A multi-objective optimization method for soil moisture sensor layout considering the probability of soil moisture change.

Abstract

This invention discloses a multi-objective optimization method for soil moisture sensor deployment, considering the probability of soil moisture changes, relating to the fields of smart agriculture and soil moisture monitoring. The purpose of this invention is to address the problems of current sensor deployment methods, such as their inability to simultaneously consider multiple practical indicators, and the lack of comprehensiveness and adaptability in practical applications. The method divides the monitoring demand area into candidate deployment locations and constructs a comprehensive risk function; it constructs a multi-objective sensor deployment optimization model and establishes constraints on the model, solving the model to obtain multiple deployment schemes; it employs a non-dominated sorting genetic algorithm with an elitist strategy to select multiple preferred deployment schemes; and it uses a preference order structure evaluation method to select a predetermined number of optimal deployment schemes from these preferred schemes. This method is used for the rational deployment of sensors in farmland.

Description

Technical Field

[0001] This invention relates to the fields of smart agriculture and soil moisture monitoring. Background Technology

[0002] In recent years, modern agriculture has been continuously developing towards intelligence and digitalization. Advanced concepts such as smart agriculture and precision agriculture have been widely disseminated and deeply applied. Against this backdrop, agricultural production places higher demands on environmental perception capabilities, particularly the perception and management of soil information. Soil moisture is one of the core indicators reflecting the moisture status of the crop root zone. Its dynamic changes not only directly affect the water supply to crops but also provide guidance for the scientific arrangement and implementation timing of agronomic measures, and further relate to the optimization of irrigation system operation strategies and the rationality of resource allocation. Traditional soil moisture monitoring methods mostly adopt fixed-point deployment, relying on manual experience to select monitoring locations, resulting in problems such as numerous coverage blind spots, slow monitoring response, and low information utilization. With the development of wireless communication technology, IoT devices, and low-power sensors, soil moisture sensors now have the conditions for high-density and wide-coverage deployment. However, due to the high unit price of soil moisture sensors and constraints such as power supply, maintenance, and signal reception, it is still difficult to fully promote the uniform and dense deployment of sensors in large areas of farmland.

[0003] Given limited resources and budgets, scientifically planning and effectively configuring a limited number of soil moisture sensors to monitor the most representative and variable soil moisture information in key areas has become a crucial issue that current smart agriculture monitoring systems urgently need to address. Current research has attempted to introduce optimization algorithms for sensor deployment design, but most methods still suffer from problems such as failing to consider the varying agricultural importance of different monitoring areas, and failing to balance multiple practical indicators such as sensor quantity, information value, and response efficiency. This results in poor overall sensing performance, and often treats every field as equal, leading to a singular optimization objective and a lack of comprehensiveness and adaptability in practical applications. However, in agricultural production practice, different plots have different crop types, growth stages, and irrigation needs, resulting in significant differences in the impact of soil moisture changes on yield and quality. Summary of the Invention

[0004] The purpose of this invention is to address the problems of poor overall sensing effect and lack of comprehensiveness and adaptability in practical applications of current sensor deployment methods. This invention proposes a multi-objective optimization deployment method for soil moisture sensors that considers the probability of soil moisture change.

[0005] A multi-objective optimization method for the placement of soil moisture sensors considering the probability of soil moisture change, the method comprising the following:

[0006] Step 1: Divide the monitoring requirement area into For each candidate deployment location, three types of soil moisture monitoring risk functions are established. By combining the three types of soil moisture monitoring risk functions, a comprehensive risk function is constructed.

[0007] Step 2: Construct a multi-target sensor deployment optimization function, including a sensor quantity objective function, a monitoring coverage objective function for high-risk areas established based on the comprehensive risk function, an average detection delay time objective function, and an information entropy objective function. Establish constraints to minimize the total number of sensors objective function, maximize the monitoring coverage objective function for high-risk areas, minimize the average detection delay time objective function, and maximize the information entropy objective function. Solve the multi-target sensor deployment optimization function to obtain multiple deployment schemes.

