Farmland carbon and nitrogen cycle adjusting system based on ecological model and implementation method thereof

Abstract

The invention discloses a farmland carbon and nitrogen cycle adjusting system based on an ecological model and an implementation method of the farmland carbon and nitrogen cycle adjusting system, and relates to the technical field of ecological adjustment. The system comprises a farmland information database, a visual circulating system and an adjusting system. The farmland information database comprises fertilization amount, climate conditions, soil characteristics and farmland management; the visual circulating system is used for simulating element migration among soil, plants and atmosphere, emission of greenhouse gases, organic matter analysis, CO2 generated by root respiration and dynamic behaviors of a soil carbon library and a nitrogen library; the system comprises a carbon nitrogen cycle adjustment analyzer and an early warning module. Soil data is collected in real time and imported into the mixed neural network model for carbon nitrogen cycle adjustment analysis, the optimal fertilizer usage amount and irrigation amount of the corresponding soil are calculated, the adjustment scheme is generated according to the analysis result, and the accuracy and efficiency of ecological system carbon nitrogen cycle monitoring are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of ecological regulation, and in particular relates to a farmland carbon-nitrogen cycle regulation system based on an ecological model and a method for realizing the same. Background Art

[0002] In recent years, the global carbon balance problem caused by global climate change has attracted more and more attention. Soil organic matter is an important carbon pool in the global carbon balance process. The soil carbon pool of agricultural ecosystems is the most active component of the global carbon pool. Under the influence of human cultivation, fertilization, irrigation and climate change, not only its quantity is changing rapidly, but also its status as a source and sink of global CO is constantly changing, thus affecting the regional and global environment, and also affecting the food production on which human beings depend for survival. Therefore, the study of carbon and nitrogen cycle issues in agricultural ecosystems has become a hot topic.

[0003] The dynamics of soil carbon and nitrogen cycles in agricultural ecosystems is a complex biogeochemical process that includes the production, decomposition, nitrification, denitrification and fermentation of organic matter. This process is controlled by many factors, such as soil temperature, humidity, redox potential, plant growth and soil erosion. These environmental conditions are further subject to ecological driving forces such as climate, soil quality, vegetation and human activities. Any change in the combination of these influencing factors will change the quality and quantity of soil carbon and nitrogen. In recent years, various ecosystem carbon and nitrogen cycle mechanism models have been published internationally, as many as a dozen. These models have played a unique role in further revealing the ecosystem carbon and nitrogen cycle mechanism, quantifying carbon and nitrogen cycle fluxes and greenhouse gas emissions, and are increasingly being used around the world. It has become the main means of studying global change and terrestrial ecosystems.

[0004] Since the carbon and nitrogen dynamics of farmland soil are closely related to soil productivity and changes in greenhouse gas concentrations in the atmosphere, the study of soil carbon and nitrogen cycles has always been a key area of ​​focus in the fields of soil science, agronomy, and ecology. Therefore, this application document provides a farmland carbon and nitrogen cycle regulation system based on an ecological model and its implementation method to improve the carbon and nitrogen cycle efficiency of the ecosystem. Summary of the invention

[0005] The purpose of the present invention is to provide a farmland carbon-nitrogen cycle regulation system based on an ecological model and its implementation method, which collects soil data in real time and imports it into a hybrid neural network model to perform carbon-nitrogen cycle regulation analysis, and generates a regulation plan based on the analysis results, thereby solving the problems of inaccurate carbon-nitrogen cycle monitoring of existing ecosystems and low carbon-nitrogen regulation efficiency.

