Engineering crop disaster prevention measure formulating method, device, equipment, medium and product based on waterlogging disaster

Abstract

The invention discloses a method, a device, equipment, a medium and a product for formulating flood disaster engineering crop disaster prevention measures, and relates to the field of farmland disaster prevention and disaster loss reduction, and the method comprises the steps: inputting environment monitoring data, a flood disaster grade and a crop growth stage into a crop yield prediction model, and determining the crop yield; on the basis of the crop yield prediction model, according to a necessary disaster prevention technology, predicting a baseline yield of the crops under a non-intervention condition; constructing a target function based on the baseline yield in combination with an optional disaster prevention technology; the target function is solved according to an NSGA-III algorithm, at least one selectable disaster prevention technology is screened, the screened selectable disaster prevention technology and the necessary disaster prevention technology are integrated, the flood disaster engineering crop disaster prevention measures are determined, and the accurate response requirements of the disaster prevention measures under different disaster levels can be met.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of farmland disaster prevention and disaster mitigation, and in particular to a method and device for formulating a crop disaster prevention measure based on waterlogging disaster engineering. BACKGROUND

[0002] In recent years, global climate change has led to an increase in extreme weather events, and natural disasters such as farmland flooding and soil waterlogging have had a serious impact on agricultural production. Traditional disaster prevention measures mainly rely on experience or a single technology, which often cannot meet the precise response needs under different disaster levels. For example, under mild disaster conditions, low-cost and light drainage measures can be used, while under extreme disaster conditions, more complex and costly disaster prevention technologies such as enhanced drainage and subsurface pipe waterlogging reduction are required. However, how to ensure basic disaster prevention response while automatically selecting appropriate technology combinations through scientific data to maximize crop yield while reducing technology investment costs has been a technical problem that needs to be solved in the field of agricultural disaster prevention. SUMMARY

[0003] The purpose of the present application is to provide a method and device for formulating a crop disaster prevention measure based on waterlogging disaster engineering, which solves the problem that traditional disaster prevention measures cannot meet the precise response needs under different disaster levels.

[0004] To achieve the above-mentioned purpose, the present application provides the following solutions:

[0005] In a first aspect, the present application provides a method for formulating a crop disaster prevention measure based on waterlogging disaster engineering, comprising:

[0006] inputting environmental monitoring data, waterlogging disaster levels and crop growth stages into a crop yield prediction model to determine the crop yield;

[0007] based on the crop yield prediction model, predicting the baseline yield of the crop under no intervention according to a mandatory disaster prevention technology; the mandatory disaster prevention technology is a disaster prevention measure pre-configured for different disaster levels of waterlogging disasters;

[0008] based on the baseline yield, combining an optional disaster prevention technology to build a target function; the optional disaster prevention technology includes foliar fertilization, intercropping, deep ripping, soil remediation, biodegradation and combination of planting and breeding;

[0009] solving the target function according to the NSGA-III algorithm, screening at least one optional disaster prevention technology, and integrating the screened optional disaster prevention technology with the mandatory disaster prevention technology to determine the waterlogging disaster engineering crop disaster prevention measure.

[0010] In an embodiment, the crop yield prediction model comprises an input data layer, an encoder layer, a feature concatenation layer, a full connection layer and an output layer connected in sequence; the data input by the input data layer comprises environment monitoring data, waterlogging disaster parameters and crop growth stages; the encoder layer comprises a double-flow spatio-temporal attention network and an Embedding network; the double-flow spatio-temporal attention network comprises an LSTM network and a Transformer network connected in sequence;

[0011] The environment monitoring data is input into the double-flow spatio-temporal attention network to output environment monitoring data results;

[0012] The waterlogging disaster grade and the crop growth stage are respectively input into the Embedding network to output waterlogging disaster grade results and crop growth stage results;

[0013] The environment monitoring data results, the waterlogging disaster grade results and the crop growth stage results are input into the feature concatenation layer to be concatenated to determine the concatenated features;

[0014] The concatenated features are input into the full connection layer to determine the fused features;

[0015] The fused features are input into the output layer to output the crop yield.

[0016] In an embodiment, the mandatory disaster prevention technology is determined in advance for different disaster grades, and is fixedly configured through the waterlogging disaster grade and expert evaluation.

[0017] In an embodiment, the optional disaster prevention technology is preset based on the crop yield prediction model.

