Land use pattern prediction and optimization method considering non-point source pollution control

Abstract

The invention discloses a method for predicting and optimizing land use patterns considering non-point source pollution control, belonging to the technical field of non-point source pollution control methods. Through the dynamic feedback relationship between social and economic indicators and land use changes, the invention constructs a system dynamics model of land use changes, analyzes and predicts the characteristics of land use pattern changes under the influence of social and economic indicators, introduces fuzzy numbers on the basis of the system dynamics model of land use changes to represent the different pollutant output coefficients of different land use types in the study area, reflects the uncertainty characteristics of non-point source pollution output, estimates the area of different land uses and the total load of non-point source pollutants; couples the interval linear programming and fuzzy parameter programming methods to construct an interval fuzzy linear programming model, sets constraints on the basis of maximizing the economic benefits of the land system, and obtains an optimized land use pattern plan that meets the needs of regional social and economic development and achieves specific non-point source pollution control reduction targets.

Description

Technical Field

[0001] The invention relates to the technical field of non-point source pollution control methods, in particular to a land use pattern prediction and optimization method considering non-point source pollution control. Background Art

[0002] At present, non-point source pollution caused by land use and its changes can no longer be ignored. Non-point source pollution has become one of the main water environment problems. Changes in land use patterns affect the output load of sediments and nutrients. Inappropriate land use allocation may cause the nutrient load to exceed the environmental carrying capacity, causing the non-point source pollution to continue to deepen, seriously threatening the safety of water resources and water environment. In particular, the land use pattern in the water source area will directly affect the water quality in the downstream area. The land use pattern needs to be optimized in combination with the reduction of non-point source pollution and the protection of the water environment.

[0003] On the one hand, in the actual control and management of non-point source pollution and the process of land use change, there are complex and multiple uncertainties, including uncertainties in system parameters, model methods and the relationships between various elements of the system. The sources of parameter uncertainty are very extensive. For example, under the interaction of multiple factors such as precipitation, temperature, soil and human activities, non-point source pollution has the characteristics of randomness, extensiveness and latency. The pollutant output coefficient expressed by a certain empirical value is often difficult to reflect the actual characteristics and conditions of non-point source pollution.

[0004] On the other hand, in the process of optimizing land use pattern, there are multiple uncertainties from various social and economic factors. For example, the regional fiscal situation changes differently every year, and people's investment preferences are also mainly affected by subjectivity. At the same time, parameters such as the output load of non-point source pollution per unit area are affected by natural conditions and human activities, such as temperature, rainfall, soil physical and chemical properties, planting patterns, population and road hardening, which have multiple uncertainties that cannot be reflected by constants. Many uncertain factors make the changes in economic output and input costs of various land use types have complex uncertainties.

[0005] There are complex and multiple uncertainties in the process of land use change and non-point source pollution output. At present, the research on land use pattern prediction and optimization considering non-point source pollution control has not paid enough attention to the uncertainty issue; in addition, many land use pattern prediction studies use models and algorithms such as Markov chains, FLUES and machine learning, or use time series remote sensing data such as MODIS to predict and analyze future land use patterns; however, these methods have a high time cost, the required data are difficult to obtain, and the dynamic feedback relationship between land use change and socio-economic factors is rarely reflected. Summary of the invention

[0006] The purpose of the present invention is to provide a method for predicting and optimizing land use patterns considering non-point source pollution control, so as to solve the problems raised in the above-mentioned background technology.

[0007] To solve the above technical problems, the present invention provides the following technical solutions: A method for predicting and optimizing land use patterns considering non-point source pollution control, the method comprising: a land use change system dynamics model, a non-point source pollution estimation module, and an interval fuzzy linear programming model. The land use change system dynamics model is based on the system dynamics model. By determining the linear fitting mathematical relationship between social and economic indicators and land use transfer probabilities, a land use change system dynamics model is constructed. After obtaining the land use areas under different scenarios, they are input into the non-point source pollution estimation module. The non-point source pollution estimation module estimates the total load of land pollutants of different land use types. The interval fuzzy linear programming model aims to maximize the economic benefits of the land use system, and under the constraint conditions, solves the interval fuzzy linear programming model through a two-step interactive algorithm to obtain the optimized land use pattern and decision-making suggestions for different scenarios.

[0008] The land use change system dynamics model includes state variables, rate variables, and auxiliary variables. The state variables are the main output results of the model. The rate variables represent the change rates of the state variables in each time period. The auxiliary variables then cause corresponding changes in the output values of the state variables by affecting the rate variables. The state variables are the areas of various land use types at a certain moment. The rate variables are the land use transfer probabilities, and the auxiliary variables are social and economic indicators closely related to land use changes.

