Agricultural management measure scheme simulation method, system and equipment based on improved agent model

Abstract

The invention provides an agricultural management measure scheme simulation method, system and equipment based on an improved agent model, and belongs to the technical field of agricultural non-point source pollution treatment. The method comprises the following steps: acquiring various data of a research basin, constructing an SWAT model of the research basin, and calculating monthly-scale landscape yields of hydrological variables of sub-basins under different agricultural management measures; constructing a corresponding improved response matrix according to the monthly-scale landscape yield of the hydrological variable of each sub-basin under different agricultural management measures, and performing proxy calculation on the outlet load of each sub-basin by using an improved response matrix method; adopting an improved response matrix method as an agent model; and performing agricultural management measure scheme optimization by using the agent model, and determining an optimal agricultural management measure scheme. According to the method, agricultural management measures can be simulated and optimized more quickly and finely, and reliable guarantee is provided for regulation and control of watershed non-point source pollution.

Description

Technical Field

[0001] The invention belongs to the technical field of agricultural non-point source pollution control, and relates to an agricultural management measure simulation method, system and equipment based on an improved agent model. Background Art

[0002] At present, the SWAT (Soil and Water Assessment Tool) model is a representative of distributed watershed hydrological models. The model divides the watershed into several subbasins according to the watershed elevation. Each subbasin is further divided into different hydrological response units (HRUs) according to land use, soil and slope. The SWAT model can be used to simulate water quality and quantity, and predict the impact of different agricultural management measures on the hydrology, sediment and nitrogen and phosphorus nutrients in the watershed. It is one of the more commonly used simulation models for watershed non-point source pollution prevention and control measures.

[0003] However, the simulation of measures based on physical process models is computationally complex and time-consuming, and the required computational cost is high. For example, it may take several minutes to hours to simulate hydrology and nutrient loads using models such as SWAT and MODFLOW. Especially when physical process models are coupled with various optimization models, the models need to be run thousands of times (for example, when searching for the best management decision), and the calculations are very complex and time-consuming, resulting in slow screening of agricultural management measures and difficult optimization processes. Therefore, it is necessary to use proxy models to replace traditional basin hydrological models.

[0004] Proxy modeling can reduce the computational complexity of the model and alleviate the problem of heavy load by directly capturing the relationship between the input and output of the physical mechanism model without considering the complex process mechanism within the watershed system. However, existing proxy modeling techniques, such as the response matrix method RM (Receptance Matrix) and machine learning methods, essentially replace the original physical mechanism model with a simpler response relationship, which is a high-efficiency calculation at the expense of the model calculation fidelity. In some sub-basins with strong interference from human activities such as reservoir interception, the response relationship is different from other natural basins, and the existing proxy modeling methods are no longer effective for simulation. This has restricted the application of proxy modeling in watershed pollution simulation research to a certain extent.

[0005] Therefore, there is an urgent need to improve the process and structure of the proxy model to take into account both the high fidelity and computational efficiency of the simulation model, so as to ensure accurate and rapid simulation and evaluation of watershed non-point source pollution prevention and control measures. Summary of the invention

[0006] The purpose of the present invention is to provide a response matrix proxy modeling method that takes into account improvement strategies, which can be used to evaluate and decide on agricultural best management practices (Best Management Practices, hereinafter referred to as BMPs) to serve non-point source pollution control. The agricultural management measures simulation method, system and equipment based on the improved response matrix proxy model can simulate and optimize agricultural management measures more quickly and accurately, providing reliable guarantees for achieving watershed non-point source pollution control.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A method for simulating agricultural management measures based on an improved response matrix proxy model. The method first obtains a variety of data from the study basin, constructs a SWAT model for the study basin, and calculates the monthly landscape yield of hydrological variables in each sub-basin under different agricultural management measures; secondly, a corresponding improved response matrix is ​​constructed based on the monthly landscape yield of hydrological variables in each sub-basin under different agricultural management measures, and the improved response matrix is ​​used to preliminarily calculate the export load of each sub-basin; thirdly, based on the preliminary load calculation, for sub-basins with strong human activity interference such as reservoir interception, the in-flow reaction formula and the reservoir interception formula are used to obtain the updated load calculation results; finally, the agricultural management measures are selected based on the updated load results to determine the optimal agricultural management measures. The following steps are included:

[0009] Obtain various data of the study basin and build the spatial database and attribute database required for the operation of the SWAT model; the various data include DEM elevation, land use, soil type, meteorological and hydrological data of the study basin;

[0010] The first step is to build a SWAT model of the study basin based on the spatial database and attribute database; the study basin includes multiple sub-basins and hydrological response units; specifically:

[0011] 1.1) Based on the spatial database and attribute database of the basin, the study basin is divided into several sub-basins and hydrological response units, and a preliminary basin SWAT model is established;

[0012] 1.2) Use the global sensitivity analysis method (the SUFI-2 algorithm built into the model) to perform parameter sensitivity analysis and parameter calibration on the preliminary SWAT model of the basin, and update the optimal parameter values ​​of the SWAT model of the basin;

[0013] 1.3) The simulation results of the basin SWAT model with the optimal parameter values ​​are evaluated by the Nash efficiency coefficient and the determinism coefficient evaluation index. If the Nash efficiency coefficient and the determinism coefficient are both greater than 0.8, the SWAT model with updated parameters is considered to be completed.