[0018] Based on the baseline yield, the overall yield gain of the crop is determined in combination with the additional gain of the preset optional disaster prevention technology.

[0019] In an embodiment, the objective function comprises an overall yield gain objective function and an overall cost objective function.

[0020] The overall yield gain objective function is determined by f1(xopt)=-(Y baseline (E,d,t,T m )+ΔY opt (xopt)); wherein, f1(xopt) is the overall yield gain; Y baselinea baseline yield of the crop under the current environmental and disaster conditions using the selected disaster prevention technology; E is a set of environmental monitoring data, including air temperature, air humidity, soil humidity, rainfall, and ground water level; d is a disaster grade, including slight, light to moderate, moderate, severe, and extremely severe; t is a disaster duration; T m a selected disaster prevention technology determined in advance for the disaster grade d; ΔY opt (xopt) is an additional gain combined with the selected disaster prevention technology combination xopt;

[0021] f2(xopt) = C m (d) + C opt (xopt) is determined; wherein f2(xopt) is the total cost; C m (d) is the fixed cost of the selected disaster prevention technology; C opt (xopt) is the total cost of the selected disaster prevention technology combination.

[0022] In an embodiment, an initial population is randomly generated, and a set of reference points is generated using the Das-Dennis method; the initial population includes a plurality of crops;

[0023] The initial population is evaluated based on the set of reference points, non-dominated sorting, and target normalization, and an environmental selection is performed using a niche preservation mechanism;

[0024] In the environmental selection process, a sub-population is generated using crossover and mutation operations, and the sub-population is updated until a termination condition is met, and the selected disaster prevention technology is screened out;

[0025] The selected disaster prevention technology is integrated with the selected disaster prevention technology to determine the disaster prevention measures for the crop engineering under waterlogging disaster.

[0026] In a second aspect, the application provides a device for formulating disaster prevention measures for crop engineering under waterlogging disaster, comprising:

[0027] A crop yield determination module for inputting environmental monitoring data, waterlogging disaster grade, and crop growth stage into a crop yield prediction model to determine the crop yield;

[0028] A baseline yield prediction module for predicting a baseline yield of the crop under no intervention based on the crop yield prediction model and the selected disaster prevention technology; the selected disaster prevention technology is a disaster prevention measure configured in advance for different disaster grades of waterlogging disaster;

[0029] A target function construction module for constructing a target function based on the baseline yield and the selected disaster prevention technology; the selected disaster prevention technology includes foliar fertilization, intercropping, deep ploughing, soil remediation, biodegradation, and combination of planting and breeding;

[0030] The disaster prevention measure determination module is configured to solve the target function according to the NSGA-III algorithm, screen at least one optional disaster prevention technology, and integrate the screened optional disaster prevention technology and the mandatory disaster prevention technology to determine the waterlogging disaster engineering crop disaster prevention measure.

[0031] In a third aspect, the present application provides a computer device, comprising: a memory, a processor to store a computer program on the memory and executable on the processor, and the processor executes the computer program to implement the steps of the method for determining the waterlogging disaster engineering crop disaster prevention measure according to any one of the above.

[0032] In a fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the method for determining the waterlogging disaster engineering crop disaster prevention measure according to any one of the above.

[0033] In a fifth aspect, the present application provides a computer program product, comprising a computer program, and the computer program is executed by a processor to implement the steps of the method for determining the waterlogging disaster engineering crop disaster prevention measure according to any one of the above.

[0034] According to the embodiments provided in the present application, the following technical effects are disclosed:

[0035] The present application provides a method, device, equipment, medium and product for determining a waterlogging disaster engineering crop disaster prevention measure. The environmental monitoring data, waterlogging disaster grade and crop growth stage are input into a crop yield prediction model to determine the crop yield. Based on the crop yield prediction model, the baseline yield of the crop under no intervention is predicted according to the mandatory disaster prevention technology. The target function is constructed based on the baseline yield and in combination with the optional disaster prevention technology. The target function is solved according to the NSGA-III algorithm, at least one optional disaster prevention technology is screened, and the screened optional disaster prevention technology is integrated with the mandatory disaster prevention technology. The mandatory disaster prevention technology is pre-set for different disaster grades, and the optional disaster prevention technology is optimized based thereon. The waterlogging disaster engineering crop disaster prevention measure is determined to meet the precise response requirement under different disaster grades. The integration of the mandatory disaster prevention technology and the optional disaster prevention technology is automatically selected by the NSGA-III algorithm to maximize the crop yield and reduce the technical investment cost. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0037] Figure 1 A flooding disaster engineering-agricultural ecology synergistic method process schematic diagram provided for an embodiment of the present application;

