Regional agricultural planting structure planning system and method considering multiple indexes

Abstract

The invention discloses a regional agricultural planting structure planning system and method giving consideration to multiple indexes, and belongs to the field of agricultural planting systems and method.The regional agricultural planting structure planning technology of intelligent sorting logic is used for reference, and structured identity tags including geographical, soil and topographic features are established for land parcels; according to the method, a'demand label 'containing growth demand parameters is constructed for crops, and automatic and precise matching of the'demand label' and the growth demand parameters is realized by utilizing an intelligent algorithm, so that the problems of high subjectivity, low resource allocation efficiency and insufficient multi-target collaboration in a traditional planning mode are solved, and the planning target of'precise adaptation of plot conditions and crop demands' is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of agricultural planting systems and methods, more particularly, to a regional agricultural planting structure planning system and method considering multiple indexes. BACKGROUND

[0002] Agricultural planting structure planning is a core link to realize the optimal allocation of regional agricultural resources, improve production efficiency and ecological sustainability, and its rationality directly affects land utilization rate, crop yield and regional ecological balance. Traditional regional agricultural planting structure planning mostly relies on the experience of grass-roots agricultural technicians, and the basic situation of the land and the growth habit of crops are investigated manually to determine the planting allocation scheme. Although this mode has certain feasibility in small-scale planting scenarios, it gradually exposes significant limitations under the demand of large-scale and fine agriculture development:

[0003] For example, the matching process is subjective, and the matching of crops and land mostly depends on manual experience judgment, lacking unified quantitative standards and operation logic. For example, high-quality land may be allocated to crops with lower growth condition requirements due to experience bias, causing resource waste, while poor-quality land may be allocated to high-demand crops, resulting in low yield or even planting failure, forming the contradiction of "high-quality land waste and poor-quality land forced".

[0004] In recent years, with the development of smart agriculture technology, although some research attempts to introduce information means to assist planning, the existing technology has not formed a systematic "land-crop" precise matching mechanism: on the one hand, the land data collection has not realized the structured association and unique identification of geographic information, soil properties and topographic features, lacking a standardized information carrier like "express delivery slip"; on the other hand, the crop demand has not been converted into a calculable feature vector, and the matching process has not introduced an automatic comparison and optimization algorithm like "express sorting machine", still relying on manual experience.

[0005] Therefore, there is an urgent need for a regional agricultural planting structure planning technology that can draw on the "intelligent sorting" logic, by establishing a structured "identity tag" for land containing geographic, soil and topographic features, and constructing a "demand tag" for crops containing growth demand parameters, and using intelligent algorithms to realize the automatic and precise matching of the two, so as to solve the problems of strong subjectivity, inefficient resource allocation and insufficient multi-target coordination in the traditional planning mode, and achieve the planning goal of "precise adaptation of land conditions and crop demand", so we propose a regional agricultural planting structure planning system and method considering multiple indexes to solve the above problems. SUMMARY

[0006] 1. Technical problems to be solved

[0007] The "plot-crop" matching process is subjective, and the matching of crops and plots depends on artificial experience judgment, lacks unified quantitative standards and operation logic, and high-quality plots may be allocated to crops with lower requirements for growth conditions due to experience bias, resulting in resource waste, while poor plots may be allocated to high-demand crops, resulting in low yield or even planting failure, forming the contradiction of "high-quality land waste and poor land forced".

[0008] 2. Technical solution

[0009] To solve the above problems, the technical scheme adopted by the present application is as follows.

[0010] The regional agricultural planting structure planning system considering multiple indicators comprises:

[0011] The plot information acquisition module comprises a geographic information acquisition unit, a soil detection unit and a data integration unit, the geographic information acquisition unit acquires the longitude and latitude coordinates and terrain elevation data of the plot through a satellite positioning device, and sends the original data to the data integration unit, the soil detection unit is composed of a distributed soil sensor group, the sensor group comprises a soil humidity sensor, a pH sensor and a nutrient sensor, and the collected soil organic matter content, pH value, nitrogen, phosphorus and potassium concentration and field water holding capacity data are transmitted to the data integration unit through a wireless communication network, and the data integration unit is based on a microprocessor, associates the collected data to a plot unique code, generates structured plot identification data containing plot geographic location, soil properties and terrain characteristics, and stores them in a local storage array;

[0012] The crop information database pre-stores a crop characteristic data set, the crop characteristic data set contains crop variety code, soil adaptation parameter range, water critical value, photoperiod requirement, accumulated temperature threshold, unit area yield and ecological adaptability parameter, a data model is established through Python programming language, each parameter is associated to form a crop demand characteristic vector through JSON standardized data format, and is stored in a Redis in-memory database;

