample plant growth feature set sequence is reached to obtain the interactive fusion sample plant growth feature set.

[0074] Further, feature correlation interactive fusion analysis is performed on the first sample plant growth feature set and the second sample plant growth feature set of the sample plant growth feature set sequence to obtain a first correlation interactive fusion sample plant growth feature set. The step S600 of the embodiment of the present application includes:

[0075] performing inner product calculation on the first sample plant growth characteristic set and the second sample plant growth characteristic set to determine a first correlation interaction fusion characteristic similarity set;

[0076] performing similarity normalization processing on the first correlation interaction fusion characteristic similarity set, and embedding the processing result in a matrix to obtain a first correlation interaction fusion matrix;

[0077] performing convolution calculation on the first correlation interaction fusion matrix and the second sample plant growth characteristic set to determine a first correlation interaction fusion sample plant growth characteristic set.

[0078] In one possible embodiment, M receptive fields of M adaptive convolutional network layers are determined according to the sizes of the M environmental deviation factors. The receptive field is the extraction width when the adaptive convolutional network layer extracts features from the sequence of sample plant growth monitoring data mean sets. The greater the environmental deviation factor, the greater the deviation of the type of planting environment monitoring data in the actual planting process, and the more detailed analysis is needed, so the smaller the corresponding receptive field.

[0079] Optionally, the ratio of each of the M environmental deviation factors to the sum of the M environmental deviation factors is calculated, and the calculation result is multiplied by a preset receptive field to obtain the M receptive fields. The preset receptive field is the extraction width when the adaptive convolutional network layer extracts features from the sequence of sample plant growth monitoring data mean sets. Then, the parameters of the initial adaptive convolutional network layer are configured based on the M receptive fields to obtain the M adaptive convolutional network layers configured.

[0080] Optionally, a plurality of historical sample plant growth monitoring data mean set sequences and a plurality of historical sample plant growth characteristic sets are obtained as training data, and the adaptive convolutional network layer constructed based on the convolutional neural network is supervised trained to learn the mapping relationship between the sample plant growth monitoring data mean set sequence and the sample plant growth characteristic set until the training converges, and the initial adaptive convolutional network layer trained is obtained.

[0081] The M adaptive convolutional network layers are used to perform multi-scale feature analysis on the sequence of sample plant growth monitoring data mean sets, and the analysis result is subjected to feature correlation interaction fusion analysis to obtain an interaction fusion sample plant growth characteristic set. The interaction fusion sample plant growth characteristic set reflects the growth state of the strawberry plants in the target strawberry growth area. Thus, the correlation analysis of the plant growth state in the preset monitoring window is realized, the growth state of the strawberry plants is deeply mined, and the technical effect of providing reliable analysis data support for subsequent analysis of the growth and development of the strawberries is achieved.

[0082] In one embodiment, the M adaptive convolutional network layers are used to perform multi-scale data feature extraction on the sample plant growth monitoring data mean set sequence, and M sample plant growth feature sets are obtained. According to the order of the M receptive fields from large to small, the corresponding M sample plant growth feature sets are mapped and sorted to obtain a sample plant growth feature set sequence.

[0083] Further, the first sample plant growth feature set and the second sample plant growth feature set of the sample plant growth feature set sequence are subjected to feature correlation interaction fusion analysis to obtain a first correlation interaction fusion sample plant growth feature set. Specifically, the inner product of the first sample plant growth feature set and the second sample plant growth feature set is calculated using the cosine similarity formula to determine a first correlation interaction fusion feature similarity set. Each first correlation interaction fusion feature similarity reflects the similarity between a growth feature in the first sample plant growth feature set and a corresponding growth feature in the second sample plant growth feature set.

[0084] The first correlation interaction fusion feature similarity set is traversed using the softmax formula to perform similarity normalization processing, determine the normalized value of each first correlation interaction fusion feature similarity, and embed the processing result in a matrix to obtain a first correlation interaction fusion matrix. The first correlation interaction fusion matrix is a matrix after normalization processing and elimination of data dimension differences. Further, the first correlation interaction fusion matrix and the second sample plant growth feature set are subjected to convolution calculation using a convolutional network to determine the first correlation interaction fusion sample plant growth feature set. This achieves the technical effect of deep fusion of the first sample plant growth feature set and the second sample plant growth feature set, and further extracts more refined growth features.