[0038] Figure 2 A flooding disaster engineering-agricultural ecology synergistic method process schematic diagram provided for an embodiment of the present application;

[0039] Figure 3 A crop yield prediction model schematic diagram provided for an embodiment of the present application;

[0040] Figure 4 A computer device structure schematic diagram provided for an embodiment of the present application. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments only constitute some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0042] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0043] The present application relates to the field of farmland disaster prevention and disaster mitigation, and specifically relates to a method and system for comprehensively optimizing and recommending farmland disaster prevention technology using a multi-objective optimization algorithm NSGA-III. The present application is particularly suitable for combining environmental monitoring data with a hierarchical response technology library, automatically recommending an optimal disaster prevention technology combination according to different disaster levels, and achieving the dual goals of crop yield increase and cost control. Specifically, the present application includes engineering-agricultural-ecological technologies (normal drainage, enhanced drainage, subsurface drainage, foliar fertilization, intercropping, deep ripping, soil remediation, biodegradation, and crop-livestock integration), and five flooding disaster levels (slight, light to moderate, moderate, severe, and extremely severe).

[0044] Traditional waterlogging prevention techniques rely on fixed thresholds (e.g., soil moisture > 90% triggers drainage), lacking adaptability to dynamic changes in crop growth stages (e.g., water needs differ between tillering and booting stages) and regional climate characteristics. For example, Patent No. CN202010123456.7 uses a single soil moisture threshold, without considering the spatial and temporal distribution of rainfall, leading to frequent misjudgments during the Meiyu period in the middle and lower reaches of the Yangtze River. Existing technologies focus on a single objective (e.g., minimizing cost or maximizing yield), ignoring the need for engineering-agronomy-ecology multi-objective technology coordination. For example, Patent No. US20220056789A1 proposes a drainage scheduling model that reduces energy consumption by 15%, but increases the risk of soil salinization by 23%, damaging the farmland ecosystem. Current systems use independent sensor networks, failing to effectively integrate multi-dimensional physiological and ecological data such as soil conductivity, root layer oxygen content, and leaf temperature. For example, the published literature "Farmland Drainage Control System Based on Internet of Things" only collects water level data without coupling rainfall prediction information from weather radars, resulting in a delay of more than 30 minutes in emergency response.

[0045] Currently, the multi-objective optimization algorithm NSGA-III has achieved excellent results in high-dimensional and multi-objective optimization problems in various fields, but its application in farmland disaster prevention technology recommendation is still in the exploratory stage. Meanwhile, the idea of a hierarchical response technology library, which sets mandatory technologies for different disaster levels and then optimizes the selection of optional technologies, has not yet been implemented in a mature automated decision-making system. Therefore, this application proposes a method for developing waterlogging disaster engineering and crop disaster prevention measures, aiming to use real-time environmental monitoring data, disaster levels, and duration parameters to automatically recommend an optimal combination of disaster prevention technologies that balances yield increase benefits and cost savings, providing scientific decision-making support for farmland disaster prevention and control.

[0046] For example, Figure 1As shown, the present application discloses a kind of flood disaster engineering agronomic ecological synergy method.The method first collects real-time environmental monitoring data and disaster parameters, according to disaster grade, select optional disaster prevention technology from pre-designed hierarchical response technology library, then combine crop yield prediction model and each disaster prevention technology benefit model, construct overall objective function (including crop yield gain and technical cost two goals).Optimize the combination of selectable technology using NSGA-III multi-objective evolutionary algorithm, adopt reference point generation, target normalization and niche preservation mechanism, the niche preservation mechanism is in multi-objective optimization, when multiple solutions have the same non-dominated level, the mechanism is used to maintain the diversity of population, avoid solution excessive aggregation in a local area, ensure that the final solution is evenly distributed to cover the entire Pareto frontier.A set of uniformly distributed reference points (direction vectors) are generated by Das-Dennis method to guide the distribution direction of solutions;Each solution is associated with the nearest reference point, and the perpendicular projection distance (or angle distance after target normalization) of the solution is calculated;The number of solutions associated with each reference point is counted, and the reference point with fewer associated solutions (sparse area) is preferentially selected, so as to retain the solution with sparse distribution.Finally, multiple Pareto optimal solutions are output, and the optional technology is integrated to form a complete disaster prevention technology recommendation scheme.The method realizes the goal of maximizing crop yield and reducing technical input cost under the premise of ensuring basic disaster prevention response, has the advantages of data-driven, flexible response and intelligent decision-making, and provides a scientific and intelligent decision-making support tool for farmland disaster prevention and control.