[0014] In the second step, the watershed SWAT model constructed in the first step was used to simulate the monthly landscape yield of hydrological variables in each sub-watershed under different agricultural management measures (BMPs), and the response matrix corresponding to each agricultural management measure was calculated; specifically:

[0015] 2.1) Using the SWAT model constructed in the study basin, different agricultural management measures under consideration were set up for simulation. During the simulation, each agricultural management measure was applied to all agricultural type hydrological response units in all sub-basins; the agricultural management measures included non-engineering measures and engineering measures; the non-engineering measures included different fertilization systems, irrigation strategies and tillage methods; the engineering measures included vegetation filter strips and grassed waterways;

[0016] 2.2) The monthly landscape yield of hydrological variables in each sub-basin obtained by different agricultural management measures, which will be recorded as y m,t , represents the landscape yield in the tth month when the mth agricultural management measure BMP is applied to 100% of the land area in the whole basin; the hydrological variables include: flow, total phosphorus, total nitrogen and nitrate.

[0017] 2.3) Monthly landscape yield y of hydrological variables in each sub-basin under different agricultural management measures m,t is the response matrix Y m,t The constituent elements of Y m,t It is a two-dimensional matrix with the month as the action and the type of agricultural management measures as the column, as shown in the attached manual. Figure 2 (b) as shown.

[0018] The third step is to calculate the response matrix Y of different agricultural management measures (BMPs) obtained in the second step. m,t , preliminarily calculate the hydrological landscape yield load value Q at the outlet of each sub-basin after hydrological confluence t , specifically:

[0019] 3.1) Constructing the decision vector F representing the weights of agricultural management measures m , represents the application area ratio of the mth agricultural management measure BMP in a sub-basin, ranging from 0-100%, 100% means that the entire sub-basin is fully used;

[0020] 3.2) Calculate the total area A of each sub-basin based on the basin spatial database, which is a vector of all sub-basin areas;

[0021] 3.3) Construct a connection matrix W that describes the upstream and downstream relationships of each sub-basin to approximate the process of the SWAT model hydrological basin. i,j , if sub-basin j is located upstream of sub-basin i, the off-diagonal element wi,j|i≠j =1, otherwise w i,j|i≠j = 0; when i = j, it represents the diagonal element w of each sub-basin i,j = 1. As shown in the instruction manual Figure 2 (d) as shown.

[0022] 3.4) Based on the response matrix Y m,t , decision vector F m , total area A and connection matrix W, and use the following formula to calculate the outlet load of each sub-basin:

[0023]

[0024] Among them, q t ∈R N is the hydrological landscape yield load vector generated in N sub-basins in the tth month; (Y m,t ) Q ∈R N is the response matrix in month t when the mth BMPs is applied to 100% of the land area in N sub-basins; A∈R N is the vector of N sub-basin areas; F m ∈R N is the vector of the area ratio of the mth BMP in N sub-basins; diag(·) is the function of diagonalizing the vector; Q t ∈R N is the vector of the hydrological landscape yield load at the outlet of N sub-basins in month t. The hydrological landscape yield load includes flow, total nitrogen, total phosphorus and nitrate nitrogen.

[0025] In the fourth step, the load calculation results of each sub-basin outlet obtained in the third step are updated using the in-stream reaction formula. In addition, for sub-basins with strong human activity interference such as reservoir interception, the load calculation results are updated using the reservoir interception formula. Specifically;

[0026] 4.1) Extract the pollutant-related items (including total nitrogen, total phosphorus and nitrate nitrogen values) and sub-basin flow values ​​at each sub-basin outlet (except for sub-basins with reservoirs, which are calculated independently according to Section 4.2) in the basin SWAT model, establish the regression relationship between the pollutant-related items and the flow, and obtain the regression coefficient. Then, based on the in-flow reaction formula (3), calculate the updated value of the pollution load, and then add the updated value to the preliminary load calculation results of each sub-basin outlet obtained in the third step to obtain the updated load result. Taking total phosphorus TP as an example, the in-flow reaction formula is as follows:

[0027]

[0028] Among them, ΔTP t ∈RN TP_sum is the vector of the total phosphorus loss in the river at the outlet of N sub-basins, which is obtained by adding a total phosphorus related term and a flow related term; t ∈R N is the total phosphorus load generated for each sub-basin in month t obtained by preliminary calculation in step 3; is the vector of regression coefficients of total phosphorus lost due to sedimentation in N sub-basins and total phosphorus production in the inflowing river; is a vector of the inverse regression coefficients of total phosphorus and water yield in N sub-basins lost due to algae absorption, decomposition and benthic absorption. The calculation process of the above total phosphorus index can be replaced by pollutant indicators such as total nitrogen and nitrate nitrogen, so as to realize the calculation process of the reaction of different pollutants in the flow.

[0029] 4.2) In particular, for sub-basins with strong human activity interference such as reservoir interception, the monthly inflow and outflow hydrological landscape production loads of the sub-basin where the reservoir is located in the SWAT model are extracted, and a linear regression is established, and then the regression coefficient is obtained to update the preliminary load calculation results. Taking total phosphorus TP as an example, the calculation formula is as follows:

[0030]

[0031] in, and are the inflow and outflow loads of the sub-basin where the reservoir is located in the tth month obtained by SWAT model simulation of the baseline scenario; and are the linear regression coefficients of the incoming and outgoing loads under the baseline scenario; (w' i,j ) TP To consider the total phosphorus interception effect of the reservoir, the variables of each element in the improved connection matrix are: D is the set of all sub-basins downstream of the reservoir, and U is the set of all sub-basins upstream of the reservoir; TP min ∈R N is the vector of the minimum intercepted total phosphorus load of each sub-basin after being intercepted by the reservoir. The value of the sub-basin immediately downstream of the reservoir is In other sub-basins, the values ​​are all 0; TP t ∈R N is the total phosphorus load generated by each sub-basin in month t obtained by preliminary calculation in the third step; TP t,end ∈R N It is the updated total phosphorus load of each sub-basin in the tth month after the reservoir interception.