[0085] Optionally, a third sample plant growth feature set of the sample plant growth feature set sequence is extracted, and subjected to feature correlation interaction fusion analysis with the first correlation interaction fusion sample plant growth feature set based on the same principle as obtaining the first correlation interaction fusion sample plant growth feature set to determine a second correlation interaction fusion sample plant growth feature set.

[0086] Optionally, based on the same principle as obtaining the first correlation interaction fusion sample plant growth feature set, the second correlation interaction fusion sample plant growth feature set is used to continue feature correlation interaction fusion analysis on the sample plant growth feature set sequence until the end of the sample plant growth feature set sequence is reached, and the interaction fusion sample plant growth feature set is obtained.

[0087] Through utilizing M adaptive convolution network layers, multi-scale feature extraction is performed on sample plant growth monitoring data, and through interactive fusion and inner product calculation processing, deep analysis of plant growth characteristics is realized. Through this process, a comprehensive interactive fusion sample plant growth characteristic set is finally obtained, and a technical effect of providing an important basis for subsequent growth trend analysis is achieved.

[0088] Step S700: utilizing the strawberry growth and development model to analyze the interactive fusion sample plant growth characteristic set, determining a strawberry growth and development evaluation factor, and determining a planting data extraction frequency for the target strawberry growth area according to the size of the strawberry growth and development evaluation factor, and performing integrated planting data management according to the planting data extraction frequency.

[0089] In one possible embodiment, the strawberry growth and development model is a functional model that can reliably analyze the growth and development of strawberry plants according to strawberry plant growth characteristics. Optionally, a plurality of historical strawberry plant growth characteristics and a plurality of historical strawberry growth and development evaluation factors are obtained as training data, and a framework based on a convolutional neural network is supervised trained using the training data. In the training, the mapping relationship between the strawberry plant growth characteristics and the strawberry growth and development evaluation factor is learned until the training converges, and the trained strawberry growth and development model is obtained.

[0090] Optionally, the interactive fusion sample plant growth characteristic set is input into the strawberry growth and development model for analysis to obtain a strawberry growth and development evaluation factor. The strawberry growth and development evaluation factor reflects the growth and development state of the strawberry. A strawberry growth and development evaluation factor-extraction frequency mapping relationship pre-set by a person skilled in the art is obtained, and the strawberry growth and development evaluation factor is used as an index to search the strawberry growth and development evaluation factor-extraction frequency mapping relationship to obtain the planting data extraction frequency. Preferably, if the evaluation factor is close to the ideal value range, it indicates that the growth state of the strawberry is stable, and the data extraction frequency can be reduced (such as once a day). If the evaluation factor deviates from the ideal value range, it indicates that the strawberry has growth abnormalities, and the data extraction frequency needs to be increased (such as once an hour). According to the obtained planting data extraction frequency, the plant growth and planting environment data of the target strawberry growth area are managed. A technical effect of improving the efficiency and quality of integrated planting data management according to the analysis result of the strawberry growth and development model to determine the planting data extraction frequency is achieved.

[0091] In summary, the embodiments of the present application have at least the following technical effects:

[0092] The application realizes scientific conditional sampling method, ensures the representativeness of sample plants, reduces data bias, optimizes data by using central value analysis and mean value processing, eliminates random fluctuations in environmental data, extracts the most representative data characteristics, determines M environmental deviation factors, constructs M adaptive convolution network layers, and performs multi-scale feature analysis on the mean value set sequence of sample plant growth monitoring data, realizes in-depth correlation analysis of capturing strawberry growth, determines the interactive fusion sample plant growth feature set, and then uses the strawberry growth and development model to analyze the interactive fusion sample plant growth feature set, determines the strawberry growth and development evaluation factor, and determines the frequency of planting data extraction of the target strawberry growth area according to the size of the strawberry growth and development evaluation factor, and performs integrated planting data management according to the planting data extraction frequency. The technical effect of managing the planting data according to the corresponding extraction frequency according to the actual growth state of the strawberries is achieved.