[0008] Step 3: Use a non-dominated sorting genetic algorithm with an elitist strategy to select T preferred layout schemes from M layout schemes, where M > T;

[0009] Step 4: Use the preference order structure evaluation method to select the preset number of optimal deployment schemes from the T preferred deployment schemes.

[0010] Preferably, the three types of soil moisture monitoring risk functions include spatial impact function, water deficit function, and crop sensitivity function.

[0011] Preferably, the comprehensive risk function is expressed as:

[0012] Formula 1,

[0013] In the formula, For the comprehensive risk function, For the first The weights of the soil moisture monitoring risk function, ,and , , For spatial influence functions, It is a water deficit function. For crop sensitivity functions;

[0014] Spatial influence function Represented as:

[0015] Formula 2,

[0016] In the formula, Indicates the node to be deployed. Abnormal soil moisture can affect nodes Otherwise, it is 0. ;

[0017] Water deficit function Represented as:

[0018] Formula 3,

[0019] In the formula, , Indicates the node to be deployed. In time Crop water shortage, Nodes for candidate deployment locations In time Crop water requirements Nodes for candidate deployment locations In time Soil moisture content, Total time;

[0020] Crop sensitivity function Represented as:

[0021] Formula 4,

[0022] In the formula, Nodes for candidate deployment locations The sensitivity score of the crop to water at its current growth stage. This represents the maximum sensitivity score among all nodes.

[0023] Preferably, in step 2, the objective function for the number of sensors is... for:

[0024] Formula 5,

[0025] In the formula, , ;

[0026] Objective function for monitoring coverage of high-risk areas for:

[0027] Formula 6,

[0028] In the formula, For the covering function, if the node The value is 1 if the node is deployed or covered by an adjacent node, and 0 otherwise.

[0029] Average detection delay time objective function for:

[0030] Formula 7,

[0031] In the formula, For nodes The detection delay time function;

[0032] Information entropy objective function for:

[0033] Formula 8,

[0034] In the formula, This indicates the number of intervals into which the residuals of the monitoring data are divided. Indicates the first The number of records in each interval This represents the total amount of data recorded by all sensors.

[0035] Preferably, the constraint conditions are:

[0036] Formula 9,

[0037] In the formula, Indicates the maximum number of sensors;

[0038] Formula 10,

[0039] In the formula, Represents a node The set of adjacent nodes, For nodes The set of neighboring nodes The first in Each node.

[0040] Preferably, in step 3, a non-dominated sorting genetic algorithm with an elitist strategy is used to select T preferred layout schemes from M layout schemes. The specific process is as follows:

[0041] Step 31: Take the M layout schemes as the initial population, and use the non-dominated sorting method to sort the M layout schemes in the initial population by non-dominated level to obtain the sorted initial population.

[0042] Step 32: Calculate the crowding degree of each layout scheme in the initial population after sorting. Based on the non-dominance level and crowding degree, select T layout schemes from M layout schemes as the parent population of the h-th generation using the binary tournament method. The initial value of h is 1.

[0043] Step 33: Perform crossover and mutation operations on the h-th generation parent population to generate new candidate schemes, which are used as the h-th generation offspring population. Then merge the h-th generation parent population and the h-th generation offspring population into a new h-th generation population.

[0044] Step 34: Perform non-dominance ranking and crowding calculation on the h-th generation new population, and select T deployment schemes from the h-th generation new population based on the results of non-dominance ranking and crowding calculation.

[0045] Step 35: Determine whether h is equal to the preset number of iterations. If not, select T layout schemes from the new population of generation h as the parent population of generation h+1, and set h=h+1, then execute step 33. If yes, select T layout schemes from the new population of generation h as the T preferred layout schemes.

[0046] Preferably, in step 4, the preferred order structure evaluation method is used to select a preset number of optimal layout schemes from the T preferred layout schemes. The specific process is as follows:

[0047] Using a weighted pairwise comparison preference function, the overall superiority of each preferred deployment scheme relative to other preferred deployment schemes is calculated. Combined with the net flow calculation formula, the overall superiority value of each preferred deployment scheme is calculated. The overall superiority values ​​of the T preferred deployment schemes are sorted from largest to smallest, and the top W optimal deployment schemes are selected, where W is the preset number.