[0013] The intelligent matching unit comprises a parameter comparison submodule, a weight calculation submodule and a multi-objective optimization submodule, the parameter comparison submodule comprises a parallel operation program written in Java, and the parallel operation program has a preset parameter comparison logic, the weight calculation submodule has a fuzzy comprehensive evaluation algorithm program built-in, and the fuzzy comprehensive evaluation algorithm program comprises a three-level evaluation index system, calculates the scores of each index through a membership function, determines the index weight and generates a weight coefficient table through an analytic hierarchy process, and transmits the weight coefficient table to the multi-objective optimization submodule through an in-memory sharing mode, and the multi-objective optimization submodule comprises a genetic algorithm program written in MATLAB;

[0014] The planning output unit comprises a data visualization component, a scheme generation component and an output management component, the data visualization component comprises a layer rendering engine developed by ECharts, the engine reads the optimal matching combination data table through a RESTful API interface, associates the plot code to the electronic map vector layer, allocates the RGB color value according to the crop type, generates a spatial distribution heat map with a transparency gradient, the scheme generation component comprises a report generation program developed by C++, the program calls the plot area data table in the SQL Server database, calculates the planting area proportion of each crop in combination with the optimal matching combination, generates a structured Excel report template comprising the plot code, the crop variety, the planting area and the expected yield value, and the output management component comprises a USB interface controller and a network communication module, the controller supports the printer driver protocol, the network module is built-in with an SSL encryption protocol, and the corresponding data output operation can be performed according to the user input instruction.

[0015] Further, the preset parameter comparison logic in the parallel operation program is to read the soil property field and the terrain feature field in the plot identification data, and to perform numerical comparison with the corresponding parameter threshold range in the crop demand feature vector, when all field values are within the corresponding threshold range, a candidate matching marking program is triggered to generate a candidate matching pair data table comprising the plot code and the crop code, which is transmitted to the weight calculation submodule through a Socket communication protocol.

[0016] The three-level evaluation index system comprises a first-level index, a second-level index and a third-level index, the first-level index is land utilization rate, economic benefit index and ecological compatibility, the second-level index corresponds to plot soil fertility grade, irrigation guarantee coefficient and crop unit area yield value, and the third-level index is a specific detection parameter.

[0017] Further, the genetic algorithm program inputs the candidate matching pair data table and the weight coefficient table, and outputs the matching combination data table with the highest fitness through the following calculation process:

[0018] The comprehensive benefit value calculation formula is C = ω1 x L + ω2 x E + ω3 x S,

[0019] Wherein, C is the comprehensive benefit value, ω1, ω2 and ω3 are the weight coefficients of land utilization rate, economic benefit index and ecological compatibility respectively, and ω1 + ω2 + ω3 = 1, L is the land utilization coefficient, which is calculated according to the ratio of the actual plot area to the total area, E is the economic benefit index, and the calculation formula is Wherein, V is the crop unit area yield value, Y is the expected yield, R is the commodity rate, and C im S is the ecological adaptation coefficient, which is evaluated according to the adaptation degree of the crop to the local ecological environment and the influence degree on the ecological system;

[0020] By setting the population size, crossover probability, mutation probability and other parameters, iterative calculation is performed, and each iteration is evaluated and screened according to the comprehensive benefit value until the preset termination condition is met, and the optimal matching combination data table is output.

[0021] Further, the crop information database further comprises a planting mode association table, and the planting mode association table structure comprises a rotation period, an intercropping ratio and a crop combination code. The multi-objective optimization submodule of the intelligent matching unit calls the data of the planting mode association table through an SQL query statement, and generates a matching combination data table of a composite planting scheme through combination operation combined with a pandas library.

[0022] The regional agricultural planting structure planning method considering multiple indicators comprises the following steps:

[0023] S1, block identification construction: the satellite positioning device outputs latitude and longitude coordinates and terrain elevation data, which are transmitted to the data integration unit through the NMEA0183 protocol. The soil sensor group collects soil parameters, which are transmitted to the data integration unit after being gathered through the ZigBee network. The microprocessor runs the MySQL database program to associate the above data to the unique code of the land block, generate structured data containing 12 parameters and store them;

[0024] S2, crop demand modeling: read the crop characteristic data table through the Python script, define 8 critical parameters in JSON format, generate a crop demand feature vector and write it into the Redis database;

[0025] S3, intelligent matching operation:

[0026] S31, basic matching: the Java multithreaded program concurrently reads the land block data and crop demand vector, performs parameter comparison, and for each land block i and crop j, traverses the soil pH value, organic matter content and other parameters. When all parameters meet the threshold range of j, mark (i,j) as a candidate pair and write it into the candidate matching data table;

[0027] S32, weight determination: the fuzzy comprehensive evaluation program reads the associated data of the candidate matching pair, calculates the membership degree of each index, and calculates the weight coefficients ω1, ω2 and ω3 through the analytic hierarchy process, and ω1+ω2+ω3=1;

[0028] S33, multi-index optimization: the MATLAB genetic algorithm program initializes the population, and calculates the fitness value according to the comprehensive benefit value formula C=ω1×L+ω2×E+ω3×S, wherein After cross and mutation operations, iterative calculation is performed, and when the fitness value of continuous generations fluctuates less than a set threshold, the optimal matching combination is output;

[0029] S4, plan generation: ECharts engine parses the optimal matching data, maps the plot code to the electronic map coordinates, renders the heat map according to the crop type, the C++ report program calls the plot area data in SQLServer, calculates the planting area proportion of each crop, generates an Excel report, and outputs the management component according to the user's instructions, through the USB interface to drive the printer to output the report, or through SSL encryption and transmission to the remote terminal through TCP / IP protocol, the formula for calculating the planting area proportion of each crop is: proportion = Σ plot area / total area.

