CN109376955B - Agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function - Google Patents

Agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function Download PDF

Info

Publication number
CN109376955B
CN109376955B CN201811443371.1A CN201811443371A CN109376955B CN 109376955 B CN109376955 B CN 109376955B CN 201811443371 A CN201811443371 A CN 201811443371A CN 109376955 B CN109376955 B CN 109376955B
Authority
CN
China
Prior art keywords
target
research area
management
swat
cost
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811443371.1A
Other languages
Chinese (zh)
Other versions
CN109376955A (en
Inventor
王晓燕
庞树江
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Capital Normal University
Original Assignee
Capital Normal University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Capital Normal University filed Critical Capital Normal University
Priority to CN201811443371.1A priority Critical patent/CN109376955B/en
Publication of CN109376955A publication Critical patent/CN109376955A/en
Application granted granted Critical
Publication of CN109376955B publication Critical patent/CN109376955B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/02—Agriculture; Fishing; Mining
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

The invention discloses an agricultural non-point source optimal management measure combination optimization configuration method based on an ecological service function, which comprises the following steps: 1) collecting data of a target research area, and constructing a watershed SWAT model of the target research area; 2) screening agricultural non-point source management measures of a target research area according to the gradient condition and the soil hydrological group type; 3) calculating the cost of the existing management measures of the drainage basin where the target research area is located and the cost of the selected agricultural non-point source management measures, and constructing a cost database of the management measures; 3) adjusting key characteristic parameters of the watershed SWAT model, respectively simulating the reduction efficiency of different management measures on target pollutants, and generating a measure implementation ecological benefit database; 4) and calculating the cost database and the measure implementation ecological benefit database by adopting a self-adaptive multi-target genetic algorithm, obtaining the optimal cost benefit curves of different target pollutants, and obtaining the optimal configuration scheme of agricultural non-point source management measures in the target research area.