[0048] Preferably, the weighted combined pairwise comparison preference function is:

[0049] Formula 11,

[0050] In the formula, The preferred deployment scheme Relative preferred deployment scheme The overall superiority For the first The weights of each objective function, , The preferred deployment scheme Compared with the preferred deployment scheme In the target The relative advantage in terms of preference, for an objective function with decreasing preference, is given by the preference function as follows: For an objective function exhibiting increasing preferences, its preference function is: .

[0051] Preferably, the net flow calculation formula is as follows:

[0052] Formula 12,

[0053] In the formula, , , The preferred deployment scheme The average superiority of other preferred deployment options. Other preferred deployment options relative to the preferred deployment option The average level of superiority The total number of preferred deployment schemes.

[0054] The beneficial effects of this invention are:

[0055] This invention proposes a multi-objective optimization deployment method for soil moisture sensors that considers the probability of soil moisture changes. It can perform differentiated weight assessments based on the soil moisture monitoring risk of different plots, thereby achieving the scientific goal of prioritizing deployment in key monitoring areas. By constructing multiple risk functions to measure the sensing value of soil moisture changes in different plots, and combining a multi-objective optimization evolutionary algorithm and a preference order structure evaluation method, an optimal deployment scheme that balances deployment cost, monitoring coverage, response capability, and information quality is generated, providing a scientifically sound deployment strategy for soil moisture monitoring systems.

[0056] This invention constructs a multi-dimensional risk assessment function that fully considers factors such as terrain diffusion capacity, soil moisture deficit, and crop sensitivity, enabling the scientific quantification of monitoring priorities for each region under varying soil moisture conditions. Simultaneously, it introduces a multi-objective optimization evolutionary algorithm, comprehensively considering multiple practical indicators such as the number of sensors, risk coverage, response speed, and information value, allowing a limited number of sensors to more efficiently cover key areas and improve the overall system perception effect. Furthermore, this invention incorporates a preference order structure evaluation method to flexibly output recommended solutions based on preferences for different objectives such as deployment cost and monitoring performance, demonstrating good adaptability and scalability. Attached Figure Description

[0057] Figure 1 This is a flowchart of a multi-objective optimization layout method for soil moisture sensors that considers the probability of soil moisture change. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention.

[0060] Example:

[0061] The main objective of this embodiment is to provide a multi-objective optimization method for the deployment of soil moisture sensors that considers the probability of soil moisture changes. This method aims to address the problems of blind sensor deployment, low efficiency, slow response, and insufficient regional value differentiation in existing soil moisture monitoring systems. It prioritizes the deployment of soil moisture sensors in areas with higher risks and greater sensitivity to agricultural production, given a limited number of sensors. This improves the coverage, response speed, and information effectiveness of soil moisture monitoring. Furthermore, it balances the economic efficiency, scientific rigor, and applicability of sensor deployment, supporting precision agriculture applications in various scenarios such as smart irrigation and crop management.

[0062] A multi-objective optimization method for the placement of soil moisture sensors considering the probability of soil moisture change, the method comprising the following:

[0063] Step 1: Divide the monitoring requirement area into For each candidate deployment location, three types of soil moisture monitoring risk functions are established. By combining the three types of soil moisture monitoring risk functions, a comprehensive risk function is constructed.

[0064] Step 2: Construct a multi-target sensor deployment optimization function, including a sensor quantity objective function, a monitoring coverage objective function for high-risk areas established based on the comprehensive risk function, an average detection delay time objective function, and an information entropy objective function. Establish constraints to minimize the total sensor quantity objective function, maximize the monitoring coverage objective function for high-risk areas, minimize the average detection delay time objective function, and maximize the information entropy objective function. Solve the multi-target sensor deployment optimization function to obtain M deployment schemes, where M is a positive integer.

