A method and system for identifying crops for agricultural irrigation

Abstract

The present invention provides a crop identification method and system for agricultural irrigation, belonging to the field of crop identification. The method comprises: obtaining data of each well user of each well in a target area, wherein the data of the well user includes the power consumption data of the well user and the planting area data of the user; removing the non-agricultural power consumption data from the power consumption data of each well user, and obtaining the agricultural power consumption data of each well; obtaining the time series data of the water consumption of each well according to the agricultural power consumption data of each well and the planting area data of the user corresponding to each well; inputting the time series data of the water consumption of each well into a pre-trained classification model, and obtaining the type of crops corresponding to each well in the target area. The present invention can make the identification result of crops more accurate, improve efficiency and save costs, which is of great significance for planning irrigation volume.

Description

Technical Field

[0001] The present invention relates to the technical field of crop identification, and in particular to a crop identification method and system for agricultural irrigation. Background Art

[0002] Irrigation is one of the technical measures to supplement the water needed by crops. In order to ensure the normal growth of crops and obtain high and stable yields, crops must be supplied with sufficient water. Under natural conditions, the water requirements of crops are often not met due to insufficient precipitation or uneven distribution. Therefore, artificial irrigation must be carried out to make up for the lack of natural rainfall.

[0003] In North my country, water for crop irrigation mainly comes from groundwater. Therefore, the irrigation characteristics of users can be identified and analyzed by statistically analyzing the temporal and spatial distribution of water consumption in agricultural wells.

[0004] At present, when irrigation is carried out, the irrigation amount is usually planned based on the planting area of ​​crops, planting conditions, and power consumption of wells. However, statistics on the planting area and planting types of farmland crops require a lot of resources to investigate. At present, it is generally based on satellite remote sensing images and image recognition methods, which is costly and inefficient.

[0005] That is to say, at the current stage, there are problems of high cost and low efficiency when counting the cultivated area of ​​farmland crops, planting types, etc. Summary of the invention

[0006] The embodiment of the present invention provides a crop identification method and system for agricultural irrigation, so as to solve the problems of high cost and low efficiency in counting the planting area and planting types of farmland crops at the current stage.

[0007] In a first aspect, an embodiment of the present invention provides a method for identifying crops for agricultural irrigation, comprising:

[0008] Acquire the data of the well users of each well in the target area, wherein the data of the well users include the electricity consumption data of the well users and the planting area data of the users;

[0009] Eliminate the non-agricultural electricity consumption data from the electricity consumption data of each well user to obtain the agricultural electricity consumption data of each well;

[0010] The time series data of water consumption of each well is obtained based on the agricultural electricity consumption data of each well and the planting area data of the users corresponding to each well;

[0011] The time series data of water consumption of each well is input into the pre-trained classification model to obtain the type of crops corresponding to each well in the target area.

[0012] In a possible implementation, before removing the non-agricultural electricity consumption data from the electricity consumption data of each well user and obtaining the agricultural electricity consumption data of each well, the following is further included:

[0013] The data of each well user is input into the Gaussian mixture model to obtain the well user category to which each well user belongs; the well user category includes pure agricultural irrigation and drainage well users, mixed agricultural irrigation and drainage well users and non-agricultural irrigation and drainage well users.

[0014] In a possible implementation, non-agricultural electricity consumption data is removed from the electricity consumption data of each well user to obtain agricultural electricity consumption data of each well, including:

[0015] For each well, the electricity consumption data of non-agricultural irrigation and drainage well users and the non-agricultural electricity consumption data of mixed agricultural irrigation and drainage well users are eliminated from the well users of the well to obtain the agricultural electricity consumption data of the well.

[0016] In a possible implementation, the time series data of the water consumption of each well is obtained according to the agricultural electricity consumption data of each well and the planting area data of the user corresponding to each well, including:

[0017] The agricultural water consumption corresponding to the agricultural electricity consumption data of each well is calculated by using the electricity-to-water conversion coefficient;

[0018] The time series data of the water consumption of each machine-mounted well is obtained according to the agricultural water consumption of each machine-mounted well and the planting area data of the users corresponding to each machine-mounted well.

[0019] In a possible implementation, the time series data of the water consumption of each well includes time series at multiple moments; the time series data of the water consumption of each well is input into a pre-trained classification model to obtain the types of crops corresponding to each well in the target area, including:

[0020] For each pumped well, taking each moment as a reference, the time series data of the water consumption of the pumped well is intercepted forward by a signal segment window data of a preset width to obtain multi-dimensional signal vector data of the water consumption of the pumped well; and according to the multi-dimensional signal vector data of the water consumption of the pumped well, an input data matrix of the water consumption of the pumped well is obtained;

[0021] The input data matrix of water consumption of each well is input into the pre-trained classification model to obtain the type of crops corresponding to each well in the target area.