Description

Agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function
Technical Field
The invention relates to the field of environmental science, in particular to a space combination optimal configuration method of agricultural non-point source optimal management measures, which takes water resources, water quality and economic benefits into consideration.
Background
With the effective control of point source pollution, agricultural non-point source pollution has become an important factor for the water quality of surface water bodies to decline worldwide (Huang J J J et al, 2015; Minet E et al, 2017;). Practice shows that the implementation of the agricultural optimal management measure is an effective means for effectively reducing the agricultural non-point source pollution load and promoting the quality of the watershed water environment to be obviously improved (Sharpley A et al, 2004; Rocha J et al, 2015; Haas M B et al, 2017). However, agricultural non-point source pollution is not only controlled by natural factors such as topography, hydrology, land utilization and soil background values, but also comprehensively acted by agricultural planting, social economy and market factors (Azzellino A et al, 2006; Krause S et al, 2008; Ouyang W et al, 2018). Therefore, the characteristics of the target area to be implemented (including natural factors and socioeconomic factors) are the main reference conditions for selecting and configuring the optimal management measures for agriculture. Therefore, the agricultural optimal management measure configuration needs to comprehensively consider environmental benefits, ecological benefits and social and economic benefits, and the river basin agricultural optimal management measure is changed into a multi-objective comprehensive optimization problem (Maringenti et al, 2009; Panagopoulos et al, 2012; Chen L et al, 2016; Wu H et al, 2017). At present, common methods for space configuration of optimal management measures include an objective configuration method based on expert experience and an optimal configuration method based on an intelligent algorithm. The optimization configuration based on the intelligent algorithm becomes the mainstream practice of the optimal management measure control (Kao J et al, 2003; Maringenti et al, 2011; Qiu J et al, 2018). In the optimal management space optimization configuration, the optimization objective and the optimization algorithm are the key points of whether an effective management measure scheme can be obtained. It has been investigated that the pollutant abatement effects and deployment costs (including implementation costs, maintenance costs and opportunity costs, etc.) of a facility after its implementation are often targeted for optimal management plan deployment (Veith, T.L et al, 2003; Kaini, P et al, 2012;). However, the space-time change of the ecological service function after the measures are implemented is not considered from the whole watershed in the true sense, and some measures (such as terraced fields, returning to cultivation, protective farming and the like) can actually reduce the pollutant load and play a role in purifying the water quality, but at the same time, the measures can intercept the upstream water resource quantity after the measures are implemented, and can reduce the industrial and agricultural water, the domestic water, the ecological landscape water and the like in the downstream area, and further can influence the social and economic development and the ecological quality reduction of the downstream area. Therefore, from the overall benefits of the drainage basin, the ecological appeal of different benefit bodies on the upstream and the downstream must be brought into the optimal agricultural management measure space optimization configuration link, the ecological service function (mainly comprising water quality purification and water yield) is taken as the potential environmental benefit target implemented by the measure optimization configuration scheme, and the two types of target comprehensive optimization of water resource increase and water quality improvement of the drainage basin can be realized.
On the other hand, the selection of the optimization algorithm and the verification of key parameters are important guarantees for obtaining the configuration scheme of the agricultural optimal management measures. The optimization algorithms used in the existing research include fuzzy interval programming algorithm (Dai, C et al 2016), dynamic differential programming algorithm (Hsieh P H et al 2007), genetic algorithm (Ciou, S.K et al 2012) and the like, wherein the genetic algorithm is a biological evolution process evolution model simulating natural selection and genetic mechanism in darwinian biological evolution theory, and the optimal solution of the problem is obtained through generating feasible solutions representing the problem and through multiple iterative operations of selection operators, crossover operators, mutation operators, recombination operators and the like. The algorithm is not limited by a specific problem form (such as continuity and conductability of a model), can directly operate a structural object, and has an internal latent parallel characteristic and good global optimization capability; by adopting a probabilistic optimization mechanism, the search direction can be adaptively adjusted without exact design rules. Genetic algorithms have been widely used in the fields of machine learning, signal processing, adaptive control, combinatorial optimization, artificial intelligence, and the like. The multi-objective combination optimization is a mathematical problem with wide application, and the phenomenon that mutual conflict among targets is difficult to solve by a traditional mathematical method due to the fact that an unavoidable contradiction exists among a plurality of objective functions. The non-dominated sorting genetic algorithm (NSGA-II) introduced with the Pareto solution set concept can properly process mutual exclusivity among different targets (Deb K et al, 2002), and is widely applied to the engineering field, environmental planning governance and water resource scheduling (Bekele E G et al, 2007; Dhanalaksmi S et al, 2011). However, at present, NSGA-ii has a certain problem, mainly focusing on the following 3 aspects, that is, (1) the algorithm adopts the same cross probability and mutation probability, and the NSGA-ii algorithm based on a fixed probability mechanism may cause the algorithm to be premature or difficult to converge; (2) the maximum evolution algebra of the algorithm is a fixed value, so that when the evolution algebra is too small, the algorithm is difficult to obtain an optimal solution, otherwise, excessive optimization is possible to waste computing resources; (3) the population scale is a fixed value, and the population is too small and is easy to fall into a local solution, so that the optimal solution is difficult to find; the search speed is easy to reduce when the population is too large, and the operation load of a computer is increased. Therefore, in order to effectively reduce the influence of human factors on the optimization model, it is necessary to adjust the main parameters (population scale, evolution algebra, cross probability and variation probability) of the genetic algorithm to an adaptive mode which changes along with the individual fitness of the population.
Disclosure of Invention