[0065] Step 3: Use a non-dominated sorting genetic algorithm with an elitist strategy to select T preferred layout schemes from M layout schemes, where M > T;

[0066] Step 4: Use the preference order structure evaluation method to select the preset number of optimal deployment schemes from the T preferred deployment schemes.

[0067] Specifically, this embodiment adopts a two-step optimization decision mechanism, namely, step 3 and step 4. The first step is step 3, which generates a non-dominated solution set that satisfies multiple optimization objectives through a multi-objective optimization evolutionary algorithm. The second step is step 4, which introduces the preference order structure evaluation method to analyze and sort the solution set and select the optimal layout scheme.

[0068] In step 2, a multi-objective nonlinear programming method can be used to solve the multi-objective sensor deployment optimization function.

[0069] Further defining the three types of soil moisture monitoring risk functions are the spatial impact function, the water deficit function, and the crop sensitivity function.

[0070] Further specifying, the comprehensive risk function is expressed as:

[0071] Formula 1,

[0072] In the formula, For the comprehensive risk function, ,and , , For spatial influence functions, It is a water deficit function. For crop sensitivity functions;

[0073] Spatial influence function Represented as:

[0074] Formula 2,

[0075] In the formula, Indicates the node to be deployed. Abnormal soil moisture can affect nodes Otherwise, it is 0. ;

[0076] Water deficit function Represented as:

[0077] Formula 3,

[0078] In the formula, , Indicates the node to be deployed. In time Crop water shortage, Nodes for candidate deployment locations In time Crop water requirements Nodes for candidate deployment locations In time Soil moisture content, Total time;

[0079] Crop sensitivity function Represented as:

[0080] Formula 4,

[0081] In the formula, Nodes for candidate deployment locations The sensitivity score of the crop to water at its current growth stage. This represents the maximum sensitivity score among all nodes.

[0082] Specifically, step 1 is to comprehensively evaluate the importance of soil moisture perception at each location, propose three types of soil moisture monitoring risk functions, and thus obtain a comprehensive risk function to measure the possible agricultural production consequences of failing to monitor soil moisture changes in a certain area from different perspectives.

[0083] Among the three types of soil moisture monitoring risk functions, the spatial impact function represents the spatial propagation range of the impact of abnormal soil moisture conditions such as drought and waterlogging in the plot; the water deficit function represents the degree to which soil moisture content cannot meet the needs of crops; the crop sensitivity function is based on the sensitivity of crops to changes in soil moisture at different growth stages; and the comprehensive risk function is a weighted linear combination of the first three types of risk functions, comprehensively reflecting the multidimensional risk impact.

[0084] Further defining the construction of the multi-objective optimization model: In step 2, the objective function for the number of sensors is:

[0085] Formula 5,

[0086] In the formula, , ;

[0087] Objective function for monitoring coverage of high-risk areas for:

[0088] Formula 6,

[0089] In the formula, For the covering function, if the node The value is 1 if the node is deployed or covered by an adjacent node, and 0 otherwise.

[0090] Average detection delay time objective function for:

[0091] Formula 7,

[0092] In the formula, For nodes The detection delay time function;

[0093] Information entropy objective function for:

[0094] Formula 8,

[0095] In the formula, This indicates the number of intervals into which the residuals of the monitoring data are divided. Indicates the first The number of records in each interval This represents the total amount of data recorded by all sensors.

[0096] Specifically, in step 1, the monitoring requirement area is divided into... There are 1 candidate deployment locations, and each location node is numbered as follows: .

[0097] Minimize the total number of sensors to reduce system construction and maintenance costs; maximize monitoring coverage in high-risk areas to improve response capabilities in key areas; minimize average detection latency to enable rapid identification of soil anomalies; and maximize information entropy to ensure the comprehensiveness and independence of the spatial distribution of monitoring data.

[0098] Further, the constraints are as follows:

[0099] Formula 9,

[0100] In the formula, Indicates the maximum number of sensors;

[0101] Formula 10,

[0102] In the formula, Represents a node The set of adjacent nodes, For nodes The set of neighboring nodes The first in Each node.