[0009] The grey relational analysis method, principal component analysis, and linear regression method are introduced to screen social and economic indicators, calculate land use transfer probabilities, determine the linear fitting mathematical relationship between social and economic indicators and land use transfer probabilities, construct a land use change system dynamics model, and output the land use areas under different scenarios.

[0010] The land use transfer probability is calculated using the raster calculation tool of ArcGIS, and the raster data resolution is 30m. The land use transfer probability calculation formula:

[0011]

[0012]

[0013] P is the land use transfer probability; P ij is the transfer probability of converting the i-th land use type to the j-th land use type in a certain year; x ij is the area of converting the i-th land use type to the j-th land use type in a certain year; x′ iis the area of the \(i\)-th land use type before conversion in a certain year.

[0014] There are many socio-economic indicators affecting land use change. A total of 59 socio-economic indicators were collected in the early stage of the case study. The grey relational analysis method was introduced to identify the key socio-economic indicators with greater influence on land use change. On the premise of satisfying the correlation between variables, the number of socio-economic indicators was reduced. Appropriate variables were selected according to the problem to achieve appropriate simplification, so that the model was easy to understand and analyze, and the modeling cost was reduced.

[0015] The grey relational analysis method was introduced to identify the key socio-economic indicators with greater influence on land use change. The standardized comparison sequences were calculated through the selected socio-economic indicator comparison sequences. The comprehensive land use degree index was calculated as the reference sequence. The correlation coefficients were obtained by calculating the standardized comparison sequences and the reference sequence. Finally, the grey relational degree was obtained. The greater the grey relational degree, the higher the correlation degree between the comprehensive land use degree index and the socio-economic indicators. The socio-economic indicators with a high grey relational degree with the comprehensive land use degree index were used as the auxiliary variables of the system dynamics model. The socio-economic indicators were further screened by the methods of principal component analysis and linear regression, and the collinearity within the indicators was eliminated at the same time.

[0016] Taking the socio-economic indicators as the comparison sequences, the standardized comparison sequences of the socio-economic indicators were calculated. The calculation formula for the standardized comparison sequences:

[0017]

[0018] d l (k) is the standardized comparison sequence; d′ l (k) is the comparison sequence, and the comparison sequence takes the value of the socio-economic indicator; l is the number of indicators in the comparison sequence;

[0019] Calculate the comprehensive land use degree index. The calculation formula:

[0020]

[0021] La is the comprehensive land use degree index; B i is the land use degree classification index of the \(i\)-th level; C i is the percentage of the area classified at the \(i\)-th level of land use degree;

[0022] The land use degree in the present invention is divided into 3 levels, namely Level 1: forest land, water area and grassland; Level 2: cultivated land; Level 3: construction land;

[0023] Calculate the standardized reference sequence of the comprehensive land use degree index. The calculation formula:

[0024]

[0025] d 0 (k) is the standardized reference sequence of the comprehensive land use degree index; La(k) is the reference sequence, and the reference sequence takes the value of the comprehensive land use degree index.

[0026] Calculate the correlation coefficient, and the calculation formula is:

[0027]

[0028] ξ 0l (k) is d l For d 0 The correlation coefficient at time k, p is the resolution coefficient, and its value range is 0 to 1. The larger p is, the smaller the resolution is, and vice versa;

[0029] Calculate the grey correlation degree through the correlation coefficient, and the calculation formula is:

[0030]

[0031] r 0l is the grey correlation degree. The larger the correlation degree value is, the higher the correlation degree between the reference sequence and the comparison sequence is.

[0032] Perform principal component analysis on the selected socio-economic indicators. The KOM value is 0.786, and the Bartlett's spherical test is 0.000, indicating that the selected indicators are suitable for principal component analysis. Further screen the socio-economic indicators through principal component analysis and linear regression methods, and at the same time eliminate the collinearity within the indicators.

[0033] Lookup F1=(2000, 38), (2005, 40), (2010, 42), (2015, 38), (2020, 59)

[0034] Lookup F2=(2000, 6.15), (2005, 4.34), (2010, 4.42), (2015, 4.72), (2020, 5.65)

[0035] Lookup F3=(2000, 1860), (2005, 3684), (2010, 4658), (2015, 4812), (2020, 3436)

[0036] Lookup F4=(2000, 2836), (2005, 2370), (2010, 3317), (2015, 4263), (2020, 5850)

[0037] Lookup F5 = (2000, 13339), (2005, 12476), (2010, 12063), (2015, 11970), (2020, 8700)

[0038] F1 to F5 are five socioeconomic indicators. The annual data of socioeconomic indicators are sourced from statistical yearbooks and statistical bulletins on national economic and social development.