In order to overcome the defects that the current situation of the ecological service function of the drainage basin level is difficult to effectively improve by the existing agricultural non-point source optimal management measures and the optimization algorithm parameters have large uncertainty, the invention provides an agricultural non-point source optimal management measure combination optimization configuration method based on the ecological service function, so that the integral improvement of the drainage basin ecological service function is realized, and powerful technical support is provided for the comprehensive regulation and control of the water quality and the water quantity of the drainage basin.
The invention comprehensively considers the water quality purification function and the water production capacity, couples the ecological service function with the improved self-adaptive multi-target genetic algorithm (ANSGA-II), is applied to the space optimization configuration of the non-point source optimal management measure of the drainage basin agriculture, and simultaneously realizes the maximization of the water quality purification and water production efficiency and the minimization of the management measure cost.
The technical scheme adopted by the invention for solving the technical problems is as follows:
an agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function comprises the following specific implementation steps:
1) collecting geographic spatial data, meteorological hydrological data, pollution source data and agricultural management information data of a target research area, and constructing a river basin SWAT model of the target research area; collecting water quality sensitivity parameters and range information in a SWAT model in related existing documents by taking water quantity and water quality data as a reference, using the parameters as a parameter database for SWAT model sensitivity parameter analysis, and then performing Latin Hypercube Sampling (LHS) on parameters in the parameter database by using a SWAT-CUP model to obtain random parameter combinations; then substituting the random parameter combination into the SWAT model to observe the change of the analog value and the observed value and the disturbance difference to obtain the global sensitivity grade of the SWAT model, and selecting the significance level (p value)<0.05) as the basis for determining the sensitivity parameters, so as to screen out the sensitivity parameters of the SWAT model; dividing the SWAT simulation period of the target research area into a preheating period, a calibration period and a verification period, setting the theoretical range of the sensitivity parameters, and performing Latin hypercube sampling on the sensitivity parameters by using the SUFI-2 algorithm of the SWAT-CUP modelThen substituting the obtained data into the SWAT model of the target research area to judge the matching degree of the analog value and the observed value, repeating multiple iterations to obtain the parameter combination with the optimal calibration period (the judgment criterion: Nash-Sutcliffe efficiency, NSE) is more than 0.50, and determining the coefficient (R)2) Greater than 0.65); finally, substituting the optimal parameters of the calibration period into the verification period of the SWAT model of the target research area, and when the simulation effect meets the judgment criterion, indicating that the calibration and verification of the sensitivity parameters of the SWAT model are completed;
2) extracting main gradient conditions and soil hydrologic group types in a research area, taking the main gradient conditions and the soil hydrologic group types as query conditions, adopting an agricultural non-point source optimal management measure efficiency evaluation toolbox (Geng R et al, 2015) which is internally provided with published relevant information (including average gradient of implementation point positions, soil hydrologic group, measure radial flow, sediment, granular nitrogen and phosphorus reduction efficiency and dissolved nitrogen and phosphorus reduction efficiency) about agricultural optimal management measures of different terrain units in China, statistically describing the total reduction efficiency of measures of the measure load query conditions (average gradient and soil hydrologic group) (number of measure types meeting the conditions, reduction efficiency (maximum value, minimum value, median, average value, variance and the like), sequencing the reduction efficiency of all measures from high to low, taking the measure with the highest reduction efficiency as an alternative scheme of measure space configuration, preliminarily screening to obtain agricultural non-point source management measures with strong applicability in a research area;
3) combining the actual characteristics of a research area, including agricultural cultivation management modes (the existing agricultural field water conservation measures, crop types and distribution, common fertilizer types, fertilization time and fertilization amount, irrigation water sources and modes, irrigation water consumption, straw management modes and the like), livestock and poultry cultivation management modes (livestock and poultry cultivation types, cultivation water supply modes, livestock and poultry sewage and excrement treatment modes and the like) and domestic sewage management modes (the total area population, the number of agricultural population, the number of urban population, the discharge amount of sewage per capita, the domestic sewage treatment mode and the like); taking a union set of the existing management measures of the drainage basin and the management measures obtained by screening the tool boxes to obtain a drainage basin management measure total library (comprising various measure types and corresponding pollutant reduction efficiency), carrying out research activities on the acceptable degree of drainage basin farmers about the measure types to obtain the acceptable degree indexes of the farmers about different measure types, and selecting measures with high reduction efficiency and strong acceptable degree through Non-dominant Sorting (Non-dominant Sorting) of the pollutant reduction efficiency and the acceptable degree indexes; calculating the implementation cost, maintenance cost and opportunity cost of the selected agricultural optimal management measure unit according to the actual commodity price and labor cost of the implementation area, and constructing various optimal management measure cost information databases;
4) the SWAT model can simulate the reduction efficiency of the management measures on the target pollutants by adjusting key characteristic parameters of the agricultural management measures. Referring to the relevant literature, the characteristic parameters of engineered BMPs commonly used in SWAT models include CN values, USLE-P, USLE-C, OV-N, SLSIBUBBSN, FILTERW, CH-W2, CH-D, CH-N2, CH-COV, CH-EROD, and CH-S2. On the basis of the key characteristic parameters, acquiring ecological efficiency data after different measures are implemented by adjusting parameter values in a SWAT model, for example, the slope is used for realizing the simulation of high-altitude planting (Contour planting) measures of 3.00-8.00% of land parcels, the calibrated CN needs to be reduced by 3 units, and the USLE-P is adjusted to be 0.60; the characteristic parameter corresponding to the terrace (Parallel Terraces) measure is that the CN value is reduced by 6 units, and the USLE-P is adjusted to be 0.10; for other types of BMPs, reference is made to the literature (Arabi M et, 2008; Panagopoulos Y et, 2012), which is not repeated here. Repeatedly executing the SWAT model to complete the construction of the measure implementation ecological benefit database;