[0047] As Figure 2 shown, the present application embodiment provides a kind of method for formulating flood disaster engineering crop disaster prevention measures, and specific steps are as follows.

[0048] S1: environmental monitoring data, flood disaster grade and crop growth stage are input into crop yield prediction model, and the yield of crop is determined.

[0049] S2: according to the crop yield prediction model, the baseline yield of crop under no intervention is predicted according to the optional disaster prevention technology;The optional disaster prevention technology is the disaster prevention measure configured in advance for different disaster grades of flood disaster.

[0050] S3: based on the baseline yield, combined with selectable disaster prevention technology, construct objective function;The selectable disaster prevention technology includes leaf surface fertilization, intercropping, deep ripping, soil remediation, biodegradation and combination of planting and breeding.

[0051] S4: according to NSGA-III algorithm, at least one selectable disaster prevention technology is screened out by solving the objective function, and the selected selectable disaster prevention technology is integrated with the optional disaster prevention technology to determine the flood disaster engineering crop disaster prevention measure.

[0052] As Figure 3As shown, in one exemplary embodiment, S1 can be replaced with the following steps.

[0053] S101: The crop yield prediction model comprises an input data layer, an encoder layer, a feature concatenation layer, a full connection layer and an output layer connected in sequence; the data input by the input data layer comprises environment monitoring data, waterlogging disaster parameters and crop growth stages; the encoder layer comprises a double-flow spatio-temporal attention network and an Embedding network; the double-flow spatio-temporal attention network comprises an LSTM network and a Transformer network connected in sequence.

[0054] S102: The environment monitoring data is input into the double-flow spatio-temporal attention network to output environment monitoring data results.

[0055] S103: The waterlogging disaster grade and the crop growth stage are respectively input into the Embedding network to output waterlogging disaster grade results and crop growth stage results.

[0056] S104: The environment monitoring data results, the waterlogging disaster grade results and the crop growth stage results are input into the feature concatenation layer for concatenation to determine the concatenated features.

[0057] S105: The concatenated features are input into the full connection layer to determine the fused features.

[0058] S106: The fused features are input into the output layer to output the crop yield.

[0059] The crop yield prediction model is based on the input environment monitoring data (air temperature, air humidity, soil humidity, rainfall, ground water level), disaster parameters (waterlogging disaster grade d and duration t, the collected waterlogging disaster grade d and duration t are known quantities), and crop growth stages. The environment monitoring data (air temperature, air humidity, soil humidity, rainfall, ground water level), waterlogging disaster parameters (waterlogging disaster grade d and duration t), and crop growth stages are input, a double-flow spatio-temporal attention network based on long short-term memory network (Long Short-Term Memory, LSTM) and Transformer is used, and the predicted yield is output.

[0060] Firstly, the environmental monitoring data contains spatio-temporal data, and LSTM is used to mine the spatio-temporal association between the data, and then a Transformer encoder is used to combine multi-source environmental monitoring data. The waterlogging disaster grade and the crop growth stage are encoded by one-hot embedding. The encoded features of the three types of data are concatenated, and finally input into the full connection layer for yield prediction. The loss function of the model uses mean squared error (MSE) to measure the yield prediction error. The training optimizer is AdamW (learning rate 3e-4, weight decay 1e-5), and the training number of rounds is 500 epochs. (The training data comes from public crop yield data and data recorded in journal articles.)

[0061] Further, in an exemplary embodiment, S2 further comprises, before S2:

[0062] For different disaster grades, the mandatory disaster prevention technologies are determined in advance, and are fixedly configured through the waterlogging disaster grade and expert evaluation.