[0039] By determining the linear fitting mathematical relationship between socioeconomic indicators and land use transfer probabilities, a system dynamics model of land use change is constructed. The mathematical function relationships in the model are as follows:

[0040] CRL(t) = CRL(0) + [T5(t) × FL(0) + T9(t) × GL(0) + T13(t) × WA(0) + T17(t) × COL(0) - (T1(t) + T2(t) + T3(t) + T4(t)) × CRL(0)] × dt

[0041] FL(t) = FL(0) + [T1(t) × CRL(0) + T10(t) × GL(0) + T14(t) × WA(0) + T18(t) × COL(0) - (T5(t) + T6(t) + T7(t) + T8(t)) × FL(0)] × dt

[0042] GL(t) = GL(0) + [T2 × CRL(0) + T6 × FL(0) + T15 × WA(0) + T19 × COL(0) - (T9 + T10 + T11 + T12) × GL(0)] × dt

[0043] WA(t) = WA(0) + [T3 × CRL(0) + T7 × FL(0) + T11 × GL(0) + T20 × COL(0) - (T13 + T14 + T15 + T16) × WA(0)] × dt

[0044] COL(t) = COL(0) + [T4 × CRL(0) + T8 × FL(0) + T12 × GL(0) + T16 × WA(0) - (T17 + T18 + T19 + T20) × COL(0)] × dt

[0045] T1 = 0.1785 × F2 - 0.7786

[0046] T2 = 0.0064 × F1 - 0.2476

[0047] T3 = 0.0003 × F1 - 0.0116

[0048] T4 = 0.0022 × F1 - 0.0754

[0049] T5 = 0.0286×F2 - 0.1247

[0050] T6 = 0.0037×F1 - 0.1428

[0051] T7 = 0.0001×F1 - 0.0049

[0052] T8 = -7.876×10 -7 ×F5 + 0.0106

[0053] T9 = 0.0041×F1 - 0.1467

[0054] T10 = 0.0294×F1 - 1.0583

[0055] T11 = 0.0001×F1 - 0.0057

[0056] T12 = 3.4971×10 -6 ×F4 - 0.0074

[0057] T13 = 0.0941×F2 - 0.4123

[0058] T14 = 0.1139×F2 - 0.0001×F3

[0059] T15 = 0.0051×F1 - 0.1935

[0060] T16 = 0.0027×F1 - 0.0982

[0061] T17 = 0.2219×F2 - 0.9736

[0062] T18 = 0.0476×F2 - 0.2056

[0063] T19 = 0.0047×F1 - 0.1811

[0064] Lookup T20 = (2000, 0.0010), (2005, 0), (2010, 0.0121), (2015, 0), (2020, 0.0046)

[0065] CRL(t), FL(t), GL(t), WA(t) and COL(t) are the areas of cultivated land, forest land, grassland, water area and construction land at time t respectively, and CRL(0), FL(0), GL(0), WA(0) and COL(0) are the initial areas of cultivated land, forest land, grassland, water area and construction land respectively. T1 to T20 are the land use transfer probabilities;

[0066] The non-point source pollution estimation module is based on the output coefficient model, introducing fuzzy numbers to represent the different pollutant output coefficients of different land use types in the study area to reflect the uncertainty characteristics of non-point source pollution output, and introducing interval numbers to represent the area of different land uses and the total load of non-point source pollutants. The calculation formula is:

[0067]

[0068] is the total load of the rth pollutant; i is the land use type; is the output coefficient of the rth pollutant in the ith land use type; is the area of the ith land use type;

[0069] The interval fuzzy linear programming model aims to maximize the economic benefits of the land use system;

[0070] The calculation formula of the objective function is:

[0071]

[0072] is the economic output per unit area of the ith land use type in the sth sub-region; is the input cost per unit area of the ith land use type in the sth sub-region.

[0073] The constraint conditions of the interval fuzzy linear programming model include non-point source pollution load constraint, food production constraint, land use area supply constraint and non-negativity constraint;

[0074] The calculation formula of the non-point source pollution load constraint is:

[0075]

[0076] is the output coefficient of the rth pollutant; is the total load of the rth pollutant;

[0077] The calculation formula of the food production constraint is:

[0078]

[0079] is the food crop yield per unit area; OGP ± is the total food crop production.

