A rice water-saving intelligent irrigation decision-making method and system

Abstract

The present invention relates to the field of agricultural irrigation technology, and in particular to a rice water-saving intelligent irrigation decision method and system, the method comprising the following steps: determining the basic information of a target rice field, and then using a random roadmap method and an adaptive heuristic optimization algorithm to obtain a rice intelligent water-saving irrigation path; obtaining the water demand of rice in the target rice field in real time, and then building a water demand prediction model based on CEEMD and LSTM; using the water demand prediction model to predict the water demand of the target rice field, and combining rainfall information and rice growth stages to obtain rice water-saving irrigation parameters of each irrigation point on the rice intelligent water-saving irrigation path; according to the rice water-saving irrigation parameters, using an irrigation system to complete the irrigation of the target rice field. The present invention can improve the accuracy and intelligence of rice irrigation decisions, reduce the complexity of rice irrigation decisions and the dependence on expert experience, and provide more options for rice irrigation control.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural irrigation, and in particular to a rice water-saving intelligent irrigation decision-making method and system. Background Art

[0002] As one of the most important food crops in the world, rice has a huge demand for water resources during its cultivation. Traditional rice irrigation methods often rely on manual experience and fixed irrigation plans, which may not only lead to serious waste of water resources, but also affect the normal growth and final yield of rice. In recent years, with the rapid development of intelligent technology, intelligent irrigation systems have gradually been applied to agricultural production. These systems achieve precision and automation of irrigation by real-time monitoring of soil moisture, meteorological conditions and other information, combined with advanced algorithms and control strategies.

[0003] Among the existing control strategies of intelligent irrigation systems, intelligent irrigation decisions based on soil moisture and meteorological conditions, intelligent irrigation decisions based on crop water status models, and intelligent irrigation decisions based on fuzzy control are several common and effective irrigation control methods. However, there are also some disadvantages in using these intelligent irrigation decision schemes to irrigate rice, resulting in unsatisfactory irrigation effects on rice fields. Among them, intelligent irrigation decisions based on soil moisture and meteorological conditions usually adopt unified irrigation rules for different growth stages of different crops, and the degree of intelligence is low; intelligent irrigation decisions based on crop water status models require a large amount of experimental data to calibrate and verify model parameters, which is costly and may not fully reflect the actual situation; intelligent irrigation decisions based on fuzzy control require the design of complex fuzzy control algorithms, the determination of appropriate fuzzy sets, membership functions, and fuzzy rules, and are highly dependent on expert experience.

[0004] In summary, in order to improve the precision and intelligence of rice irrigation decisions, reduce the complexity of rice irrigation decisions and dependence on expert experience, and provide more optional control schemes for rice irrigation, it is necessary to explore a new rice water-saving intelligent irrigation decision-making method. Summary of the invention

[0005] In view of the defects in the prior art, the present invention provides a rice water-saving intelligent irrigation decision-making method and system.

[0006] In order to achieve the above-mentioned purpose, in a first aspect, the present invention provides a rice water-saving intelligent irrigation decision-making method, the method comprising the following steps: determining the basic information of the target rice field, and then using the random roadmap method and the adaptive heuristic optimization algorithm to obtain the rice intelligent water-saving irrigation path; obtaining the environmental information of the target rice field, and then using the Penman formula to calculate the water requirement of rice in the target rice field at different time points to obtain a rice water requirement sequence; using CEEMD to decompose the rice water requirement sequence multiple times to obtain multiple intrinsic mode function sub-components; using LSTM to construct an independent water requirement prediction model for each intrinsic mode function sub-component; using the water requirement prediction model to predict the water requirement of the target rice field, and combining the rainfall information and the rice growth stage to obtain the rice water-saving irrigation parameters of each irrigation point on the rice intelligent water-saving irrigation path; according to the rice water-saving irrigation parameters, using an irrigation system to complete the irrigation of the target rice field. The present invention can improve the accuracy and intelligence of rice irrigation decision-making, reduce the complexity of rice irrigation decision-making and the dependence on expert experience, and provide more optional solutions for rice irrigation control.

[0007] Optionally, determining the basic information of the target rice field and then using the random roadmap method and the adaptive heuristic optimization algorithm to obtain the intelligent water-saving irrigation path for rice includes the following steps:

[0008] Determining basic information of the target rice field, wherein the basic information of the target rice field includes the location of the rice field, the location of the water source, and the irrigation demand of the rice field;

[0009] Based on the basic information of the rice field, a random roadmap method is used to obtain an initial irrigation path for the rice;

[0010] The adaptive ant colony optimization algorithm is improved to obtain an irrigation path planning algorithm, and then the initial rice irrigation path is optimized to obtain the rice intelligent water-saving irrigation path.