[0063] As shown in Table 1, for the five different grades of waterlogging disasters (1st grade mild, 2nd grade mild to moderate, 3rd grade moderate, 4th grade severe, 5th grade extremely severe), the mandatory disaster prevention technologies corresponding to each grade are determined in advance in the technology library. The mandatory technologies have the following characteristics: they can meet the basic disaster prevention needs under the disaster grade, ensuring that the farmland has the most basic emergency response capability when the disaster occurs. The cost is relatively fixed and has been verified by a large number of practices, and has stability and reliability. For example, under mild disaster conditions, conventional drainage measures can be selected; under extremely severe disaster conditions, the mandatory technology may be enhanced drainage or subsurface pipe subsidence measures (mandatory technology is related to disaster grade, not related to region, and the current disaster prevention technology recommendation method is mainly for southern grain production areas). These mandatory technologies will be fixed response measures in the algorithm, ensuring that no matter how changes occur, the system can always provide a basic disaster prevention scheme.

[0064] Table 1 Graded Response Technology Library

[0065]

[0066] Further, in an exemplary embodiment, S2 can be replaced with the following steps.

[0067] S201: Based on the crop yield prediction model, preset optional disaster prevention technologies.

[0068] On the basis of the mandatory technology, one or more technologies are allowed to be selected from the remaining 8 disaster prevention technologies for combination to further improve the crop yield increase effect or reduce the overall technology cost. The optional technologies include but are not limited to: foliage fertilization, intercropping, deep ploughing, soil remediation, biodegradation, and combination of planting and breeding, etc. Through benefit evaluation of different combinations, the system can automatically recommend the optimal optional technology combination according to the specific environment and disaster parameters.

[0069] S202: Based on the baseline yield, the overall yield gain of the crop is determined in combination with the additional gain of the preset optional disaster prevention technology.

[0070] Under the action of the mandatory technology Tm(d), the baseline yield Ybaseline(E, d, t, Tm) of the crop under no technical intervention is predicted. Further, in combination with the additional gain ΔYopt(xopt) of the optional technology combination xopt, the overall yield gain can be calculated.

[0071] Ybaseline represents the baseline yield of the crop under the current environment and disaster conditions after using the mandatory technology (unit: tons / acre or other applicable units); E represents a set of environmental monitoring data, including air temperature, air humidity, soil humidity, rainfall, ground water level, etc.; d represents the waterlogging disaster level, with a value range of 1st, 2nd, 3rd, 4th, and 5th; t represents the disaster duration (unit: hours or days, defined according to actual conditions). Tm represents the mandatory disaster prevention technology determined in advance for the disaster level d, which ensures the basic disaster prevention response under this level; the negative sign indicates that the larger the gain, the smaller the objective function value.

[0072] Further, in an exemplary embodiment, S3 can be replaced by the following steps.

[0073] S301: The objective function includes an overall yield gain objective function and an overall cost objective function.

[0074] S302: The overall yield gain objective function is determined using f1(xopt) = -(Ybaseline(E, d, t, Tm) + ΔYopt(xopt)), where f1(xopt) is the overall yield gain; Ybaseline is the baseline yield of the crop under the current environment and disaster conditions using the mandatory disaster prevention technology; E is a set of environmental monitoring data, including air temperature, air humidity, soil humidity, rainfall, and ground water level; d is the waterlogging disaster level, including slight, light to moderate, moderate, moderate, and extremely heavy; t is the disaster duration; Tm is the mandatory disaster prevention technology determined in advance for the disaster level d; and ΔYopt(xopt) is the additional gain in combination with the optional disaster prevention technology combination xopt.

[0075] S303: Determine the overall cost objective function by f2(xopt) = Cm(d) + Copt(xopt), wherein f2(xopt) is the total cost, Cm(d) is the fixed cost of the mandatory disaster prevention technology, and Copt(xopt) is the total cost of the optional disaster prevention technology combination.

[0076] The technical investment cost is divided into the fixed cost Cm(d) of the mandatory technology and the total cost Copt(xopt) of the optional technology combination. Among them, xi is an optional technology decision variable (0 or 1), and ci is the cost of the ith optional technology. By constructing the above objective function, the algorithm can optimize the optional technology combination on the basis of ensuring the basic disaster prevention response of the mandatory technology, and balance the crop yield gain and the technology cost.

[0077] Further, in an exemplary embodiment, S4 can be replaced with the following steps.

[0078] S401: Randomly generate an initial population, and generate a reference point set by using the Das-Dennis method; the initial population includes a plurality of crops.