[0022] In a possible implementation, the input data matrix of the water consumption of each well is input into a pre-trained classification model to obtain the types of crops corresponding to each well in the target area, including:

[0023] Input the input data matrix of water consumption of each well into the pre-trained classification model to obtain the expected number of classifications;

[0024] The Toeplitz matrix is ​​introduced into the classification model to obtain the inverse covariance matrix; wherein the inverse covariance matrix is ​​used to describe the correlation of the time series data between the water consumption of each well;

[0025] Generate the objective function based on the inverse covariance matrix;

[0026] Solve the objective function to obtain the clustering result;

[0027] According to the clustering results, the types of crops corresponding to each well in the target area are obtained.

[0028] In a possible implementation, the types of crops corresponding to each well in the target area are obtained according to the clustering results, including:

[0029] The F1 score was used as a metric to analyze the clustering results and obtain the types of crops corresponding to each well in the target area.

[0030] In a possible implementation, after obtaining the user data of each well, the method further includes:

[0031] The data of each machine-well user of each machine-well in the target area are preprocessed; and the preprocessed data of each machine-well user are standardized using the z-score standardization method.

[0032] In a possible implementation, after inputting the time series data of water consumption of each well into a pre-trained classification model to obtain the types of crops corresponding to each well in the target area, the following is further included:

[0033] The types of crops and planting proportions corresponding to each well in the target area are visualized according to the target area or each well.

[0034] In a second aspect, an embodiment of the present invention provides a crop identification system for agricultural irrigation, comprising:

[0035] A data processing module is used to obtain the data of each well user of each well in the target area, wherein the data of the well user includes the power consumption data of the well user, the rainfall data of the area where the user is located and the planting area data of the user;

[0036] The non-irrigation electricity consumption identification module is used to remove the non-agricultural electricity consumption data from the electricity consumption data of the pumped well users to obtain the agricultural electricity consumption data of each pumped well;

[0037] The irrigation characteristic analysis module is used to obtain the time series data of the water consumption of each well according to the agricultural electricity consumption data of each well and the planting area data of the user corresponding to each well;

[0038] The irrigation feature analysis module is also used to input the time series data of the water consumption of each well into a pre-trained classification model to obtain the type of crops corresponding to each well in the target area.

[0039] The embodiment of the present invention provides a crop identification method and system for agricultural irrigation. By acquiring the data of each well user of each well in the target area, the power consumption data of each well and the planting area data of the user corresponding to each well can be clarified. In order to make the identification result more accurate, the non-agricultural power consumption data in the power consumption data of each well user can be eliminated to obtain the agricultural power consumption data of each well. Since different crops require different irrigation amounts and have different time series, the time series data of the water consumption of each well can be obtained according to the agricultural power consumption data of each well and the planting area data of the user corresponding to each well; the time series data of the water consumption of each well is input into the pre-trained classification model to obtain the type of crops corresponding to each well in the target area. In the embodiment of the present invention, when performing crop identification, it is not necessary to use satellites to identify the types of crops, and the types of crops irrigated by each well can be identified only through the data of each well user. In the process of identification, the non-agricultural power consumption data is eliminated, so that the identification result is more accurate, the efficiency is improved and the cost is saved. The embodiment of the present invention identifies the types of crops irrigated by each well in the target area, which is of great significance for planning the irrigation amount. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 is a flow chart of an implementation method of a crop identification method for agricultural irrigation provided by an embodiment of the present invention;

[0042] Figure 2 This is a system architecture diagram for implementing a crop identification method for agricultural irrigation provided by an embodiment of the present invention;

[0043] Figure 3 It is a structural schematic diagram of a crop identification system for agricultural irrigation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0045] In order to make the purpose, technical solutions and advantages of the present invention more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0046] Figure 1 FIG. 1 is a flow chart of a method for identifying crops for agricultural irrigation provided by an embodiment of the present invention. Figure 1 As shown:

[0047] Step 110: Acquire the data of the well users of each well in the target area, wherein the data of the well users include the electricity consumption data of the well users and the planting area data of the users.

[0048] In this embodiment, there is at least one well in the target area, and each well corresponds to a user. Accordingly, each well has data of the corresponding well user. The data of the well user can be stored in the account corresponding to each well. Among them, the data of the well user can include the annual electricity consumption data of the well user, the rainfall data of the area where the well (or the planting area) is located, the longitude and latitude data of the area where the well is located, the planting area data of the user, etc. Among them, the electricity consumption data can also include electricity consumption time, electricity power, load, etc., which are not limited here.