[0032] The fifth step is to regard the updated hydrological landscape yield load results (including total nitrogen, total phosphorus, nitrate nitrogen, etc.) obtained in the fourth step as the ecological scores of different agricultural management measures (BMPs) at the outlet of the basin, and calculate the total cost of different agricultural management measures (BMPs) as the evaluation criteria. The optimal spatial combination of basin agricultural management measures (BMPs) is selected based on the comprehensive ecological score and the evaluation criteria, providing a scientific basis for the management decision of basin non-point source pollution. Specifically:

[0033] 5.1) The calculation formula for the ecological scores and evaluation criteria of different agricultural management measures (BMPs) is as follows: 1-3 is the ecological scoring formula, and f4 is the judging standard formula:

[0034]

[0035] Where t is the calculation step, month; T is the total number of simulation periods, month; TN outlet_avg TP outlet_avg and NO3-N outlet_avg TN are the multi-year average total nitrogen, total phosphorus and nitrate nitrogen loads at the basin outlet; mod_outlet,t TP mod_outlet,t and NO3-N mod_outlet,t are the updated total nitrogen, total phosphorus and nitrate nitrogen load values ​​at the watershed outlet calculated in the fourth step, kg; G is the cost under different BMPs spatial allocation combinations, 10,000 yuan; for the study of M BMPs scenarios in N sub-basins, A∈R N is the vector of agricultural land area in each sub-basin, hm 2 ; F∈R N×M is the spatial allocation matrix, which represents the proportion of the application area of ​​each sub-basin allocated to each BMPs scenario; C∈R M is the vector of unit area costs corresponding to the scenario of applying M BMPs to 100% of the land area in the whole basin, RMB / hm 2 , x is the application area ratio of different BMPs measures in each sub-basin, that is, F in formula (1) m .

[0036] 5.2) The weights of the three ecological scores in the optimization target are set to be the same as those of the evaluation criteria. The combination of agricultural management measures with the highest final score is the optimal governance decision for non-point source pollution in the watershed. The optimization target formula is as follows:

[0037] F(x)=0.25f1(x)+0.25f2(x)+0.25f3(x)+0.25f4(x) (8)

[0038] In the formula, the letters represent the same as formula (7).

[0039] A simulation system for agricultural management measures based on an improved agent model, comprising:

[0040] The data acquisition module is used to acquire various data of the study basin and build the spatial database and attribute database required for the operation of the SWAT model; the various data include DEM elevation, land use, soil type, meteorological and hydrological data of the study basin;

[0041] A SWAT model building module is used to build a SWAT model of a study basin based on a spatial database and an attribute database; the study basin includes multiple sub-basins and hydrological response units;

[0042] The monthly scale landscape yield calculation module is used to simulate different agricultural management measures using the SWAT model and calculate the monthly scale landscape yield of hydrological variables in each sub-basin under different agricultural management measures;

[0043] The module for calculating the export load of each sub-basin is used to construct the corresponding response matrix according to the monthly landscape yield of the hydrological variables of each sub-basin under different agricultural management measures, and use the improved response matrix method to calculate the export load of each sub-basin;

[0044] An improved proxy model determination module is used to adopt an improved response matrix method as an improved proxy model according to proxy calculation results of the outlet load of each sub-basin;

[0045] The optimal agricultural management measure program determination module is used to utilize the improved agent model to optimize the agricultural management measure program and determine the optimal agricultural management measure program.

[0046] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for simulating agricultural management measures based on an agent model is implemented.

[0047] Optionally, the memory is a non-transitory computer-readable storage medium.

[0048] The beneficial effects of the present invention are:

[0049] The present invention provides a method, system and device for simulating agricultural management measures based on an improved proxy model. The method includes: constructing a SWAT model of a research basin based on a variety of data in an acquired spatial database and an attribute database, using a proxy model based on an improved response matrix method to calculate the monthly scale landscape yield of hydrological variables in each sub-basin under different agricultural management measures, further considering the in-flow reaction and the impact of reservoir interception, and finally determining the optimal agricultural management measures. The present invention achieves high-fidelity and rapid simulation and optimization of agricultural management measures, overcomes the problems of high complexity, long time consumption and difficulty in nested optimization of the SWAT model in simulating agricultural management measures for non-point source pollution in the basin, and also overcomes the problem that the previous proxy modeling method cannot be effectively simulated in sub-basins with strong interference from human activities such as reservoir interception by improving the corresponding matrix. Thereby taking into account the high fidelity and computational efficiency of the simulation model, providing a reliable guarantee for the accurate and rapid evaluation of the optimal non-point source pollution prevention and control measures in the basin. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flow chart of a method for simulating agricultural management measures based on an improved agent model provided by the present invention;

[0051] Figure 2 A schematic diagram of the process of determining the proxy model provided by the present invention;

[0052] Figure 3 Schematic diagram of the principle of considering in-flow reactions in the improved response matrix method provided by the present invention. DETAILED DESCRIPTION

[0053] 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.

[0054] The purpose of the present invention is to provide a method, system and equipment for simulating agricultural management measures based on an improved agent model, which can simulate and optimize agricultural management measures more quickly and accurately, and provide reliable guarantee for achieving non-point source pollution control in watersheds.