[0079] An initial population of size N is randomly generated, and each individual is a binary vector xopt representing an optional technology combination. Reference point generation: a set of reference points Z = {z1, z2, …, zK} is uniformly generated in the normalized target space according to the target number m (2 in this application) by using the Das-Dennis method, to ensure the uniformity of the solution distribution in the search process.

[0080] S402: Perform objective function evaluation, non-dominated sorting, and target normalization on the initial population based on the reference point set, and perform environmental selection by using the niche preservation mechanism.

[0081] S403: In the environmental selection process, a sub-population is generated by using crossover and mutation operations, and the sub-population is updated until the termination condition is met, and the optional disaster prevention technology is screened out.

[0082] S404: Integrate the screened optional disaster prevention technology with the mandatory disaster prevention technology to determine the disaster prevention measures for crops in waterlogging disaster engineering.

[0083] The values of the objective functions f1(xopt) and f2(xopt) are calculated for each individual xopt in the population by using the established crop yield prediction model and technology benefit model.

[0084] To avoid the influence of the scale difference of each objective function on the selection process, the objective values are normalized.

[0085] Specific steps include calculating the ideal point (minimum value vector) of each objective function in the population, the extreme point, and constructing a normalized hyperplane to normalize all objective functions to a similar scale.

[0086] Non-dominated sorting and environmental selection:

[0087] After merging the current population with the last generation population, use the non-dominated sorting method to divide different non-dominated layers F1, F2, ….

[0088] When the number of individuals in the last non-dominated layer exceeds the remaining space, use the niche preservation mechanism specific to NSGA-III for environmental selection: correlate the normalized individuals with reference points and calculate the perpendicular distance of individuals to each reference direction. The environmental selection step in the NSGA-III algorithm includes.

[0089] a) Calculate the ideal point and extreme point of each objective in the population, and construct a normalized hyperplane.

[0090] b) Correlate the normalized individuals with the preset reference points and calculate the perpendicular distance of each individual to the nearest reference direction.

[0091] c) According to the reference point correlation count and perpendicular distance, use the niche preservation mechanism to select individuals into the next generation.

[0092] According to the correlation count of each reference point, prefer those individuals that are optimal in perpendicular distance and balance the correlation number of each reference point, ensuring that the selected population has good diversity.

[0093] Crossover, mutation and population update:

[0094] Apply crossover (such as uniform crossover or single-point crossover) and bit flip mutation operations suitable for binary encoding to the selected individuals to generate a new generation of sub-population.

[0095] Merge the parent population and the child population, update the population through non-dominated sorting and environmental selection until the predetermined number of iterations or convergence conditions are met, and generate the next generation population.

[0096] Repeat the above process until the predetermined maximum iteration number Tmax or other convergence conditions are met.

[0097] Result output:

[0098] Finally output the non-dominated solution set (Pareto front), and combine it with the mandatory technology Tm(d) to form a complete disaster prevention technology recommendation scheme for farmers or management departments to refer to.

[0099] The application combines the hierarchical response technology library and the NSGA-III multi-objective optimization algorithm to propose a new integrated technology recommendation method for farmland waterlogging reduction and waterlogging disaster reduction. The method ensures that there is a mandatory technology as a basic response under each disaster level, and dynamically optimizes the optional technology combination according to real-time environmental data, achieving the optimal balance between crop yield increase and technology cost. The systematic process and multi-objective optimization mechanism not only improve the decision-making efficiency, but also provide multiple representative alternative solutions for practical application, having high popularization and application value.

[0100] The technical scheme of the application can obtain the following beneficial effects:

[0101] Hierarchical response: The mandatory technology and the optional technology are organically combined to ensure that there is a basic disaster prevention response capability under different disaster levels, and the overall system benefit is further improved through the optional technology.

[0102] Data-driven decision-making: Real-time environmental monitoring data and disaster parameters are fully utilized to dynamically adjust the crop yield prediction model and the technology benefit evaluation model, improving the accuracy and adaptability of the recommended solution.

[0103] Multi-objective balance: The NSGA-III multi-objective evolutionary algorithm is adopted to achieve the dynamic balance between crop yield increase and technology cost, so that the final recommended solution can maximize economic benefit and effectively reduce investment risk.

[0104] Population diversity: The NSGA-III ensures uniform distribution of population solution set through reference point generation and niche preservation mechanism, providing multiple representative Pareto optimal solutions for decision makers, thereby meeting the needs of different application scenarios.

[0105] The embodiment of the application provides a device for formulating crop disaster prevention measures based on waterlogging disaster engineering, and specific modules are as follows.

[0106] The crop yield determination module is used to input environmental monitoring data, waterlogging disaster level and crop growth stage into the crop yield prediction model to determine the crop yield.

[0107] The baseline yield prediction module is used to predict the baseline yield of the crop under no intervention based on the crop yield prediction model and the mandatory disaster prevention technology; the mandatory disaster prevention technology is a disaster prevention measure configured in advance for different disaster levels of waterlogging disasters.

[0108] The objective function construction module is used to construct an objective function based on the baseline yield and in combination with the optional disaster prevention technology; the optional disaster prevention technology includes leaf fertilization, intercropping, deep ripping, soil remediation, biodegradation and combination of planting and breeding.

[0109] The disaster prevention measure determination module is configured to solve the objective function according to the NSGA-III algorithm, screen at least one optional disaster prevention technology, and integrate the screened optional disaster prevention technology and the mandatory disaster prevention technology to determine the waterlogging disaster engineering crop disaster prevention measure.

[0110] In an exemplary embodiment, a computer device, which can be a server or a terminal, is provided, and an internal structure diagram of the computer device can be as shown in Figure 4 The computer device includes a processor, a memory, an input / output interface (I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is configured to formulate a waterlogging disaster engineering crop disaster prevention measure. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to communicate with external terminals through a network connection. The computer program is executed by the processor to realize the formulation of the waterlogging disaster engineering crop disaster prevention measure.

[0111] Those skilled in the art can understand that Figure 4 The structure shown in the above

[0112] In an exemplary embodiment, a computer device is also provided, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to realize the steps in the above method embodiments.

[0113] In an exemplary embodiment, a computer readable storage medium is provided, which stores a computer program. The computer program is executed by a processor to realize the steps in the above method embodiments.

[0114] In an exemplary embodiment, a computer program product is provided, which includes a computer program. The computer program is executed by a processor to realize the steps in the above method embodiments.

[0115] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0116] It can be understood by those skilled in the art that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing related hardware through a computer program, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments of each method. In the embodiments provided in the present application, any reference to memory, database or other medium can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric memory (Ferroelectric Random Access Memory, FRAM), phase change memory (Phase Change Memory, PCM), graphene memory, etc. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc.

[0117] The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a blockchain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0118] The technical features of the above embodiments can be combined arbitrarily. In order to make the description simple, not all possible combinations of the technical features in the above embodiments are described, but as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0119] The principles and implementations of the present application are described in detail with specific examples in this paper, and the above examples are only used to help understand the method of the present application and its core idea; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation and application range will be changed. Therefore, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A method for developing a disaster prevention measure for crops based on waterlogging disaster engineering, characterized by, The method for formulating the crop disaster prevention measures based on waterlogging disaster engineering comprises the following steps: inputting environmental monitoring data, waterlogging disaster grade and crop growth stage into a crop yield prediction model to determine crop yield; based on the crop yield prediction model, predicting baseline yield of crops without intervention according to mandatory disaster prevention technology; the mandatory disaster prevention technology is a disaster prevention measure configured in advance for different disaster grades of waterlogging disasters; based on the baseline yield, combining optional disaster prevention technology to build a target function; the optional disaster prevention technology includes leaf fertilization, intercropping, deep ploughing, soil remediation, biodegradation and combination of planting and breeding; solving the target function according to NSGA-III algorithm, screening at least one optional disaster prevention technology, and integrating the screened optional disaster prevention technology with the mandatory disaster prevention technology to determine waterlogging disaster engineering crop disaster prevention measures.

2. The method of claim 1, wherein the method is characterized by, inputting environmental monitoring data, waterlogging disaster grade and crop growth stage into a crop yield prediction model to determine crop yield, which specifically comprises the following steps: the crop yield prediction model comprises an input data layer, an encoder layer, a feature splicing layer, a full connection layer and an output layer connected in sequence; the data input by the input data layer includes environmental monitoring data, waterlogging disaster parameters and crop growth stage; the encoder layer includes a double-flow spatio-temporal attention network and an Embedding network; the double-flow spatio-temporal attention network includes an LSTM network and a Transformer network connected in sequence; inputting the environmental monitoring data into the double-flow spatio-temporal attention network to output environmental monitoring data results; inputting the waterlogging disaster grade and the crop growth stage into the Embedding network respectively to output waterlogging disaster grade results and crop growth stage results; inputting the environmental monitoring data results, the waterlogging disaster grade results and the crop growth stage results into the feature splicing layer for splicing to determine spliced features; inputting the spliced features into the full connection layer to determine fused features; inputting the fused features into the output layer to output crop yield.

3. The method of claim 1, wherein the method is characterized by, based on the crop yield prediction model, predicting baseline yield of crops without intervention according to mandatory disaster prevention technology, which further comprises the following steps: for different disaster grades, the mandatory disaster prevention technology is determined in advance respectively, and is fixedly configured through waterlogging disaster grade and expert evaluation.

4. The method of claim 1, wherein the method is characterized by, based on the crop yield prediction model, predicting baseline yield of crops without intervention according to mandatory disaster prevention technology, which further comprises the following steps: based on the crop yield prediction model, presetting optional disaster prevention technology; based on the baseline yield, combining the additional gain of the preset optional disaster prevention technology to determine the overall yield gain of crops.

5. The method of claim 1, wherein the method is characterized by, based on the baseline yield, combining optional disaster prevention technology to build a target function, which specifically comprises the following steps: the target function includes an overall yield gain target function and an overall cost target function; using f1(xopt) = -(Y baseline (E, d, t, T m )+ ΔY opt (xopt)), determine an overall yield gain objective function; wherein f1(xopt) is an overall yield gain; Y baseline is a baseline yield of the crop under the current environmental and disaster conditions using the mandatory disaster prevention technology; E is a set of environmental monitoring data including air temperature, air humidity, soil humidity, rainfall, and ground water level; d is a waterlogging disaster grade including slight, light-moderate, moderate, severe, and extremely severe; t is a disaster duration; T m is a mandatory disaster prevention technology predetermined for the disaster grade d; ΔY opt (xopt) is an additional gain combined with the optional disaster prevention technology combination xopt. f2(xopt) = C m (d) + C opt (xopt), determine the overall cost objective function; wherein f2(xopt) is the total cost; C m (d) is the fixed cost of the mandatory disaster prevention technology; C opt (xopt) is the total cost of the optional disaster prevention technology combination.

6. The method of claim 1, wherein the method is characterized by, Solving the objective function according to the NSGA-III algorithm, screening at least one optional disaster prevention technology, and integrating the screened optional disaster prevention technology and the mandatory disaster prevention technology to determine the waterlogging disaster engineering crop disaster prevention measure, specifically comprising: Randomly generating an initial population, and generating a reference point set by using the Das-Dennis method; the initial population includes multiple crops; Based on the reference point set, the initial population is evaluated for the objective function, non-dominated sorting and target normalization, and the environment is selected by using the niche preservation mechanism; In the environment selection process, a sub-population is generated by using crossover and mutation operations, and the sub-population is updated until the termination condition is met, and the optional disaster prevention technology is screened out; The screened optional disaster prevention technology and the mandatory disaster prevention technology are integrated to determine the waterlogging disaster engineering crop disaster prevention measure.

7. A device for formulating disaster prevention measures for crops based on waterlogging disaster engineering, characterized in that: The waterlogging disaster engineering crop disaster prevention measure determination device comprises: A crop yield determination module for inputting environmental monitoring data, waterlogging disaster grade and crop growth stage into a crop yield prediction model to determine crop yield; A baseline yield prediction module for predicting the baseline yield of crops under no intervention based on the crop yield prediction model according to the mandatory disaster prevention technology; the mandatory disaster prevention technology is a disaster prevention measure configured in advance for different disaster grades of waterlogging disasters; An objective function construction module for constructing an objective function based on the baseline yield combined with optional disaster prevention technologies; the optional disaster prevention technologies include foliar fertilization, intercropping, deep ripping, soil remediation, biodegradation and combination of planting and breeding; A disaster prevention measure determination module for solving the objective function according to the NSGA-III algorithm, screening at least one optional disaster prevention technology, and integrating the screened optional disaster prevention technology and the mandatory disaster prevention technology to determine the waterlogging disaster engineering crop disaster prevention measure.

8. A computer device comprising: A memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the waterlogging disaster engineering crop disaster prevention measure determination method according to any one of claims 1-6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the waterlogging disaster engineering crop disaster prevention measure determination method according to any one of claims 1-6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the waterlogging disaster engineering crop disaster prevention measure determination method according to any one of claims 1-6.

Images