[0080] The calculation formula of the land use area supply constraint is:

[0081]

[0082]

[0083] MIA is and MAA is are the minimum and maximum available supplies of the areas of different land use types respectively, and this value is determined by the land use area prediction results and the land use current situation area in 2020 (the base year); TA ± is the total area of the study area;

[0084] Non - negative constraint calculation formula:

[0085]

[0086] is the area of the i - th land use type in the s - th sub - region.

[0087] Solve the interval - valued fuzzy linear programming model through a two - step interactive algorithm to obtain the optimized land use pattern schemes and decision - making suggestions under different scenarios; the fuzzy parameters in the model are determined by the triangular fuzzy membership function. Denote the triangular fuzzy number where b is the most likely value, and a and c are the minimum and maximum possible values respectively. In this technical solution, by introducing the α - cut set, the fuzzy parameters are transformed into interval values, The α - cut set of

[0088] A(α) = [AL(α), AR(α)], 0 ≤ α ≤ 1

[0089] A L (α) = a+(b - a)×α is the left - hand endpoint of the α - cut set; A R (α) = c-(c - b)×α is the right - hand endpoint of the α - cut set.

[0090] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0091] Through the dynamic feedback relationship between social - economic indicators and land use change, construct a system dynamics model of land use change, analyze and predict the characteristics of land use pattern change under the influence of social - economic indicators, making the simulation and prediction process of land use pattern more in line with the actual regional development. The social - economic data required in the model is usually easy to obtain, and the system dynamics model of land use change is easy to operate and has high practicability;

[0092] The non - point source pollution estimation module introduces fuzzy numbers on the basis of the system dynamics model of land use change to represent the different pollutant output coefficients of different land use types in the study area to reflect the uncertainty characteristics of non - point source pollution output, and introduces interval numbers to represent the areas of different land uses and the total load of non - point source pollutants;

[0093] Based on the system dynamics model of land use change, the interval linear programming and fuzzy parameter programming methods are coupled to construct an interval fuzzy linear programming model. Constraints are set on the basis of maximizing the economic benefits of the land system, including the non-point source pollution load reduction target, regional grain output, and the supply area of various land uses. The optimized land use pattern obtained by this model can simultaneously meet the needs of regional social and economic development and achieve specific non-point source pollution control and reduction targets;

[0094] The interval fuzzy linear programming model introduces intervals and fuzzy parameters, effectively reflecting the uncertainty characteristics of various variables in actual case studies. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0096] Figure 1 is a schematic structural diagram of the method for predicting and optimizing the land use pattern considering non-point source pollution control of the present invention;

[0097] Figure 2 is the land use transfer table of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0098] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the 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 of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0099] Please refer to Figure 1 , the present invention provides the following technical solutions:

[0100] Embodiment 1: A method for predicting and optimizing the land use pattern considering non-point source pollution control, the method includes: a system dynamics model of land use change, a non-point source pollution estimation module, and an interval fuzzy linear programming model. The system dynamics model of land use change is based on the system dynamics model. By determining the linear fitting mathematical relationship between social and economic indicators and land use transfer probabilities, a system dynamics model of land use change is constructed. After obtaining the land use areas under different scenarios, they are input into the non-point source pollution estimation module. The non-point source pollution estimation module estimates the total load of land pollutants of different land use types. The interval fuzzy linear programming model aims to maximize the economic benefits of the land use system and solves the interval fuzzy linear programming model through a two-step interactive algorithm under the constraint conditions to obtain the optimized land use pattern and decision-making suggestions for the region under different scenarios.

[0101] The system dynamics model of land use change includes state variables, rate variables, and auxiliary variables. State variables are the main output results of the model. Rate variables represent the change rate of state variables in each time period. Auxiliary variables, by affecting rate variables, cause corresponding changes in the output values of state variables. State variables are the areas of various land use types at a certain moment. Rate variables are the land use transfer probabilities. Auxiliary variables are socio-economic indicators closely related to land use change.

[0102] The grey relational analysis method, principal component analysis, and linear regression method are introduced to screen socio-economic indicators, calculate land use transfer probabilities, and construct a system dynamics model of land use change by determining the linear fitting mathematical relationship between socio-economic indicators and land use transfer probabilities, and output the land use areas under different scenarios.