[0030] Further, the step S33 also includes the calculation of economic benefit index, through the formula Calculate the economic benefit index, where V reads the crop unit area value data from the Redis database, C im Read the plot improvement cost data from the plot improvement cost table, and the calculation process is performed by a MATLAB script.

[0031] Further, step S4 also includes scheme verification:

[0032] The Python program reads the historical yield data table, calculates the deviation rate of the predicted yield and the actual yield of the planning scheme, and the deviation rate calculation formula is Where P pre is the predicted yield, P act is the actual yield, when the deviation rate δ>15%, send adjustment instructions to the intelligent matching unit through Socket communication, and re-execute the matching operation of step S3.

[0033] 3. Advantage

[0034] Compared with the prior art, the advantages of the present application are:

[0035] (1), this scheme establishes a structured "identity tag" containing geographic coordinates, soil properties, and terrain characteristics for each plot through the plot information acquisition module, and converts crop requirements into a computable feature vector through the crop information database, solving the problem of "plot data fragmentation and crop requirement fuzzification" in traditional planning, and combining parallel operation of the intelligent matching unit and genetic algorithm, realizing automatic comparison and optimization of plot conditions and crop requirements, avoiding the "high-quality land waste, poor-quality land forced" phenomenon caused by manual experience, making land and soil resources accurately matched, and significantly improving resource utilization rate;

[0036] (2) The traditional planning relies on artificial experience judgment, and there are problems of subjective deviation and non-uniform standard. The scheme converts the matching process into quantitative weight calculation and multi-objective optimization through a three-level evaluation index system and a fuzzy comprehensive evaluation algorithm, and combines the iterative screening of the genetic algorithm, so that the planning result is free from the dependence on artificial experience. At the same time, the information of the land and the crops is stored and transmitted through a standardized format, forming a standardized process similar to "express face sheet + sorting machine", which improves the objectivity and replicability of the planning;

[0037] (3) The existing technology is difficult to balance the synergistic relationship of yield, benefit and ecology. The scheme integrates land utilization rate, economic benefit index and ecological compatibility into a unified optimization model through a comprehensive benefit value calculation formula. The economic benefit index is quantitatively calculated through crop output value, expected yield, improvement cost and other parameters, and the ecological compatibility is evaluated based on the crop ecological adaptability coefficient, realizing the multi-objective synergy of "high yield, high efficiency and ecological friendliness";

[0038] (4) The planning output unit generates a crop spatial distribution heat map through the ECharts engine, which intuitively displays the planting area and adaptability of different crops. At the same time, it provides detailed data of land coding, planting area and expected output value through a structured Excel report. In addition, the scheme verification mechanism dynamically adjusts the matching results through deviation rate calculation, further improving the reliability and landing of the planning scheme. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 is a schematic diagram of the system architecture of the present application;

[0040] Figure 2 is a schematic diagram of the function principle of the land information acquisition module of the present application;

[0041] Figure 3 is a schematic diagram of the function principle of the crop information database of the present application;

[0042] Figure 4 is a schematic diagram of the function principle of the intelligent matching unit of the present application;

[0043] Figure 5 is a schematic diagram of the function principle of the planning output unit of the present application;

[0044] Figure 6 is a schematic diagram of the method flow of the present application.

[0045] Explanation of figure numbers:

[0046] 1, land information acquisition module; 101, geographic information acquisition unit; 102, soil detection unit; 103, data integration unit;

[0047] 2, crop information database; 201, crop characteristic data set; 202, planting mode association table;

[0048] 3, intelligent matching unit; 301, parameter comparison sub-module; 302, weight calculation sub-module; 303, multi-objective optimization sub-module;

[0049] 4, planning output unit; 401, data visualization component; 402, scheme generation component; 403, output management component. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the present application specification; obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments, and all other embodiments obtained by those skilled in the art without creative labor based on the embodiments in the present application shall fall within the scope of protection of the present application.