[0103] Specifically, the constraint in Formula 9 is that the total number of sensors does not exceed the upper limit; the constraint in Formula 10 is that each node is covered by a sensor.

[0104] Further specifying, in step 3, a non-dominated sorting genetic algorithm with an elitist strategy is used to select T preferred layout schemes from M layout schemes. The specific process is as follows:

[0105] Step 31: Take the M layout schemes as the initial population, and use the non-dominated sorting method to sort the M layout schemes in the initial population by non-dominated level to obtain the sorted initial population.

[0106] Step 32: Calculate the crowding degree of each layout scheme in the initial population after sorting. Based on the non-dominance level and crowding degree, select T layout schemes from M layout schemes as the parent population of the h-th generation using the binary tournament method. The initial value of h is 1.

[0107] Step 33: Perform crossover and mutation operations on the h-th generation parent population to generate new candidate schemes, which are used as the h-th generation offspring population. Then merge the h-th generation parent population and the h-th generation offspring population into a new h-th generation population.

[0108] Step 34: Perform non-dominance ranking and crowding calculation on the h-th generation new population, and select T deployment schemes from the h-th generation new population based on the results of non-dominance ranking and crowding calculation.

[0109] Step 35: Determine whether h is equal to the preset number of iterations. If not, select T layout schemes from the new population of generation h as the parent population of generation h+1, and set h=h+1, then execute step 33. If yes, select T layout schemes from the new population of generation h as the T preferred layout schemes.

[0110] Specifically, the iterative process of step 35 is explained below with examples. If not, it means that further iteration is needed. In this case, the T layout schemes obtained from the first generation of the new population are used as the second generation parent population. The second generation parent population is subjected to crossover and mutation operations to generate new candidate schemes, which are used as the second generation offspring population. Then, the second generation parent population and the second generation offspring population are merged into the second generation new population. T layout schemes are selected from the second generation new population. Then, it is checked again whether the second time is equal to the preset number of iterations. If it is equal, the iteration stops, and T layout schemes are selected from the second generation new population as the T preferred layout schemes. If it is not equal, the process is repeated to continue iterating.

[0111] Steps 31 to 34 constitute a multi-objective optimization evolutionary algorithm. Step 31 specifically involves:

[0112] ① Chromosome coding. Deployment scheme for each sensor. Encoded as a binary chromosome; where Indicates the first Sensors are deployed at each candidate location. This indicates that no deployment will be made.

[0113] ② Initial population generation. Multiple different deployment schemes are randomly generated as the initial population; each scheme is a feasible solution that satisfies the sensor quantity constraint.

[0114] ③ Non-dominated ranking. By comparing multiple objective function values, the population is divided into multiple non-dominated levels; the solution in the first level is the current optimal solution, that is, it is not dominated by other solutions on any objective.

[0115] Step 32 specifically involves:

[0116] ④ Crowding distance calculation. Calculate the degree of dispersion of individuals on the frontier for each solution; retain solutions with wider distribution and richer information to avoid premature population convergence.

[0117] Step 33 is as follows:

[0118] ⑤ Elite selection and renewal. The parent and offspring generations are merged using an elite retention mechanism, and the top few solutions with high fitness and even distribution are retained to form a new generation of population.

[0119] Steps 34 and 35 are as follows:

[0120] Repeated iterative selection, crossover, mutation, and merging operations eventually converge to a set of Pareto optimal solutions.

[0121] Iteration termination condition: The maximum number of iterations is reached, or the Pareto front changes very little over several generations.

[0122] Further specifying step 4, the preferred order structure evaluation method is used to select a predetermined number of optimal deployment schemes from the T preferred deployment schemes. The specific process is as follows:

[0123] Using a weighted pairwise comparison preference function, the overall superiority of each preferred deployment scheme relative to other preferred deployment schemes is calculated. Combined with the net flow calculation formula, the overall superiority value of each preferred deployment scheme is calculated. The overall superiority values ​​of the T preferred deployment schemes are sorted from largest to smallest, and the top W optimal deployment schemes are selected, where W is the preset number.