[0011] Furthermore, the present invention improves the adaptive ant colony optimization algorithm to obtain the intelligent water-saving irrigation path for rice, which can optimize the layout of the irrigation system, help reduce the length of the pipeline, reduce the loss during the water delivery process, and provide a basis for realizing intelligent irrigation and precision irrigation.

[0012] Optionally, the step of improving the adaptive ant colony optimization algorithm to obtain an irrigation path planning algorithm, and then obtaining the rice intelligent water-saving irrigation path based on the basic information of the rice field comprises the following steps:

[0013] Considering the positional relationship between paddy field blocks, the heuristic function of the adaptive ant colony optimization algorithm is optimized to obtain the first irrigation path planning algorithm;

[0014] Introducing a pheromone concentration threshold, a pheromone reward and punishment mechanism, and a dynamic disturbance mechanism into the first irrigation path planning algorithm to obtain the irrigation path planning algorithm;

[0015] Dividing the initial rice irrigation path into a plurality of sub-paths according to the inflection points on the initial rice irrigation path, and then initializing the grid pheromone using a grid pheromone initialization model;

[0016] After initializing the grid pheromone, the irrigation path planning algorithm is used to optimize the initial rice irrigation path to obtain the rice intelligent water-saving irrigation path;

[0017] Rice irrigation facilities are installed according to the rice intelligent water-saving irrigation path.

[0018] Furthermore, the present invention optimizes the heuristic function of the adaptive ant colony optimization algorithm, and introduces the pheromone concentration threshold, the pheromone reward and punishment mechanism and the dynamic disturbance mechanism into the algorithm to obtain the irrigation path planning algorithm, thereby optimizing the initial irrigation path of rice, and optimizing the layout of the irrigation system, which is beneficial to reducing the length of the pipeline and reducing the loss in the water delivery process, and providing a basis for realizing intelligent irrigation and precision irrigation.

[0019] Optionally, the heuristic function and pheromone update formula of the irrigation path planning algorithm are respectively as shown below:

[0020]

[0021]

[0022] in, is the heuristic information from grid i to grid j, is the potential field constant, is the Euclidean distance between grid i and grid j, is the pheromone concentration from grid i to grid j at the k+1th iteration step, is the pheromone concentration from grid i to grid j at the kth iteration step, is the maximum pheromone concentration threshold, is the minimum pheromone concentration threshold, is the minimum pheromone volatility factor, is the maximum change of pheromone volatility factor, is the pheromone reward and penalty value associated with the length L of the path traversed by the ants, is a random number that follows a t-distribution with w degrees of freedom.

[0023] Optionally, the grid pheromone initialization model satisfies the following relationship:

[0024]

[0025] in, is the initialization grid pheromone of grid i, is the distance between grid i and the end point in the initial irrigation path of the rice, is the minimum distance between grid i and all the sub-paths.

[0026] Optionally, the step of using CEEMD to decompose the rice water requirement series multiple times to obtain multiple intrinsic mode function subcomponents comprises the following steps:

[0027] CEEMD was used to perform the first decomposition of the rice water requirement series in the experimental field, and multiple intrinsic mode function components were obtained;

[0028] Based on the energy contribution rate, a plurality of main intrinsic mode function components are selected and a second decomposition is performed using CEEMD to obtain a plurality of intrinsic mode function sub-components.

[0029] Furthermore, the present invention uses CEEMD to decompose the rice water requirement series multiple times to obtain multiple intrinsic mode function sub-components to reflect the fluctuation components of different frequencies and scales in the series, which helps to reveal the inherent laws of water requirement changes, improve the accuracy of rice water requirement prediction, and further improve the accuracy of irrigation control.

[0030] Optionally, the rice water-saving irrigation parameters include irrigation water volume, irrigation time and irrigation rate;

[0031] The method of using the water demand prediction model to predict the water demand of the target rice field and obtaining the rice water-saving irrigation parameters of each irrigation point on the rice intelligent water-saving irrigation path in combination with rainfall information and rice growth stage comprises the following steps:

[0032] Using the water demand prediction model respectively, predict the predicted value of the corresponding intrinsic mode function sub-component at a certain moment in the future;

[0033] The predicted values ​​of each of the intrinsic mode function subcomponents at a certain time in the future are summed to obtain the water requirement of the target rice field at a certain time in the future, and recorded as the predicted water requirement value;

[0034] Determine whether irrigation is needed according to the water demand prediction value and rainfall information, and if irrigation is needed, calculate the irrigation water volume of each irrigation point on the rice intelligent water-saving irrigation path in the target rice field;

[0035] The irrigation time and irrigation rate of each irrigation point are set according to the irrigation water volume, rainfall information and rice growth stage.