[0103] The land use transfer probability is calculated using the raster calculation tool of ArcGIS. The raster data resolution is 30m. The formula for calculating the land use transfer probability is as follows:

[0104]

[0105]

[0106] P is the land use transfer probability; P ij is the transfer probability of the i-th land use type converted to the j-th land use type in a certain year; x ij is the area of the i-th land use type converted to the j-th land use type in a certain year; x′ i is the area of the i-th land use type before conversion in a certain year.

[0107] There are many socio-economic indicators affecting land use change. A total of 59 socio-economic indicators were collected in the early stage of the case study. The grey relational analysis method was introduced to identify the key socio-economic indicators that have a greater impact on land use change. On the premise of meeting the correlation between variables, the number of socio-economic indicators was reduced, and appropriate variables were selected according to the problem to achieve appropriate simplification, so as to make the model easy to understand and analyze and reduce the modeling cost.

[0108] The grey relational analysis method is introduced to identify the key socio-economic indicators that have a greater impact on land use change. The standardized comparison sequences are calculated from the selected socio-economic indicator comparison sequences, and the comprehensive land use degree index is calculated as the reference sequence. The correlation coefficients are obtained by calculating the standardized comparison sequences and the reference sequence, and finally the grey relational degree is obtained. The greater the grey relational degree, the higher the correlation between the comprehensive land use degree index and the socio-economic indicators. The socio-economic indicators with a high grey relational degree with the comprehensive land use degree index are used as auxiliary variables in the system dynamics model. The socio-economic indicators are further screened by principal component analysis and linear regression methods, and the collinearity within the indicators is eliminated at the same time.

[0109] Taking the socio-economic indicators as the comparison sequences, calculate the standardized comparison sequences of the socio-economic indicators. The calculation formula for the standardized comparison sequences is:

[0110]

[0111] d l (k) is the standardized comparison sequence; d l (k) is the comparison sequence, and the comparison sequence takes the values of the socio-economic indicators; l is the number of indicators in the comparison sequence;

[0112] Calculate the comprehensive land use degree index. The calculation formula is:

[0113]

[0114] La is the comprehensive land use degree index; B i is the land use degree classification index of the i-th level; C i is the percentage of the area classified at the i-th level of land use degree; The land use degree in the present invention is divided into 3 levels, namely Level 1: forest land, water area and grassland; Level 2: cultivated land; Level 3: construction land;

[0115] Calculate the standardized reference sequence of the comprehensive land use degree index. The calculation formula is:

[0116]

[0117] d 0 (k) is the standardized reference sequence of the comprehensive land use degree index; La(k) is the reference sequence, and the reference sequence takes the values of the comprehensive land use degree index.

[0118] Calculate the correlation coefficient. The calculation formula is:

[0119]

[0120] ξ 0l (k) is d l For d0 The correlation coefficient at time k, where p is the discrimination coefficient with a value range of 0 to 1. The larger p is, the smaller the resolution, and vice versa.

[0121] Calculate the grey relational degree through the correlation coefficient. The calculation formula is:

[0122]

[0123] r 0l is the grey relational degree. The larger the relational degree value, the higher the degree of correlation between the reference sequence and the comparison sequence.

[0124] Perform principal component analysis on the selected socio-economic indicators. The KOM value is 0.786, and Bartlett's spherical test is 0.000, indicating that the selected indicators are suitable for principal component analysis. Further screen the socio-economic indicators through principal component analysis and linear regression methods, and at the same time eliminate the collinearity within the indicators.

[0125] Lookup F1 = (2000, 38), (2005, 40), (2010, 42), (2015, 38), (2020, 59)

[0126] Lookup F2 = (2000, 6.15), (2005, 4.34), (2010, 4.42), (2015, 4.72), (2020, 5.65)

[0127] Lookup F3 = (2000, 1860), (2005, 3684), (2010, 4658), (2015, 4812), (2020, 3436)

[0128] Lookup F4 = (2000, 2836), (2005, 2370), (2010, 3317), (2015, 4263), (2020, 5850)

[0129] Lookup F5 = (2000, 13339), (2005, 12476), (2010, 12063), (2015, 11970), (2020, 8700)

[0130] F1 to F5 are five socio-economic indicators. The annual data of the socio-economic indicators are sourced from the statistical yearbook and the statistical communiqué of the national economic and social development.