[0124] The weighted combined pairwise comparison preference function is as follows:

[0125] Formula 9,

[0126] In the formula, The preferred deployment scheme Relative preferred deployment scheme The overall superiority For the first The weights of each objective function, , The preferred deployment scheme Compared with the preferred deployment scheme In the target The relative advantage in terms of preference, for an objective function with decreasing preference, is given by the preference function as follows: For an objective function exhibiting increasing preferences, its preference function is: .

[0127] The formula for calculating net cash flow is:

[0128] Formula 10,

[0129] In the formula, , , The preferred deployment scheme The average superiority of other preferred deployment options. Other preferred deployment options relative to the preferred deployment option The average level of superiority The total number of preferred deployment schemes.

[0130] Specifically, step 4, the preference order structure evaluation method, first constructs a preference evaluation index system:

[0131] The multi-objective optimization model uses four objective functions as evaluation dimensions: fewer total sensors, greater coverage of high-risk areas, shorter average detection time, and greater monitoring information entropy. Let the first function be... The weights of each objective are: ,satisfy The specific weight values ​​are set according to the actual preference requirements of prioritizing saving, coverage, fast response, and high information entropy;

[0132] Calculate the preference function:

[0133] a. Let the set of all available sensor deployment schemes be . , Indicates the first Layout plan.

[0134] b. Calculate the pairwise comparison preference function for each objective. , indicating the comparison scheme and In the target The relative advantage in terms of the objective. For an objective with decreasing preference, the preference function is: For an objective whose optimization goal exhibits increasing preference, its preference function is: .

[0135] c. Weighted composite pairwise comparison preference function

[0136] ,in Representation scheme relatively The overall superiority.

[0137] Calculate net flow:

[0138] mainstream , indicating the scheme The average degree of superiority compared to other options.

[0139] Negative current This indicates that other options are relative to option 1. The average level of superiority.

[0140] Net flow , indicating the scheme The greater the net flow, the better the overall performance of the solution.

[0141] The final deployment plan was determined as follows:

[0142] All deployment schemes are sorted from largest to smallest net flow value. The larger the value, the better the overall evaluation. Finally, the scheme with the highest net flow value is selected as the recommended deployment configuration.

[0143] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A multi-objective optimization method for the placement of soil moisture sensors considering the probability of soil moisture change, characterized in that, The method includes the following: Step 1: Divide the monitoring requirement area into For each candidate deployment location, three types of soil moisture monitoring risk functions are established. By combining the three types of soil moisture monitoring risk functions, a comprehensive risk function is constructed. Step 2: Construct a multi-target sensor deployment optimization function, including a sensor quantity objective function, a monitoring coverage objective function for high-risk areas established based on the comprehensive risk function, an average detection delay time objective function, and an information entropy objective function. Establish constraints to minimize the total sensor quantity objective function, maximize the monitoring coverage objective function for high-risk areas, minimize the average detection delay time objective function, and maximize the information entropy objective function. Solve the multi-target sensor deployment optimization function to obtain M deployment schemes, where M is a positive integer. Step 3: Use a non-dominated sorting genetic algorithm with an elitist strategy to select T preferred layout schemes from M layout schemes, where M > T; Step 4: Use the preference order structure evaluation method to select a preset number of optimal deployment schemes from the T preferred deployment schemes; The three types of soil moisture monitoring risk functions include spatial impact function, water deficit function, and crop sensitivity function; The overall risk function is expressed as: Official 1, In the formula, For the comprehensive risk function, For the first The weights of the soil moisture monitoring risk function, ,and , , For spatial influence functions, It is a water deficit function. For crop sensitivity functions; Spatial influence function Represented as: Official 2, In the formula, Indicates the node to be deployed. Abnormal soil moisture can affect nodes Otherwise, it is 0. ; Water deficit function Represented as: Official 3, In the formula, , Indicates the node to be deployed. In time Crop water shortage, Nodes for candidate deployment locations In time Crop water requirements Nodes for candidate deployment locations In time Soil moisture content, Total time; Crop sensitivity function Represented as: Official 4, In the formula, Nodes for candidate deployment locations The sensitivity score of the crop to water at its current growth stage. This represents the maximum sensitivity score among all nodes.