[0036] Furthermore, based on the accurate prediction of rice water demand, the present invention sets the irrigation water volume, irrigation rate and irrigation time of each irrigation point according to the irrigation water volume, rainfall information and rice growth stage, thereby achieving precise irrigation.

[0037] Optionally, the rainfall information includes rainfall time and rainfall per unit time;

[0038] The step of setting the irrigation time and irrigation rate of each irrigation point according to the irrigation water volume, rainfall information and rice growth stage comprises the following steps:

[0039] If the rice needs to be flooded during its growth period, directly open each of the irrigation points according to the irrigation water volume and irrigate at the set maximum irrigation rate;

[0040] If the rice does not need to be flooded during its growth period and meets , P is the rainfall per unit area per unit time, V is the irrigation water volume of the irrigation point, T is the time interval between the current moment and a certain moment in the future, t is the rainfall duration between the current moment and a certain moment in the future, s is the irrigation area that the irrigation point is responsible for, then the irrigation time is the time period without rainfall between the current moment and a certain moment in the future, and the irrigation rate is ;

[0041] If the rice does not need to be flooded during its growth period and does not meet , then the irrigation time is the entire time period between the current moment and a certain moment in the future, where the irrigation rate in the time period with rainfall is , the irrigation rate during the period without rainfall is .

[0042] In a second aspect, the present invention also provides a rice water-saving intelligent irrigation decision system, the rice water-saving intelligent irrigation decision system comprising: a data acquisition device, a data output device, a processor and a storage device, the storage device comprising a computer-readable storage medium, the computer-readable storage medium storing a computer program, the computer program comprising program instructions, the program instructions when executed by the processor enable the processor to implement a rice water-saving intelligent irrigation decision method provided by the present invention. The system can not only improve the efficiency of rice water-saving intelligent irrigation, but also improve the practicality of the method. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A schematic diagram of a process flow of a rice water-saving intelligent irrigation decision-making method according to an embodiment of the present invention;

[0045] Figure 2 The present invention is a schematic diagram of a framework of a rice water-saving intelligent irrigation decision system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The specific embodiments of the present invention will be described in detail below. It should be noted that the embodiments described herein are only for illustration and are not intended to limit the present invention. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present invention. However, it is obvious to those of ordinary skill in the art that these specific details do not need to be adopted to implement the present invention. In other examples, in order to avoid confusing the present invention, known circuits, software or methods are not specifically described.

[0047] Throughout the specification, references to "one embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in conjunction with the embodiment or example is included in at least one embodiment of the present invention. Therefore, the phrases "in one embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily all refer to the same embodiment or example. In addition, particular features, structures, or characteristics may be combined in one or more embodiments or examples in any suitable combination and / or subcombination. In addition, it should be understood by those of ordinary skill in the art that the figures provided herein are for illustrative purposes and that the figures are not necessarily drawn to scale.

[0048] It should be noted in advance that, in an optional embodiment, except for independent explanations, the same symbols or letters appearing in all formulas have the same meanings and values.

[0049] In an alternative embodiment, see Figure 1 The present invention provides a rice water-saving intelligent irrigation decision-making method, the method comprising the following steps:

[0050] S1. Determine the basic information of the target rice field, and then use the random landmark graph method and adaptive heuristic optimization algorithm to obtain the intelligent water-saving irrigation path of rice.

[0051] Based on the use of random roadmap method to obtain the initial rice irrigation path, this embodiment improves the adaptive ant colony optimization algorithm to optimize the initial rice irrigation path, and then obtains the intelligent water-saving rice irrigation path, which is conducive to optimizing the layout of the irrigation system, reducing the length of the pipeline, and reducing the loss during the water delivery process, providing a basis for realizing intelligent irrigation and precision irrigation. Step S1 specifically includes the following steps:

[0052] S11. Determine basic information of the target rice field, where the basic information of the target rice field includes the location of the rice field, the location of the water source, and the irrigation demand of the rice field.

[0053] Specifically, in the present embodiment, the location of the rice field is the geographical location of the target rice field, which usually involves recording the longitude and latitude coordinates of the target rice field. Knowing the specific location of the rice field is helpful for subsequent water resource allocation and route planning, and also provides basic data for farmland management. The location of the water source is the starting point of rice field irrigation, so it is very important to determine the exact location of the water source. This includes the type of water source (such as rivers, lakes, and wells), the amount of water, and the relative position relationship between the water source and the target rice field. Clarifying the location of the water source helps to plan an economical and efficient irrigation route. The rice field irrigation demand is whether the target rice field needs to be irrigated. Determining the rice field irrigation demand is also helpful for planning the irrigation route.

[0054] More specifically, the target rice field described in this embodiment includes a plurality of rice field blocks separated from each other. In other optional embodiments, the target rice field can also be a whole rice field.

[0055] S12. Based on the basic information of the rice field, an initial rice irrigation path is obtained using a random roadmap method.

[0056] Specifically, in this embodiment, the random landmark graph method is an algorithm for path planning in the prior art, which is particularly suitable for path search problems in complex environments. It usually includes three steps: constructing a random landmark graph, searching for an initial irrigation path, and optimizing and adjusting. For the sake of brevity, the specific content will not be described in detail here.

[0057] S13, improving the adaptive ant colony optimization algorithm to obtain an irrigation path planning algorithm, and then optimizing the initial rice irrigation path to obtain the intelligent water-saving rice irrigation path.

[0058] Wherein, step S13 specifically includes the following steps:

[0059] S131. Considering the positional relationship between rice field blocks, the heuristic function of the adaptive ant colony optimization algorithm is optimized to obtain a first irrigation path planning algorithm.

[0060] Specifically, in this embodiment, the heuristic function in the traditional ant colony algorithm usually evaluates the attractiveness of the node based on the Euclidean distance. This attractiveness evaluation method mainly relies on the direct distance information between nodes, which limits the global search ability of the algorithm to a certain extent. Therefore, this embodiment introduces the attractiveness distribution of the node into the heuristic function to reflect more complex spatial relationships and dynamic changes, thereby improving the intelligence and globality of the ant search path.

[0061] More specifically, each paddy field block is a node, and the optimized heuristic function is as follows:

[0062]

[0063] in, is the heuristic information from grid i to grid j, is the potential field constant, is the Euclidean distance between grid i and grid j. Represents the attractive force distribution of nodes, by setting different The performance of the first irrigation path planning algorithm can be verified by using the value, and then the optimal value.

[0064] S132. Introducing a pheromone concentration threshold, a pheromone reward and punishment mechanism, and a dynamic disturbance mechanism into the first irrigation path planning algorithm to obtain the irrigation path planning algorithm.

[0065] Specifically, in this embodiment, the pheromone concentration threshold is introduced, that is, the limit of the pheromone concentration is set, that is, the maximum pheromone concentration threshold and the minimum pheromone concentration threshold; the pheromone reward and punishment mechanism is introduced, that is, when updating the pheromone, a larger pheromone increment is given to the shortest path taken by the ant, and a smaller pheromone increment is given to the longest path taken by the ant, that is, the pheromone reward and punishment value given later; the dynamic perturbation mechanism is introduced, that is, the pheromone volatility factor is dynamically updated as the number of iterations increases, and the t distribution is introduced to perturb the pheromone concentration. Introducing the pheromone concentration threshold in the first irrigation path planning algorithm can prevent ants from excessively concentrating on the local optimal solution; introducing the pheromone reward and punishment mechanism in the first irrigation path planning algorithm can improve the efficiency of the algorithm's global search and local optimization; introducing the dynamic perturbation mechanism in the first irrigation path planning algorithm can accelerate the convergence speed of the algorithm and prevent the algorithm from falling into the local optimum too early. These improvement measures can improve the adaptability and flexibility of the algorithm for different rice field environments.

[0066] More specifically, the pheromone update formula of the irrigation path planning algorithm is as follows:

[0067]

[0068] in, is the pheromone concentration from grid i to grid j at the k+1th iteration step, is the pheromone concentration from grid i to grid j at the kth iteration step, is the maximum pheromone concentration threshold, is the minimum pheromone concentration threshold, is the minimum pheromone volatility factor, is the maximum change of pheromone volatility factor, is the pheromone reward and penalty value associated with the length L of the path traversed by the ants, is a random number that follows a t-distribution with w degrees of freedom. , is the maximum pheromone volatility factor, and The values ​​of are usually around 0.2 and 0.5 respectively, and can be adjusted according to actual needs as long as the algorithm can obtain the optimal solution. , is the number of ants that have traveled the shortest path, is the number of ants that have walked the longest path, Q is the pheromone deposition intensity constant, is the length of the shortest path taken by the ant, is the length of the longest path taken by the ant.

[0069] S133, dividing the initial rice irrigation path into a plurality of sub-paths according to the inflection points on the initial rice irrigation path, and then initializing grid pheromones using a grid pheromone initialization model.

[0070] Specifically, in this embodiment, firstly, all inflection points on the initial rice irrigation path, i.e., points where the direction of the path changes, need to be determined by analyzing the geometric characteristics of the path. Then, the initial rice irrigation path is divided into multiple sub-paths according to the inflection points on the initial rice irrigation path, and the path between two adjacent inflection points is a sub-path.

[0071] Furthermore, the grid method is used as the environment representation method of the irrigation path planning algorithm, and the grid pheromone initialization model is used to initialize the grid pheromone. The grid pheromone initialization model satisfies the following relationship:

[0072]

[0073] in, is the initialization grid pheromone of grid i, is the distance between grid i and the end point in the initial irrigation path of rice, is the minimum distance between grid i and all sub-paths. The grid pheromone initialization model can make the initial pheromone concentration of the grid distributed according to the distance difference between the sub-path and the end point, so that the ants can explore near the initial irrigation path of rice and reduce the blindness of ant search.

[0074] S134. After initializing the grid pheromone, the irrigation path planning algorithm is used to optimize the initial rice irrigation path to obtain the intelligent water-saving rice irrigation path.

[0075] Specifically, in this embodiment, after initializing the grid pheromone, the parameters of the irrigation path planning algorithm are further initialized, which is actually to initialize the parameters of the adaptive ant colony optimization algorithm, and then let the ants start searching from the starting point of the initial rice irrigation path, and then continuously optimize the initial rice irrigation path during the algorithm iteration process, and output the intelligent water-saving rice irrigation path when the maximum number of iterations is reached.

[0076] S135. Install rice irrigation facilities according to the intelligent water-saving rice irrigation path.

[0077] Specifically, in this embodiment, rice irrigation facilities are installed according to the intelligent water-saving rice irrigation path, wherein the rice irrigation facilities include sensors, intelligent monitoring boxes, communication networks, irrigation equipment, etc. For more specific content of the rice irrigation facilities, reference can be made to the prior art.

[0078] More specifically, on the intelligent water-saving irrigation path for rice, the setting of irrigation points can be determined according to the size of the rice field block. It is only necessary to ensure that there is at least one irrigation point in each rice field block.

[0079] S2. Obtaining environmental information of the target rice field, and then using the Penman formula to calculate the water requirement of rice in the target rice field at different time points to obtain a rice water requirement sequence.

[0080] Specifically, in the present embodiment, the environmental information of the target paddy field includes the average daily temperature, the average daily air humidity and the solar radiation, etc., so as to use the Penman formula to calculate the water requirement of the rice in the target paddy field at different time points, and then form a rice water requirement sequence. These environmental information can be collected by existing technical means, such as using corresponding sensors to collect, so they are not described in detail here. In other optional embodiments, the modified Penman formula existing in the prior art can also be used to more accurately calculate the water requirement of the rice in the target paddy field at different time points, but the modified Penman formula often involves more complex calculations and needs to collect more environmental information, so relevant personnel can choose which method to use to calculate the water requirement of the rice in the target paddy field according to actual conditions.

[0081] More specifically, in the rice water requirement sequence, the time interval between two adjacent water requirements is denoted as T.

[0082] S3. Use CEEMD to decompose the rice water requirement series multiple times to obtain multiple intrinsic mode function sub-components.

[0083] This embodiment uses CEEMD, i.e. fully integrated empirical mode decomposition, to decompose the rice water requirement sequence multiple times to obtain multiple intrinsic mode function subcomponents to reflect the fluctuation components of different frequencies and scales in the sequence, which helps to reveal the inherent law of water requirement changes, improve the accuracy of rice water requirement prediction, and thus improve the accuracy of irrigation control. Step S3 specifically includes the following steps:

[0084] S31. Use CEEMD to perform the first decomposition of the rice water requirement series in the experimental field to obtain multiple intrinsic mode function components.

[0085] S32. Select a plurality of main intrinsic mode function components based on energy contribution rate and use CEEMD to perform a second decomposition to obtain a plurality of intrinsic mode function sub-components.

[0086] Specifically, in this embodiment, the energy contribution rate of each intrinsic mode function component is calculated, and then they are sorted in order of energy contribution rate from large to small, and then the first 80% or so of the intrinsic mode function components are selected for a second decomposition to obtain multiple intrinsic mode function sub-components.

[0087] S4. Use LSTM to build an independent water demand prediction model for each of the intrinsic mode function sub-components.

[0088] Specifically, in this embodiment, a clipping window is set. For any intrinsic mode function subcomponent, the clipping window is used to clip it multiple times to obtain multiple component sequences, and then these component sequences are used to construct a data set, and the training and verification of LSTM are completed to obtain the corresponding water demand prediction model.

[0089] Furthermore, this embodiment uses CEEMD to decompose the rice water requirement series multiple times to obtain multiple intrinsic mode function sub-components and use LSTM to construct independent water requirement prediction models respectively, which can more carefully capture the changing characteristics of the water requirement series. The obtained model can better adapt to the prediction of rice water requirement in different regions and under different conditions, thereby improving the accuracy of irrigation.

[0090] S5. Use a water demand prediction model to predict the water demand of the target rice field, and obtain rice water-saving irrigation parameters of each irrigation point on the rice intelligent water-saving irrigation path in combination with rainfall information and rice growth stage.

[0091] The rice water-saving irrigation parameters include irrigation water volume, irrigation time and irrigation rate. Step S5 specifically includes the following steps:

[0092] S51. Use the water demand prediction model to predict the predicted value of the corresponding intrinsic mode function sub-component at a certain moment in the future.

[0093] Specifically, in this embodiment, the current time is recorded as , and record a future moment as ,and and If the time interval between them is T, the water demand prediction model can be used to predict the predicted value of the corresponding intrinsic mode function sub-component at a certain moment in the future.

[0094] S52, summing the predicted values ​​of the eigenmode function subcomponents at a certain moment in the future to obtain the water requirement of the target rice field at a certain moment in the future, and recording it as the predicted water requirement value.

[0095] S53, judging whether irrigation is needed according to the predicted water demand value and rainfall information, and if irrigation is needed, calculating the irrigation water volume of each irrigation point on the intelligent water-saving irrigation path of rice in the target rice field.

[0096] Specifically, in this embodiment, the rainfall information includes the rainfall time and the rainfall amount per unit time, and the rainfall information is obtained from the local meteorological station. First, the rainfall time is determined based on the rainfall information. and Is there rainfall between the two? If so, calculate and The total rainfall between. and The total rainfall between , is the rainfall per unit time, is the total area of ​​the target rice field, t is the duration of rainfall between the current moment and a certain moment in the future. According to the calculation relationship of the total rainfall, the total rainfall here is and The volume of water accumulated in the target rice field caused by rainfall when the rainfall duration between is t. Then, the relationship between the predicted water demand and the total rainfall is determined. If the predicted water demand is not greater than the total rainfall, it means that the rice field does not need to be irrigated, otherwise it needs to be irrigated. In addition, if and The absence of rainfall in between also indicates the need for irrigation of the rice fields.

[0097] Furthermore, if irrigation is required, the irrigation water volume of each irrigation point on the rice intelligent water-saving irrigation path in the target rice field is calculated. For a certain irrigation point H, its irrigation water volume Satisfies the following relationship:

[0098]

[0099] in, is the area of ​​the paddy field where the irrigation point H is located, is the predicted water demand, is the number of irrigation points in the paddy field block where irrigation point H is located.

[0100] S54, setting the irrigation time and irrigation rate of each irrigation point according to the irrigation water volume, rainfall information and rice growth stage.

[0101] Based on the accurate prediction of rice water demand, this embodiment sets the irrigation water volume, irrigation rate and irrigation time of each irrigation point according to the irrigation water volume, rainfall information and rice growth stage, thereby achieving precise irrigation.

[0102] It should be noted that rice has different requirements for water at different stages of its growth cycle. Generally, in the early stages of the greening period, tillering period, panicle development period, panicle flowering period, and filling and fruiting period, rice needs a flooded environment, and a certain water layer will be maintained in the paddy field. This is because the growth and development of rice in these stages require a relatively stable water environment. However, in the late stage of the field drying period and the filling and fruiting period, rice does not need to be flooded, and the accumulated water in the paddy field is usually discharged, and there is no obvious water layer in the paddy field at this time. In these stages where flooding is not required, the water required for rice growth mainly comes from the soil. Therefore, irrigation can be used to ensure the moisture content of the soil to meet the needs of normal rice growth. Based on the above considerations, this embodiment does not take the residual water in the paddy field as the main consideration when making irrigation decisions.

[0103] Step S54 specifically includes the following steps:

[0104] S541. If the rice needs to be flooded during its growth period, directly open each irrigation point according to the irrigation water volume and irrigate at the set maximum irrigation rate.

[0105] Specifically, in this embodiment, the set maximum irrigation rate is the maximum irrigation rate of the irrigation equipment.

[0106] S542. If the rice does not need to be flooded during its growing season and meets , P is the rainfall per unit area per unit time, V is the irrigation water volume of the irrigation point, T is the time interval between the current moment and a certain moment in the future, t is the rainfall duration between the current moment and a certain moment in the future, s is the irrigation area that the irrigation point is responsible for, then the irrigation time is the time period without rainfall between the current moment and a certain moment in the future, and the irrigation rate is .

[0107] Specifically, in this embodiment, different from the traditional rainfall per unit time, the rainfall per unit area per unit time here is the volume of water accumulated by rain falling on a unit area per unit time. The irrigation area that an irrigation point is responsible for is the ratio of the area of ​​the rice field block where the irrigation point is located to the number of irrigation points in the rice field block.

[0108] Furthermore, if , when it rains, the irrigation equipment at each irrigation point will not be turned on to irrigate the rice to avoid excessive water accumulation in the rice fields during rainfall; when there is no rainfall, the irrigation equipment will be turned on according to The rice fields are irrigated at an irrigation rate so that the rice fields receive the same amount of water per unit time.

[0109] S543. If the rice does not need to be flooded during its growth period and does not meet , then the irrigation time is the entire time period between the current moment and a certain moment in the future, where the irrigation rate in the time period with rainfall is , the irrigation rate during the period without rainfall is .

[0110] Specifically, in this embodiment, if the rice does not need to be flooded during its growth period and does not meet , then in the period of rainfall, The rice fields are irrigated at an irrigation rate of Irrigation rate to irrigate the rice fields to ensure to During this period, the rice fields receive the same amount of water per unit time.

[0111] S6. According to the rice water-saving irrigation parameters, use an irrigation system to irrigate the target rice field.

[0112] Specifically, in this embodiment, the rice water-saving irrigation parameters are calculated in step S5, and the target rice field is irrigated using an irrigation system, wherein the irrigation system is composed of various rice irrigation facilities installed in step S135.

[0113] It should be noted that, in some cases, the actions described in the specification can be performed in a different order and still achieve the desired results. In this embodiment, the order of steps given is only to make the embodiment appear clearer and easier to explain, rather than to limit it.

[0114] In an alternative embodiment, see Figure 2 In order to improve the practicality and rice irrigation efficiency of the method, the present invention also provides a rice water-saving intelligent irrigation decision-making system, which includes: a data acquisition device 1, a data output device 2, a processor 3 and a storage 4, wherein the storage 4 includes a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, wherein the computer program includes program instructions, and when the program instructions are executed by the processor 3, the processor 3 implements the contents described in steps S1 to S6.

[0115] Specifically, in this embodiment, the data acquisition device 1 includes various sensors for collecting target rice field environmental information and rainfall. The data output device 2 is used to output rice water-saving irrigation parameters, so that the irrigation system can irrigate the rice field according to the rice water-saving irrigation parameters.

[0116] In summary, first, this method uses the random roadmap method to obtain the initial irrigation path of rice, and improves the adaptive ant colony optimization algorithm to optimize the initial irrigation path of rice, thereby obtaining the intelligent water-saving irrigation path of rice, optimizing the layout of the irrigation system, reducing the length of the pipeline, reducing the loss in the water delivery process, and providing a basis for realizing intelligent irrigation and precision irrigation. Secondly, this method uses CEEMD to decompose the rice water demand sequence multiple times, obtains multiple intrinsic mode function subcomponents, and uses LSTM to construct independent water demand prediction models respectively, which can capture the changing characteristics in the water demand sequence more carefully, and the obtained model can better adapt to the prediction of rice water demand in different regions and under different conditions, thereby improving the accuracy of irrigation. Finally, on the basis of accurately predicting the water demand of rice, this method sets the irrigation water volume, irrigation rate and irrigation time of each irrigation point according to the irrigation water volume, rainfall information and rice growth stage, thereby realizing precision irrigation. In addition, the present invention also provides a system adapted to this method to improve the efficiency of rice water-saving intelligent irrigation and the practicality of this method.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A rice water-saving intelligent irrigation decision-making method, characterized in that: The steps include: Determining basic information of the target rice field, wherein the basic information of the target rice field includes the location of the rice field, the location of the water source, and the irrigation demand of the rice field; Based on the basic information of the target rice field, a random roadmap method is used to obtain an initial irrigation path for the rice; Considering the positional relationship between paddy field blocks, the heuristic function of the adaptive ant colony optimization algorithm is optimized to obtain the first irrigation path planning algorithm; Introducing a pheromone concentration threshold, a pheromone reward and punishment mechanism, and a dynamic disturbance mechanism into the first irrigation path planning algorithm to obtain an irrigation path planning algorithm; The heuristic function and pheromone update formula of the irrigation path planning algorithm are shown as follows: ， ， in, is the heuristic information from grid i to grid j, is the potential field constant, is the Euclidean distance between grid i and grid j, is the pheromone concentration from grid i to grid j at the k+1th iteration step, is the pheromone concentration from grid i to grid j at the kth iteration step, is the maximum pheromone concentration threshold, is the minimum pheromone concentration threshold, is the minimum pheromone volatility factor, is the maximum change of pheromone volatility factor, is the pheromone reward and penalty value associated with the length L of the path traversed by the ants, is a random number that follows a t-distribution with w degrees of freedom; ， in, is the number of ants that have traveled the shortest path, is the number of ants that have walked the longest path, Q is the pheromone deposition intensity constant, is the length of the shortest path taken by the ant, is the length of the longest path taken by the ant; Dividing the initial rice irrigation path into a plurality of sub-paths according to the inflection points on the initial rice irrigation path, and then initializing the grid pheromone using a grid pheromone initialization model; The grid pheromone initialization model satisfies the following relationship: ， in, is the initialization grid pheromone of grid i, is the distance between grid i and the end point in the initial irrigation path of the rice, is the minimum value of the distance between grid i and all the sub-paths; After initializing the grid pheromone, the irrigation path planning algorithm is used to optimize the initial rice irrigation path to obtain the rice intelligent water-saving irrigation path; Installing rice irrigation facilities according to the rice intelligent water-saving irrigation path; Obtaining environmental information of a target rice field, and then using the Penman formula to calculate the water requirement of rice in the target rice field at different time points, to obtain a rice water requirement sequence; CEEMD was used to perform the first decomposition of the rice water requirement series in the experimental field, and multiple intrinsic mode function components were obtained; Selecting a plurality of main intrinsic mode function components based on energy contribution rate and performing a second decomposition using CEEMD to obtain a plurality of intrinsic mode function subcomponents; Using LSTM to construct independent water demand prediction models for each of the intrinsic mode function sub-components; Using the water demand prediction model respectively, predict the predicted value of the corresponding intrinsic mode function sub-component at a certain moment in the future; The predicted values ​​of each of the intrinsic mode function subcomponents at a certain time in the future are summed to obtain the water requirement of the target rice field at a certain time in the future, and recorded as the predicted water requirement value; Determine whether irrigation is needed according to the water demand prediction value and rainfall information, and if irrigation is needed, calculate the irrigation water volume of each irrigation point on the rice intelligent water-saving irrigation path in the target rice field; Setting the irrigation time and irrigation rate of each irrigation point according to the irrigation water volume, rainfall information and rice growth stage; The target rice field is irrigated using an irrigation system according to the rice water-saving irrigation parameters, wherein the rice water-saving irrigation parameters include the irrigation water volume, the irrigation time and the irrigation rate.

2. A rice water-saving intelligent irrigation decision-making method according to claim 1, characterized in that: The rainfall information includes rainfall time and rainfall per unit time; The step of setting the irrigation time and irrigation rate of each irrigation point according to the irrigation water volume, rainfall information and rice growth stage comprises the following steps: If the rice needs to be flooded during its growth period, directly open each of the irrigation points according to the irrigation water volume and irrigate at the set maximum irrigation rate; If the rice does not need to be flooded during its growth period and meets , P is the rainfall per unit area per unit time, V is the irrigation water volume of the irrigation point, T is the time interval between the current moment and a certain moment in the future, t is the rainfall duration between the current moment and a certain moment in the future, s is the irrigation area that the irrigation point is responsible for, then the irrigation time is the time period without rainfall between the current moment and a certain moment in the future, and the irrigation rate is ; If the rice does not need to be flooded during its growth period and does not meet , then the irrigation time is the entire time period between the current moment and a certain moment in the future, where the irrigation rate in the time period with rainfall is , the irrigation rate during the period without rainfall is .

3. A rice water-saving intelligent irrigation decision-making system, characterized in that: The rice water-saving intelligent irrigation decision system comprises: a data acquisition device, a data output device, a processor and a storage device, wherein the storage device comprises a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program comprises program instructions, and when the program instructions are executed by the processor, the processor implements the rice water-saving intelligent irrigation decision method as described in any one of claims 1 to 2.

Images