[0049] Step 120: Eliminate the non-agricultural electricity consumption data from the electricity consumption data of each well user to obtain the agricultural electricity consumption data of each well.

[0050] In this embodiment, in some scenarios, not all wells are used for irrigation, or not all electricity consumption data of each well is generated by irrigation. This embodiment is to identify the type of crops corresponding to each well when it is irrigated. Therefore, in order to ensure the accuracy of identification, it is necessary to eliminate interference data, that is, non-agricultural electricity consumption data in the electricity consumption data of each well user, to obtain the agricultural electricity consumption data of each well.

[0051] Step 130: Obtain the time series data of the water consumption of each machine-mounted well according to the agricultural electricity consumption data of each machine-mounted well and the planting area data of the users corresponding to each machine-mounted well.

[0052] In this embodiment, since the agricultural electricity consumption data of each well can be the annual electricity consumption data of each well, which has a certain time series, then correspondingly, according to the agricultural electricity consumption data with time series and the planting area data of the user corresponding to the well, the water consumption of each well also has a time series, that is, the time series data of the water consumption of each well can be obtained. The time series data of the water consumption of each well can describe the water consumption and irrigation time of the crops irrigated by each well per unit area.

[0053] Step 140: Input the time series data of the water consumption of each well into a pre-trained classification model to obtain the type of crops corresponding to each well in the target area.

[0054] In this embodiment, there are significant differences in irrigation time and water consumption for different types of crops. On the basis of obtaining the irrigation characteristics of each type of crop, the classification model can be used to obtain the type of crop corresponding to each well in the target area according to the input time series data of the water consumption of each well.

[0055] In summary, in North my country, water for crop irrigation mainly comes from groundwater. Therefore, the identification and analysis of user irrigation characteristics can be performed by statistically analyzing the temporal and spatial distribution of agricultural well water consumption. The embodiment of the present invention does not need to use satellites to identify crop types when identifying crops. The types of crops irrigated by each well can be identified only through the data of each well user. Non-agricultural electricity consumption data is eliminated during the identification process to make the identification result more accurate, improve efficiency and save costs. The embodiment of the present invention is of great significance for planning irrigation volume by identifying the types of crops irrigated by each well in the target area.

[0056] In some embodiments, before removing the non-agricultural electricity consumption data from the electricity consumption data of each well user in step 120 and obtaining the agricultural electricity consumption data of each well, the following may also be included:

[0057] The data of each well user is input into the Gaussian mixture model to obtain the well user category to which each well user belongs; the well user category includes pure agricultural irrigation and drainage well users, mixed agricultural irrigation and drainage well users and non-agricultural irrigation and drainage well users.

[0058] In this embodiment, before eliminating the non-agricultural electricity consumption data from the electricity consumption data of each motor well user, the electricity consumption data of each motor well user needs to be classified to determine which motor well users' electricity consumption data need to be presented.

[0059] There are certain differences in the time distribution of electricity consumption behaviors of different types of users. For example, the electricity consumption data of pure agricultural irrigation and drainage well users generally show that it is more concentrated during the irrigation period, and the electricity consumption data during the irrigation period is relatively balanced; the electricity consumption data of mixed agricultural irrigation and drainage well users generally show that it is more concentrated during the irrigation period and there is also electricity consumption data during the non-irrigation period; the electricity consumption data of non-agricultural irrigation and drainage well users generally show that the electricity consumption data is uncertain.

[0060] Then, based on these characteristics, the machine well user category to which each machine well user belongs can be determined. Specifically, the power consumption data of each machine well user is input into the Gaussian mixture model as a group of data according to the machine well; the Gaussian mixture model outputs three machine well user categories, each of which includes a certain number of power consumption data of each machine well user in units of machine wells, so as to determine the machine well user category to which each machine well user belongs.

[0061] In some embodiments, in step 120, non-agricultural electricity consumption data is removed from the electricity consumption data of each well user to obtain agricultural electricity consumption data of each well, which may include:

[0062] For each well, the electricity consumption data of non-agricultural irrigation and drainage well users and the non-agricultural electricity consumption data of mixed agricultural irrigation and drainage well users are eliminated from the well users of the well to obtain the agricultural electricity consumption data of the well.

[0063] In this embodiment, by eliminating the non-agricultural electricity consumption data of each well user, the agricultural electricity consumption data of each well can be obtained, thereby making it more accurate to identify the type of crops irrigated by each well.