2. The method for multi-objective optimization of soil moisture sensor layout considering the probability of soil moisture change, as described in claim 1, is characterized in that... In step 2, the objective function for the number of sensors for: Official 5, In the formula, , ; Objective function for monitoring coverage of high-risk areas for: Official 6, In the formula, For the covering function, if the node The value is 1 if the node is deployed or covered by an adjacent node, and 0 otherwise. Average detection delay time objective function for: Official 7, In the formula, For nodes The detection delay time function; Information entropy objective function for: Official 8, In the formula, This indicates the number of intervals into which the residuals of the monitoring data are divided. Indicates the first The number of records in each interval This represents the total amount of data recorded by all sensors.

3. The method for multi-objective optimal layout of soil moisture sensors considering the probability of soil moisture change, as described in claim 2, is characterized in that... The constraints are: Official 9, In the formula, Indicates the maximum number of sensors; Official 10, In the formula, Represents a node The set of adjacent nodes, For nodes The set of neighboring nodes The first in Each node.

4. The method for multi-objective optimal layout of soil moisture sensors considering the probability of soil moisture change, as described in claim 1, is characterized in that... In step 3, a non-dominated sorting genetic algorithm with an elitist strategy is used to select T preferred layout schemes from M layout schemes. The specific process is as follows: Step 31: Take the M layout schemes as the initial population, and use the non-dominated sorting method to sort the M layout schemes in the initial population by non-dominated level to obtain the sorted initial population. Step 32: Calculate the crowding degree of each layout scheme in the initial population after sorting. Based on the non-dominance level and crowding degree, select T layout schemes from M layout schemes as the parent population of the h-th generation using the binary tournament method. The initial value of h is 1. Step 33: Perform crossover and mutation operations on the h-th generation parent population to generate new candidate schemes, which are used as the h-th generation offspring population. Then merge the h-th generation parent population and the h-th generation offspring population into a new h-th generation population. Step 34: Perform non-dominance ranking and crowding calculation on the h-th generation new population, and select T deployment schemes from the h-th generation new population based on the results of non-dominance ranking and crowding calculation. Step 35: Determine whether h is equal to the preset number of iterations. If not, select T layout schemes from the new population of generation h as the parent population of generation h+1, and set h=h+1, then execute step 33. If yes, select T layout schemes from the new population of generation h as the T preferred layout schemes.

5. A multi-objective optimization layout method for soil moisture sensors considering the probability of soil moisture change, as described in claim 1, 3, or 4, characterized in that, In step 4, the preferred order structure evaluation method is used to select a predetermined number of optimal deployment schemes from the T preferred deployment schemes. The specific process is as follows: Using a weighted pairwise comparison preference function, the overall superiority of each preferred deployment scheme relative to other preferred deployment schemes is calculated. Combined with the net flow calculation formula, the overall superiority value of each preferred deployment scheme is calculated. The overall superiority values ​​of the T preferred deployment schemes are sorted from largest to smallest, and the top W optimal deployment schemes are selected, where W is the preset number.

6. A multi-objective optimization layout method for soil moisture sensors considering the probability of soil moisture change, as described in claim 5, is characterized in that... The weighted combined pairwise comparison preference function is as follows: Official 11, In the formula, The preferred deployment scheme Relative preferred deployment scheme The overall superiority For the first The weights of each objective function, , The preferred deployment scheme Compared with the preferred deployment scheme In the target The relative advantage in terms of preference, for an objective function with decreasing preference, is given by the preference function as follows: For an objective function exhibiting increasing preferences, its preference function is: .

7. A multi-objective optimization layout method for soil moisture sensors considering the probability of soil moisture change, as described in claim 6, is characterized in that... The formula for calculating net cash flow is: Official 12, In the formula, , , The preferred deployment scheme The average superiority of other preferred deployment options. Other preferred deployment options relative to the preferred deployment option The average level of superiority The total number of preferred deployment schemes.

Images