[0055] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] See also Figure 1 The present invention discloses a method for simulating agricultural management measures based on an improved agent model, comprising:

[0057] Step 1: Obtain various data of the study basin and build the spatial database and attribute database required for the operation of the SWAT model.

[0058] The various data include data such as DEM elevation, land use, soil type, meteorological and hydrological data of the study basin. Among them, the spatial database of the study basin is usually determined according to the natural geographical characteristics of the study basin. The spatial database mainly includes data such as digital elevation map (DEM), land use distribution map, soil type distribution map and meteorological station distribution map of the study basin. The attribute database mainly includes data such as land use type, soil properties, various meteorological element values, hydrological data and water quality data of the study basin. Among them, the meteorological element values ​​include: rainfall, temperature, solar radiation, wind speed and relative humidity.

[0059] In a specific embodiment, the SWAT model needs to input a lot of data during simulation calculation, which is mainly divided into two categories: spatial data and attribute data. The spatial data includes DEM, land use, soil type, slope, slope aspect, etc., and the attribute data includes meteorological data such as daily relative humidity and daily evaporation, soil physical and chemical properties, and hydrological data such as rainfall, sediment, and runoff at the hydrological station, and water quality data such as total nitrogen, total phosphorus, and nitrate during the actual monitoring process, providing a data basis for building a distributed SWAT model (abbreviated as SWAT model).

[0060] Step 2: Construct a SWAT model of the study basin based on the spatial database and the attribute database; the study basin includes multiple sub-basins and hydrological response units.

[0061] Specifically, the SWAT model was combined with ArcGIS software to extract the water system of the study basin based on DEM. The study basin was divided into multiple sub-basins by setting the sub-basin division threshold and the sub-basin exit point. The plots with the same land use type, soil type and slope value in each sub-basin were further divided into a hydrological response unit. Then, the parameter sensitivity analysis and parameter calibration of the SWAT model were performed based on the global sensitivity analysis method of the SUFI-2 algorithm built into SWAT-CUP to determine the optimal parameter values ​​of the SWAT model. The Nash coefficient (NSE) and the coefficient of determination (R 2 )Evaluation index is used to evaluate the simulation result accuracy of the SWAT model with the optimal parameter value, and the SWAT model with higher evaluation result is taken as the constructed SWAT model.

[0062] Step 3: Use the SWAT model to simulate different agricultural management measures, preliminarily calculate the monthly landscape yield of hydrological variables in each sub-basin under different agricultural management measures, and calculate the response matrix corresponding to each agricultural management measure.

[0063] Step 3.1: Using the SWAT model constructed in the study basin, different agricultural management measures under consideration are set up for simulation. The management measures include non-engineering measures of source control such as different fertilization systems, irrigation strategies, and tillage methods, as well as engineering measures such as vegetation filter strips and grassed waterways. The simulation of the setting of different agricultural management measures is to apply each agricultural management measure to all agricultural types of HRUs in all sub-basins. The preliminary monthly scale landscape yield of hydrological variables in each sub-basin obtained by the output of different agricultural management measures, where the hydrological variables include: flow, total phosphorus, total nitrogen and nitrate.

[0064] Step 3.2: Use the monthly landscape yield composition response matrix Y of each sub-basin hydrological variable under different agricultural management measures m,t , so that Y m,t A two-dimensional matrix with months as the rows and types of agricultural management measures as the columns, as shown in the attached manual. Figure 2 (b) as shown.

[0065] Step 4: Response matrix Y for different agricultural management measures m,t , preliminarily calculate the hydrological landscape yield load value at the outlet of each sub-basin after hydrological confluence.

[0066] Step 4.1: Construct a decision vector F representing the weights assigned to agricultural management measures m , represents the proportion of the application area of ​​the mth agricultural management measure BMP in a sub-basin;

[0067] Step 4.2: Calculate the total area A of each sub-basin based on the basin spatial database;

[0068] Step 4.3: Construct a connection matrix W that describes the upstream and downstream relationships of each sub-basin to approximate the process of the SWAT model hydrological basin. i,j , if sub-basin j is located upstream of sub-basin i, the off-diagonal element w i,j|i≠j =1, otherwise w i,j|i≠j = 0; when i = j, it represents the diagonal element w of each sub-basin i,j = 1. As shown in the instruction manual Figure 2 (d) as shown.

[0069] Step 4.4: Based on the response matrix Y m,t , decision vector F m , total area A and connection matrix W, and use formulas (1) and (2) to calculate the outlet load of each sub-basin:

[0070]

[0071] Among them, q t ∈R Nis the hydrological landscape yield load vector generated in N sub-basins in the tth month; (Y m,t ) Q ∈R N is the response matrix in month t when the mth BMPs is applied to 100% of the land area in N sub-basins; A∈R N is the vector of N sub-basin areas; F m ∈R N is the vector of the area ratio of the mth BMP in N sub-basins; diag(·) is the function of diagonalizing the vector; Q t ∈R N is the vector of the hydrological landscape yield load at the outlet of N sub-basins in month t. The hydrological landscape yield load includes flow, total nitrogen, total phosphorus and nitrate nitrogen.

[0072] See the above steps for Figure 2 Based on the improved response matrix method, the hydrological landscape yield load value at the outlet of each sub-basin after hydrological confluence is preliminarily calculated:

[0073] (a) Using a well-constructed and calibrated watershed SWAT model, the agricultural management measures taken into consideration are set on the HRUs of all agricultural types in all sub-watersheds. The output load results can be regarded as the load values ​​when each measure is applied to 100% of the area of ​​each sub-watershed.