[0064] In some embodiments, in step 130, the time series data of the water consumption of each well is obtained according to the agricultural electricity consumption data of each well and the planting area data of the user corresponding to each well, including:

[0065] The agricultural water consumption corresponding to the agricultural electricity consumption data of each well is calculated based on the electricity-to-water coefficient.

[0066] The time series data of the water consumption of each machine-mounted well is obtained according to the agricultural water consumption of each machine-mounted well and the planting area data of the users corresponding to each machine-mounted well.

[0067] In this embodiment, "electricity to water" is a method of analyzing the hydropower conversion coefficient by studying the relationship between power consumption and water intake, and then calculating the water intake by multiplying the power consumption by the hydropower conversion coefficient. The agricultural electricity consumption data of each well can be the electricity consumption data for the whole year, which has temporal and spatial characteristics. Since the statistics of agricultural water use generally adopt the "electricity to water" method, the agricultural water consumption corresponding to the agricultural electricity consumption data of each well can be obtained by multiplying the agricultural electricity consumption data of each well by the "electricity to water coefficient". Therefore, the electricity consumption data characteristics of each well can be converted into water consumption characteristics. That is, the agricultural water consumption of each well also has temporal and spatial characteristics.

[0068] Since the non-agricultural electricity consumption data is not used for irrigation, the agricultural electricity consumption data of each well has the corresponding user's planting area data. Therefore, the time series data of the water consumption of each well is obtained based on the agricultural water consumption of each well and the planting area data of the user corresponding to each well. Among them, the time series data of the water consumption of the well can represent the water consumption of the crops per unit planting area irrigated by the well and the irrigation time.

[0069] In some embodiments, the time series data of the water consumption of each well includes a time series of multiple moments; in step 140, the time series data of the water consumption of each well is input into a pre-trained classification model to obtain the types of crops corresponding to each well in the target area, which may include:

[0070] Step 141: for each pumped well, based on each moment, intercept the time series data of the water consumption of the pumped well forward by a signal segment window data of a preset width to obtain multi-dimensional signal vector data of the water consumption of the pumped well; and obtain the input data matrix of the water consumption of the pumped well according to the multi-dimensional signal vector data of the water consumption of the pumped well;

[0071] Step 142: Input the input data matrix of the water consumption of each well into the pre-trained classification model to obtain the type of crops corresponding to each well in the target area.

[0072] In this embodiment, the pre-trained classification model can be a time series segmentation clustering algorithm model (Toeplitz Inverse Covariance-based Clustering, TICC). Based on this model, the TICC method is used to identify and analyze the irrigation characteristics of agricultural well users, targeting the time and spatial distribution characteristics of electricity consumption of agricultural well users, as well as the geographical location and irrigation area of ​​the wells, which can effectively support the identification of irrigation crops for well users, water demand analysis, and other tasks.

[0073] Since the traditional statistical methods of farmland crop planting area and planting types have certain limitations, this embodiment takes into account the low cost of obtaining electricity characteristics and the easy online acquisition and analysis of data through the well power data collection system, so the identification of irrigation characteristics based on electricity characteristics can effectively reduce costs and improve analysis efficiency.

[0074] Considering that the time series data of irrigation electricity consumption has problems of inconsistent time intervals and inconsistent durations, and that it is affected by precipitation and irrigation cycles and has periodic changes; there is a strong correlation between precipitation and irrigation water, geographic location and other dimensions. Traditional clustering methods examine the absolute values ​​of each dimension of the signal to determine the similarity between signals. The TICC method adopted in the embodiment of the present invention examines the correlation between the dimensions of the signal to determine the similarity between the signals, which is more compatible with business problems.

[0075] In this embodiment, the preset width may be w, and the time series data of water consumption of each pumped well user may be: X i =[X i-w+1 ,……,X i-1 ,X i ], i = 1, 2, ..., T. Among them, the data (also signal) at each moment is X i is an n-dimensional vector.

[0076] In order to examine the correlation between signals, we take a signal segment window X with a width of w at each moment. i , the signal segment is an nw-dimensional vector.

[0077] Therefore, the input data matrix of any well can be expressed as:

[0078]

[0079] in, Represents the field of real numbers.

[0080] The input data matrix of the water consumption of each well including agricultural water consumption in the target area is input into the pre-trained classification model to obtain the types of crops corresponding to each well including agricultural water consumption in the target area.

[0081] The present embodiment is described below using a specific embodiment:

[0082] Well user A is a pure agricultural irrigation and drainage well user, and the electricity consumption data is 364 days a year (for the convenience of example calculation, the actual data volume can be set as needed). Take each day as a moment and intercept the signal segment window X with a width of 7 forward. i , and obtain the multi-dimensional signal vector data of well user A:

[0083] The input data matrix of machine well user A is expressed as:

[0084]

[0085] According to the input data matrix method of well user A, the input data matrix of other well users with agricultural electricity consumption in the target area can be obtained. These input data matrices are input into the pre-trained classification model in units of wells to obtain the types of crops corresponding to each well with agricultural electricity consumption in the target area.