[0093] In the second embodiment, based on the same inventive concept as the integrated planting data management method combined with the strawberry growth and development model in the preceding embodiments, as shown in the following Figure 2 The application provides an integrated planting data management system combined with a strawberry growth and development model, and the system and method embodiments in the application embodiment are based on the same inventive concept. The system comprises:

[0094] The planting environment monitoring data set sequence obtaining module 11 is used for conditionally sampling the strawberry plants in the target strawberry growth area, obtaining K sample plants, and performing plant growth monitoring and planting environment monitoring on the K sample plants in a preset monitoring window to obtain K sample plant growth monitoring data set sequences and K planting environment monitoring data set sequences, wherein K is an integer greater than or equal to 1.

[0095] The data central value set obtaining module 12 is used for performing sequence internal data central value analysis on the K planting environment monitoring data set sequences to obtain K planting environment monitoring data central value sets.

[0096] The planting environment monitoring data central value mean value set obtaining module 13 is used for taking the planting environment monitoring data type as an index, respectively performing mean value processing on the same type of planting environment monitoring data in the K planting environment monitoring data central value sets, and obtaining a planting environment monitoring data central value mean value set.

[0097] The environment deviation factor determining module 14 is used for determining an environment deviation factor set according to the deviation degree of the planting environment monitoring data central value mean value set and a preset planting environment data set, wherein the environment deviation factor set comprises M environment deviation factors, and each environment deviation factor corresponds to a planting environment monitoring data type.

[0098] The monitoring data mean set sequence obtaining module 15 is configured to calculate the cross-sample data mean of the K sample plant growth monitoring data set sequences to obtain a sample plant growth monitoring data mean set sequence.

[0099] The interactive fusion sample plant growth feature set obtaining module 16 is configured to construct M adaptive convolutional network layers based on the M environmental deviation factors, perform multi-scale feature analysis on the sample plant growth monitoring data mean set sequence by using the M adaptive convolutional network layers, and perform feature correlation interactive fusion analysis on the analysis result to obtain an interactive fusion sample plant growth feature set.

[0100] The data management module 17 is configured to analyze the interactive fusion sample plant growth feature set by using a strawberry growth and development model, determine a strawberry growth and development evaluation factor, determine a planting data extraction frequency for the target strawberry growth area according to the size of the strawberry growth and development evaluation factor, and perform integrated planting data management according to the planting data extraction frequency.

[0101] Further, the interactive fusion sample plant growth feature set obtaining module 16 is configured to perform the following steps:

[0102] The M adaptive convolutional network layers perform data feature extraction on the sample plant growth monitoring data mean set sequence according to M receptive fields to obtain M sample plant growth feature sets, wherein the receptive field is the extraction width of the adaptive convolutional network layer when performing feature extraction on the sample plant growth monitoring data mean set sequence.

[0103] The M sample plant growth feature sets are mapped and sorted according to the order from large to small of the M receptive fields to obtain a sample plant growth feature set sequence.

[0104] The sample plant growth feature sets in the sample plant growth feature set sequence are sequentially subjected to feature correlation interactive fusion analysis to obtain an interactive fusion sample plant growth feature set.

[0105] Further, the interactive fusion sample plant growth feature set obtaining module 16 is configured to perform the following steps:

[0106] The first sample plant growth feature set and the second sample plant growth feature set of the sample plant growth feature set sequence are subjected to feature correlation interactive fusion analysis to obtain a first correlation interactive fusion sample plant growth feature set.

[0107] extracting a third sample plant growth feature set of the sample plant growth feature set sequence, and performing feature correlation interactive fusion analysis on the first correlation interactive fusion sample plant growth feature set to determine a second correlation interactive fusion sample plant growth feature set;

[0108] Based on the second correlation interactive fusion sample plant growth feature set, continue to perform feature correlation interactive fusion analysis on the sample plant growth feature set sequence until the end of the sample plant growth feature set sequence is reached, and obtain the interactive fusion sample plant growth feature set.

