A multi-objective and multi-scale optimization method for crop planting structure

Abstract

The present invention discloses a method for optimizing crop planting structure based on multi-scale and all-round evaluation. The SWAT-WF distributed hydrological model is used to simulate the regional crop water productivity and production water footprint, and a multi-objective spatial crop planting structure optimization model with the goals of maximizing regional economic water productivity, minimizing blue water dependence, and minimizing grey water footprint is established. Combining multi-level benefit indicators, a comprehensive evaluation index system based on regional, provincial, and national scales is constructed. Through the entropy weight-TOPSIS comprehensive evaluation model, the optimization of multiple crop planting structure schemes is realized. The method of the present invention considers the comprehensive benefits within the region and the impact on the grain trade pattern outside the region, provides reasonable suggestions for the adjustment of regional planting layout, and has important significance for the sustainable development of multi-scale agricultural systems.

Description

Technical Field

[0001] The present invention belongs to the technical field of agriculture, and relates to a method for optimizing the spatial layout of crop planting, and particularly to a method for optimizing crop planting structure based on multi-objectives and multi-scales. Background Art

[0002] Water is the foundation of production. Against the backdrop of water resource shortage, agricultural water consumption in China accounts for more than 60% of the total water consumption, and 85% of agricultural water is used to meet the irrigation needs of crops. The uneven distribution of water resources has led to prominent contradictions between agricultural water use and water use in other industries, especially in arid northern regions. Population growth and economic development have brought greater pressure to China's food security. Optimizing the crop planting structure can bring multiple benefits to the region and play an important role in ensuring the sustainable development of regional agriculture. The optimization of the planting structure aims to establish an optimal allocation between water resources and agricultural production, and maximize regional benefits by reasonably adjusting the crop planting ratio and spatial distribution.

[0003] The concept of virtual water trade provides a new perspective for solving water resource shortage problems in water-scarce regions. Virtual water represents the total amount of water resources consumed in the production process of goods or services, and it exists in goods or services in an "invisible" form. Virtual water trade is a method for water-scarce regions to reduce their own water resource consumption by importing water-intensive products, and it is a commodity strategy to ensure regional water resource security.

[0004] In the prior art, most studies maximize regional production benefits by introducing the concept of water footprint into the regional planting structure optimization model. However, the adjustment of the regional crop planting structure will change the actual yields of various local crops, affect the regional food supply and demand balance, and further affect inter-regional food trade. The lack of consideration of the associated effects outside the region will lead to one-sided optimization of the regional planting structure. Therefore, it is of great practical significance to research a method for optimizing crop planting structure based on a multi-objective and multi-scale evaluation index system, taking the virtual water trade of different regional crops as the link, and comprehensively considering the optimization of the regional crop planting structure and the associated benefits at multiple scales for evaluation.

[0005] Object of the Invention

[0006] The object of the present invention is to solve the above problems existing in the prior art, provide a method for optimizing crop planting structure based on a multi-objective and multi-scale evaluation index system, give an optimization plan for the planting structure that is suitable for improving the comprehensive regional benefits and has the lowest negative external benefits, and provide theoretical suggestions for the adjustment of the regional planting layout. Summary of the Invention

[0007] The present invention provides a method for optimizing crop planting structure based on multi-objectives and multi-scales, comprising the following steps:

[0008] Step 1: Select the area where the crops to be planted are located as the research area for optimizing the crop planting structure, determine the main crop planting types, and construct a simulation model for regional crop growth and production water footprint;

[0009] Step 2: Determine the virtual water trade objects based on the analysis of crop production water footprint and regional supply-demand balance, set the trade coefficient, and determine the optimization plan for the regional crop planting structure;

[0010] Step 3: Construct a regional multi-objective crop planting structure optimization model to achieve the optimal layout adjustment of the regional planting structure under each optimization plan;

[0011] Step 4: Quantify the impact of regional planting structure changes on the balance state of provincial food supply and demand and the change of the inter-provincial food trade pattern;

[0012] Step 5: Combine multi-level benefit indicators to construct a comprehensive evaluation index system based on regional, provincial, and national scales;

[0013] Step 6: Based on the constructed comprehensive evaluation index system and entropy weight-TOPSIS comprehensive evaluation model, select the optimization plan for the planting structure and determine the optimal planting layout suitable for regional development.

[0014] Preferably, Step 1 includes selecting the research area for optimizing the crop planting structure, determining the main crop planting types and research years in the research area; collecting and sorting out the spatial attribute data and agricultural management measure data in the research area, constructing a SWAT-WF distributed hydrological model that can simultaneously simulate crop growth and crop production water footprint, and using a multi-process calibration program based on the NSGA-II algorithm to calibrate and verify the hydrological process and crop growth process parameters of the SWAT model; the spatial attribute data includes digital elevation data, soil data, land use data, and meteorological data; the agricultural management measure data includes crop planting systems and irrigation schedules.

[0015] Preferably, in step 2, the supply and demand status of crops is determined based on the supply and actual demand of the main food crops in the study area; where the supply is obtained by simulating with the SWAT model, and the demand is calculated based on the local actual population and food security guarantee settings; according to the simulation results of the crop production water footprint, select the crop types in the study area with a supply greater than demand and a relatively high production water footprint as the objects of virtual water trade; on the basis of ensuring regional food security, by setting the trade coefficients of each trade object, multiple trade combination plans are formed; the trade coefficient varies between 0 and 1, with an increment of 0.05, which refers to the reduction rate of the planting area exceeding the regional demand, that is, as the trade coefficient of a crop increases, the planting proportion of this crop in the study area decreases, and at the same time the yield decreases, but the state of oversupply in the region will not change until the trade coefficient equals 1, reaching the supply and demand balance; by adjusting the trade coefficients, the planting proportion of crops with a high regional production water footprint is reduced, and the crop planting structure of the region is changed.

[0016] Preferably, step 3 includes two processes: the analysis of the suitability of crop growth space and the construction of a regional multi-objective crop planting structure optimization model; among them, the analysis of the suitability of crop growth space is a process of making qualitative and quantitative evaluations of the suitability of growing a certain crop in a specific area, and is used to predict the crop production potential and limitations in the region; the suitability of crop growth is measured by a suitability index, which is the product of the membership degree values of each evaluation index and the index weight, and the value ranges from 0 to 1; the evaluation indexes include eight soil factors and two terrain factors, and the eight soil factors include soil texture, bulk density, pH value, total nitrogen, available nitrogen, available phosphorus, available potassium, and organic matter; the two terrain factors include elevation and slope; according to the S-type function, the membership degree values are determined for multiple rating indexes of various crops, and the contribution rate of each evaluation index to the change of crop water productivity is calculated by the partial least squares regression method, and the contribution rate is used as the index weight of each index, and the calculation formula is shown as formula (1):

[0017]

[0018] In the formula: W i is the contribution rate of the i-th evaluation index factor to the change of crop water productivity; VIP i is the variable projection importance of the i-th index factor; R represents the multiple correlation coefficient of the PLS regression model, which is calculated by SPSS software; n represents the number of evaluation indexes in the PLS regression model;

[0019] Using the natural breakpoint method, the suitability of each crop is divided into 5 grades, and the higher the grade, the lower the suitability.

[0020] Further preferably, the regional multi-objective crop planting structure optimization model is constructed based on the cellular automaton and the SWAT model. The grid cells for crop planting are used as cells, the grid cell matrix of the entire study area is used as the cell space, and the Moore-type neighbors, that is, the grids in 8 directions around the central cell, are used as cell neighbors. The conversion rules are used as the basis for determining the next state of each cell unit, considering the constraints set in the model. The regional multi-objective crop planting structure optimization model takes improving the economic water productivity EWP as the optimization objective at the economic level, reducing the blue water footprint BWF rate as the optimization objective at the resource level, and reducing the grey water footprint GWF grey as the optimization objective at the ecological benefit level; where BWF rate is the ratio of the blue water footprint of crop production to the total crop water footprint. The smaller BWF rate is, the smaller the degree of dependence of crop production on agricultural irrigation, that is, the higher the utilization efficiency of water resources. The normalization process is used to solve the problem of inconsistent dimensions of multiple objective functions in multi-objective programming.

[0021] Further preferably, in the regional multi-objective crop planting structure optimization model, the expressions of the multi-objective optimization equation and the normalization process are shown in formulas (2)-(3):

[0022]

[0023]

[0024] In the formula: f k,min and f k,max represent the minimum and maximum values of the objective function f(x) respectively; λ k represents the weight of F k (x), which is determined by the analytic hierarchy process; the value of n is equal to 3, where the objective function of the economic water productivity EWP seeks the maximum value, and the objective functions of the blue water footprint BWF rate and the grey water footprint GWF grey seek the minimum values;

[0025] The grey water footprint GWF grey is obtained by simulating with the SWAT-WF model; the calculation formulas of the two indicators of the economic water productivity EWP and the blue water footprint BWF rate are shown in formulas (4)-(5):

[0026] s

[0027]

[0028] In the formula, the unit of EWP is yuan / m 3; Y is the crop yield per unit area simulated by the model, with the unit of kg / ha; AET is the actual evapotranspiration simulated by the model, with the unit of mm; NP is the net profit per unit yield, with the unit of yuan / kg, and the data is from the statistical yearbook; BWF blue represents the blue water footprint of crop production, and WF represents the total water footprint of crop production, both of which are obtained by simulating with the SWAT-WF model.

[0029] Further preferably, the regional multi-objective crop planting structure optimization model takes the crop planting area, irrigation water demand, and crop planting suitability as constraint conditions; among them, the crop planting area is limited by the minimum planting area of various crops in the past 10 years and food security as the lowest crop planting area limit, and at the same time, the total crop planting area does not exceed the total area of existing cultivated land, which is expressed as shown in formulas (6)-(7):

[0030]

[0031]

[0032] In the formula, CN x is the total number of grids of crop x, X represents the total number of crop types, TN is the total number of planting grids of all crops in the existing crop planting structure, TN x is the total number of grids of crop x in the existing planting structure, α x is the allowable change range of the crop planting area of crop x;

[0033] Set the irrigation constraint so that the irrigation water demand of the crops in the optimized planting structure does not exceed the irrigation water demand of the crops in the existing planting structure. The irrigation water demand of the crops is the difference between the water evapotranspiration during the crop growth period and the effective precipitation. The evapotranspiration refers to evaporation and transpiration. Express the irrigation constraint as shown in formula (8):

[0034]

[0035] In the formula, M and N are the total number of rows and columns of the grids generated by the planting distribution in the study area; K mn is the crop planting coefficient, indicating whether there is a crop planted in a grid cell. If a crop is planted in a unit, the crop planting coefficient is 1, otherwise it is 0; I x mn represents the irrigation water demand of crop x at the cell of row m and column n; TI is the irrigation water demand of the crops in the existing planting structure;

[0036] Set the crop suitability constraint so that each crop must be planted on the cultivated land grid where the crop suitability level is less than 4. At the same time, take the condition that only one crop is planted in each grid cell as the basic constraint condition, which is expressed as shown in formula (9):

[0037]

[0038] where N mn is the suitability level of crop x at the m-th row and n-th column;

[0039] The conversion rules of the multi-objective crop planting structure optimization model for the said area include two parts: judgment and conversion. A random function is used to ensure that the crop conversion probability is equal for each grid in the study area. Specifically, first, it is judged whether a grid can be converted into another crop, and this process is based on the minimum requirements for the planting area of a specific crop, suitability, and neighborhood status; without considering the minimum requirements for the planting area, a grid can be converted into any crop that meets the suitability conditions; then, the conversion priority of these crops in the same unit is determined through the objective function, that is, converting the crops planted in the grid into the crop with the highest comprehensive benefit; according to the random order, the conversion rules are iterated to each grid unit until all cells are iterated, and the actual irrigation water requirement after optimizing the global crop planting structure is calculated. When the actual irrigation water requirement after optimization is less than the irrigation water requirement of the existing planting structure, the optimization is completed; otherwise, the crops in the grid with a large irrigation water requirement are replaced. The replacement principle is to rank the comprehensive benefits of the crops with a smaller irrigation amount than the currently planted crops, and select the crop with the highest comprehensive benefit among them to replace the existing grid; the grid crops are replaced in the order of decreasing irrigation water requirement until the irrigation water requirement is less than the irrigation water requirement of the actual planting structure;

[0040] Bring the various trade combination plans generated in step 2 into the multi-objective crop planting structure optimization model for the area to obtain the optimal planting structure and the corresponding regional optimal benefit index under each combination plan.

[0041] Preferably, in step 4, different food outputs are generated for various planting structure optimization plans, and it is necessary to re-quantify the food supply-demand balance state of the provincial region and the inter-provincial food trade pattern. Specifically, the food supply-demand balance quantification formula for each provincial administrative region is expressed as shown in formula (10):

[0042] T i,j = G i,j +(im i,j - ex i,j ) - C i,j (10),

[0043] where i and j represent the crop type and the provincial administrative region respectively; T i,j represents the total surplus of food i in province j, T i,j > 0 indicates that province j has an output of crop i, representing a supply surplus, T i,j < 0 indicates that province j has an input of crop i, representing a supply shortage, while T i,j= 0 indicates that there is no inter-provincial trade of crop i in province j, representing the balance between supply and demand; G i,j and C i,j respectively represent the total production and consumption of crop i in province j; im i,j and ex i,j respectively represent the total import and export volumes of crop i in province j.

[0044] Preferably, the total crop production in each province of China is taken as the supply volume of the crop, and the data is sourced from the China Statistical Yearbook; the consumption data of each crop is sourced from the BRIC Agricultural Database, and the data is downscaled to the provincial level using the consumption proportion of each region; the grain consumption is divided into food ration consumption, feed grain consumption, industrial grain consumption, seed grain consumption, and grain loss, and the supply and demand status of each grain crop in each province is determined through the calculation of supply and consumption data;

[0045] Preferably, the inter-provincial grain trade pattern is quantified through a multi-objective linear optimization model. Specifically, the objective function is to minimize the global grain transportation cost and dietary structure difference, and the constraint condition is the conservation of grain surplus and deficit. Among them, the dietary structure difference is characterized by the Euclidean distance of the per capita food ration and animal product consumption in each province, and the grain transportation cost takes into account both the transportation distance and transportation mode; the objective function and constraint conditions of the multi-objective linear optimization model are expressed as shown in Equation (11):

[0046]

[0047] In the formula, t i,a,b and s i,a,b respectively represent the transportation cost and dietary structure difference between province a and province b; M and N respectively represent the number of export provinces and import provinces of crop i; x i,a,b represents the total transportation volume of crop i between province a and province b.

[0048] Preferably, in step 5, a multi-level benefit evaluation index system based on three scales of region, province, and country is established to select the optimal solution from different trade combinations, that is, to select the trade coefficient combination with the maximum benefit within the region and the minimum negative impact on external grain trade; the top layer of the evaluation index system is the target layer, that is, different trade combination schemes in the region; the second layer is the standard layer, including regional benefit, provincial benefit, and national benefit; the third layer is the index layer, which is composed of relevant benefit indicators at each scale. Specifically, the regional benefit indicators are represented by three indicators: economic income, blue water dependence degree, and grey water footprint; the provincial benefit indicators are the grain trade income and environmental sustainability index; the national benefit indicators are the three indicators of total transportation cost, trade emergy saving, and trade non-renewable emergy saving; the economic income is calculated from the crop yield and the net income per unit yield, and the grey water footprint GWF greyObtained by simulation of the SWAT-WF model, the blue water dependence is calculated according to Equation (5); the grain trade income is quantified by the difference in grain trade turnover of the provincial region, and the total transportation cost is calculated by the product of the freight rate and the volume of transportation; the environmental sustainability index, trade emergy savings, and trade non-renewable emergy savings are related to the calculation of the emergy system. The specific index meanings and calculation formulas are shown in Equations (12)-(14):

[0049]

[0050]

[0051]

[0052] In the formula: ESI represents the environmental sustainability index, EYR and ELR represent the system output rate and environmental load rate respectively, U, P, and S represent the total emergy consumption, purchased resources, and service resources of crop production respectively, R and N represent the renewable part and non-renewable part of the emergy input; U saving and UN saving represent the trade emergy savings and trade non-renewable emergy savings of the trade pattern of multiple crops respectively. When crops are traded from a region with low production emergy to a region with high emergy, it represents emergy savings; X represents the total amount of crops, UEV i,a and UEV i,b represent the emergy consumption of producing unit yield of crop i in Province a and Province b respectively, UEVN represents the non-renewable emergy amount of unit yield of crop; TV i,a,b represents the trade volume of crop i from Province a to Province b.

[0053] Preferably, in Step 6, various evaluation indicators are calculated based on the regional supply-demand differences and changes in trade patterns generated by various planting structure optimization schemes, and combined with the entropy weight-TOPSIS comprehensive evaluation model, the optimal planting layout suitable for regional development is determined; according to the value of the relative closeness, various planting structure optimization schemes are sorted, and the scheme with the highest relative closeness is selected as the optimal scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a flow chart of the multi-objective and multi-scale crop planting structure optimization method of the present invention.

[0055] Figure 2 is a schematic diagram of the regional crop planting structure optimization model.

[0056] Figure 3 is the crop planting structure optimization evaluation index system. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0058] Those skilled in the art should understand that the step numbers used in the text are only for convenience of description and do not limit the execution order of the steps. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" may include the plural forms. The term "and / or" refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations. In the description of the present invention, the meaning of "plural" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0059] Embodiment

[0060] Figure 1 is a flowchart of the multi-objective and multi-scale crop planting structure optimization method described in the present invention. According to Figure 1 shown, in the embodiment, a specific implementation example of a crop planting structure optimization method based on a multi-objective and multi-scale evaluation index system is provided, including the following steps:

[0061] Step 1: Select the middle and upper reaches of the Heihe River Basin as the study area in the implementation case, and use 2015 as the study year. The main crops planted in the study area are corn, wheat, barley and rapeseed. Based on the digital elevation data, soil data, land use data, meteorological data of the study area, as well as the crop planting systems and irrigation schedules of the main crops, a SWAT-WF distributed hydrological model that can simultaneously simulate crop growth and crop production water footprint is constructed. The SWAT multi-process calibration program of the NSGA-II algorithm is used to calibrate and verify the simulation processes of runoff at regional hydrological stations, evapotranspiration of sub-basins and soil water content, and to simulate the crop yield, crop water productivity and production water footprint of the region.

[0062] Step 2: Determine the supply and demand status of crops based on the supply and actual demand of the main food crops in the study area. Among them, the supply is obtained by simulating with the SWAT model, and the demand is calculated based on the local actual population and the setting of food security guarantee. A trade coefficient is set for the crop types with high virtual water for crop production and oversupply. The trade coefficient varies between 0 and 1, with an increment of 0.05, to form various crop planting area change schemes in the study area.

[0063] Step 3: Construct an optimization model for the regional multi-objective crop planting structure and achieve the optimal layout adjustment under each planting structure optimization plan.

[0064] The evaluation of the suitability of crop growth space is the main limiting condition of the spatial optimization model of the planting structure. The suitability of crop growth is measured by the suitability index, which is the product of the membership degree values of each evaluation index and the index weight, and the value ranges from 0 to 1. The evaluation indexes mainly include eight soil factors (soil texture, bulk density, pH value, total nitrogen, available nitrogen, available phosphorus, available potassium, organic matter) and two terrain factors (elevation and slope). The membership degree values of various rating indexes of various crops are obtained according to the growth habit data of each crop. The contribution rate of each evaluation index to the change of crop water productivity is calculated by the partial least squares regression method, and the contribution rate is used as the index weight of each index. The suitability of each crop is divided into 5 levels by the natural breakpoint method. The larger the level, the lower the suitability. The suitability level less than 4 represents suitable for crop growth. Based on this data, the suitable growth areas of various crops in the study area are divided. Among them, the index weights of each evaluation index can be calculated by combining two important parameters, the variable projection importance value of each evaluation index and the multiple correlation coefficient of the partial least squares regression model. The calculation formula is as follows:

[0065]

[0066] In the formula: W i is the contribution rate of the i-th evaluation index factor to the change of crop water productivity, VIP i is the variable projection importance of the i-th index factor, R represents the multiple correlation coefficient of the PLS regression model (calculated by SPSS software), and n represents the number of evaluation indexes in the PLS regression model.

[0067] Figure 2 is the schematic diagram of the optimization model for the regional crop planting structure. According to Figure 2 as shown, construct an optimization model for the regional crop planting structure. Taking the economic water productivity, blue water dependence and grey water footprint to represent the economic, resource and environmental benefits of the agricultural system respectively, construct a multi-objective optimization model. Use the normalization process to solve the problem of inconsistent dimensions of multiple objective functions in multi-objective programming. The expressions of the multi-objective optimization equation and the normalization process are as follows:

[0068]

[0069]

[0070] In the formula: f k,min and f k,maxrepresent the minimum and maximum values of the objective function f(x); the value of n is equal to 3, where the objective function of EWP seeks the maximum value, and those of BWF rate and GWF grey seek the minimum value; λ k represents the weight of F k (x), which is determined by the analytic hierarchy process. In this embodiment, the index weights of EWP, BWF rate and GWF grey are 0.20, 0.31, and 0.49 respectively.

[0071] The grey water footprint is obtained by simulating with the SWAT-WF model. The calculation formulas for the other two indicators are as follows:

[0072]

[0073]

[0074] In the formula: EWP is the economic water productivity, with the unit of yuan / m 3 ; Y is the crop yield per unit area simulated by the model, with the unit of kg / ha; AET is the actual evapotranspiration simulated by the model, with the unit of mm; NP is the net profit per unit yield, with the unit of yuan / kg, and the data is from the statistical yearbook; BWF rate represents the blue water dependence; BWF blue represents the blue water footprint of crop production; WF represents the total water footprint of crop production, all of which are obtained by simulating with the SWAT-WF model.

[0075] The entire optimization model is constructed using MATLAB. The model takes the crop planting structure in the Heihe River Basin in 2015 as the starting state of the optimization model. The data to be loaded includes the crop planting structure, the suitability distribution of crops, the number of neighborhoods, the spatial distribution of the comprehensive benefits of different crops, and the irrigation water demand. The model first uses the suitability grade distribution of each crop to judge the types of crops suitable for planting in each grid. When the crop suitability grade is less than 4, it indicates that the crop is suitable for growth. Next, based on the neighbor constraint, it is judged whether the crop can be planted in this grid. If the constraint is met, it means that the cell can plant the crop; otherwise, the cell cannot plant the crop. Then, on the premise of the planting area constraint, the conversion rule is randomly executed on each cell until all cells are iterated. The actual irrigation water demand after optimizing the global crop planting structure is calculated. When the actual irrigation water demand after optimization is less than the irrigation water demand of the existing planting structure, the optimization is completed. Otherwise, the crops in the grid with a large irrigation water demand are replaced. The replacement principle is to sort the water productivity of the crops with a smaller irrigation amount than the currently planted crops and select the crop with the highest water productivity to replace it in the existing grid. The grid crop replacement is carried out in the order of decreasing irrigation water demand until the irrigation water demand is less than the irrigation water demand of the actual planting structure. The proportion of crop planting changes determined by various trade combination schemes generated in step 2 is brought into the regional multi-objective crop planting structure optimization model to obtain the optimal planting structure and the corresponding regional optimal benefit indicators under each combination scheme.

[0076] Step 4: To realize the optimization of each crop planting plan, it is necessary to re-quantify the grain supply-demand balance state of the provincial region and the grain trade pattern between provinces under various planting structure optimization plans.

[0077] Based on the Brick Agricultural Big Data Network, we obtain the import proportion of each crop in each province. The import quantity of each province is calculated based on the national import-export difference, and combined with the actual output of each province's crops, the actual output quantity of 31 provincial regions in the country is obtained. The consumption data of crops is at the national scale and needs to be downscaled to the provincial scale. The downscaling method is completed in the following four steps:

[0078] (1) Collect the total national consumption of crops, calculate the proportion of the total grain consumption data in the central, western, and eastern regions according to the grain consumption data in the central, western, and eastern regions, and calculate the total consumption of crops in the three economic regions in combination with the total national consumption.

[0079] (2) Calculate the total consumption of different consumption types in each economic region according to the proportion of the consumption data of the three economic regions of east, west, and central.

[0080] (3) Calculate the proportion of edible consumption in each province according to the urban and rural population numbers in each provincial region; calculate the proportion of feed demand in each province according to the actual output of poultry, livestock products, eggs, milk and aquatic products and the feed conversion ratio in each provincial region; calculate the proportion of industrial demand in each province according to the quantity of industrial output products related to each crop in each province; calculate the proportion of consumption demand for seeds according to the planting area of each crop in each province; calculate the proportion of losses according to the actual output of each crop in each province.

[0081] (4) According to the proportion of the consumption data of each crop in each province in Step 3, divide the different crop consumption data in the three major economic regions to the provincial scale, and obtain the actual consumption data of the crops in the province by summing up the consumption data of multiple industries in a single province. Among them, the three major economic regions are the regional division results implemented by the "Annual Rural Household Survey in China" according to the geographical distribution and economic development level.

[0082] Calculate the production surplus or production gap of each provincial region based on the supply and demand data of each province. Use a multi-objective linear optimization model for the inter-provincial grain trade pattern. The model takes the global minimum of grain transportation costs and dietary structure differences as the objective function, and the conservation of grain surplus and deficit as the constraint condition. Among them, the dietary structure difference is characterized by the Euclidean distance of the per capita grain and animal product consumption in each province. The grain transportation cost takes into account both the transportation distance and the transportation mode.

[0083] Step 5: Establish a multi-level benefit evaluation index system based on three scales of region, province and country. See the index system diagram in Figure 3 . The top layer of the evaluation index system is the target layer, that is, different trade combination schemes in the region. The second layer is the standard layer, including regional benefits, provincial benefits and national benefits. The third layer is the index layer, which consists of relevant benefit indicators at each scale. Use three indicators of economic income, blue water dependence degree and grey water footprint to represent the regional benefit indicators. Select grain trade income and environmental sustainability index as the provincial benefit indicators. Incorporate three indicators of total transportation cost, trade emergy saving and non-renewable emergy saving in trade into the evaluation index system as the national benefit indicators. The economic income is calculated by the crop yield and the net income per unit yield. The calculation methods of blue water dependence degree and grey water footprint are shown in Step 3. The grain trade income is quantified by the difference in grain trade turnover in the provincial region, and the transportation cost is calculated by the product of the freight rate and the volume of transported goods. The regional emergy sustainability index, trade emergy saving and non-renewable emergy saving in trade are related to the calculation of the emergy system. The specific indicator meanings and calculation formulas are as follows:

[0084]

[0085]

[0086]

[0087] Where: ESI represents the regional emergy sustainable index; EYR and ELR represent the system yield rate and environmental loading rate respectively; U, P, and S represent the total emergy consumption, purchased resources, and service resources of crop production; R and N represent the renewable and non-renewable parts of emergy input. U saving and UN saving represent the emergy savings and non-renewable emergy savings in the trade pattern of multiple crops respectively. When crops are traded from regions with low emergy production to regions with high emergy, it represents emergy savings; X represents the total amount of crops; UEV i,a and UEV i,b represent the emergy consumption of producing unit yield of crop i in Province a and Province b respectively, and UEVN represents the non-renewable emergy amount of unit yield of crop. TV i,a,b represents the trade volume of crop i from Province a to Province b.

[0088] The UEV global emergy baseline adopted by the emergy system of food production and trade is 12.0×10 24 sej / year, and part of the data based on the baseline of 15.83×10 24 sej / year is converted to the baseline selected in this paper by multiplying by 0.76. According to the renewable coefficient (RNFs), the input emergy of the system is divided into a renewable part (R) and a non-renewable part (N). The numerical data of unit emergy and renewable factor are as follows in the table:

[0089]

[0090]

[0091] Step 6: Calculate each evaluation index based on the regional supply-demand differences and changes in trade patterns generated by various optimized planting structure plans, and combine the entropy weight-TOPSIS comprehensive evaluation model to realize the selection of plans and determine the optimal planting layout suitable for regional development. Among them, the specific implementation steps of the entropy weight-TOPSIS comprehensive evaluation model include (1) constructing an initial decision matrix; (2) standardizing the decision matrix; (3) calculating the entropy value of each evaluation index; (4) calculating the weight of each index; (5) constructing a weighted decision matrix; (6) determining the positive ideal solution and negative ideal solution of the evaluation plan; (7) calculating the Euclidean distance from the evaluation plan to the positive and negative ideal solutions; (8) calculating the relative closeness of each evaluation plan. The greater the relative closeness, the closer the evaluation plan is to the best ideal plan. Sort the plans according to the value of the relative closeness. Select the plan with the highest relative closeness as the optimal plan, and the comprehensive benefit is the greatest.

[0092] Based on the optimization method proposed in the present invention, an optimized planting structure plan suitable for improving the comprehensive benefit of the region and minimizing the external negative benefit is given, providing theoretical suggestions for the adjustment of the regional planting layout.

[0093] Advantages of the present invention:

[0094] Based on the deficiencies of the regional crop planting structure optimization method in the prior art of the present invention, taking virtual water trade as the connection point, an evaluation index system covering multiple levels of economy, resources and ecology at multiple scales of region, province and country is constructed, comprehensively making up for the deficiencies of the existing regional planting structure optimization research from the perspectives of grain supply and demand balance and trade pattern benefit changes, and providing a method reference for the subsequent research on the adjustment of regional planting layouts.

Claims

1. A method for optimizing crop planting structure based on multi-objectives and multi-scales, characterized in that, It includes the following steps: Step 1: Select the area where the crops are to be planted as the research area for optimizing the crop planting structure, determine the main types of crops to be planted, and construct a simulation model for the growth and production water footprint of regional crops. Step 2: Determine the virtual water trade partners based on the analysis of the crop production water footprint and the regional supply-demand balance, set the trade coefficient, and determine the optimization plan for the regional crop planting structure. Step 3: Construct a regional multi-objective crop planting structure optimization model to achieve the optimal layout adjustment of the regional planting structure under each optimization plan. Step 3 includes two processes: the analysis of the suitability of crop growth space and the construction of a regional multi-objective crop planting structure optimization model. Among them, the analysis of the suitability of crop growth space is a process of making qualitative and quantitative evaluations on the suitability of planting a certain crop in a specific area, which is used to predict the crop production potential and limitations of the region. The suitability of crop growth is measured by the suitability index, which is the product of the membership degree values of each evaluation index and the index weight, and the value ranges from 0 to 1. The evaluation indexes include eight soil factors and two topographic factors. The eight soil factors include soil texture, bulk density, pH value, total nitrogen, available nitrogen, available phosphorus, available potassium, and organic matter. The two topographic factors include elevation and slope. The membership degree values are determined for multiple rating indexes of various crops according to the S-type function, and the contribution rate of each evaluation index to the change of crop water productivity is calculated by the partial least squares regression method, and the contribution rate is used as the index weight of each index. The calculation formula is shown as formula (1). Where: W i is the contribution rate of the i-th evaluation index factor to the change in crop water productivity; VIP i is the variable projection importance of the i-th index factor; R represents the multiple correlation coefficient of the PLS regression model, calculated by SPSS software; n represents the number of evaluation indexes in the PLS regression model; The suitability of each crop is divided into 5 grades by the natural breakpoint method. The higher the grade, the lower the suitability. The regional multi-objective crop planting structure optimization model is constructed based on the cellular automata and SWAT models. The grid cells for crop planting are used as cells, and the grid cell matrix of the entire study area is used as the cell space. Moore-type neighbors, that is, the grids in 8 directions around the central cell, are used as cell neighbors. The conversion rules are used as the basis for determining the next state of each cell unit, considering the constraints set in the model. The regional multi-objective crop planting structure optimization model takes enhancing the economic water productivity EWP as the optimization objective at the economic level, reducing the blue water footprint BWF rate as the optimization objective at the resource level, and reducing the grey water footprint GWF grey as the optimization objective at the ecological benefit level; where BWF rate is the ratio of the blue water footprint of crop production to the total crop water footprint. The smaller the BWF rate the less the dependence of crop production on agricultural irrigation, that is, the higher the utilization efficiency of water resources. Normalization processing is used to solve the problem of inconsistent dimensions of multiple objective functions in multi-objective programming; Step 4: Quantify the impact of the change in the regional planting structure on the provincial food supply-demand balance state and the change in the inter-provincial food trade pattern. Step 5: Combine multi-level benefit indexes to construct a comprehensive evaluation index system based on the scales of regions, provinces, and the country. In step 5, a multi-level benefit evaluation index system based on three scales of region, province, and country is established to select the optimal solution from different trade combinations, that is, to select the trade coefficient combination with the greatest benefit within the region and the least negative impact on food trade outside the region; the top layer of the evaluation index system is the target layer, that is, different trade combination schemes in the region; the second layer is the criterion layer, including regional benefit, provincial benefit, and national benefit; the third layer is the index layer, which is composed of relevant benefit indicators at each scale. Specifically, three indicators of economic income, blue water dependence degree, and grey water footprint are used to represent regional benefit indicators; food trade income and environmental sustainability index are used as provincial benefit indicators; three indicators of total transportation cost, trade emergy savings, and trade non-renewable emergy savings are used as national benefit indicators; the economic income is calculated by the crop yield and the net income per unit yield, and the grey water footprint GWF grey is obtained by simulating with the SWAT-WF model, and the blue water dependence degree is calculated according to formula (5); the food trade income is quantified by the difference in food trade turnover in the provincial region, and the total transportation cost is calculated by the product of the freight rate and the volume of goods transported; the environmental sustainability index, trade emergy savings, and trade non-renewable emergy savings are related to the calculation of the emergy system. The specific indicator meanings and calculation formulas are shown in formulas (12)-(14): Where: ESI represents the environmental sustainability index, EYR and ELR represent the system output rate and environmental load rate respectively, U, P, and S represent the total emergy consumption, purchased resources, and service resources of crop production, R and N represent the renewable and non-renewable parts of emergy input; U saving and UN saving represent the emergy savings and non-renewable emergy savings of the trade pattern of multiple crops respectively. When crops are traded from regions with low emergy production to regions with high emergy, it represents emergy savings; X represents the total crop quantity, UEV i,a and UEV i,b represent the emergy consumption of producing unit yield of crop i in province a and province b respectively, UEVN represents the non-renewable emergy quantity of unit yield of crop; TV i,a,b represents the trade volume of crop i from province a to province b. Step 6: Based on the constructed comprehensive evaluation index system and the entropy weight-TOPSIS comprehensive evaluation model, select the optimized planting structure plan and determine the optimal planting layout suitable for regional development.

2. The multi-objective and multi-scale based crop planting structure optimization method according to claim 1, characterized in that Step 1 includes selecting the research area for optimizing the crop planting structure, determining the main crop planting types and research years in the research area; collecting and sorting out the spatial attribute data and agricultural management measure data in the research area, constructing a SWAT-WF distributed hydrological model that simultaneously simulates crop growth and crop production water footprint, and using a multi-process calibration program based on the NSGA-II algorithm to calibrate and verify the hydrological process and crop growth process parameters of the SWAT model. The spatial attribute data includes digital elevation data, soil data, land use data, and meteorological data. The agricultural management measure data includes crop planting systems and irrigation schedules.

3. The multi-objective and multi-scale based crop planting structure optimization method according to claim 1, wherein, Step 2 determines the supply-demand status of crops based on the supply and actual demand of the main food crops in the study area; where the supply is obtained by simulating with the SWAT model, and the demand is calculated based on the local actual population and food security guarantee settings; according to the simulation results of the crop production water footprint, select the crop types with oversupply and high production water footprint in the study area as the objects of virtual water trade; on the basis of ensuring regional food security, by setting the trade coefficients of each trade object, multiple trade combination schemes are formed; the trade coefficient varies between 0 and 1, with an increment of 0.05, which refers to the reduction rate of the planting area exceeding the regional demand, that is, as the trade coefficient of a crop increases, the planting proportion of this crop in the study area decreases, and at the same time the yield decreases, but the oversupply state in the region will not change until the trade coefficient equals 1, reaching the supply-demand balance; by adjusting the trade coefficient, the planting proportion of crops with high regional production water footprint is reduced, and the crop planting structure in the region is changed.

4. The multi-objective and multi-scale based crop planting structure optimization method according to claim 1, wherein In the regional multi-objective crop planting structure optimization model, the expressions of the multi-objective optimization equation and normalization are shown in formulas (2)-(3): where: f k,min and f k,max represent the minimum and maximum values of the objective function f(x), respectively; λ k represents the weight of F k (x), which is determined by the analytic hierarchy process; the value of n is equal to 3, where the objective function of the economic water productivity EWP seeks the maximum value, and the objective functions of the blue water footprint BWF rate and the grey water footprint GWF grey seek the minimum value; The grey water footprint GWF grey is obtained by simulating with the SWAT-WF model; the economic water productivity EWP and the blue water dependence BWF rate The calculation formulas for these two indicators are shown in Formulas (4)-(5): where the unit of EWP is yuan / m 3 ; Y is the crop yield per unit area simulated by the model, with the unit of kg / ha; AET is the actual evapotranspiration simulated by the model, with the unit of mm; NP is the net profit per unit of output, with the unit of yuan / kg, and the data is from the statistical yearbook; BWF blue represents the blue water footprint of crop production, and WF represents the total water footprint of crop production, both of which are obtained by simulating with the SWAT-WF model.

5. The method for optimizing crop planting structure based on multi-objective and multi-scale according to claim 4, characterized in that The regional multi-objective crop planting structure optimization model takes the crop planting area, irrigation water demand, and crop planting suitability as constraint conditions; among them, the crop planting area takes the minimum planting area of each type of crop and food security limit as the minimum crop planting area limit, and at the same time the total crop planting area does not exceed the total area of existing cultivated land, which is expressed as shown in formulas (6)-(7): TN x ×(1-α x )≤CN x (7), Where, CN x is the total number of grids of crop x, X represents the total number of crop types, TN is the number of planting grids of all crops in the existing crop planting structure, and TN x is the total number of grids of crop x in the existing planting structure, and α x is the allowable change range of the crop planting area of crop x; Set the irrigation constraint so that the irrigation water demand of the crops in the optimized planting structure cannot exceed the irrigation water demand of the crops in the existing planting structure. The irrigation water demand of the crops is the difference between the water evapotranspiration during the crop growth period and the effective precipitation, and the evapotranspiration refers to evaporation and transpiration; express the irrigation constraint as shown in formula (8): Where M and N are the total number of rows and columns of the grid generated by the planting distribution in the study area; K mn is the crop planting coefficient, indicating whether there is crop planting in a grid cell. If there is a crop planted in a cell, the crop planting coefficient is 1, otherwise it is 0; represents the irrigation water demand of crop x at the cell of row m and column n; TI is the crop irrigation water demand of the existing planting structure; Set the crop suitability constraint so that each crop must be planted on the cultivated land grid with a crop suitability level less than 4. At the same time, take the basic constraint condition that only one crop is planted in each grid cell, which is expressed as shown in formula (9): In the formula, is the suitability level of crop x at the m-th row and n-th column; The conversion rules of the regional multi-objective crop planting structure optimization model include two parts: judgment and conversion. A random function is used to ensure that the crop conversion probability is equal for each grid in the study area. Specifically, first, it is judged whether the grid is converted into another crop, and this process is based on the minimum requirements for the planting area of a specific crop, suitability, and neighborhood status; without considering the minimum requirements for the planting area, the grid is converted into any crop that meets the suitability conditions; then, the conversion priority of these crops in the same unit is determined through the objective function, that is, the crop planted in the grid is converted into the crop with the highest comprehensive benefit; according to the random order, the conversion rules are iterated to each grid unit until all cells are iterated, and the actual irrigation water requirement after the global crop planting structure is optimized is calculated. When the actual irrigation water requirement after optimization is less than the irrigation water requirement of the existing planting structure, the optimization is completed; otherwise, the crops in the grid with a large irrigation water requirement are replaced. The replacement principle is to rank the comprehensive benefits of the crops with a smaller irrigation amount than the currently planted crop, and select the crop with the largest comprehensive benefit among them to replace it in the existing grid; the grid crop replacement is carried out in the order of the irrigation water requirement from large to small until the irrigation water requirement is less than the irrigation water requirement of the actual planting structure. Bring the various trade combination plans generated in step 2 into the regional multi-objective crop planting structure optimization model to obtain the optimal planting structure and the corresponding regional optimal benefit indicators under each combination plan.

6. The multi-objective and multi-scale based crop planting structure optimization method according to claim 1, characterized in that In step 4, different food outputs are generated under various planting structure optimization plans, and it is necessary to re-quantify the food supply-demand balance state of the provincial region and the inter-provincial food trade pattern. Specifically, the food supply-demand balance quantification formula for each provincial administrative region is expressed as shown in formula (10): T i,j = G i,j + (im i,j - ex i,j ) - C i,j (10), where \(i\) and \(j\) represent crop type and provincial administrative region respectively; \(T\) i,j represents the total surplus of crop \(i\) in province \(j\), and \(T\) i,j > 0 indicates that there is an output of crop \(i\) in province \(j\), representing supply exceeding demand, and i,j < 0 indicates that there is an input of crop \(i\) in province \(j\), representing demand exceeding supply, while i,j = 0 indicates that there is no inter-provincial trade of crop \(i\) in province \(j\), representing supply-demand balance; \(G\) i,j and \(C\) i,j represent the total production and consumption of crop \(i\) in province \(j\) respectively; im i,j and ex i,j represent the total import volume and total export volume of crop i in Province j, respectively.

7. A multi-objective and multi-scale based crop planting structure optimization method according to claim 6, characterized in that, Take the total crop production of each province in China as the supply of the crop, and the data source is the "China Statistical Yearbook"; the consumption data of each crop comes from the Brick Agricultural Database, and the data is downscaled to the provincial level using the consumption proportion of each region; the food consumption is divided into rations consumption, feed grain consumption, industrial grain consumption, seed grain consumption, and food loss, and the supply-demand state of each grain crop in each province is determined through the calculation of supply and consumption data.

8. A method for optimizing crop planting structure based on multi-objective and multi-scale according to claim 7, characterized in that, Quantify the inter-provincial food trade pattern through a multi-objective linear optimization model. Specifically, the global minimum of food transportation costs and dietary structure differences is used as the objective function, and the conservation of food surplus and deficit is used as the constraint condition. Among them, the dietary structure difference is characterized by the Euclidean distance between the rations and animal product consumption in each province, and the food transportation cost takes into account both the transportation distance and the transportation method; the objective function and constraint conditions of the multi-objective linear optimization model are expressed as shown in formula (11): where t i,a,b and s i,a,b represent the transportation cost and dietary structure difference between Province A and Province B respectively; M and N represent the number of exporting provinces and importing provinces of Crop i respectively; x i,a,b represents the total transportation volume of Crop i between Province A and Province B.

9. A method for optimizing crop planting structure based on multi-objectives and multi-scales according to claim 1, characterized in that, In step 6, calculate various evaluation indicators based on the regional supply-demand differences and trade pattern changes generated by various planting structure optimization plans, and combine the entropy weight-TOPSIS comprehensive evaluation model to determine the optimal planting layout suitable for regional development; according to the value of the relative closeness, rank various planting structure optimization plans, and select the plan with the highest relative closeness as the optimal plan.

Images