[0074] (b) The response matrix Y is obtained by arranging the monthly landscape yields of each sub-basin obtained by simulating various agricultural management measures. m,t , where the indicators are flow, total nitrogen, total phosphorus and nitrate.

[0075] (c) Construct a decision vector F representing the proportion of sub-catchment area allocated to agricultural management measure m m , the load generated by each measure is reduced by the proportion of the decision vector.

[0076] (d): Construct a connection matrix W that describes the upstream and downstream relationships, where the diagonal elements w i,i =1; if sub-basin j is upstream of sub-basin i, then the off-diagonal element w i,j|i≠j =1, otherwise w i,j|i≠j =0.

[0077] (e): Calculate the total area A of each sub-basin.

[0078] (f): Use formula (9) and formula (10) to preliminarily calculate the in-river load of each sub-basin outlet for each indicator under the decision vector allocation scenario.

[0079] Step 5: Based on the preliminary load calculation results, the in-flow reaction formula is used to update the load calculation results. In addition, for sub-basins with strong human activity interference such as reservoir interception, the reservoir interception formula is used to update the load calculation results.

[0080] Step 5.1: Extract the pollutant-related items (including total nitrogen, total phosphorus and nitrate nitrogen values) and sub-basin flow values ​​at each sub-basin outlet in the basin SWAT model (except for sub-basins with reservoirs, which are calculated independently according to Section 4.2), establish the regression relationship between the pollutant-related items and the flow, and obtain the regression coefficient. Then, based on the in-flow reaction formula (3), calculate the updated value of the pollution load, and then add the updated value to the preliminary load calculation results of each sub-basin outlet obtained in the third step to obtain the updated load result. Taking total phosphorus TP as an example, the in-flow reaction formula is as follows:

[0081]

[0082] ΔTP t ∈R N TP_sum is the vector of the total phosphorus loss in the river at the outlet of N sub-basins, which is obtained by adding a total phosphorus related term and a flow related term; t ∈R N is the total phosphorus load generated for each sub-basin in month t obtained by preliminary calculation in step 3; is the vector of regression coefficients of total phosphorus lost due to sedimentation in N sub-basins and total phosphorus production in the inflowing river; is a vector of the inverse regression coefficients of total phosphorus and water yield in N sub-basins lost due to algae absorption, decomposition and benthic absorption. The calculation process of the above total phosphorus index can be replaced by pollutant indicators such as total nitrogen and nitrate nitrogen, so as to realize the calculation process of the reaction of different pollutants in the flow.

[0083] Step 5:.2: In particular, for sub-basins with strong human activity interference such as reservoir interception, extract the monthly inflow and outflow hydrological landscape production loads of the sub-basin where the reservoir is located in the SWAT model, establish a linear regression, and then obtain its regression coefficient to update the preliminary load calculation results. Taking total phosphorus TP as an example, the calculation formula is as follows:

[0084]

[0085] in, and are the inflow and outflow loads of the sub-basin where the reservoir is located in the tth month obtained by SWAT model simulation of the baseline scenario; and are the linear regression coefficients of the incoming and outgoing loads under the baseline scenario; (w' i,j )TP To consider the total phosphorus interception effect of the reservoir, the variables of each element in the improved connection matrix are: D is the set of all sub-basins downstream of the reservoir, and U is the set of all sub-basins upstream of the reservoir; TP min ∈R N is the vector of the minimum intercepted total phosphorus load of each sub-basin after being intercepted by the reservoir. The value of the sub-basin immediately downstream of the reservoir is In other sub-basins, the values ​​are all 0; TP t ∈R N is the total phosphorus load generated by each sub-basin in month t obtained by preliminary calculation in the third step; TP t,end ∈R N It is the updated total phosphorus load of each sub-basin in the tth month after the reservoir interception.

[0086] Appendix 3 shows the schematic diagram of the total phosphorus flow reaction, which represents the process of phosphorus sedimentation, mineralization, exchange with algae and release from benthic organisms, which can be expressed as:

[0087] (a): First-order reaction ΔP = k P ·P, k P is the phosphorus loss coefficient.

[0088] (b): Zero-order reaction ΔP = k depending on flow depth or algae concentration P f(depth) or ΔP = k P ·f(algae).

[0089] At the same time, the amount of phosphorus transformed or exchanged during the transport process is also related to the residence time in the river channel. Its general form is shown in formulas (9) and (10):

[0090] ΔP=k P,P ·ΔT·P (9)

[0091] ΔP=k P,Q f(depth,algae) ΔT (10)

[0092] In the formula, ΔP represents the amount of phosphorus transformed or exchanged during transport, k P,P represents the phosphorus loss coefficient in the zero-order reaction, k P,Q represents the phosphorus loss coefficient in the first-order reaction, ΔT represents the time that phosphorus stays in the river channel, and P represents the amount of phosphorus in the river channel. It is worth noting that the SWAT model assumes that the transport of phosphorus and sediment is completely decoupled once it reaches the river network. Although the loss of sediment in the landscape has a great impact on the loss of phosphorus in the landscape in reality and SWAT, sediment deposition and erosion have no effect on phosphorus transport in SWAT. Since the goal of the present invention is to proxy the SWAT model, sediment is not taken into account in the phosphorus treatment process of the in-stream reaction.