[0006] To solve the above technical problems, the present invention is achieved through the following technical solutions:

[0007] The present invention is a farmland carbon and nitrogen cycle regulation system based on an ecological model, comprising a farmland information database, a visual circulation system and a regulation system;

[0008] The farmland information database includes fertilizer application amount, climate conditions, soil characteristics and farmland management;

[0009] The visualized cycle system is used to simulate the element migration between soil, plant and atmosphere, greenhouse gas emission, organic matter analysis and CO produced by root respiration. 2 and simulate the dynamic behavior of soil carbon and nitrogen pools;

[0010] The regulation system includes a carbon-nitrogen cycle regulation analyzer and an early warning module; the carbon-nitrogen regulation analyzer is used to identify and coordinate the carbon-nitrogen exchange of different soils, and by designing a complex network structure of water-fertilizer balance and carbon-nitrogen cycle, a hybrid neural network model of convolutional neural network CNN and recurrent neural network RNN ​​is used to perform carbon-nitrogen cycle regulation analysis; the early warning module is used to perform early warning processing on abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle in the soil, and generate an early warning plan for display in a visual circulation system.

[0011] As a preferred technical solution, the climate conditions are collected using meteorological sensors, including the daily maximum and minimum temperatures and daily precipitation; the soil properties are collected using soil sensors, including soil bulk density, soil texture and clay content, organic matter and pH; the farmland management includes land use type, crop yield, crop sowing and harvesting period, fertilizer usage, fertilization time, farming time and method, irrigation time and water volume.

[0012] As a preferred technical solution, the soil sensor is deployed at different depths of the soil according to the sampling method, and the collected data is transmitted to the visualization circulation system in real time; the visualization circulation system analyzes the collected soil bulk density, soil texture and clay content, organic matter and pH, and each visualization circulation system independently calculates the optimal fertilizer usage and irrigation amount of the corresponding soil, and calculates the optimal fertilizer usage V and irrigation amount T for each piece of soil according to the soil bulk density, soil texture and clay content, organic matter and pH; the specific calculation formula is as follows:

[0013]

[0014] In the formula, k 1 represents the soil irrigation regulation coefficient, k 2 represents the soil clay content adjustment coefficient, k 3 represents the soil organic matter adjustment coefficient, T est Indicates the set irrigation amount, H estrepresents the set soil bulk density, f(AQI) represents the organic matter and pH adjustment function; α represents the soil bulk density coefficient; T target represents the optimal irrigation amount; β represents the soil texture coefficient, Q heat Indicates soil bulk density.

[0015] As a preferred technical solution, the hybrid neural network model of the convolutional neural network CNN and the recurrent neural network RNN ​​includes:

[0016] Convolutional neural network (CNN): used to process the characteristics of each piece of soil, including soil bulk density, soil texture and clay content, organic matter and pH, and identify the characteristics and abnormal areas of each piece of soil;

[0017] Recurrent neural network (RNN) part: used to process the interaction between water-fertilizer balance and carbon-nitrogen cycle;

[0018] The real-time monitoring data of each piece of soil is used as the input of CNN to build a soil property recognition model, and the historical data and current status of soil properties are used as the input of RNN;

[0019] Output layer: The output layer is designed to output the amount of fertilizer used and irrigation for each piece of soil.

[0020] As a preferred technical solution, the specific calculation form of the hybrid neural network model is as follows:

[0021] For the CNN part, it is expressed as: C i =f CNN (X i ,θ cnn ), where X i is the spatial data input of the i-th layer, C i is the feature extracted by CNN, f CNN is the calculation function of CNN, θ cnn are the parameters of CNN;

[0022] For the RNN part, it is expressed as: R t =f RNN (C i,t ,θ rnn ), where C i,t is the CNN feature sequence from layer 1 to layer t, R t is the output of the RNN at time t, f RNN is the calculation function of RNN, θ rnn are the parameters of RNN.

[0023] As a preferred technical solution, when the carbon-nitrogen regulation analyzer determines that water-fertilizer balance and carbon-nitrogen cycle are abnormal, it extracts abnormal soil properties such as soil bulk density, soil texture and clay content, organic matter and pH information or manually inputs soil bulk density, soil texture and clay content, organic matter and pH information. At the same time, the database also stores user data, document data, relationship data, system data and temporary data generated during system operation.