[0109] Further, the interactive fusion sample plant growth feature set obtaining module 16 is configured to perform the following steps:

[0110] Performing inner product calculation on the first sample plant growth feature set and the second sample plant growth feature set to determine a first correlation interactive fusion feature similarity set;

[0111] Traversing the first correlation interactive fusion feature similarity set to perform similarity normalization processing, and embedding the processing result in a matrix to obtain a first correlation interactive fusion matrix;

[0112] Performing convolution calculation on the first correlation interactive fusion matrix and the second sample plant growth feature set to determine the first correlation interactive fusion sample plant growth feature set.

[0113] Further, the data central value set obtaining module 12 is configured to perform the following steps:

[0114] Extracting a first planting environment monitoring data set sequence from the K planting environment monitoring data set sequences, wherein the first planting environment monitoring data set sequence includes a plurality of first planting environment monitoring data subsequences of a plurality of planting environment monitoring data types;

[0115] Traversing the plurality of first planting environment monitoring data subsequences to perform data mean calculation to determine a plurality of first data means, taking the plurality of first data means as a plurality of central value analysis starting points, and performing iterative analysis in the plurality of first planting environment monitoring data subsequences according to a preset iteration step length to determine a plurality of first planting environment monitoring data central values, and then collecting the plurality of first planting environment monitoring data central values to obtain the first planting environment monitoring data central value set;

[0116] Performing sequence internal data central value analysis on the K planting environment monitoring data set sequences to determine the K planting environment monitoring data central value sets.

[0117] Further, the data central value set obtaining module 12 is configured to perform the following steps:

[0118] The plurality of concentration value analysis starting points are iterated in the plurality of first planting environment monitoring data subsequences according to the preset iteration step length, and a plurality of iteration data are obtained;

[0119] It is judged whether the concentration density of the plurality of iteration data is greater than or equal to the concentration density of the plurality of concentration value analysis starting points. If yes, the plurality of iteration data are updated as the plurality of concentration value analysis starting points, and the plurality of first planting environment monitoring data subsequences are iterated according to the preset iteration step length.

[0120] Until a preset iteration number is met, the plurality of first planting environment monitoring data corresponding to the plurality of iteration data obtained by the last iteration are taken as the plurality of first planting environment monitoring data central values.

[0121] Further, the planting environment monitoring data set sequence obtaining module 11 is configured to perform the following steps:

[0122] A preset sampling number is obtained, K random sample plants are obtained by randomly sampling the plants in the target strawberry growth area based on the preset sampling number, and the preset sampling number is K.

[0123] The K random sample plants are enumerated two by two to obtain a plurality of plant enumeration combinations, and the distribution position similarity of the plurality of plant enumeration combinations is calculated by using cosine similarity to obtain a plurality of position combination similarities.

[0124] It is judged whether the plurality of position combination similarities meet a preset sampling condition. If yes, the K random sample plants are taken as K sample plants through sampling authentication, and the preset sampling condition is that the number of combinations greater than or equal to a preset position combination similarity in the plurality of position combination similarities is less than a preset number threshold.

[0125] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. Moreover, the above-mentioned embodiments are described in the specification. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

[0126] The above only describes the preferred embodiments of the present application, and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0127] The specification and drawings are, of course, subject to various interpretations and should not be viewed in any limiting sense. It will be understood that various modifications and changes can be made to the application disclosed without departing from the scope of the application. It is therefore intended that the application be limited only by the scope of the appended claims, including any amendments thereof, and their equivalents.

Claims

1. An integrated planting data management method combining a strawberry growth and development model, characterized in that, The method includes: Conditional sampling is performed on strawberry plants in the target strawberry growing area to obtain K sample plants. Plant growth and planting environment are monitored on the K sample plants in a preset monitoring window to obtain a set of K sample plant growth monitoring data and a set of K planting environment monitoring data, where K is an integer greater than or equal to 1. Traverse the K sets of planting environment monitoring data to perform intra-sequence data central value analysis, and obtain the K sets of planting environment monitoring data central values. Using the data type of planting environment monitoring as an index, the mean values ​​of planting environment monitoring data of the same type in the central value sets of the K planting environment monitoring datasets are processed to obtain the central value mean set of planting environment monitoring datasets. Based on the degree of deviation between the mean set of the planting environment monitoring data set and the preset planting environment data set, a set of environmental deviation factors is determined, wherein the set of environmental deviation factors includes M environmental deviation factors, and each environmental deviation factor corresponds to a planting environment monitoring data type; The mean of the sample plant growth monitoring data set is calculated across the K sample plant growth monitoring data set sequence to obtain the mean set sequence of sample plant growth monitoring data; Based on the M environmental deviation factors, M adaptive convolutional network layers are constructed. The M adaptive convolutional network layers are used to perform multi-scale feature analysis on the mean set sequence of the sample plant growth monitoring data. The analysis results are then subjected to feature association and interactive fusion analysis to obtain an interactive fusion set of sample plant growth features. The growth characteristics set of the interactive fusion sample plants are analyzed using a strawberry growth and development model to determine strawberry growth and development evaluation factors. Based on the magnitude of the strawberry growth and development evaluation factors, the frequency of planting data extraction for strawberry plants in the target strawberry growth area is determined, and integrated planting data management is carried out based on the planting data extraction frequency.

2. The method as described in claim 1, characterized in that, Based on the M environmental deviation factors, M adaptive convolutional network layers are constructed. These M layers are then used to perform multi-scale feature analysis on the mean set of the sample plant growth monitoring data. The analysis results are then subjected to feature association and interactive fusion analysis to obtain an interactive fusion set of sample plant growth features, including: The M adaptive convolutional network layers extract data features from the mean set sequence of the sample plant growth monitoring data based on the M receptive fields to obtain M sample plant growth feature sets, wherein the receptive field is the extraction width of the adaptive convolutional network layer when extracting features from the mean set sequence of the sample plant growth monitoring data. Based on the M receptive fields in descending order, the M sample plant growth feature sets are mapped and sorted to obtain a sample plant growth feature set sequence. The growth feature sets of the sample plants in the sequence of the sample plant growth feature set are sequentially subjected to feature association and interaction fusion analysis to obtain the interactive fusion sample plant growth feature set.

3. The method as described in claim 2, characterized in that, The growth feature sets of the sample plants in the sequence of the sample plant growth feature set are sequentially subjected to feature association and interaction fusion analysis to obtain the interactive fusion sample plant growth feature set, including: The first sample plant growth feature set and the second sample plant growth feature set of the sample plant growth feature set sequence are subjected to feature association and interaction fusion analysis to obtain the first association and interaction fusion sample plant growth feature set. Extract the third set of sample plant growth features from the sequence of sample plant growth features, and perform feature association and interaction fusion analysis on it with the first set of associated and interactive fusion sample plant growth features to determine the second set of associated and interactive fusion sample plant growth features. Based on the second associated interactive fusion sample plant growth feature set, the feature association interactive fusion analysis is continued on the sample plant growth feature set sequence until the end of the sample plant growth feature set sequence is reached, thus obtaining the interactive fusion sample plant growth feature set.

4. The method as described in claim 3, characterized in that, A feature association and interaction fusion analysis is performed on the first sample plant growth feature set and the second sample plant growth feature set of the sample plant growth feature set sequence to obtain the first associated and fused sample plant growth feature set, including: The inner product of the first sample plant growth feature set and the second sample plant growth feature set is calculated to determine the first associated interactive fusion feature similarity set. Traverse the first set of similarity of related interaction fusion features, perform similarity normalization processing, and embed the processing result into the matrix to obtain the first related interaction fusion matrix; The first association-interaction fusion matrix and the second sample plant growth feature set are convolved to determine the first association-interaction fusion sample plant growth feature set.

5. The method as described in claim 1, characterized in that, By traversing the K planting environment monitoring data sets and performing intra-sequence data central value analysis, K planting environment monitoring data central value sets are obtained, including: Extract a first planting environment monitoring data set sequence from the K planting environment monitoring data set sequences, wherein the first planting environment monitoring data set sequence includes multiple first planting environment monitoring data sub-sequences of multiple planting environment monitoring data types; The mean of the data is calculated by traversing the multiple first planting environment monitoring data subsequences to determine multiple first data means. The multiple first data means are used as the starting points for multiple central value analysis. Iterative analysis is performed on the multiple first planting environment monitoring data subsequences according to a preset iteration step size to determine multiple central values ​​of the first planting environment monitoring data. The multiple central values ​​of the first planting environment monitoring data are then summarized to obtain the set of central values ​​of the first planting environment monitoring data. Intra-sequence data central value analysis is performed on the K sets of planting environment monitoring data to determine the central value set of the K sets of planting environment monitoring data.

6. The method as described in claim 5, characterized in that, include: The multiple lumped value analysis starting points are iterated in the multiple first planting environment monitoring data subsequences according to the preset iteration step size to obtain multiple iterative data; Determine whether the concentration density of the multiple iterative data is greater than or equal to the concentration density of the multiple ensemble analysis starting points. If so, update the multiple iterative data to multiple ensemble analysis starting points, and iterate in the multiple first planting environment monitoring data subsequences according to the preset iteration step size. Until the preset number of iterations is met, the multiple first planting environment monitoring data corresponding to the multiple iteration data obtained in the last iteration are used as the central value of the multiple first planting environment monitoring data.

7. The method as described in claim 1, characterized in that, Conditional sampling was performed on plants within the target strawberry growing area to obtain K sample plants, including: A preset sampling quantity is obtained, and plants in the target strawberry growing area are randomly sampled based on the preset sampling quantity to obtain K random sample plants, wherein the preset sampling quantity is K; The K random sample plants are enumerated pairwise to obtain multiple plant enumeration combinations. The similarity of the distribution positions of the multiple plant enumeration combinations is calculated using cosine similarity to obtain the similarity of multiple position combinations. Determine whether the similarity of the multiple position combinations meets the preset sampling conditions. If so, pass the sampling authentication and take the K random sample plants as K sample plants. The preset sampling conditions are that the number of combinations with similarity greater than or equal to the preset position combination similarity is less than the preset number threshold.

8. An integrated planting data management system combining a strawberry growth and development model, characterized in that, The system is used to execute the integrated planting data management method combining a strawberry growth and development model as described in any one of claims 1-7, and the system includes: The planting environment monitoring data set sequence acquisition module is used to conditionally sample strawberry plants in the target strawberry growing area to obtain K sample plants. The module then monitors the plant growth and planting environment of the K sample plants in a preset monitoring window to obtain a K sample plant growth monitoring data set sequence and a K planting environment monitoring data set sequence, where K is an integer greater than or equal to 1. The data central value set acquisition module is used to traverse the K planting environment monitoring data set sequences to perform intra-sequence data central value analysis and obtain the K planting environment monitoring data central value set; The module for obtaining the mean set of planting environment monitoring data is used to perform mean processing on planting environment monitoring data of the same type in the K sets of planting environment monitoring data, using the data type of planting environment monitoring as an index, to obtain the mean set of planting environment monitoring data. The environmental deviation factor determination module is used to determine the set of environmental deviation factors based on the degree of deviation between the mean set of the planting environment monitoring data set and the preset planting environment data set. The set of environmental deviation factors includes M environmental deviation factors, and each environmental deviation factor corresponds to a planting environment monitoring data type. The monitoring data mean set sequence acquisition module is used to perform cross-sample data mean calculation on the K sample plant growth monitoring data set sequence to obtain the sample plant growth monitoring data mean set sequence; The interactive fusion sample plant growth feature set acquisition module is used to construct M adaptive convolutional network layers based on the M environmental deviation factors, use the M adaptive convolutional network layers to perform multi-scale feature analysis on the mean set sequence of the sample plant growth monitoring data, and perform feature association interactive fusion analysis on the analysis results to obtain the interactive fusion sample plant growth feature set. The data management module is used to analyze the growth characteristic set of the interactive fusion sample plants using the strawberry growth and development model, determine the strawberry growth and development evaluation factors, and determine the frequency of planting data extraction for strawberry plants in the target strawberry growth area based on the magnitude of the strawberry growth and development evaluation factors, and perform integrated planting data management based on the planting data extraction frequency.

Images