[0093] Step 6: Use the updated hydrological landscape yield load results (including total nitrogen, total phosphorus, nitrate nitrogen, etc.) as the ecological score of different agricultural management measures at the outlet of the basin, and calculate the total cost of different agricultural management measures as the evaluation criterion. The optimal spatial combination of basin agricultural management measures is selected based on the comprehensive ecological score and the evaluation criteria.

[0094] Step 6.1: The calculation formula for the ecological scores and evaluation criteria of different agricultural management measures (BMPs) is as follows: 1-3 is the ecological scoring formula, and f4 is the judging standard formula:

[0095]

[0096] Where t is the calculation step, month; T is the total number of simulation periods, month; TN outlet_avg TP outlet_avg and NO3-N outlet_avg TN are the multi-year average total nitrogen, total phosphorus and nitrate nitrogen loads at the basin outlet; mod_outlet,t TP mod_outlet,t and NO3-N mod_outlet,t are the updated total nitrogen, total phosphorus and nitrate nitrogen load values ​​at the watershed outlet calculated in the fourth step, kg; G is the cost under different BMPs spatial allocation combinations, 10,000 yuan; for the study of M BMPs scenarios in N sub-basins, A∈R N is the vector of agricultural land area in each sub-basin, hm 2 ; F∈R N×M is the spatial allocation matrix, which represents the proportion of the application area of ​​each sub-basin allocated to each BMPs scenario; C∈R M is the vector of unit area costs corresponding to the scenario of applying M BMPs to 100% of the land area in the whole basin, RMB / hm 2 , x is the application area ratio of different BMPs measures in each sub-basin, that is, F in formula (1) m .

[0097] Step 6.2: Set the weights of the three ecological scores in the optimization target to be the same as the evaluation criteria, and finally calculate the combination of agricultural management measures with the highest score as the optimal governance decision for non-point source pollution in the watershed. The optimization target formula is as follows:

[0098] F(x)=0.25f1(x)+0.25f2(x)+0.25f3(x)+0.25f4(x) (8)

[0099] The specific implementation process of step 6 is as follows:

[0100] (a): Simulate all agricultural management measures and select measures with better non-point source reduction effects. In this embodiment, 5 typical measures are selected.

[0101] (b): The selected typical measures are allocated in different combinations in a certain (or multiple) sub-basins as different agricultural management measures.

[0102] In this embodiment, the number of schemes for combining and allocating the five typical measures is Species, among which It represents the number of options when x measures are selected from 5 typical measures for setting, and it is considered that the measures selected from 5 typical measures for setting in this sub-basin evenly distribute the area of ​​this part of the sub-basin, that is, the decision vector of the selected x measures The decision vector F of the 5-x measures that were not selected m =0.

[0103] (c): Calculate the ecological scores and evaluation criteria of each plan after different combination allocations, and obtain the measure plan with the best nutrient reduction effect and the lowest cost as the optimal agricultural management measure plan.

[0104] Applying the optimal agricultural management measure scheme selected by the present invention to the prevention and control of watershed non-point source pollution can greatly improve the effect of watershed non-point source pollution regulation and control.

[0105] The method, system and device for simulating agricultural management measures based on improved proxy model provided by the present invention can simulate and optimize agricultural management measures more quickly and precisely according to the characteristics of independent calculation of HRU units in SWAT basin hydrological model and the use of different types of measures in different units, through data acquisition, construction of distributed SWAT hydrological model, proxying SWAT using improved response matrix method, considering the improvement process of in-flow reaction and reservoir interception, and optimizing agricultural management measures scenario, so as to provide reliable guarantee for realizing the regulation of basin non-point source pollution. Compared with the prior art, the method of the present invention not only overcomes the difficulties of high complexity, long time consumption and difficulty in optimizing the scenario simulation of agricultural management measures for basin non-point source pollution in SWAT hydrological model, but also greatly improves the regulation efficiency of basin non-point source pollution management measures by using the idea of ​​proxy model, and compared with the existing proxy model, it considers the processes such as the reaction of pollutants in the river channel and the interception effect of reservoirs, so as to realize more accurate proxy of SWAT model, which is of great significance for the technology of regulating and controlling agricultural non-point source pollution by using model.

[0106] Based on the method provided by the present invention, the present invention also provides an agricultural management measure simulation system based on an improved agent model, comprising:

[0107] The data acquisition module is used to obtain various data of the study basin and build the spatial database and attribute database required for the operation of the SWAT model; the various data include DEM elevation, land use, soil type, meteorological and hydrological data of the study basin.

[0108] The SWAT model building module is used to build a SWAT model of a study basin based on a spatial database and an attribute database; the study basin includes multiple sub-basins and hydrological response units.

[0109] The monthly scale landscape yield calculation module is used to simulate different agricultural management measures using the SWAT model and calculate the monthly scale landscape yield of hydrological variables in each sub-basin under different agricultural management measures.

[0110] The module for calculating the export load of each sub-basin is used to construct the corresponding response matrix according to the monthly landscape yield of the hydrological variables of each sub-basin under different agricultural management measures, and use the improved response matrix method to calculate the export load of each sub-basin.

[0111] The proxy model determination module is used to adopt the improved response matrix method as the proxy model according to the proxy calculation results of the outlet load of each sub-basin.

[0112] The optimal agricultural management measure scheme determination module is used to utilize the agent model to optimize the agricultural management measure scheme and determine the optimal agricultural management measure scheme.