[0051] Embodiment 1:

[0052] Please refer to Figures 1-6 , the regional agricultural planting structure planning system considering multiple indicators includes a land parcel information acquisition module 1, a crop information database 2, an intelligent matching unit 3, and a planning output unit 4, which cooperatively realize the precise matching of "land parcel-crop" and the generation of planning schemes, and the specific principles are as follows:

[0053] Land parcel information acquisition module 1: a unique and quantifiable "identity tag" is established for each land parcel, and the technical principle is based on the "multi-source data space-time association" technology, specifically:

[0054] Geographic information acquisition unit 101: differential GPS technology is adopted, satellite positioning equipment is used to receive Beidou-3 satellite signals in real time, combined with differential correction data of ground reference stations, the latitude and longitude coordinates and terrain elevation data of land parcel boundaries are calculated, the original data are packaged into standardized messages according to NMEA0183 protocol, and are wirelessly transmitted to the data integration unit;

[0055] Soil detection unit 102: a distributed sensor network architecture is adopted, and the sensor groups are arranged in a "checkerboard" layout, and the functions and principles of each sensor are as follows:

[0056] Soil moisture sensor: based on the frequency domain reflection method, the linear relationship between soil dielectric constant and water content is measured, and the humidity value of 0-100% is output;

[0057] pH sensor: based on the glass electrode method, the potential difference between the electrode and the soil solution is converted into the pH value of 0-14;

[0058] Nutrient sensor: using near-infrared spectroscopy technology, by analyzing the soil absorption characteristics of 700-2500nm band light, calculate the concentration of organic matter, nitrogen, phosphorus, potassium;

[0059] Data integration unit 103: based on the principle of edge computing, using ARM Cortex-A9 processor to realize the localization of data processing, the core logic is to establish data correlation index through the unique encoding of the plot (composed of "region ID + latitude and longitude hash value + timestamp", to ensure global uniqueness), integrate geographic information (latitude and longitude, elevation), soil properties (humidity, pH, nutrients) into structured data sets according to the preset data structure (such as the following table 1 form), and eliminate outliers through the verification algorithm, and finally store in the local SSD array.

[0060] Table 1: Plot structured data index rules

[0061] Data Type Field Name Data Format Association Key Geographical Information lon / lat Decimal Float Plot Code Topographical Feature elevation Integer (meters) Plot Code Soil Property moisture Percentage (%) Plot Code + Time Soil Property ph_value Float Plot Code + Time Soil Property nitrogen Integer (mg / kg) Plot Code + Time

[0062] Crop information database 2: convert crop growth requirements from qualitative description to quantifiable quantitative features, realize "digitalization of demand", specifically:

[0063] Crop feature data set 201: using "feature vector modeling method", the crop growth demand is decomposed into an 8-dimensional parameter vector, each dimension contains a threshold range or a specific value, the decomposition formula is: crop demand vector = [crop code, soil pH range, water critical value, nitrogen concentration range, phosphorus concentration range, potassium concentration range, accumulated temperature threshold, ecological coefficient], among them, the ecological adaptability coefficient is calculated by AHP, considering the crop's ability to prevent and control water and soil erosion (such as weight 0.3), the tolerance to soil salinization (such as weight 0.3) and the compatibility with the local ecosystem (such as weight 0.4), the value range is 0-1 (1 represents the best ecological adaptability);

[0064] Planting mode correlation table 202: based on the principle of "crop ecological niche complementation", for example, in the "corn-soybean" intercropping mode, corn (high stalk, high nitrogen demand) and soybean (short stalk, nitrogen fixation) form resource complementation, the correlation table records crop combination codes, rotation period (such as "wheat-corn" rotation period 1 year), intercropping row ratio (such as 6:4) and other parameters, providing data support for compound planting schemes;

[0065] Storage and indexing: using Redis in-memory database to realize efficient access, using "crop code + parameter type" as the key value to establish index, ensuring that the parameter query response time is less than 10ms.

[0066] Intelligent matching unit 3: core decision module, based on the fusion of "multi-objective optimization algorithm" and "fuzzy logic reasoning", specifically:

[0067] Parameter comparison sub-module 301: adopts a "parallel pipeline" operation architecture, realizes batch data processing based on Java multi-threading technology, and its comparison logic is "full field matching": for each land i (feature vector X i ) and crop j (demand vector Y j ), verify whether the parameters of X i fall within the threshold range of Y j one by one, and all satisfy the verification formula:

[0068] Candidate pair condition:

[0069] (k is the parameter dimension, n = 8; X ik is the kth parameter of land i; Y jk lover , Y jk hi9h is the upper and lower limit of the kth parameter of crop j);

[0070] Weight calculation sub-module 302: based on fuzzy comprehensive evaluation method, the core principle is to convert fuzzy evaluation index into quantitative weight value:

[0071] Index stratification: a three-level index system is constructed, where the third-level index is a directly measurable physical quantity, the second-level index is a comprehensive of the third-level index, and the first-level index is the final evaluation dimension;

[0072] Membership degree calculation: for each index x, calculate its membership degree to "optimal" through membership function μ(x), and the calculation formula of membership function μ(x) is:

[0073] μ(x) = 1, x ≥ high fertility threshold (x - medium threshold) / (high threshold - medium threshold), medium threshold < x < high threshold (formula 3) 0, x ≤ medium threshold;

[0074] Weight determination: adopt AHP method, invite at least 5 agricultural experts to compare the importance of first-level indexes two by two, construct judgment matrix and pass consistency test (CR < 0.1), and finally determine weight coefficients ω1, ω2, ω3 (satisfy ω1+ω2+ω3=1);

[0075] Multi-objective optimization sub-module 303: realizes global optimization based on genetic algorithm, and the core principle is to simulate the "natural selection" mechanism in biological evolution:

[0076] Encoding: adopt "real number encoding", each chromosome represents a planting scheme, and the gene value is the area proportion of land i planting crop j (such as 0.8 represents 80% of the area of land i planting crop j).