5) and calculating an optimal management measure cost information database and a measure implementation ecological benefit database by adopting an improved self-adaptive multi-target genetic algorithm ANSGA-II according to a set target function optimization criterion to obtain optimal cost benefit curves of different target pollutants, obtain an optimal agricultural non-point source pollution measure combination scheme, and visually express the optimal scheme configuration by combining a geographic information system. In the invention, the improved calculation flow of the self-adaptive ANSGA-II algorithm is as follows: firstly, randomly generating potential feasible solutions (namely, firstly, carrying out chromosome coding on various measures, adopting an integer coding mode, namely, each measure is represented by an integer to generate a plurality of potential feasible solutions), then substituting the feasible solutions into a fitness function to evaluate the goodness and badness of the fitness value, executing a championship selection operator to select excellent individuals, then utilizing the fitness function value to adjust the cross probability and the variation probability of each individual (the principle is that the excellent individuals are endowed with smaller cross probability and variation probability to ensure the stability of the excellent individuals, the inferior individuals are endowed with larger cross probability and variation probability to ensure that the individual structure can be quickly changed and the whole optimization process is accelerated), then executing cross and variation operations to obtain a new generation population which is crossed and varied, and continuously repeating the selection-cross-variation operation, enabling the generated population individuals to continuously approach the optimal solution; when the distance between Pareto front-end individuals separated by 10 generations is less than a given threshold (0.001), it can be considered that no more optimal individuals are produced, and the algorithm terminates. The algorithm can automatically give out cost-benefit schemes of measure configuration under different scenes and measure types configured by each corresponding land parcel.
Compared with the prior art system, the invention has the beneficial effects that:
the invention effectively relieves the contradiction between agricultural upstream non-point source pollution control and the conflict between industrial, agricultural and ecological water in downstream areas from the perspective of different types of benefits involving upstream and downstream of the drainage basin, and can promote the overall improvement of the ecological environmental benefit of the drainage basin. In addition, the NSGA-II algorithm commonly used in the existing research adopts a fixed probability mechanism to carry out cross and mutation operations, so that the algorithm is easy to mature early and falls into a local optimal solution, and an optimal agricultural non-point source management measure configuration scheme is difficult to obtain; in addition, the convergence of the algorithm is difficult to judge based on fixed evolution algebra; in the invention, the cross probability and the variation probability are adjusted based on a fitness function, the evolution intergeneration difference of a Pareto solution is used as a criterion for judging whether the NSGA-II algorithm completes convergence (the Pareto optimal solution intergeneration difference is less than 0.001 and is used as a threshold value for stopping algorithm evolution), the traditional fixed parameter NSGA-II algorithm is changed into an ANSGA-II algorithm based on parameter self-adaptation, and the specific execution flow of the algorithm is shown as the attached figure 2. As can be seen from the figure, firstly, a swap model is utilized to construct an ecological benefit database and a cost database of measures required for optimization, then an initial feasible solution population (also called chromosome) is generated based on an MATLAB software package, and an objective function is compiled to evaluate the quality of population individuals; and selecting the dominant individual by adopting a championship selection algorithm according to the fitness value, executing crossover and mutation operations (in the crossover and mutation operations, the crossover probability and the mutation probability of the whole body are self-adapted by utilizing the fitness value of the individual), and then continuously repeating the calculation steps until the Pareto front-end individual is not generated by the optimization algorithm, namely when the distance between the Pareto front-end individuals separated by 10 generations is smaller than a given threshold value (0.001), indicating that the optimization process is converged and the algorithm is terminated. Therefore, by changing the working mechanism of the optimization algorithm, the acquisition of the agricultural optimal management measure space configuration scheme is realized, the calculation time in the optimization process is effectively reduced, the configuration influence on the optimal scheme of the drainage basin management measure due to the improper parameter acquisition is effectively reduced, and the reliable technical support is provided for the drainage basin water ecological protection.
Drawings
FIG. 1 is a system diagram of the present invention for the combination and optimization of agricultural non-point source optimal management measures based on ecological service function;
fig. 2 is a flow chart of an improved parameter adaptation-based ANSGA-ii algorithm.
Detailed Description
The specific implementation of the present invention is specifically described below with reference to fig. 2. From the whole, the method comprises three main parts, namely watershed ecological service function evaluation, agricultural non-point source optimal management measure initial selection and optimal management measure multi-objective optimization configuration based on ecological service function.
(1) Watershed ecological service function assessment
The water producing function and the water quality purifying function are two most important types in the drainage basin ecological service function. In the invention, a SWAT model is adopted to evaluate the water yield of a drainage basin block unit and the space-time difference of agricultural non-point source pollutant load. This link can be roughly divided into the following 4 links.
Figure BDA0001885208490000051
FoundationPreparing data, selecting a specific case research area, and determining a specific space boundary range of a watershed; constructing a spatial database containing information such as the landform and the landform of a drainage basin, the land utilization type, the soil type and the like; establishing a meteorological hydrological database of the research area by collecting historical meteorological data (air temperature, precipitation, relative humidity, solar radiation and relative humidity), surface hydrological data (daily surface runoff) and continuous water quality data (silt, tri-state nitrogen and total phosphorus concentration) of the research area day by day; and (4) constructing an agricultural management database of the research area, wherein the agricultural management database comprises detailed information of main crop planting types, cultivation systems, fertilization methods, fertilization time, fertilization amount, irrigation modes, harvesting modes and the like.
Figure BDA0001885208490000052