[0086] In some embodiments, in step 142, the input data matrix of the water consumption of each well is input into a pre-trained classification model to obtain the types of crops corresponding to each well in the target area, which may include:

[0087] Input the input data matrix of water consumption of each well into the pre-trained classification model to obtain the expected number of classifications;

[0088] The Toeplitz matrix is ​​introduced into the classification model to obtain the inverse covariance matrix; wherein the inverse covariance matrix is ​​used to describe the correlation of the time series data between the water consumption of each well;

[0089] Generate the objective function based on the inverse covariance matrix;

[0090] Solve the objective function to obtain the clustering result;

[0091] According to the clustering results, the types of crops corresponding to each well in the target area are obtained.

[0092] In this embodiment, after the input data matrix of the water consumption of each well is input into the pre-trained classification model, all signals (data) in each well can be divided into K classes based on the well as expected, that is, P j ,j=1,2,3,…,k,P j Indicates category.

[0093] Assume that each cluster obtained by clustering the TICC model represents a specific "state" in the original time series data of each well. When the original time series data is in this specific state, it will retain a certain time-invariant structure, which will persist in the time window covered by the entire cluster, regardless of the starting point of the time window. Therefore, the Toeplitz matrix is ​​introduced, and the inverse covariance matrix Θ is used i Describe the correlation between the series:

[0094]

[0095] Θ iIt is composed of w×w sub-matrices, which are matrices of size n×n, where Represents the dependency relationship between sensor i and sensor j at the current moment, and so on. Represents the dependency between sensor i at the current moment and sensor j at the next moment. Sensor i and sensor j represent adjacent signals (data) in the matrix.

[0096] By inputting the above matrix into the classification model, we can get K Gaussian inverse covariance solutions Θ = {Θ1, Θ2, …, Θ K} and its distribution set P = {P1,…,P K},in

[0097] Therefore, the overall objective function can be defined as:

[0098]

[0099] In the above formula, we try to ensure that each θ is sparse as much as possible to increase interpretability, so the first penalty term is added In order to make adjacent time series data X i-1 and X i As much as possible, group them into one category and add the last item of the function Take the minimum value (where β is a parameter).

[0100] Use the maximum likelihood estimation method to ensure that the middle term of the function Take the minimum value, where l(X t ,Θ i ) is defined as:

[0101]

[0102] After obtaining the objective function, the expectation-maximization (EM) algorithm can be used to solve the objective function, and the objective function solution process is transformed into solving the following two problems in turn:

[0103] That is, given Θ i , solve the classification P i And given P i , solve for Θ i, to obtain the clustering results. Then, according to the clustering results, the types of crops corresponding to each well in the target area are obtained. Among them, when determining the types of crops corresponding to each well in the target area, the crops can be first divided into field crops and cash crops, and then the preset width w is changed to obtain a new input data matrix for each well user, and the new input data matrix for each well user is input into the pre-trained classification model, and then the two crop categories of field crops and cash crops are refined, and the specific analysis is made on which crop is irrigated by each well. For example, cash crops can be further divided into fruit trees, cotton, etc.

[0104] In addition, considering that different rainfall will also affect the irrigation amount, the data of each well user can include the rainfall data of the area where the well is located. While inputting the input data matrix of the water consumption of each well into the pre-trained classification model, the rainfall data corresponding to each well can also be input into the pre-trained classification model to ensure the accuracy of the final classification result.

[0105] In some embodiments, according to the clustering results, the types of crops corresponding to the wells in the target area may be obtained, which may include:

[0106] The F1 score was used as a metric to analyze the clustering results and obtain the types of crops corresponding to each well in the target area.

[0107] In this embodiment, the F1 score can measure the accuracy of the model. After solving the objective function and obtaining the clustering results. The F1 score can be used as a metric to analyze the clustering results, that is, to evaluate the classification model. When the accuracy meets the requirements, the types of crops corresponding to each well in the corresponding target area are obtained; when the accuracy does not meet the requirements, a new objective function needs to be generated, and the objective function needs to be re-solved to obtain a new clustering result, and the new clustering result needs to be re-analyzed until the accuracy meets the requirements, and the types of crops corresponding to each well in the corresponding target area are obtained. In some embodiments, after obtaining the user data of each well in step 110, it can also include:

[0108] The data of each machine-well user of each machine-well in the target area are preprocessed; and the preprocessed data of each machine-well user are standardized using the z-score standardization method.