[0077] Fitness function: take the comprehensive benefit value C as the fitness, the higher the value, the better the scheme, and the formula is:

[0078] C = ω1L + ω2E + ω3S

[0079] Wherein:

[0080] L = Σ (planting area) / Σ (total area of plots);

[0081] E = (Σ (V x Y x R) - Σ (Cim)) / Σ (Cim) (E is an economic benefit index, V is unit output value, Y is yield, R is commodity rate, and Cim is improvement cost);

[0082] S = Σ (crop ecological coefficient x planting area proportion) (ecological compatibility);

[0083] Genetic manipulation:

[0084] Selection: tournament selection method (randomly select 5 individuals, and select the one with the highest fitness);

[0085] Crossover: arithmetic crossover (parent genes a, b, offspring genes = 0.7a + 0.3b);

[0086] Mutation: random disturbance to 1% of the genes (e.g. original value 0.8 is mutated to 0.75-0.85);

[0087] Termination condition: when the maximum fitness value of 20 consecutive generations fluctuates by <1%, output the optimal solution.

[0088] Planning output unit 4: based on the principles of "data visualization" and "secure communication", efficient delivery of planning results is achieved, specifically:

[0089] Data visualization component 401: using the layer superposition technology of ECharts, associate the plot vector data (.shp format) with the matching results, generate a heat map through RGB color mapping and transparency gradient, and intuitively display the spatial distribution of crops;

[0090] Scheme generation component 402: report engine developed based on C++, reads plot area data in SQLServer through ODBC interface, calculates planting area proportion by "crop type" grouping, and automatically fills in Excel template, the grouping calculation formula for planting area proportion is: planting area proportion of crop k = Σ (plot area of planting crop k) / total area of the region x 100%;

[0091] Output management component 403: uses a "dual-mode output" mechanism, local output drives the printer through USB HID protocol, and remote transmission realizes data security upload through SSL / TLS1.3 encryption and TCP / IP protocol.

[0092] Example 2:

[0093] In view of the above embodiment 1, for further description, please refer to Figures 1-6 , the regional agricultural planting structure planning method considering multiple indicators includes the following steps:

[0094] S1, plot identification construction

[0095] Data collection synchronization: geographic information and soil data are spatio-temporally associated through time stamping to ensure that the geographic and soil data of the same plot match in the time dimension;

[0096] Outlier processing: 3σ criterion (eliminate data exceeding ±3 times the standard deviation of the mean) and adjacent interpolation method (replace outliers with the mean of the surrounding three sensors);

[0097] Unique code generation: convert latitude, longitude, region ID, and collection time into a 32-bit unique code (e.g. A1B3C5...) through SHA-256 hashing algorithm, as the unique identifier for subsequent matching.

[0098] S2, crop demand modeling

[0099] Parameter extraction: determine the threshold range of each parameter (e.g. the water critical value for rice tillering period is 70%-80% of field water capacity) through literature research (e.g. "China's Major Crop Cultivation") and field trials (record crop growth status under different conditions for 3 consecutive years);

[0100] Vector normalization: use min-max normalization to convert parameters to standardized values between 0 and 1, facilitating subsequent comparison and calculation. The min-max normalization formula is: x' = (x-x_min) / (x_max-x_min);

[0101] S3, intelligent matching operation

[0102] S31, basic matching: based on the "full field threshold verification" principle, use Java's ExecutorService thread pool to concurrently process 1000+ plot data, and perform one-by-one comparison of 8 parameters for each plot-crop pair (e.g. whether soil pH is within the crop demand range), retaining only the candidate pairs that meet all the thresholds (screening pass rate about 30%-50%);

[0103] S32, weight determination: convert "multiple indicators" to "single weight" through fuzzy comprehensive evaluation to solve the "incomparability" between indicators, for example, a plot with high soil fertility but poor irrigation conditions can balance its comprehensive score through weight calculation;

[0104] S33, Multi-index optimization: Genetic algorithm jumps out of local optimal solution through "iterative evolution", for example, the initial scheme may give priority to economic benefits and ignore ecology, but after 50 generations of evolution, the algorithm will find a global optimal solution that takes all three into account.