After the ArcSWAT 2012 software package is adopted for initializing the basin model to execute the links of sub-basin division, hydrological response unit definition, meteorological database and management information loading and the like, an input file required by the SWAT model simulation is formed, the simulation date of the model is set, and the initialization operation of the SWAT model is completed.
Figure BDA0001885208490000061
The model parameter calibration utilizes a pollutant evaluation tool loadEST model developed by the American geological survey to convert discrete water quality data into continuous daily scale water quality data and obtain pollution load of basin monitoring points under different time scales (daily scale, monthly scale and annual scale). Identifying key sensitivity parameters influencing model output by utilizing a SUFI-2 algorithm of a SWAT-CUP program, wherein the significance level is less than 0.05 as the selection standard of the sensitivity parameters; dividing the whole simulation period into a preheating period, a calibration period and a verification period, wherein the pollutant sequence of model verification is surface runoff, sediment, total nitrogen and total phosphorus; the spatial sequence is upstream and downstream, and the branch flow is followed by the main flow; the time sequence is year scale, month scale and day scale. The judgment of the model simulation precision depends on the selection of Nash efficiency coefficient and decision coefficient R2By continuously adjusting the parameter calibration model, the simulation value is made as much as possibleCan be matched with the observed value. When the Nash coefficient is larger than 0.50, the coefficient R is determined2Greater than 0.65, indicating that the model has passed the verification, the next stage of basin simulation can be performed.
Figure BDA0001885208490000062
And (3) simulating the load and water yield of non-point source pollution generation of the drainage basin under different spatial scales (sub drainage basins and HRUs) by using the verified SWAT model as the basic scene of the evaluation work of the agricultural non-point source pollution optimal management measure.
(2) Initial selection of agricultural non-point source optimal management measures
Taking a BMP Tool box as a basis (Geng R et al, 2015), extracting and calculating the average gradient and the hydrological soil group information of a target drainage basin by using a GIS technology, taking the average gradient and the hydrological soil group information as screening conditions for initial selection of optimal management measures of the drainage basin, screening to obtain management measure types suitable for the drainage basin, sequencing the reduction efficiency of target pollutants by using a statistical analysis function in the BMP Tool box, and selecting the measure type with higher reduction efficiency as a potential selection measure for optimal configuration of next measures.
In order to enhance the applicability of management measures in regions, questionnaires are researched on the basis of efficiency evaluation, the cognitive degree and the acceptable degree of target watershed beneficiaries on optimal management measures and main influence factors are analyzed, the reduction efficiency and the acceptable degree of the management measures are comprehensively weighed, and finally a measure type database which is included in optimal management measure space optimization configuration is determined.
(3) Optimal management measure multi-objective optimization configuration based on ecological service function
The part of the content needs to be developed step by step in 4 parts, namely the construction of an ecological service function effect database of an optimal management measure, the construction of a measure implementation cost database, the improvement and debugging of an ANSGA-II algorithm of self-adaptive parameters and the acquisition of an optimal management measure multi-objective optimization configuration scheme based on an ecological service function, and the specific implementation method is detailed as follows.
Figure BDA0001885208490000063
Ecological effect database the ecological effect data types constructed by the present invention include total nitrogen reduction efficiency, total phosphorus reduction efficiency and water yield increase for optimal management measures. And simulating ecological effects generated by different types of measures by modifying the characteristic parameters of the characterization measures by using the verified SWAT model, and constructing an ecological effect database of the measures, wherein the database comprises information of the number (M) of the land parcels and the types (N) of the measures to form a data table file with M rows and N columns, and the data table file is used as the ecological effect database for optimizing the optimal management measures.
Figure BDA0001885208490000064
The measure cost database takes the commodity price level and the labor cost of the area to be implemented as references, the total cost (including construction cost, maintenance cost and opportunity cost) of the unit measure in the life cycle is determined, and the actual implementation total cost database of each type of measure is obtained by combining the implementation scale level (such as the implementation area, the configuration length and the like) of each measure.
Figure BDA0001885208490000071
The improvement of the optimization algorithm is that in the invention, the water yield maximization, the total phosphorus and total nitrogen reduction maximization and the implementation cost minimization of all land parcel units (which can be regarded as hydrological response units of a SWAT model) in a research area are taken as fitness functions of the optimization algorithm, and the fitness functions are taken as the basis for judging the goodness and badness of an evaluation feasible solution, so that the invention adopts fitness function values to carry out self-adaptive adjustment on the cross probability and the variation probability in a classical genetic algorithm (NSGA-II), so that individuals with higher fitness are endowed with smaller cross probability and variation probability to ensure the integrity of excellent individuals, and bad individuals with lower fitness are endowed with larger probability values to promote the algorithm to jump out of a local optimal solution and accelerate the convergence of the algorithm; secondly, taking the inter-generation distance threshold of the Pareto optimal solution as a criterion for judging the convergence of the algorithm, and when the average distance of the Pareto optimal solution generated by the algorithm is less than 0.001, the average distance is less thanThe algorithm is considered to achieve the global convergence effect, so that the influence of the subjectivity of the maximum evolution algebra setting on the optimized structure is avoided.
Figure BDA0001885208490000072
The multi-objective measure configuration inputs ecological effect database and measure implementation cost data into an improved adaptive genetic algorithm, and the improvement of an optimization system is optimized into 4 objective functions, namely total nitrogen reduction maximization, total phosphorus reduction maximization, surface runoff increase maximization and implementation cost minimization. And then driving an ANSGA-II optimization algorithm to continuously iterate to obtain a Pareto front-end solution, namely, the Pareto front-end solution can be used as an optimal cost-benefit curve of agricultural non-point source optimal management measures so that a territory management decision-making department can select optimal space configuration schemes of measures under different schemes.