[0131] By determining the linear fitting mathematical relationship between socio-economic indicators and land use transfer probabilities, construct a system dynamics model of land use change. The mathematical function relationship in the model is as follows:

[0132] CRL(t) = CRL(0) + [T5(t)×FL(0) + T9(t)×GL(0) + T13(t)×WA(0) + T17(t)×COL(0) - (T1(t) + T2(t) + T3(t) + T4(t))×CRL(0)]×dt

[0133] FL(t) = FL(0) + [T1(t)×CRL(0) + T10(t)×GL(0) + T14(t)×WA(0) + T18(t)×COL(0) - (T5(t) + T6(t) + T7(t) + T8(t))×FL(0)]×dt

[0134] GL(t) = GL(0) + [T2×CRL(0) + T6×FL(0) + T15×WA(0) + T19×COL(0) - (T9 + T10 + T11 + T12)×GL(0)]×dt

[0135] WA(t) = WA(0) + [T3×CRL(0) + T7×FL(0) + T11×GL(0) + T20×COL(0) - (T13 + T14 + T15 + T16)×WA(0)]×dt

[0136] COL(t) = COL(0) + [T4×CRL(0) + T8×FL(0) + T12×GL(0) + T16×WA(0) - (T17 + T18 + T19 + T20)×COL(0)]×dt

[0137] T1 = 0.1785×F2 - 0.7786

[0138] T2 = 0.0064×F1 - 0.2476

[0139] T3 = 0.0003×F1 - 0.0116

[0140] T4 = 0.0022×F1 - 0.0754

[0141] T5 = 0.0286×F2 - 0.1247

[0142] T6 = 0.0037×F1 - 0.1428

[0143] T7 = 0.0001×F1 - 0.0049

[0144] T8 = -7.876×10 -7 ×F5 + 0.0106

[0145] T9 = 0.0041×F1 - 0.1467

[0146] T10 = 0.0294×F1 - 1.0583

[0147] T11 = 0.0001×F1 - 0.0057

[0148] T12 = 3.4971×10 -6 ×F4 - 0.0074

[0149] T13 = 0.0941×F2 - 0.4123

[0150] T14 = 0.1139×F2 - 0.0001×F3

[0151] T15 = 0.0051×F1 - 0.1935

[0152] T16 = 0.0027×F1 - 0.0982

[0153] T17 = 0.2219×F2 - 0.9736

[0154] T18 = 0.0476×F2 - 0.2056

[0155] T19 = 0.0047×F1 - 0.1811

[0156] Lookup T20 = (2000, 0.0010), (2005, 0), (2010, 0.0121), (2015, 0), (2020, 0.0046)

[0157] CRL(t), FL(t), GL(t), WA(t) and COL(t) are the areas of cultivated land, forest land, grassland, water area and construction land at time t respectively, CRL(0), FL(0), GL(0), WA(0) and COL(0) are the initial areas of cultivated land, forest land, grassland, water area and construction land respectively, and T1 to T20 are the land use transfer probabilities;

[0158] Please refer to Figure 2: T1 refers to the transfer probability of cultivated land being converted into forest land; T2 refers to the transfer probability of cultivated land being converted into grassland; T3 refers to the transfer probability of cultivated land being converted into water area; T4 refers to the transfer probability of cultivated land being converted into construction land; T5 refers to the transfer probability of forest land being converted into cultivated land; T6 refers to the transfer probability of forest land being converted into grassland; T7 refers to the transfer probability of forest land being converted into water area; T8 refers to the transfer probability of forest land being converted into construction land; T9 refers to the transfer probability of grassland being converted into cultivated land; T10 refers to the transfer probability of grassland being converted into forest land; T11 refers to the transfer probability of grassland being converted into water area; T12 refers to the transfer probability of grassland being converted into construction land; T13 refers to the transfer probability of water area being converted into cultivated land; T14 refers to the transfer probability of water area being converted into forest land; T15 refers to the transfer probability of water area being converted into grassland; T16 refers to the transfer probability of water area being converted into construction land; T17 refers to the transfer probability of construction land being converted into cultivated land; T18 refers to the transfer probability of construction land being converted into forest land; T19 refers to the transfer probability of construction land being converted into grassland; T20 refers to the transfer probability of construction land being converted into water area.

[0159] The non-point source pollution estimation module is based on the output coefficient model, introducing fuzzy numbers to represent the different pollutant output coefficients of different land use types in the study area to reflect the uncertainty characteristics of non-point source pollution output, and introducing interval numbers to represent the area of different land uses and the total load of non-point source pollutants. The calculation formula is:

[0160]

[0161] is the total load of the rth pollutant; i is the land use type; is the output coefficient of the rth pollutant in the ith land use type; is the area of the ith land use type;

[0162] The interval fuzzy linear programming model aims to maximize the economic benefits of the land use system;

[0163] The calculation formula of the objective function:

[0164]

[0165] is the economic output per unit area of the ith land use type in the sth sub-region; is the input cost per unit area of the ith land use type in the sth sub-region.