[0105] S4: Plan generation and verification

[0106] Plan generation: Visualization and report generation is based on the principle of "data mapping", which converts abstract matching results into intuitive charts and structured tables.

[0107] Deviation rate calculation: Verify the accuracy of the prediction through historical data. When the deviation rate δ > 15% (indicating that there is a significant difference between the model assumption and the actual situation), trigger a re-match. The data verification formula is: δ = |P_pre-P_act| / P_act x 100%, (P_pre is the predicted yield, P_act is the average actual yield in the past 3 years);

[0108] Iterative optimization: When re-matching, increase search diversity by adjusting the crossover probability of genetic algorithm (such as from 0.7 to 0.8), or modify the weight coefficient (such as increasing the weight of the ecological index), until the deviation rate ≤ 15%.

[0109] Example 3:

[0110] In view of the above example 1 and example 2, in order to make the person skilled in the art more clearly understand the present application, the following will be described in detail in combination with a typical scene, to highlight the precision adaptation of the land condition and crop demand of the regional agricultural planting structure planning system and method considering multiple indexes in agricultural planting structure planning:

[0111] I. Implementation background

[0112] A typical plain agricultural county in China (area about 1200 square kilometers, including 5 towns, 236 administrative villages) as the implementation area, the main crops planted in this area are wheat, corn, soybean, cotton and other crops, but there has been a long-term problem of "high-quality land planting miscellaneous grains, low-yield land planting economic crops". In 2020, the local agricultural and rural bureau introduced the system described in the present application, and intelligently planned the spring planting structure in 2021. The specific implementation process is as follows.

[0113] II. System deployment and data collection

[0114] Land information collection

[0115] Geographic information collection: Through the Beidou satellite positioning equipment, the 2.8 million mu of cultivated land in the county is positioned block by block, and the latitude and longitude of each block is obtained (such as Block A: N36°25', E117°18'), terrain elevation (average elevation 45 meters, slope <3°), and low-lying waterlogging area (about 8%) is identified through terrain elevation data.

[0116] Soil detection: 3-5 distributed sensors (soil moisture, pH, nitrogen, phosphorus, and potassium sensors) are placed in each block to collect data as follows:

[0117] Soil pH: 4.8-8.2 (acidic blocks account for 12%, neutral blocks account for 65%);

[0118] Organic matter content: 1.2%-3.5% (high-fertility blocks account for 28%);

[0119] Field water holding capacity: 22%-30% (up to 35% or more in low-lying areas).

[0120] Data integration: The above 12 parameters (including latitude and longitude, elevation, 5 soil indicators, block area, etc.) are associated with a unique code (such as "DK-2024-00158") to generate a structured "block identity tag" and stored in a local database.

[0121] Crop information storage

[0122] For the four main crops in the local area, a demand feature vector (JSON format) is constructed using Python and stored in a Redis database:

[0123] Wheat: soil pH adaptation range 6.0-7.5, water critical value 25%, cumulative temperature threshold ≥2100°C, unit area yield 1800 yuan / acre, ecological adaptability coefficient 0.85;

[0124] Corn: soil pH adaptation range 5.5-8.0, nitrogen content requirement ≥120 mg / kg, cumulative temperature threshold ≥2400°C, unit area yield 2200 yuan / acre;

[0125] Soybean: soil pH adaptation range 6.5-7.0, phosphorus content requirement ≥20 mg / kg, poor waterlogging resistance (field water holding capacity ≤28%), unit area yield 2500 yuan / acre;

[0126] Cotton: soil pH adaptation range 7.0-8.5, strong drought resistance (water critical value 20%), unit area yield 3000 yuan / acre, ecological adaptability coefficient 0.7 (requires rotation to avoid continuous cropping obstacles).

[0127] III. Intelligent matching and optimization process

[0128] Basic matching

[0129] Java parallel operation program for 2.8 million mu plot and four crops parameter comparison, screening candidate matching pair:

[0130] Land A (pH = 6.8, organic matter 2.8%, nitrogen content 95 mg / kg) meets the soil requirements of wheat and soybean, marked as candidate pair (DK-2024-00158, wheat), (DK-2024-00158, soybean);

[0131] Land B (pH = 7.8, nitrogen content 130 mg / kg, field water holding capacity 22%) meets the requirements of corn and cotton, marked as candidate pair (DK-2024-00321, corn), (DK-2024-00321, cotton);

[0132] Low-lying land (field water holding capacity 35%) exceeds the waterlogging threshold of soybean, only corn (stronger waterlogging resistance) is marked as a candidate.

[0133] Weight calculation

[0134] Fuzzy comprehensive evaluation algorithm for index scoring of candidate pairs, and analytic hierarchy process to determine the weight:

[0135] The weight of the first index: land utilization rate (ω1 = 0.3), economic benefit (ω2 = 0.5), ecological compatibility (ω3 = 0.2);

[0136] The score of land A planting wheat in the third level index: soil fertility adaptation 90 points, irrigation guarantee 85 points, unit output value 1800 yuan, comprehensive score 88 points; The score of planting soybean is 82 points (because the unit output value of soybean is higher but the demand of local market is lower);

[0137] The maximum weight coefficient table shows that the matching weight of land A and wheat is higher than that of soybean.

[0138] Multi-objective optimization

[0139] Genetic algorithm takes "comprehensive benefit value C = 0.3L + 0.5E + 0.2S" as the objective function, and outputs the optimal scheme after 50 iterations:

[0140] High-fertility land (organic matter ≥3%) is preferentially allocated to cotton (highest output value) and soybean (good ecological adaptability);

[0141] Medium-fertility land is allocated to wheat and corn (to ensure food production capacity);

[0142] Low-lying land only grows corn (strong waterlogging resistance);

[0143] Introducing the rotation mode: cotton and soybean are planted alternately (using the nitrogen fixation characteristics of soybean to improve the soil), and the ecological compatibility S is improved to 0.85.

[0144] Four, planning output and verification

[0145] Result visualization

[0146] The data visualization component generates a space heat map: the red area is the cotton planting area, the green area is the wheat area, and the yellow area is the corn area, clearly showing the distribution of crops in each township;

[0147] The scheme generation component outputs an Excel report: the county's cotton planting area is 2800 mu (10%), wheat is 11200 mu (40%), corn is 8400 mu (30%), soybean is 5600 mu (20%), and the total output value is expected to be 4980 million yuan.

[0148] Scheme verification

[0149] Compared with the historical data of the past three years, the deviation rate of the predicted yield and the actual yield is 8.3% (<15%), and no adjustment is needed; if the deviation exceeds the threshold, the system will automatically trigger the intelligent matching unit to recalculate.

[0150] Five, implementation effect

[0151] Through the planning of the system, the resource utilization rate, unit area output value and ecological compatibility of the county have been improved, verifying the effectiveness of the invention in solving the problems of "subjective matching and resource mismatch" in large-scale agricultural scenarios.

[0152] The above is only the preferred specific embodiment of the present application; however, the protection scope of the present application is not limited thereto. Any skilled person in the art, according to the technical solution and improvement concept of the present application, can make equivalent replacement or change within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A regional agricultural planting structure planning system that takes into account multiple indicators, characterized by: include: The land parcel information acquisition module includes a geographic information acquisition unit, a soil detection unit, and a data integration unit. The geographic information acquisition unit sends the acquired raw data to the data integration unit, and the soil detection unit transmits the collected data to the data integration unit via a wireless communication network. The data integration unit, based on a microprocessor, associates the collected data with the land parcel's unique code, generates structured land parcel identification data containing the land parcel's geographic location, soil properties, and terrain characteristics, and stores the data in a local storage array. Crop information database: A crop feature dataset is pre-stored. The crop feature dataset contains crop variety parameters. A data model is first established using the Python programming language. Then, each parameter is associated using the JSON standardized data format to form a crop demand feature vector, which is stored in the Redis memory database. Intelligent matching unit: including a parameter comparison submodule, a weight calculation submodule and a multi-objective optimization submodule. The parameter comparison submodule includes a parallel operation program written in Java, and the parallel operation program has preset parameter comparison logic. The weight calculation submodule has a built-in fuzzy comprehensive evaluation algorithm program, and the fuzzy comprehensive evaluation algorithm program includes a three-level evaluation index system. The score of each index is calculated through the membership function, and the hierarchical analysis method is used to determine the index weight and generate a weight coefficient table, which is transmitted to the multi-objective optimization submodule through memory sharing. The multi-objective optimization submodule includes a genetic algorithm program written in MATLAB; Planning output unit: includes data visualization component, solution generation component and output management component. The data visualization component includes a layer rendering engine developed by ECharts, the solution generation component includes a report generation program developed by C++, and the output management component includes a USB interface controller and a network communication module.

2. The regional agricultural planting structure planning system taking into account multiple indicators according to claim 1 is characterized by: The geographic information acquisition unit obtains the original data of the longitude and latitude coordinates of the land parcel and the terrain elevation through a satellite positioning device; The soil detection unit is composed of a distributed soil sensor group, which includes a soil moisture sensor, a pH sensor, and a nutrient sensor, which are used to collect data on soil organic matter content, pH value, nitrogen, phosphorus, and potassium concentrations, and field water holding capacity; The crop characteristic dataset includes crop variety parameters such as code, soil adaptation parameter range, water critical value, photoperiod requirement, accumulated temperature threshold, unit area yield value and ecological adaptability parameter; The layer rendering engine reads the optimal matching combination data table through the RESTful API interface, associates the plot code with the electronic map vector layer, assigns RGB color values ​​according to crop type, and generates a spatial distribution heat map with transparency gradient; The report generation program calls the plot area data table in the SQL Server database, calculates the proportion of each crop planting area based on the optimal matching combination, and generates a structured Excel report template containing plot code, crop variety, planting area, and estimated output value.

3. The regional agricultural planting structure planning system taking into account multiple indicators according to claim 1 is characterized by: The preset parameter comparison logic in the parallel operation program is to read the soil attribute field and terrain feature field in the plot identification data, and perform numerical comparison with the corresponding parameter threshold range in the crop demand feature vector. When the values ​​of all fields are within the corresponding threshold range, the candidate matching marking program is triggered to generate a candidate matching pair data table containing plot codes and crop codes, and transmit it to the weight calculation submodule through the Socket communication protocol.

4. The regional agricultural planting structure planning system taking into account multiple indicators according to claim 1 is characterized by: The genetic algorithm program inputs the candidate matching pair data table and the weight coefficient table, and outputs the matching combination data table with the highest fitness through the following calculation process: The formula for calculating the comprehensive benefit value is: , in, is the comprehensive benefit value, 、 、 are the weight coefficients of land utilization rate, economic benefit index and ecological compatibility respectively, and , is the land use coefficient, which is calculated based on the ratio of the actual used area of ​​the plot to the total area. is the economic benefit index, and the calculation formula is ,in is the crop output per unit area, is the expected output, is the commodity rate, Cost of land improvements, The ecological adaptation coefficient is obtained by evaluating the degree of adaptation of crops to the local ecological environment and the degree of their impact on the ecosystem; By setting parameters such as population size, crossover probability, and mutation probability, iterative calculations are performed. Each iteration evaluates and screens the matching combinations based on the comprehensive benefit value until the preset termination conditions are met, and the optimal matching combination data table is output.

5. The regional agricultural planting structure planning system taking into account multiple indicators according to claim 1 is characterized by: The crop information database also includes a planting pattern association table, and the planting pattern association table structure includes crop rotation cycle, intercropping ratio, and crop combination code. The multi-objective optimization submodule of the intelligent matching unit calls the planting pattern association table data through an SQL query statement, combines with the pandas library to perform combination operations, and generates a matching combination data table of the composite planting plan.

6. A method for planning a regional agricultural planting structure taking into account multiple indicators, comprising the method for planning a regional agricultural planting structure taking into account multiple indicators as claimed in any one of claims 1 to 5, characterized in that the method comprises the following steps: S1. Block identification construction: The satellite positioning device outputs latitude and longitude coordinates and terrain elevation data, which are transmitted to the data integration unit via the NMEA0183 protocol. The soil sensor group collects soil parameters, which are aggregated via the ZigBee network and transmitted to the data integration unit. The microprocessor runs the MySQL database program, associates the above data with the unique code of the plot, generates structured data containing 12 parameters, and stores them; S2. Crop demand modeling: Use Python scripts to read the crop characteristic data table, define eight critical parameters in JSON format, generate crop demand feature vectors, and write them into the Redis database; S3, intelligent matching operation: S31. Basic matching: Java multi-threaded program concurrently reads plot data and crop demand vectors, performs parameter comparison, and With crops , traverse the soil pH value, organic matter content and other parameters, when all parameters meet When the threshold range is For candidate pairs, write the candidate matching data table; S32. Weight determination: The fuzzy comprehensive evaluation program reads the associated data of the candidate matching pairs, calculates the membership of each indicator, and calculates the weight coefficient through the hierarchical analysis method. 、 、 ,and ; S33, multi-index optimization: MATLAB genetic algorithm program initializes the population, and calculates the comprehensive benefit value according to the formula Calculate the fitness value, where ,After iterative calculation of crossover and mutation operations, when the fitness value fluctuation for multiple consecutive generations is less than the set threshold, the optimal matching combination is output; S4. Planning scheme generation: The ECharts engine parses the optimal matching data, maps the plot code to the electronic map coordinates, and renders the heat map according to the crop type. The C++ report program calls the plot area data in SQL Server, calculates the proportion of each crop planting area, and generates an Excel report. The output management component drives the printer to output the report through the USB interface according to user instructions, or transmits it to the remote terminal through the TCP / IP protocol after SSL encryption. The formula for calculating the proportion of the crop planting area is: proportion = Σplot area / total area of ​​the region.

7. The method for regional agricultural planting structure planning taking into account multiple indicators according to claim 6, characterized in that: The step S33 also includes the calculation of the economic benefit index, which is calculated by the formula Calculate the economic benefit index, where Read the crop yield per unit area data from the Redis database, The land improvement cost data is read from the land improvement cost table, and the calculation process is performed by a MATLAB script.

8. The method for regional agricultural planting structure planning taking into account multiple indicators according to claim 6, characterized in that: The step S4 also includes scheme verification: The Python program reads the historical production data table and calculates the deviation rate between the predicted production of the planning scheme and the actual production. The deviation rate calculation formula is: ,in To predict production, is the actual output, when the deviation rate When the adjustment instruction is sent to the intelligent matching unit via Socket communication, the matching operation of step S3 is re-executed.

Images