Claims (6)

1. An agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function comprises the following steps:
1) collecting geographic spatial data, meteorological hydrological data, pollution source data and agricultural management information data of a target research area, and constructing a river basin SWAT model of the target research area; the geographic space data comprises a watershed landform, a land utilization type and a soil type; the meteorological hydrological data comprise meteorological data, day-by-day surface runoff and water quality data; then, verifying the sensitivity parameters of the watershed SWAT model, and performing step 2) after the verification is passed;
the checking method comprises the following steps:
collecting water quality sensitivity parameters and range information in a SWAT model in related existing documents by taking water quantity and water quality data as a reference, using the parameters as a parameter database for SWAT model sensitivity parameter analysis, and performing Latin hypercube sampling on parameters in the parameter database by using the SWAT-CUP model to obtain a random parameter combination; then substituting the random parameter combination into the SWAT model to observe the change of the analog value and the observed value and the disturbance difference to obtain the global sensitivity grade of the SWAT model, and selecting the significance level as the basis for determining the sensitivity parameter so as to screen out the sensitivity parameter of the SWAT model;
dividing the SWAT simulation period of the target research area into a preheating period, a calibration period and a verification period, setting the theoretical range of the sensitivity parameters, performing Latin hypercube sampling on the sensitivity parameters by using the SUFI-2 algorithm of a SWAT-CUP model, substituting the sensitivity parameters into the SWAT model of the target research area to judge the matching degree of the simulation value and the observation value, and repeating multiple iterations to obtain the optimal parameter combination of the calibration period; finally, substituting the optimal parameters of the calibration period into the verification period of the SWAT model of the target research area, and when the simulation effect meets the set judgment criterion, passing the calibration;
2) screening agricultural non-point source management measures of the target research area according to the gradient condition and the soil hydrological group type of the target research area; the step 2) is based on a BMP Too1 toolbox, and the average gradient and the hydrological soil group information of the watershed where the target research area is located are extracted and calculated by utilizing a GIS technology, and the average gradient and the hydrological soil group information are used as screening conditions for initial selection of the optimal management measures of the watershed, and agricultural non-point source management measures of the target research area are screened;
3) calculating the cost of the existing management measures of the drainage basin where the target research area is located and the cost of the selected agricultural non-point source management measures, and constructing a cost database of the management measures;
4) adjusting key characteristic parameters of agricultural management measures in the watershed SWAT model, respectively simulating the reduction efficiency of different management measures on target pollutants, and generating a measure implementation ecological benefit database;
5) calculating the cost database and the measure implementation ecological benefit database by adopting a self-adaptive multi-target genetic algorithm according to a set target function optimization criterion to obtain optimal cost benefit curves of different target pollutants and obtain an optimal configuration scheme of agricultural non-point source management measures in the target research area;
the self-adaptive multi-target genetic algorithm is obtained by modifying the operation of crossing and mutation on a fixed probability mechanism in the NSGA-II algorithm into a mode of adjusting the crossing probability and the mutation probability based on a fitness function and adopting evolutionary interpersonal difference of Pareto solution as a criterion for judging whether the NSGA-II algorithm completes convergence.
2. The method of claim 1, wherein the agricultural management measures include agricultural cultivation management, livestock breeding management, and domestic sewage management.
3. The method as claimed in claim 2, wherein the cost database of each management measure is constructed by calculating the unit implementation cost, maintenance cost and opportunity cost of each management measure selected in step 2) and each management measure in the drainage basin where the target research area is located based on the actual price and labor cost of the target research area.
4. The method of claim 1, wherein the method of screening for the sensitivity parameter comprises: collecting water quality sensitivity parameters and range information in a SWAT model in related existing documents by taking water quantity and water quality data as a reference, using the parameters as a parameter database for SWAT model sensitivity parameter analysis, and performing Latin hypercube sampling on parameters in the parameter database by utilizing the SWAT-CUP model to obtain a random parameter combination; and then substituting the random parameter combination into the SWAT model to observe the change of the analog value and the observed value and the disturbance difference to obtain the global sensitivity level of the SWAT model, selecting the significance level as the basis for determining the sensitivity parameters, and screening the sensitivity parameters of the SWAT model.
5. The method of claim 1, wherein the adaptive multi-objective genetic algorithm comprises the following steps: firstly, carrying out chromosome coding on various measures, wherein each measure is represented by an integer to generate a plurality of potential feasible solutions; then substituting the feasible solution into a fitness function to select excellent individuals, adjusting the cross probability and the variation probability of each individual by using the fitness function value, endowing the excellent individuals with smaller cross probability and variation probability, and endowing the inferior individuals with larger cross probability and variation probability; then, performing crossover and variation operations to obtain a new generation of crossed and varied population, and continuously repeating the operation of selecting one crossover and one variation to ensure that the generated population individuals continuously approach to an optimal solution; when the distance of Pareto front-end individuals separated by 10 generations is less than a given threshold, the algorithm terminates.
6. The method of claim 5, wherein a Pareto optimal solution intergenerational difference of less than 0.01 is used as a threshold for algorithm evolution termination.
CN201811443371.1A 2018-11-29 2018-11-29 Agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function Active CN109376955B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811443371.1A CN109376955B (en) 2018-11-29 2018-11-29 Agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811443371.1A CN109376955B (en) 2018-11-29 2018-11-29 Agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function