[0109] In this embodiment, the data of each well user of each well in the target area contains duplicate items, erroneous items, and blank items, so it is necessary to preprocess these data. The preprocessing methods include but are not limited to data merging, data cleaning, data enhancement, etc. Finally, the preprocessed data is standardized using the z-score method. In particular, when preprocessing the data, since the data comes from multiple well users, the timestamp can be processed as a one-dimensional signal.

[0110] In some embodiments, after inputting the time series data of water consumption of each well into a pre-trained classification model to obtain the types of crops corresponding to each well in the target area, the following may also be included:

[0111] The types of crops and planting proportions corresponding to each well in the target area are visualized according to the target area or each well.

[0112] In this embodiment, the types of crops planted by each well user in the statistical area, the planting ratio, etc. can be visualized in the form of a chart. Specifically, the displayed content may include: the types of crops and planting ratios in the target area, the location information of each well in the target area, the rainfall of each well, the types of crops irrigated by each well and the corresponding planting ratios, etc.

[0113] Figure 2 is a system architecture diagram of a method for identifying crops for agricultural irrigation provided by an embodiment of the present invention; Figure 2 As shown:

[0114] In this embodiment, the system structure may include a data processing module, a pumped well user irrigation feature analysis module, a non-irrigation electricity identification module, an irrigation feature analysis module and a result visualization module.

[0115] Specifically, the data processing module is used to pre-process the original power data (ie, the power consumption data of each pumped well user) after the system obtains the original power data to obtain the effective power consumption data of each pumped well user.

[0116] The pumped well user irrigation characteristic analysis module is used to determine the pumped well user category to which each pumped well user belongs based on the effective electricity consumption data of each pumped well user.

[0117] The non-irrigation electricity consumption identification module is used to eliminate the non-agricultural electricity consumption data of the wells belonging to mixed agricultural irrigation and drainage well users and non-agricultural irrigation and drainage well users, and obtain the agricultural electricity consumption data of each well user.

[0118] The irrigation feature analysis module is used to determine the types of crops corresponding to each well in the target area through the TICC model based on the agricultural electricity consumption data of each well user.

[0119] The result visualization module is used to visualize the types of crops and planting proportions corresponding to each well in the target area according to the target area or each well in the GUI interface through the visualization of ECharts charts.

[0120] The non-irrigation electricity identification module and the irrigation feature analysis module need to be pre-trained based on the data processed by the data processing module. For the parts not fully described in this embodiment, reference can be made to the contents described in the above embodiments.

[0121] In summary, the embodiment of the present invention provides a method and system for crop identification for agricultural irrigation, which uses the TICC method to segment and cluster the corresponding electricity consumption time series data of well users; on this basis, a system is designed to achieve online automation from the acquisition of agricultural well user electricity consumption data to the analysis of user irrigation crops. The embodiment of the present invention has higher accuracy, data acquisition efficiency and operation efficiency in crop identification, and can achieve online automation from the acquisition of agricultural well user electricity consumption data to the analysis of user irrigation crops, which is of great significance for planning irrigation volume.

[0122] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean 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 embodiment of the present invention.

[0123] The following is an embodiment of the device of the present invention. For details not described in detail therein, reference may be made to the corresponding method embodiment described above.

[0124] Figure 3 1 is a schematic diagram of the structure of a crop identification system for agricultural irrigation provided by an embodiment of the present invention. For the sake of convenience, only the parts related to the embodiment of the present invention are shown, which are described in detail as follows:

[0125] like Figure 3 As shown, a crop identification system 3 for agricultural irrigation includes:

[0126] The data processing module 31 is used to obtain the data of each well user of each well in the target area, wherein the data of the well user includes the power consumption data of the well user, the rainfall data of the area where the user is located and the planting area data of the user;

[0127] The non-irrigation electricity consumption identification module 32 is used to remove the non-agricultural electricity consumption data from the electricity consumption data of the pumped well users, and obtain the agricultural electricity consumption data of each pumped well;

[0128] The irrigation characteristic analysis module 33 is used to obtain the time series data of the water consumption of each well according to the agricultural electricity consumption data of each well and the planting area data of the user corresponding to each well;

[0129] The irrigation feature analysis module 33 is also used to input the time series data of the water consumption of each well into a pre-trained classification model to obtain the type of crops corresponding to each well in the target area.

[0130] In a possible implementation, a crop identification system 3 for agricultural irrigation further includes a well user classification module 34, and the well user classification module 34 is specifically used for:

[0131] The data of each well user is input into the Gaussian mixture model to obtain the well user category to which each well user belongs; the well user category includes pure agricultural irrigation and drainage well users, mixed agricultural irrigation and drainage well users and non-agricultural irrigation and drainage well users.

[0132] In a possible implementation, the non-irrigation electricity identification module 32 is specifically used to:

[0133] For each well, the electricity consumption data of non-agricultural irrigation and drainage well users and the non-agricultural electricity consumption data of mixed agricultural irrigation and drainage well users are eliminated from the well users of the well to obtain the agricultural electricity consumption data of the well.

[0134] In a possible implementation, the irrigation feature analysis module 33 is specifically used for:

[0135] The agricultural water consumption corresponding to the agricultural electricity consumption data of each well is calculated by using the electricity-to-water conversion coefficient;

[0136] The time series data of the water consumption of each machine-mounted well is obtained according to the agricultural water consumption of each machine-mounted well and the planting area data of the users corresponding to each machine-mounted well.

[0137] In a possible implementation, the time series data of the water consumption of each pumped well includes time series at multiple moments; the irrigation feature analysis module 33 is specifically used for:

[0138] For each pumped well, taking each moment as a reference, the time series data of the water consumption of the pumped well is intercepted forward by a signal segment window data of a preset width to obtain multi-dimensional signal vector data of the water consumption of the pumped well; and according to the multi-dimensional signal vector data of the water consumption of the pumped well, an input data matrix of the water consumption of the pumped well is obtained;

[0139] The input data matrix of water consumption of each well is input into the pre-trained classification model to obtain the type of crops corresponding to each well in the target area.

[0140] In a possible implementation, the irrigation feature analysis module 33 is specifically used for:

[0141] Input the input data matrix of water consumption of each well into the pre-trained classification model to obtain the expected number of classifications;

[0142] The Toeplitz matrix is ​​introduced into the classification model to obtain the inverse covariance matrix; wherein the inverse covariance matrix is ​​used to describe the correlation of the time series data between the water consumption of each well;

[0143] Generate the objective function based on the inverse covariance matrix;

[0144] Solve the objective function to obtain the clustering result;

[0145] According to the clustering results, the types of crops corresponding to each well in the target area are obtained.

[0146] In a possible implementation, the irrigation feature analysis module 33 is specifically used for:

[0147] The F1 score was used as a metric to analyze the clustering results and obtain the types of crops corresponding to each well in the target area.

[0148] In a possible implementation, the acquisition module 31 is further used for:

[0149] The data of each machine-well user of each machine-well in the target area are preprocessed; and the preprocessed data of each machine-well user are standardized using the z-score standardization method.

[0150] In a possible implementation, a crop identification system 3 for agricultural irrigation further includes a result visualization module 36, and the result visualization module 36 is specifically used for:

[0151] The types of crops and planting proportions corresponding to each well in the target area are visualized according to the target area or each well.

[0152] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0153] Those of ordinary skill in the art will appreciate that the templates, units, and algorithm steps of each example 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 performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0154] If the module / unit is implemented in the form of 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, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of each of the above-mentioned crop identification method embodiments for agricultural irrigation. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal and software distribution medium.

[0155] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for identifying crops for agricultural irrigation, characterized in that: include: Acquire the data of the well users of each well in the target area, wherein the data of the well users include the power consumption data of the well users and the planting area data of the users; Eliminate the non-agricultural electricity consumption data from the electricity consumption data of each well user to obtain the agricultural electricity consumption data of each well; The time series data of water consumption of each well is obtained based on the agricultural electricity consumption data of each well and the planting area data of the users corresponding to each well; The time series data of water consumption of each well is input into the pre-trained classification model to obtain the types of crops corresponding to each well in the target area; The time series data of water consumption of each well is obtained according to the agricultural electricity consumption data of each well and the planting area data of the user corresponding to each well, including: The agricultural water consumption corresponding to the agricultural electricity consumption data of each well is calculated by using the electricity-to-water conversion coefficient; According to the agricultural water consumption of each well and the planting area data of the users corresponding to each well, the time series data of the water consumption of each well is obtained; The time series data of the water consumption of each well includes time series at multiple moments; the time series data of the water consumption of each well is input into a pre-trained classification model to obtain the types of crops corresponding to each well in the target area, including: For each pumped well, taking each moment as a reference, the time series data of the water consumption of the pumped well is intercepted forward by a signal segment window data of a preset width to obtain multi-dimensional signal vector data of the water consumption of the pumped well; and according to the multi-dimensional signal vector data of the water consumption of the pumped well, an input data matrix of the water consumption of the pumped well is obtained; The input data matrix of water consumption of each well is input into the pre-trained classification model to obtain the types of crops corresponding to each well in the target area; The input data matrix of the water consumption of each well is input into the pre-trained classification model to obtain the types of crops corresponding to each well in the target area, including: Input the input data matrix of water consumption of each well into the pre-trained classification model to obtain the expected number of classifications; Introducing the Toeplitz matrix into the classification model, an inverse covariance matrix is ​​obtained; wherein the inverse covariance matrix is ​​used to describe the correlation of the time series data between the water consumption of each well; generating an objective function based on the inverse covariance matrix; Solving the objective function to obtain a clustering result; According to the clustering result, the types of crops corresponding to each well in the target area are obtained.