[0113] Furthermore, the present invention also provides an electronic device, which may include: a processor, a communication interface, a memory and a communication bus. The processor, the communication interface and the memory communicate with each other via the communication bus. The processor may call a computer program in the memory to execute the agricultural management measure simulation method based on the agent model.

[0114] In addition, when the computer program in the above-mentioned memory is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a non-transitory computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories, random access memories, magnetic disks or optical disks.

[0115] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0116] The present invention uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core ideas of the present invention. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for simulating agricultural management measures based on an improved agent model, characterized in that: The agricultural management measure program simulation method comprises the following steps: The first step is to obtain various data of the study basin and build a SWAT model of the study basin based on the spatial database and the attribute database; the study basin includes multiple sub-basins and hydrological response units; In the second step, the watershed SWAT model constructed in the first step was used to simulate the monthly landscape yield of hydrological variables in each sub-watershed under different agricultural management measures BMPs, and the response matrix corresponding to each agricultural management measure was calculated; The third step is to calculate the response matrix Y of different agricultural management measures BMPs obtained in the second step. m,t , preliminarily calculate the hydrological landscape yield load value Q at the outlet of each sub-basin after hydrological confluence t ; Step 4: Based on the preliminary load calculation results of each sub-basin outlet obtained in step 3, the load calculation results are updated using the in-stream reaction formula. For sub-basins with strong human activity interference such as reservoir interception, the reservoir interception formula is used to update the load calculation results; In the fifth step, based on the updated hydrological landscape yield load results obtained in the fourth step, it is regarded as the ecological score of different agricultural management measures BMPs at the basin outlet. At the same time, the total cost of different agricultural management measures BMPs is calculated as the evaluation criterion. The optimal spatial combination of basin agricultural management measures BMPs is selected by combining the ecological score and the evaluation criteria, providing a basis for the governance decision-making of basin non-point source pollution.

2. The method for simulating agricultural management measures based on an improved agent model according to claim 1, characterized in that: The first step is specifically: 1.1) Based on the spatial database and attribute database of the basin, the study basin is divided into several sub-basins and hydrological response units, and a preliminary basin SWAT model is established; 1.2) Use the global sensitivity analysis method to perform parameter sensitivity analysis and parameter calibration on the preliminary SWAT model of the basin, and update the optimal parameter values ​​of the SWAT model of the basin; 1.3) The simulation results of the basin SWAT model with the optimal parameter values ​​are evaluated by the Nash efficiency coefficient and the determinism coefficient evaluation index. If the Nash efficiency coefficient and the determinism coefficient are both greater than 0.8, the SWAT model with updated parameters is considered to be completed.

3. The method for simulating agricultural management measures based on an improved agent model according to claim 2, characterized in that: The second step is specifically as follows: 2.1) Using the SWAT model constructed in the study basin, different agricultural management measures were simulated, and each agricultural management measure was applied to all agricultural types of hydrological response units in all sub-basins during the simulation; 2.2) The monthly landscape yield of hydrological variables in each sub-basin obtained by different agricultural management measures, which will be recorded as y m,t , represents the landscape yield in month t when the mth agricultural management measure BMP is applied to 100% of the land area in the whole basin; the hydrological variables include: flow, total phosphorus, total nitrogen and nitrate; 2.3) Monthly landscape yield y of hydrological variables in each sub-basin under different agricultural management measures m,t is the response matrix Y m,t The constituent elements of Y m,t It is a two-dimensional matrix with the rows being months and the columns being types of agricultural management measures.

4. The method for simulating agricultural management measures based on an improved agent model according to claim 3, characterized in that: In the above 2.1), the agricultural management measures include non-engineering measures and engineering measures; the non-engineering measures include different fertilization systems, irrigation strategies and farming methods; the engineering measures include vegetation filter belts and grassed waterways.

5. The method for simulating agricultural management measures based on an improved agent model according to claim 3, characterized in that: The third step is specifically as follows: 3.1) Constructing the decision vector F representing the weights of agricultural management measures m , represents the application area ratio of the mth agricultural management measure BMP in a sub-basin, ranging from 0-100%, 100% means that the entire sub-basin is fully used; 3.2) Calculate the total area A of each sub-basin based on the basin spatial database, which is a vector of all sub-basin areas; 3.3) Construct a connection matrix W that describes the upstream and downstream relationships of each sub-basin to approximately represent the process of the SWAT model hydrological basin; where, for each element w in the connection matrix i,j , if sub-basin j is located upstream of sub-basin i, the off-diagonal element w i,j|i≠j =1, otherwise w i,j|i≠j = 0; when i = j, it represents the diagonal element w of each sub-basin i,j =1; as shown in Figure 2(d) of the specification; 3.4) Based on the response matrix Y m,t , decision vector F m , total area A and connection matrix W, and use the following formula to calculate the outlet load of each sub-basin: Among them, q t ∈R N is the hydrological landscape yield load vector generated in N sub-basins in the tth month; (Y m,t ) Q ∈R N is the response matrix in month t when the mth BMPs is applied to 100% of the land area in N sub-basins; A∈R N is the vector of N sub-basin areas; F m ∈R N is the vector of the area ratio of the mth BMP in N sub-basins; diag(·) is the function of diagonalizing the vector; Q t ∈R N is the vector of the hydrological landscape yield load at the outlet of N sub-basins in the tth month, where the hydrological landscape yield load includes flow, total nitrogen, total phosphorus and nitrate nitrogen.