[0166] The constraint conditions of the interval fuzzy linear programming model include non-point source pollution load constraint, grain yield constraint, land use area supply constraint and non-negativity constraint;

[0167] The calculation formula of the non-point source pollution load constraint:

[0168]

[0169] is the output coefficient of r-class pollutants; is the total load of r-class pollutants;

[0170] Calculation formula for grain yield constraint:

[0171]

[0172] is the grain crop yield per unit area; OGP ± is the total grain crop yield.

[0173] Calculation formula for land use area supply constraint:

[0174]

[0175]

[0176] MIA is and MAA is are the minimum and maximum available supplies of different land use type areas respectively, and this value is determined by the land use area prediction results and the land use current situation area in 2020 (base year); TA ± is the total area of the study area;

[0177] Non-negativity constraint calculation formula:

[0178]

[0179] is the area of the i-th land use type in the s-th sub-region.

[0180] Solve the interval fuzzy linear programming model through a two-step interactive algorithm to obtain the optimized regional land use pattern and decision-making suggestions under different scenarios; the fuzzy parameters in the model are determined by the triangular fuzzy membership function, and the triangular fuzzy number is denoted as where b is the most likely value, and a and c are the minimum and maximum possible values respectively. In this technical solution, by introducing the α-cut set, the fuzzy parameters are transformed into interval values, α-cut set of:

[0181] A(α) = [AL(α), AR(α)], 0 ≤ α ≤ 1

[0182] A L (α) = a + (b - a) × α is the left endpoint of the α-cut set; A R$(α) = c - (c - b)×α$ is the right endpoint of the α-cut set.

[0183] Embodiment 2: Assume that the minimum possible value a, the most likely value b, and the maximum possible value c of the fuzzy parameter F determined according to historical data, etc. are 5, 7, and 10 respectively. Let the α-cut sets be 0.3, 0.6, and 0.9. Then

[0184] When the α-cut set is 0.3,

[0185] A L $(α) = 5 + (7 - 5)×0.3 = 5.6$, A R $(α) = 10 - (10 - 7)×0.3 = 9.1$

[0186] When the α-cut set is 0.6,

[0187] A L $(α) = 5 + (7 - 5)×0.6 = 6.2$, A R $(α) = 10 - (10 - 7)×0.6 = 8.2$;

[0188] When the α-cut set is 0.9,

[0189] A L $(α) = 5 + (7 - 5)×0.9 = 6.8$, A R $(α) = 10 - (10 - 7)×0.9 = 7.3$.

[0190] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0191] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for predicting and optimizing land use patterns considering non-point source pollution control, characterized in that: The method includes: a land use change system dynamics model, a non-point source pollution estimation module, and an interval fuzzy linear programming model. The land use change system dynamics model is based on the system dynamics model. By determining the linear fitting mathematical relationship between social and economic indicators and land use transfer probabilities, a land use change system dynamics model is constructed. After obtaining the land use areas under different scenarios, they are input into the non-point source pollution estimation module. The non-point source pollution estimation module estimates the total load of land pollutants of different land use types. The interval fuzzy linear programming model aims to maximize the economic benefits of the land use system. Under the constraint conditions, the interval fuzzy linear programming model is solved through a two-step interactive algorithm to obtain the optimized land use pattern and decision-making suggestions for the region under different scenarios; The non-point source pollution estimation module is based on the output coefficient model. Fuzzy numbers are introduced to represent the different pollutant output coefficients of different land use types in the study area to reflect the uncertainty characteristics of non-point source pollution output. Interval numbers are introduced to represent the areas of different land uses and the total load of non-point source pollutants. The calculation formula: ； is the total load of pollutant type r; i is the land use type; is the output coefficient of pollutant type r in the i-th land use type; is the area of the i-th land use type; The interval fuzzy linear programming model aims to maximize the economic benefits of the land use system; Objective function calculation formula: ； is the economic output per unit area of the i-th land use type in the s-th sub-region; is the input cost per unit area of the i-th land use type in the s-th sub-region; Non-point source pollution load constraint calculation formula: ； is the output coefficient of pollutant of type r; is the total load of pollutant of type r; Solve the interval fuzzy linear programming model through a two-step interactive algorithm to obtain the optimized regional land use pattern and decision-making suggestions under different scenarios; the fuzzy parameters in the model ( , ) are determined by the triangular fuzzy membership function. Denote the triangular fuzzy number as , where b is the most likely value, and a and c are the minimum and maximum possible values respectively; by introducing the cut set, transform the fuzzy parameters into interval values, of cut set: ； is the left endpoint of the cut set; is the right endpoint of the cut set.

2. The method for predicting and optimizing land use patterns considering non-point source pollution control according to claim 1, characterized in that: The land use change system dynamics model includes state variables, rate variables, and auxiliary variables. The state variables are the areas of various land use types at a certain moment. The rate variables are the land use transfer probabilities. The auxiliary variables are social and economic indicators closely related to land use changes; The grey relational analysis method, principal component analysis, and linear regression method are introduced to screen social and economic indicators, calculate the land use transfer probabilities. By determining the linear fitting mathematical relationship between social and economic indicators and land use transfer probabilities, a land use change system dynamics model is constructed and the land use areas under different scenarios are output.

3. The method for predicting and optimizing land use patterns considering non-point source pollution control according to claim 2, characterized in that: The land use transfer probability calculation formula: i = 1, 2, …, n; j = 1, 2, …, m ； P is the land use transfer probability; is the transfer probability of the i-th land use type converted to the j-th land use type in a certain year; is the area of the i-th land use type converted to the j-th land use type in a certain year; is the area of the i-th land use type before conversion in a certain year.

4. The method for predicting and optimizing land use patterns considering non-point source pollution control according to claim 3, characterized in that: The grey relational analysis method is introduced to identify the key socio-economic indicators that have a greater impact on land use change. The standardized comparison sequences are calculated from the selected socio-economic indicator comparison sequences, and the comprehensive index of land use degree is calculated as the reference sequence. The correlation coefficients are calculated through the standardized comparison sequences and the reference sequence, and finally the grey relational degree is obtained. The greater the grey relational degree, the higher the degree of association between the comprehensive index of land use degree and the socio-economic indicators. The socio-economic indicators with a high grey relational degree with the comprehensive index of land use degree are used as the auxiliary variables of the system dynamics model. The socio-economic indicators are further screened by principal component analysis and linear regression methods, and the collinearity within the indicators is eliminated at the same time.

5. The method for predicting and optimizing land use pattern considering non-point source pollution control according to claim 4, characterized in that: Taking the socio-economic indicators as the comparison sequences, calculating the standardized comparison sequences of the socio-economic indicators, and the calculation formula for the standardized comparison sequences: ； is a standardized comparison sequence of socio-economic indicators; is a comparison sequence, and the comparison sequence takes socio-economic indicators; is the number of indicators in the comparison sequence; calculate the comprehensive index of land use degree, calculation formula: ； is the comprehensive index of land use degree; is the land use degree classification index of the i-th level; is the percentage of the area classified by the land use degree of the i-th level; Calculating the standardized reference sequence of the comprehensive index of land use degree, and the calculation formula: ; is the standardized reference sequence of the comprehensive index of land use degree; is the reference sequence, and the reference sequence takes the value of the comprehensive index of land use degree; Calculating the correlation coefficient, and the calculation formula: ； is for the correlation coefficient at time k, is the discrimination coefficient, and its value range is 0 to 1. The larger it is, the smaller the resolution is, and vice versa. Calculating the grey relational degree through the correlation coefficient, and the calculation formula: ； is the grey relational grade.

6. The method for predicting and optimizing land use pattern considering non-point source pollution control according to claim 5, characterized in that: The constraint conditions of the interval fuzzy linear programming model include non-point source pollution load constraint, food production constraint, land use area supply constraint and non-negativity constraint; The calculation formula for the food production constraint: ； is the grain crop yield per unit area; is the total output of grain crops; The calculation formula for the land use area supply constraint: ; ； and are the minimum and maximum available supplies of the areas of different land use types respectively, and the available supply of this area is determined by the land use area prediction results and the land use status area in 2020; is the total area of the study area; The calculation formula for the non-negativity constraint: ; is the area of the i-th land use type in the s-th sub-region.

7. The method for predicting and optimizing land use pattern considering non-point source pollution control according to claim 6, characterized in that: Calculating the relative error between the system dynamics model of land use change and historical data to test the accuracy of the constructed system dynamics model of land use change, and the calculation formula for the relative error: ; is the relative error between the simulated area and the actual area; is the simulated area of a certain land use type in a certain year by the system dynamics model; is the historical actual area corresponding to the land use type and year.

Images