2. A method for identifying crops for agricultural irrigation according to claim 1, characterized in that: Before removing the non-agricultural electricity consumption data from the electricity consumption data of each well user to obtain the agricultural electricity consumption data of each well, the method further includes: The data of each pumped well user is input into a Gaussian mixture model to obtain the pumped well user category to which each pumped well user belongs; wherein the pumped well user category includes pure agricultural irrigation and drainage pumped well users, mixed agricultural irrigation and drainage pumped well users and non-agricultural irrigation and drainage pumped well users.

3. A method for identifying crops for agricultural irrigation according to claim 2, characterized in that: The non-agricultural electricity consumption data in the electricity consumption data of each well user is removed to obtain the agricultural electricity consumption data of each well, including: For each well, the electricity consumption data of non-agricultural irrigation and drainage well users and the non-agricultural electricity consumption data of the mixed agricultural irrigation and drainage well users are eliminated to obtain the agricultural electricity consumption data of the well.

4. A method for identifying crops for agricultural irrigation according to claim 1, characterized in that: The types of crops corresponding to the wells in the target area are obtained according to the clustering results, including: The clustering results are analyzed using the F1 score as a metric to obtain the types of crops corresponding to each well in the target area.

5. A method for identifying crops for agricultural irrigation according to claim 1, characterized in that: After obtaining the user data of each well, the method further includes: The data of each machine-well user of each machine-well in the target area are preprocessed; and the preprocessed data of each machine-well user are standardized using the z-score standardization method.

6. A method for identifying crops for agricultural irrigation according to claim 1, characterized in that: After inputting the time series data of water consumption of each well into the pre-trained classification model to obtain the types of crops corresponding to each well in the target area, the method further includes: The types of crops and planting proportions corresponding to each well in the target area are visualized according to the target area or each well.

7. A crop identification system for agricultural irrigation, characterized in that: include: A data processing module is used to obtain the data of each well user of each well in the target area, wherein the data of the well user includes the power consumption data of the well user, the rainfall data of the area where the user is located and the planting area data of the user; A non-irrigation electricity consumption identification module is used to remove the non-agricultural electricity consumption data from the electricity consumption data of the pumped well users to obtain the agricultural electricity consumption data of each pumped well; The irrigation characteristic analysis module is used to obtain the time series data of the water consumption of each well according to the agricultural electricity consumption data of each well and the planting area data of the user corresponding to each well; The irrigation feature analysis module is also used to input the time series data of the water consumption of each well into a pre-trained classification model to obtain the type of crops corresponding to each well in the target area; Wherein, the irrigation feature analysis module is specifically used for: The agricultural water consumption corresponding to the agricultural electricity consumption data of each well is calculated by using the electricity-to-water conversion coefficient; According to the agricultural water consumption of each well and the planting area data of the users corresponding to each well, the time series data of the water consumption of each well is obtained; The time series data of water consumption of each well includes time series of multiple moments; the irrigation feature analysis module is specifically used for: For each pumped well, taking each moment as a reference, the time series data of the water consumption of the pumped well is intercepted forward by a signal segment window data of a preset width to obtain multi-dimensional signal vector data of the water consumption of the pumped well; and according to the multi-dimensional signal vector data of the water consumption of the pumped well, an input data matrix of the water consumption of the pumped well is obtained; The input data matrix of water consumption of each well is input into the pre-trained classification model to obtain the types of crops corresponding to each well in the target area; Wherein, the irrigation feature analysis module is specifically used for: Input the input data matrix of water consumption of each well into the pre-trained classification model to obtain the expected number of classifications; Introducing the Toeplitz matrix into the classification model, an inverse covariance matrix is ​​obtained; wherein the inverse covariance matrix is ​​used to describe the correlation of the time series data between the water consumption of each well; generating an objective function based on the inverse covariance matrix; Solving the objective function to obtain a clustering result; According to the clustering result, the types of crops corresponding to each well in the target area are obtained.

Images