6. The method for simulating agricultural management measures based on an improved agent model according to claim 5, characterized in that: The fourth step is specifically as follows: 4.1) First, extract the pollutant-related items and sub-basin flow values ​​at the outlet of each sub-basin in the basin SWAT model, establish the regression relationship between the pollutant-related items and the flow, and obtain the regression coefficients. The pollutant-related items include the values ​​of total nitrogen, total phosphorus and nitrate nitrogen. Then, calculate the updated value of the pollution load based on the in-flow reaction formula (3), and then add the updated value to the preliminary load calculation results of each sub-basin outlet obtained in the third step to obtain the updated load results. Taking total phosphorus TP as an example, the in-flow reaction formula is as follows: Among them, ΔTP t ∈R N TP_sum is the vector of the total phosphorus loss in the river at the outlet of N sub-basins, which is obtained by adding a total phosphorus related term and a flow related term; t ∈R N is the total phosphorus load generated for each sub-basin in month t obtained by preliminary calculation in step 3; is the vector of regression coefficients of total phosphorus lost due to sedimentation in N sub-basins and total phosphorus production in the inflowing river; is a vector of inverse regression coefficients of total phosphorus and water yield in N sub-basins, which are lost due to algal uptake, decomposition and benthic biosorption; The calculation process of total phosphorus index can be replaced by total nitrogen, nitrate nitrogen or other pollutant indexes to realize the calculation process of different pollutant flows; 4.2) For the sub-basins with disturbances, extract the monthly inflow and outflow hydrological landscape production loads of the sub-basins where the reservoir is located in the SWAT model, establish a linear regression, and obtain its regression coefficient to update the preliminary load calculation results; Taking total phosphorus TP as an example, the calculation formula is as follows: in, and are the inflow and outflow loads of the sub-basin where the reservoir is located in the tth month obtained by SWAT model simulation of the baseline scenario; and are the linear regression coefficients of the incoming and outgoing loads under the baseline scenario; (w' i,j ) TP To consider the total phosphorus interception effect of the reservoir, the variables of each element in the improved connection matrix are: D is the set of all sub-basins downstream of the reservoir, and U is the set of all sub-basins upstream of the reservoir; TP min ∈R N is the vector of the minimum intercepted total phosphorus load of each sub-basin after being intercepted by the reservoir. The value of the sub-basin immediately downstream of the reservoir is In other sub-basins, the values ​​are all 0; TP t ∈R N is the total phosphorus load generated by each sub-basin in month t obtained by preliminary calculation in the third step; TP t,end ∈R N It is the updated total phosphorus load of each sub-basin in the tth month after the reservoir interception.

7. The method for simulating agricultural management measures based on an improved agent model according to claim 6, characterized in that: The fifth step is specifically as follows: 5.1) The calculation formula for the ecological scores and evaluation criteria of BMPs of different agricultural management measures is as follows: 1-3 is the ecological scoring formula, and f4 is the judging standard formula: Where t is the calculation step, month; T is the total number of simulation periods, month; TN outlet_avg TP outlet_avg and NO3-N outlet_avg TN are the multi-year average total nitrogen, total phosphorus and nitrate nitrogen loads at the basin outlet; mod_outlet,t TP mod_outlet,t and NO3-N mod_outlet,t are the updated total nitrogen, total phosphorus and nitrate nitrogen load values ​​at the watershed outlet calculated in the fourth step, kg; G is the cost under different BMPs spatial allocation combinations, 10,000 yuan; for the study of M BMPs scenarios in N sub-basins, A∈R N is the vector of agricultural land area in each sub-basin, hm 2 ; F∈R N×M is the spatial allocation matrix, which represents the proportion of the application area of ​​each sub-basin allocated to each BMPs scenario; C∈R M is the vector of unit area costs corresponding to the scenario of applying M BMPs to 100% of the land area in the whole basin, RMB / hm 2 , x is the application area ratio of different BMPs measures in each sub-basin, that is, F in formula (1) m ; 5.2) Set the weights of the three ecological scores in the optimization target to be the same as the evaluation criteria, and calculate the combination of agricultural management measures with the highest score as the optimal governance decision for non-point source pollution in the watershed. The optimization target formula is as follows: F(x)=0.25f1(x)+0.25f2(x)+0.25f3(x)+0.25f4(x) (8) In the formula, the letters represent the same as in formula (7).

8. An agricultural management measure simulation system based on an improved agent model, characterized in that: include: The data acquisition module is used to acquire various data of the study basin and build the spatial database and attribute database required for the operation of the SWAT model; the various data include DEM elevation, land use, soil type, meteorological and hydrological data of the study basin; A SWAT model building module is used to build a SWAT model of a study basin based on a spatial database and an attribute database; the study basin includes multiple sub-basins and hydrological response units; The monthly scale landscape yield calculation module is used to simulate different agricultural management measures using the SWAT model and calculate the monthly scale landscape yield of hydrological variables in each sub-basin under different agricultural management measures; The module for calculating the export load of each sub-basin is used to construct the corresponding response matrix according to the monthly landscape yield of the hydrological variables of each sub-basin under different agricultural management measures, and use the improved response matrix method to calculate the export load of each sub-basin; An improved proxy model determination module is used to adopt an improved response matrix method as an improved proxy model according to proxy calculation results of the outlet load of each sub-basin; The optimal agricultural management measure program determination module is used to use the improved agent model to optimize the agricultural management measure program and determine the optimal agricultural management measure program.

9. An electronic device, characterized in that: The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for simulating agricultural management measures based on an agent model is implemented.

10. An electronic device according to claim 9, characterized in that: The memory is a non-transitory computer-readable storage medium.

Images