Publications (2)

Publication Number Publication Date
CN109376955A CN109376955A (en) 2019-02-22
CN109376955B true CN109376955B (en) 2021-09-21

Family

ID=65374930

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811443371.1A Active CN109376955B (en) 2018-11-29 2018-11-29 Agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function

Country Status (1)

Country Link
CN (1) CN109376955B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111899126B (en) * 2020-08-10 2023-12-15 中国水利水电科学研究院 Three red line control index dividing method based on water circulation simulation
CN113112153B (en) * 2021-04-13 2023-06-06 中国科学院生态环境研究中心 Regenerated water optimal control pollutant screening and inventory precipitation method, system, terminal and medium
CN114240257B (en) * 2022-01-30 2022-12-23 哈尔滨工业大学 River basin agricultural source nitrogen-oriented monthly-scale net input quantity acquisition system
CN116757098B (en) * 2023-08-17 2023-11-10 厦门大学 Automatic verification method based on SWAT model multi-objective optimization

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104933300A (en) * 2015-06-03 2015-09-23 中国农业科学院农业资源与农业区划研究所 Calculation method of drainage basin agricultural non-point source pollutant riverway reduction coefficient
CN106875032A (en) * 2016-12-23 2017-06-20 首都师范大学 A kind of Optimal Configuration Method of Watershed Agricultural Non-point Sources Pollution control measure
US20180181894A1 (en) * 2016-12-02 2018-06-28 Gary Michael Schneider System and method for developing multi-objective production plans for prouction agriculture
CN108536908A (en) * 2018-03-01 2018-09-14 北京师范大学 Method based on the assessment of non-point source nitrogen and phosphorus loss risk watershed water environment safety
CN108573302A (en) * 2018-05-02 2018-09-25 济南大学 A kind of simulation of basin non-point source pollution loading and Best Management Practices optimization method