[0024] The present invention is a method for implementing a farmland carbon and nitrogen cycle regulation system based on an ecological model, comprising the following steps:

[0025] Step S1: deploying soil sensors at different depths of the soil according to the sample method;

[0026] Step S2: The soil sensor collects soil bulk density, soil texture and clay content, organic matter and pH information for storage and visual display;

[0027] Step S3: importing the collected information into the trained hybrid neural network model;

[0028] Step S4: The hybrid neural network model outputs the optimal fertilizer usage and irrigation amount for each piece of soil;

[0029] Step S5: early warning of abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle in the soil;

[0030] Step S6: Match the abnormal situation with the solution library, generate corresponding solutions, and display them in the visual cycle system.

[0031] As a preferred technical solution, in step S3, the hybrid neural network model is composed of a convolutional neural network CNN and a recurrent neural network RNN. When it is judged that the soil has abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle, the water-fertilizer ratio and the carbon-nitrogen cycle ratio are obtained, and then the adjustment plan of water-fertilizer balance and carbon-nitrogen cycle is generated, and the information is stored in a database, and the visualization circulation system is started at the same time.

[0032] As a preferred technical solution, in step S5, an ecological process is selected and parameters are set in a typical ecological model within the visual circulation system, the model is calculated using data in the database, and the prediction results are intuitively displayed and analyzed to represent the optimal data of water-fertilizer balance and carbon-nitrogen cycle.

[0033] The present invention has the following beneficial effects:

[0034] The present invention collects soil data in real time and imports it into a hybrid neural network model to perform carbon-nitrogen cycle regulation analysis, calculates the optimal fertilizer usage and irrigation amount for the corresponding soil, generates a regulation plan based on the analysis results, and improves the accuracy and efficiency of ecosystem carbon-nitrogen cycle monitoring.

[0035] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0037] Figure 1 This is a schematic diagram of the structure of a farmland carbon and nitrogen cycle regulation system based on an ecological model of the present invention;

[0038] Figure 2 The present invention is a flow chart of the implementation method of a farmland carbon and nitrogen cycle regulation system based on an ecological model. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0041] In order to make the purpose, technical solution and advantages of this application clearer, the following Figure 1 The implementation methods of the present application are described in further detail.

[0042] Before introducing the embodiments of the present application, the agricultural ecosystem is first described.

[0043] In order to describe the complex biogeochemical process of organic carbon and nitrogen cycle in agricultural ecosystems, scientists have been committed to developing computer simulation models of carbon and nitrogen cycles in various ecosystems since the 1980s. Since organic matter (SOM) is the core of carbon and nitrogen dynamics, agricultural ecosystem models (or biogeochemical models) mainly refer to soil organic matter models. So far, more than a dozen such models have been published. The more mature biogeochemical models with the core of describing soil organic matter activities include CENTURY, DNDC, NCSO.IL, RothC, etc. These models have become the main means of studying global change and terrestrial ecosystems (GCTE). In order to better evaluate these models and promote the development of corresponding application research, such as evaluating the impact of land use, farmland management and climate change on soil carbon and nitrogen and greenhouse gas dynamics, the Global Soil Organic Matter Research Network (SOMNET) was established in 1995 with the support of the Terrestrial Origins in Global Environmental Research (TIGER) project under the Natural Environment Research Council (NERC) of the United Kingdom. It attracted nearly 30 of the best soil organic matter modelers and more than 70 long-term experimental data experts in the world to join, realizing the sharing of information and resources. SOMNET quickly became an important international scientific frontier and was accepted as a core project by the Global and Terrestrial Ecosystems Project (GCTE) of the International Geosphere-Biosphere Program (IGBP). The seven long-term experimental databases selected by it representing different land use types, climate zones and management measures were used to verify and compare the nine soil organic matter models published at that time. From the performance of the model simulation results, four models (RothC, NCSOIL, CENTURY, SOMM) can simulate various land use types, and the simulation results of six models (RothC, CENTURY, DAISY, CANDY, NCSOIL, DNDC) are significantly better than the other three, and the CENTURY and DNDC models are the most distinctive. The common characteristics of these models are basically to use meteorology, soil, land use and farmland management measures as driving conditions to mathematically simulate the processes of soil organic matter generation, decomposition and transformation. Finally, the model outputs soil organic carbon dynamics and the content of each component, crop yield, and greenhouse gas emissions such as CO, CH, N, NO and NO.

[0044] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figure 1-2 It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0045] Embodiment 1

[0046] See also Figure 1As shown, the present invention is a farmland carbon and nitrogen cycle regulation system based on an ecological model, including a farmland information database, a visualization cycle system and a regulation system;

[0047] The farmland information database includes fertilizer application rates, climate conditions, soil characteristics, and farmland management;

[0048] Visual cycle system is used to simulate the element transfer between soil, plant and atmosphere, greenhouse gas emission, organic matter analysis and CO produced by root respiration 2 and simulate the dynamic behavior of soil carbon and nitrogen pools;

[0049] The regulation system includes a carbon-nitrogen cycle regulation analyzer and an early warning module; the carbon-nitrogen regulation analyzer is used to identify and coordinate the carbon-nitrogen exchange of different soils. By designing a complex network structure of water-fertilizer balance and carbon-nitrogen cycle, a hybrid neural network model of convolutional neural network CNN and recurrent neural network RNN ​​is used to perform carbon-nitrogen cycle regulation analysis; the early warning module is used to warn of abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle in the soil, and generate early warning plans for display in the visual circulation system.

[0050] Climate conditions are collected using meteorological sensors, including daily maximum and minimum temperatures and daily precipitation; soil properties are collected using soil sensors, including soil bulk density, soil texture and clay content, organic matter and pH; farmland management includes land use type, crop yield, crop sowing and harvesting period, fertilizer usage, fertilization time, tillage time and method, irrigation time and water amount.

[0051] Soil sensors are deployed at different depths of the soil according to the sampling method, and the collected data are transmitted to the visual circulation system in real time; the visual circulation system analyzes the collected soil bulk density, soil texture and clay content, organic matter and pH, and each visual circulation system independently calculates the optimal fertilizer usage and irrigation amount for the corresponding soil, and calculates the optimal fertilizer usage V and irrigation amount T for each piece of soil according to the soil bulk density, soil texture and clay content, organic matter and pH; the specific calculation formula is as follows:

[0052]

[0053] In the formula, k 1 represents the soil irrigation regulation coefficient, k 2 represents the soil clay content adjustment coefficient, k 3 represents the soil organic matter adjustment coefficient, T est Indicates the set irrigation amount, H est represents the set soil bulk density, f(AQI) represents the organic matter and pH adjustment function; α represents the soil bulk density coefficient; T target represents the optimal irrigation amount; β represents the soil texture coefficient, Qheat Indicates soil bulk density.

[0054] The hybrid neural network model of convolutional neural network CNN and recurrent neural network RNN ​​includes:

[0055] Convolutional neural network (CNN): used to process the characteristics of each piece of soil, including soil bulk density, soil texture and clay content, organic matter and pH, and identify the characteristics and abnormal areas of each piece of soil;

[0056] Recurrent neural network (RNN) part: used to process the interaction between water-fertilizer balance and carbon-nitrogen cycle;

[0057] The real-time monitoring data of each piece of soil is used as the input of CNN to build a soil property recognition model, and the historical data and current status of soil properties are used as the input of RNN;

[0058] Output layer: The output layer is designed to output the amount of fertilizer used and irrigation for each piece of soil.

[0059] The specific calculation form of the hybrid neural network model is as follows:

[0060] For the CNN part, it is expressed as: C i =f CNN (X i ,θ cnn ), where X i is the spatial data input of the i-th layer, C i is the feature extracted by CNN, f CNN is the calculation function of CNN, θ cnn are the parameters of CNN;

[0061] For the RNN part, it is expressed as: R t =f RNN (C i,t ,θ rnn ), where C i,t is the CNN feature sequence from layer 1 to layer t, R t is the output of the RNN at time t, f RNN is the calculation function of RNN, θ rnn are the parameters of RNN.

[0062] When the carbon-nitrogen regulation analyzer determines that the water-fertilizer balance and carbon-nitrogen cycle are abnormal, the abnormal soil characteristics are extracted for soil bulk density, soil texture and clay content, organic matter and pH information, or the soil bulk density, soil texture and clay content, organic matter and pH information are manually input. At the same time, the database also saves user data, document data, relationship data, system data and temporary data generated during system operation.

[0063] Embodiment 2

[0064] See also Figure 2 As shown, the present invention is a method for implementing a farmland carbon and nitrogen cycle regulation system based on an ecological model, which can be used to implement the structural content of Example 1 of the present invention, including the following steps:

[0065] Step S1: deploying soil sensors at different depths of the soil according to the sample method;

[0066] Step S2: The soil sensor collects soil bulk density, soil texture and clay content, organic matter and pH information for storage and visual display;

[0067] Step S3: importing the collected information into the trained hybrid neural network model;

[0068] Step S4: The hybrid neural network model outputs the optimal fertilizer usage and irrigation amount for each piece of soil;

[0069] Step S5: early warning of abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle in the soil;

[0070] Step S6: Match the abnormal situation with the solution library, generate corresponding solutions, and display them in the visual cycle system.

[0071] In step S3, the hybrid neural network model is composed of a convolutional neural network (CNN) and a recurrent neural network (RNN). When it is determined that the soil has abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle, the water-fertilizer ratio and the carbon-nitrogen cycle ratio are obtained, and then the adjustment plan of water-fertilizer balance and carbon-nitrogen cycle is generated, and the information is stored in the database, and the visualization circulation system is started at the same time.

[0072] In step S5, an ecological process is selected and parameters are set in a typical ecological model in the visual circulation system, the model is operated using data in the database, and the prediction results are intuitively displayed and analyzed to represent the optimal data of water-fertilizer balance and carbon-nitrogen cycle.

[0073] It is worth noting that in the above system embodiment, the various units included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0074] In addition, those skilled in the art can understand that all or part of the steps in the above-mentioned embodiments can be completed by instructing related hardware through a program, and the corresponding program can be stored in a computer-readable storage medium.

[0075] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A farmland carbon and nitrogen cycle regulation system based on an ecological model, comprising a farmland information database, a visualization circulation system and a regulation system, characterized in that: The farmland information database includes fertilizer application amount, climate conditions, soil characteristics and farmland management; The visualization cycle system is used to simulate the element migration between soil, plant and atmosphere, the emission of greenhouse gases, the CO2 produced by organic matter analysis and root respiration, and the dynamic behavior of soil carbon and nitrogen pools; The regulation system includes a carbon-nitrogen cycle regulation analyzer and an early warning module; the carbon-nitrogen regulation analyzer is used to identify and coordinate the carbon-nitrogen exchange of different soils, and by designing a complex network structure of water-fertilizer balance and carbon-nitrogen cycle, a hybrid neural network model of convolutional neural network CNN and recurrent neural network RNN ​​is used to perform carbon-nitrogen cycle regulation analysis; the early warning module is used to perform early warning processing on abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle in the soil, and generate an early warning plan for display in a visual circulation system.

2. The farmland carbon and nitrogen cycle regulation system based on ecological model according to claim 1 is characterized in that: The climate conditions are collected using meteorological sensors, including the daily maximum and minimum temperatures and daily precipitation; the soil properties are collected using soil sensors, including soil bulk density, soil texture and clay content, organic matter and pH; the farmland management includes land use type, crop yield, crop sowing and harvesting period, fertilizer usage, fertilization time, tillage time and method, irrigation time and water volume.

3. The farmland carbon and nitrogen cycle regulation system based on ecological model according to claim 2 is characterized in that: The soil sensor is deployed at different depths of the soil according to the sample method, and the collected data is transmitted to the visualization circulation system in real time; the visualization circulation system analyzes the collected soil bulk density, soil texture and clay content, organic matter and pH, and each visualization circulation system independently calculates the optimal fertilizer usage and irrigation amount of the corresponding soil, and calculates the optimal fertilizer usage V and irrigation amount T for each piece of soil according to the soil bulk density, soil texture and clay content, organic matter and pH; the specific calculation formula is as follows: In the formula, k1 represents the soil irrigation adjustment coefficient, k2 represents the soil clay content adjustment coefficient, k3 represents the soil organic matter adjustment coefficient, T est Indicates the set irrigation amount, H est represents the set soil bulk density, f(AQI) represents the organic matter and pH adjustment function; α represents the soil bulk density coefficient; T target represents the optimal irrigation amount; β represents the soil texture coefficient, Q heat Indicates soil bulk density.

4. The farmland carbon and nitrogen cycle regulation system based on ecological model according to claim 1 is characterized in that: The hybrid neural network model of the convolutional neural network CNN and the recurrent neural network RNN ​​includes: Convolutional neural network (CNN): used to process the characteristics of each piece of soil, including soil bulk density, soil texture and clay content, organic matter and pH, and identify the characteristics and abnormal areas of each piece of soil; Recurrent neural network (RNN) part: used to process the interaction between water-fertilizer balance and carbon-nitrogen cycle; The real-time monitoring data of each piece of soil is used as the input of CNN to build a soil property recognition model, and the historical data and current status of soil properties are used as the input of RNN; Output layer: The output layer is designed to output the fertilizer usage and irrigation amount for each piece of soil.

5. The farmland carbon and nitrogen cycle regulation system based on ecological model according to claim 4 is characterized in that: The specific calculation form of the hybrid neural network model is as follows: For the CNN part, it is expressed as: C i =f CNN (X i ,θ cnn ), where X i is the spatial data input of the i-th layer, C i is the feature extracted by CNN, f CNN is the calculation function of CNN, θ cnn are the parameters of CNN; For the RNN part, it is expressed as: R t =f RNN (C i,t ,θ rnn ), where C i,t is the CNN feature sequence from layer 1 to layer t, R t is the output of the RNN at time t, f RNN is the calculation function of RNN, θ rnn are the parameters of RNN.

6. The farmland carbon and nitrogen cycle regulation system based on ecological model according to claim 1 is characterized in that: When the carbon-nitrogen regulation analyzer determines that water-fertilizer balance and carbon-nitrogen cycle are abnormal, it extracts abnormal soil properties such as soil bulk density, soil texture and clay content, organic matter and pH information or manually inputs soil bulk density, soil texture and clay content, organic matter and pH information. At the same time, the database also stores user data, document data, relationship data, system data and temporary data generated when the system is running.

7. A method for implementing a farmland carbon and nitrogen cycle regulation system based on an ecological model, characterized in that: The steps include: Step S1: deploying soil sensors at different depths of the soil according to the sample method; Step S2: The soil sensor collects soil bulk density, soil texture and clay content, organic matter and pH information for storage and visual display; Step S3: importing the collected information into the trained hybrid neural network model; Step S4: The hybrid neural network model outputs the optimal fertilizer usage and irrigation amount for each piece of soil; Step S5: early warning of abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle in the soil; Step S6: Match the abnormal situation with the solution library, generate corresponding solutions, and display them in the visual cycle system.

8. The method for implementing the farmland carbon and nitrogen cycle regulation system based on the ecological model according to claim 7 is characterized in that: In step S3, the hybrid neural network model is composed of a convolutional neural network CNN and a recurrent neural network RNN. When it is determined that the soil has abnormal conditions of water-fertilizer balance and carbon-nitrogen cycle, the water-fertilizer ratio and the carbon-nitrogen cycle ratio are obtained, and then the adjustment plan of water-fertilizer balance and carbon-nitrogen cycle is generated, and the information is stored in a database, and the visualization circulation system is started at the same time.

9. The method for implementing the farmland carbon and nitrogen cycle regulation system based on the ecological model according to claim 7 is characterized in that: In step S5, an ecological process is selected and parameters are set in a typical ecological model in the visual circulation system, the model is calculated using data in the database, and the prediction results are intuitively displayed and analyzed to represent the optimal data of water-fertilizer balance and carbon-nitrogen cycle.

Images