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104933300A (en) * 2015-06-03 2015-09-23 中国农业科学院农业资源与农业区划研究所 Calculation method of drainage basin agricultural non-point source pollutant riverway reduction coefficient
US20180181894A1 (en) * 2016-12-02 2018-06-28 Gary Michael Schneider System and method for developing multi-objective production plans for prouction agriculture
CN106875032A (en) * 2016-12-23 2017-06-20 首都师范大学 A kind of Optimal Configuration Method of Watershed Agricultural Non-point Sources Pollution control measure
CN108536908A (en) * 2018-03-01 2018-09-14 北京师范大学 Method based on the assessment of non-point source nitrogen and phosphorus loss risk watershed water environment safety
CN108573302A (en) * 2018-05-02 2018-09-25 济南大学 A kind of simulation of basin non-point source pollution loading and Best Management Practices optimization method

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
农业非点源污染控制最佳管理措施（BMPs）研究;梁玉好;《吉林水利》;20131031(第10期);第1-9页 *
基于模型的农业非点源污染最佳管理措施效率评估研究进展;耿润哲 等;《生态学报》;20141130;第34卷(第22期);第6398-6405页 *
非点源污染最佳管理措施成本-效益分析研究综述;耿润哲 等;《环境科学与技术》;20160930;第39卷(第9期);第103-111页 *

Also Published As

Publication number Publication date
CN109376955A (en) 2019-02-22

Similar Documents

Publication Publication Date Title
CN109376955B (en) Agricultural non-point source optimal management measure combination optimization configuration method based on ecological service function
Memmah et al. Metaheuristics for agricultural land use optimization. A review
Ghasemi-Mobtaker et al. A comparative of modeling techniques and life cycle assessment for prediction of output energy, economic profit, and global warming potential for wheat farms
Sahoo et al. GIS-based biomass assessment and supply logistics system for a sustainable biorefinery: A case study with cotton stalks in the Southeastern US
Ahmed et al. A nutrient recommendation system for soil fertilization based on evolutionary computation
Mousavi-Avval et al. Combined application of life cycle assessment and adaptive neuro-fuzzy inference system for modeling energy and environmental emissions of oilseed production
Femeena et al. Spatial optimization of cropping pattern for sustainable food and biofuel production with minimal downstream pollution
Chouhan et al. Sustainable planning and decision-making model for sugarcane mills considering environmental issues
Liu et al. A SWAT-based optimization tool for obtaining cost-effective strategies for agricultural conservation practice implementation at watershed scales
Getahun et al. Integrated modeling system for evaluating water quality benefits of agricultural watershed management practices: Case study in the Midwest
Walsh et al. Design of an agent-based model to examine population–environment interactions in Nang Rong District, Thailand
CN106779255B (en) Optimal management measure optimization method and device for river basin agricultural non-point source pollution control
CN107832892B (en) Multi-block seed selection decision optimization method based on combinatorial optimization
CN109858180B (en) Regional crop water consumption space pattern optimization design method
O'Neill et al. Integrated spatially explicit landscape and cellulosic biofuel supply chain optimization under biomass yield uncertainty
CN111882244A (en) Construction method of multi-source homeland development risk assessment system based on hierarchical framework
Zheng et al. Responses of phosphorus use efficiency to human interference and climate change in the middle and lower reaches of the Yangtze River: historical simulation and future projections
Cheng et al. Precision agriculture management based on a surrogate model assisted multiobjective algorithmic framework
Yang et al. Optimal allocation of best management practices based on receiving water capacity constraints
Ma et al. Environmental effects and their causes of agricultural production: Evidence from the farming regions of China
Giri et al. Regulators’ and stakeholders’ perspectives in a framework for bioenergy development
Sharafi et al. Estimating energy consumption and GHG emissions in crop production: A machine learning approach
Li et al. Quantitative assessment of cultivated land use intensity in Heilongjiang Province, China, 2001–2015
Ghaffari et al. Assessing the economic effects of drought using Positive Mathematical Planning model under climate change scenarios
Scheffran et al. A spatial-dynamic model of bioenergy crop introduction in Illinois

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant