Dynamic water distribution wheel irrigation marshalling method for intelligent irrigation system

Abstract

The invention discloses a dynamic water distribution wheel irrigation marshalling method for an intelligent irrigation system. The method comprises the steps that S1, design parameters of upper and lower channels of an irrigation area and the water shortage index and the water demand of crops in the irrigation area corresponding to the lower channel are obtained; s2, constructing a multi-objective optimization function considering crop water shortage indexes, channel water delivery losses, distances in the rotation irrigation groups and water diversion differences among the rotation irrigation groups in the rotation irrigation marshalling, and setting constraint conditions of lower-level channels and irrigation marshalling; s3, randomly generating an initial rotation irrigation marshalling, taking the initial rotation irrigation marshalling as an initial population of an NSGA-III algorithm, and solving the multi-objective function by adopting the NSGA-III algorithm to obtain an optimal rotation irrigation marshalling; and S4, performing dynamic water distribution on the irrigation area by adopting the optimal rotation irrigation marshalling.

Description

Technical Field

[0001] The present invention belongs to the field of dynamic water distribution technology and computational simulation technology in irrigation areas, and specifically relates to a method for dynamic water distribution and rotation irrigation grouping in an intelligent irrigation system. Background Art

[0002] With the increasing water shortage and growing demand for agricultural irrigation, efficient water utilization in irrigation areas has become a core challenge for modern agricultural development. Traditional rotation irrigation grouping methods are mostly based on experience or single-objective optimization (such as minimizing water loss or balancing water distribution), which makes it difficult to balance the multiple needs of crop water demand, water delivery efficiency, and geographical distribution. Existing technologies have the following limitations: Insufficient response to crop water demand: Traditional methods do not dynamically incorporate the crop water shortage index into the optimization target, resulting in a poor match between irrigation timing and crop water demand cycle, which can easily cause local drought or excessive irrigation, affecting crop yield and water resource utilization efficiency.

[0003] Poor multi-objective coordination: Existing rotation irrigation models mostly focus on a single objective (such as minimizing water transmission losses), ignoring the impact of upstream and downstream distances within the rotation irrigation group on water distribution efficiency (excessive distance increases water transmission delays and energy consumption), and lack effective methods to balance conflicts among multiple objectives.

[0004] Weak adaptability of optimization algorithms: Classic multi-objective optimization algorithms (such as the weighted sum method and NSGA-II) are prone to falling into local optimality or uneven distribution of solutions when dealing with high-dimensional multi-objective problems, making it difficult to generate Pareto solutions that cover a wide range of trade-off spaces.

[0005] In recent years, the rise of smart irrigation systems has promoted the development of dynamic water distribution technology, but how to achieve multi-objective collaborative optimization remains a technical difficulty. Summary of the Invention

[0006] In response to the above steps in the prior art, the dynamic water distribution and irrigation grouping method of the intelligent irrigation system provided by the present invention solves the problem that the grouping obtained by the existing irrigation is difficult to take into account the timely water demand of crops, water delivery efficiency and geographical distribution.

[0007] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is: A method for dynamically distributing water and irrigation groups in a smart irrigation system is provided, which comprises the following steps: S1. Obtain the design parameters of the upper and lower level channels of the irrigation area and the water shortage index and water demand of the crops in the irrigation area corresponding to the lower level channels; S2. Construct a multi-objective optimization function that considers the crop water shortage index, channel water loss, distance within the rotation irrigation group, and water diversion differences between rotation irrigation groups, and set constraints for the lower-level channels and irrigation groups; S3. Randomly generate an initial rotation group and use it as the initial population of the NSGA-III algorithm. Use the NSGA-III algorithm to solve the multi-objective function and obtain the optimal rotation group; S4. Use the best rotation irrigation group to dynamically distribute water to the irrigation area.

[0008] Furthermore, the expression of the multi-objective optimization function is: Among them, S is the total target value; 、 、 and They are the smallest water loss, the smallest total water shortage index, the shortest distance within the rotation irrigation group, and the smallest difference in water diversion between rotation irrigation groups; is the minimization function; described 、 、 and The expressions are: , in, and are leakage losses of upper and lower channels, respectively; is the leakage loss function; is the channel bed permeability coefficient of the upper and lower channels; are the channel bed permeability indexes of the upper and lower channels respectively; are the water diversion flows of the upper and lower channels respectively; are the water conveyance lengths of the upper and lower channels, respectively; are the water distribution times of the upper and lower channels, respectively; in, is the water shortage index of the irrigation area corresponding to the jth lower-level channel that is not irrigated; is the total number of lower-level channels that are not irrigated; is the total number of lower-level channels that have completed irrigation; N is the total number of lower-level channels; in, is the distance between the upstream channel and the downstream channel in the i-th irrigation group; m is the total number of irrigation cycles; in, Design traffic for upstream channels; is the switch status of the j-th water outlet of the i-th round filling group, To close the water outlet, Open the water outlet; is the flow rate of the j-th outlet.

[0009] Furthermore, the constraints of the lower-level channels and irrigation groups include: One-time water diversion constraint for channels: any one lower-level channel can only be opened once during the rotation irrigation period: The flow rate of the downstream channel should be constrained. The water distribution flow of any downstream channel should be 0.6-1.0 times its design flow rate: Water balance constraint: the sum of the flow rates of the lower-level channels that distribute water at the same time at any time should be equal to the flow rate of the upper-level channels ; Time constraint: the water distribution time of each irrigation group shall not exceed the maximum allowable water delivery time T of the water distribution channel: in, is the design flow of the jth downstream channel; is the water distribution time of the jth lower channel.

[0010] Furthermore, the dynamic water distribution rotation grouping method of the smart irrigation system also includes processing constraints and updating the multi-objective function: Define the constraint violation amount for each constraint: for , the violation amount is , for , the violation amount is , for , the violation amount is , for , the violation amount is ; According to the amount of constraint violation, design the penalty term of the multi-objective function: in, 、 、 and are penalty factors corresponding to the amount of constraint violation; Update the multi-objective function according to the penalty term: in, For penalty items.

[0011] Furthermore, the step S3 further includes: S31. Generate several reference points, which are combinations of different objectives. The objectives are the 、 、 and ; S32. Use the Latin hypercube sampling method to randomly generate multiple rotation groups, use the rotation groups that satisfy the constraint violation amount as individuals in the initial population, and calculate the multi-objective function value of each individual; S33, perform crossover and mutation operations on the individuals of the population, and then convert the parent population P t and the offspring population Q t Merge into population R t , the new population size is 2n; S34, the population R t Stratify by non-dominated level, and select n individuals as the new population P based on the reference point and non-dominated level t+1 ; S35, determine whether the number of population iterations reaches the maximum number of iterations, if so, go to step S36, otherwise return to step S33; S36. Select the rotation irrigation formation with the smallest multi-objective function value in the new population obtained in the last iteration as the optimal rotation irrigation formation.

[0012] Furthermore, the total number of reference points is: Where H is the total number of reference points; p is the number of divisions for each target direction; is the total number of targets.

[0013] Furthermore, during the crossover operation, the expression for generating two offspring is: in, and is the first population before the crossover operation To the parent generation; and for and The offspring obtained after the crossover operation; β is the distribution factor, which controls the degree of similarity between the offspring individuals and the parent individuals; The expression for mutation operation on individuals is: Among them, δ is the mutation probability; For the Parameters related to irrigation in each individual; for The parameter value obtained after mutation; and They are The maximum and minimum values ​​of .

[0014] Furthermore, n individuals are selected as the new population P according to the reference point and non-dominated level. t+1 The methods include: When Rank1+Rank2+...+Rank s =n, put Rank1, Rank2,...,Rank s All individuals in the new population P t+1 ,Rank1,Rank2,...,Rank s All are non-dominant hierarchical stratification; When Rank1+Rank2+...+Rank s-1 = K< n, and Rank1+Rank2+...+Rank s > n, K individuals and s n-K individuals are selected from the population and placed into the new population P t+1 .

[0015] The beneficial effects of the present invention are as follows: the multi-objective function constructed by this scheme can ensure that the irrigation timing is highly matched with the crop water requirement cycle by considering the crop water shortage index. Combined with the NSGA-III algorithm guided by the reference point, it can generate a Pareto solution set with uniform distribution and wide coverage, and can provide a variety of rotation irrigation grouping schemes, thereby ensuring that the best implementation plan is selected to reduce the risk of drought and water resource waste of crops, so as to ultimately achieve the goal of increasing crop yield and improving water utilization efficiency.

[0016] When constructing the objective function, this scheme also takes into account the constraints and their corresponding constraint violations, and introduces the constraint violations as penalty functions into the multi-objective optimization function. This can effectively solve complex constraint problems such as flow exceeding limits and water balance, ensuring the feasibility of the scheme.

[0017] This solution combines the constructed multi-objective optimization function with the NSGA-III algorithm to solve problems such as multi-objective conflicts, simplistic decision-making, and resource waste in traditional rotation irrigation groups. It provides a scientific, efficient, and scalable solution for dynamic water distribution in irrigation areas, with significant economic, ecological, and social benefits, laying a technical foundation for the large-scale application of smart irrigation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1This is a flow chart of the dynamic water distribution and irrigation grouping method for the smart irrigation system. DETAILED DESCRIPTION

[0019] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0020] refer to Figure 1 , Figure 1 The flowchart of the method for dynamic water distribution and irrigation grouping of the smart irrigation system is shown; Figure 1 As shown, the method S includes steps S1 to S4.

[0021] In step S1, the design parameters of the upper and lower level channels of the irrigation area and the water shortage index and water demand of the crops in the irrigation area corresponding to the lower level channels are obtained; In this solution, the current canal system corresponding to the irrigation system consists of two levels of channels, including one upper-level channel and N lower-level channels. In step S1, the relevant parameters of the upper-level channel and the lower-level channel that need to be obtained include: (1) Design flow of 1 upper channel and N lower channels, m 3 / s; (2) the length of one upper channel and N lower channels, m; (3) the channel bed permeability coefficient, channel bed permeability index, and seepage reduction coefficient after anti-seepage of one upper channel and N lower channels; the water shortage index of the field crops corresponding to the N lower channels; and the water demand of the field crops corresponding to the N lower channels. In step S2, a multi-objective optimization function is constructed that considers the crop water shortage index, channel water loss, distance within the rotation irrigation group, and water diversion difference between the rotation irrigation groups, and the constraints of the lower-level channels and irrigation groups are set; In one embodiment of the present invention, the expression of the multi-objective optimization function is: Among them, S is the total target value; 、 、 and They are the smallest water loss, the smallest total water shortage index, the shortest distance within the rotation irrigation group, and the smallest difference in water diversion between rotation irrigation groups; is the minimization function; described 、 、 and The expressions are: , in, and are leakage losses of upper and lower channels, respectively; is the leakage loss function; is the channel bed permeability coefficient of the upper and lower channels; are the channel bed permeability indexes of the upper and lower channels respectively; are the water diversion flows of the upper and lower channels respectively; are the water conveyance lengths of the upper and lower channels, respectively; are the water distribution times of the upper and lower channels, respectively; in, is the water shortage index of the irrigation area corresponding to the jth lower-level channel that is not irrigated; is the total number of lower-level channels that are not irrigated; is the total number of lower-level channels that have completed irrigation; N is the total number of lower-level channels; in, is the distance between the upstream channel and the downstream channel in the i-th irrigation group; m is the total number of irrigation cycles; in, Design traffic for upstream channels; is the switch status of the j-th water outlet of the i-th round filling group, To close the water outlet, Open the water outlet; is the flow rate of the j-th outlet.

[0022] During implementation, the constraints for the preferred lower-level channels and irrigation groups in this solution include: One-time water diversion constraint for channels: any one lower-level channel can only be opened once during the rotation irrigation period: ; The flow rate of the downstream channel should be constrained. The water distribution flow of any downstream channel should be 0.6-1.0 times its design flow rate: Water balance constraint: the sum of the flow rates of the lower-level channels that distribute water at the same time at any time should be equal to the flow rate of the upper-level channels ; Time constraint: the water distribution time of each irrigation group shall not exceed the maximum allowable water delivery time T of the water distribution channel: in, is the design flow of the jth downstream channel; is the water distribution time of the jth lower channel.

[0023] In this scheme, the dynamic water distribution and irrigation grouping method of the smart irrigation system also includes constraint processing and updating the multi-objective function: Define the constraint violation amount for each constraint: for , the violation amount is , for , the violation amount is , for , the violation amount is , for , the violation amount is ; According to the amount of constraint violation, design the penalty term of the multi-objective function: in, 、 、 and are penalty factors corresponding to the amount of constraint violation; Update the multi-objective function according to the penalty term: in, For penalty items.

[0024] In step S3, an initial rotation group is randomly generated and used as the initial population of the NSGA-III algorithm. The NSGA-III algorithm is used to solve the multi-objective function to obtain the optimal rotation group; In one embodiment of the present invention, step S3 further includes: S31. Generate several reference points. The reference points are used to guide the population to be evenly distributed toward the Pareto frontier. The reference points are a combination of different objectives, focusing on the minimum water shortage index or the minimum water loss. The objective is the multi-objective optimization function. 、 、 and .

[0025] Each target direction is normalized and uniformly sampled along the hyperplane in the normalized space. The total number of targets is , the number of divisions in each target direction is p, then the total number of reference points is: Where H is the total number of reference points.

[0026] Before generating reference points, initialize the parameters. The population size is the number of solutions to the round-robin grouping retained in each generation of the algorithm. The description and meaning of the initialized parameters can be found in the following table: S32. Use the Latin hypercube sampling method to randomly generate multiple rotation groups, use the rotation groups that meet the constraint violation amount as individuals of the initial population, and calculate the multi-objective function value of each individual; use Latin hypercube sampling (LHS) for initial population generation to ensure that variables are evenly distributed in space.

[0027] S33, perform crossover and mutation operations on the individuals of the population, and then convert the parent population P t and the offspring population Q t Merge into population R t , the new population size is 2n; The expression that generates two children is: in, and is the first population before the crossover operation To the parent generation; and for and The offspring obtained after the crossover operation; β is the distribution factor, which controls the degree of similarity between the offspring individuals and the parent individuals; The expression for mutation operation on individuals is: Among them, δ is the mutation probability; For the Parameters related to irrigation in each individual; for The parameter value obtained after mutation; and They are The maximum and minimum values ​​of .

[0028] In this scheme, the main purpose of the crossover operation is to combine the advantages of the two rotation irrigation grouping schemes, where β obeys a polynomial distribution; the mutation operation is mainly to fine-tune the details of the irrigation-related parameters in the rotation irrigation grouping, such as the irrigation start time and irrigation flow rate.

[0029] S34, the population R t Stratify by non-dominated level, and select n individuals as the new population P based on the reference point and non-dominated level t+1: When Rank1+Rank2+...+Rank s =n, put Rank1, Rank2,...,Rank s All individuals in the new population P t+1 ,Rank1,Rank2,...,Rank s All are non-dominant hierarchical stratification; When Rank1+Rank2+...+Rank s-1 = K< n, and Rank1+Rank2+...+Rank s > n, K individuals and s n-K individuals are selected from the population and placed into the new population P t+1 .

[0030] S35, determine whether the number of population iterations reaches the maximum number of iterations, if so, go to step S36, otherwise return to step S33; S36. Select the rotation irrigation formation with the smallest multi-objective function value in the new population obtained in the last iteration as the optimal rotation irrigation formation.

[0031] When implemented, this solution is preferred in Rank s The method of selecting n-K individuals is to fill the new population P with non-dominated ranks. t+1 , ensuring that the solution set of the rotation irrigation group can have different trade-offs such as the minimum sum of water shortage index and the minimum water transmission loss. The detailed implementation process is as follows: S341. Normalize the target: in, and is the extreme point, is the target value, is the normalized target value.

[0032] Objective normalization solves the dimensional differences and numerical scale imbalance problems between different objective functions, thereby preventing the optimization process from being dominated by a single objective.

[0033] S342, reference point association: associate each population individual with a reference point, connect the reference point with the origin, define a reference line corresponding to each reference point, calculate the vertical distance between each population individual and each reference line, and associate the reference point whose reference line is closest to the population individual with the population individual; S343, counting the number of microhabitats, that is, counting the number of individuals associated with each reference point, and selecting the reference point with the fewer associated individuals; S344. From the individuals associated with the selected reference point, select the closest individual (number of individuals is nK) to join the population P t+1 .

[0034] The NSGA-3 algorithm in this scheme uses population genetics based on reference point association and microhabitat counts to effectively solve the problem of poor diversity maintenance effect of NSGA-2 under high-dimensional objectives (objective function dimension ≥ 3), ensuring that irrigation rotation groups can achieve water distribution optimization with multiple objectives such as timeliness, efficiency, and water conservation.

[0035] In summary, this solution solves the contradictory multi-objective optimization of crop timeliness, low water loss, and short distance within the rotation irrigation group in the rotation irrigation group objectives. The established Pareto solution set of the rotation irrigation group can better guide the dynamic water distribution process of the irrigation area canal system, and provide technical support for promoting the scientific and rational water distribution of the canal system of the smart irrigation system.

Claims

1. A method for dynamic water distribution and irrigation grouping in a smart irrigation system, characterized in that: Including steps: S1. Obtain the design parameters of the upper and lower level channels of the irrigation area and the water shortage index and water demand of the crops in the irrigation area corresponding to the lower level channels; S2. Construct a multi-objective optimization function that considers the crop water shortage index, channel water loss, distance within the rotation irrigation group, and water diversion differences between rotation irrigation groups, and set constraints for the lower-level channels and irrigation groups; S3. Randomly generate an initial rotation group and use it as the initial population of the NSGA-III algorithm. Use the NSGA-III algorithm to solve the multi-objective function and obtain the optimal rotation group; S4. Use the best rotation irrigation group to dynamically distribute water to the irrigation area.

2. The method for dynamic water distribution wheel irrigation of the smart irrigation system according to claim 1 is characterized in that: The expression of the multi-objective optimization function is: Among them, S is the total target value; 、 、 and They are the smallest water loss, the smallest total water shortage index, the shortest distance within the rotation irrigation group, and the smallest difference in water diversion between rotation irrigation groups; is the minimization function; described 、 、 and The expressions are: ， in, and are leakage losses of upper and lower channels, respectively; is the leakage loss function; is the channel bed permeability coefficient of the upper and lower channels; are the channel bed permeability indexes of the upper and lower channels respectively; are the water diversion flows of the upper and lower channels respectively; are the water conveyance lengths of the upper and lower channels, respectively; are the water distribution times of the upper and lower channels, respectively; in, is the water shortage index of the irrigation area corresponding to the jth lower-level channel that is not irrigated; is the total number of lower-level channels that are not irrigated; is the total number of lower-level channels that have completed irrigation; N is the total number of lower-level channels; in, is the distance between the upstream channel and the downstream channel in the i-th irrigation group; m is the total number of irrigation cycles; in, Design traffic for upstream channels; is the switch status of the j-th water outlet of the i-th round filling group, To close the water outlet, Open the water outlet; is the flow rate of the j-th outlet.

3. The method for dynamic water distribution wheel irrigation of the smart irrigation system according to claim 1, characterized in that: The constraints on the sub-channels and irrigation groups include: One-time water diversion constraint for channels: any one lower-level channel can only be opened once during the rotation irrigation period: ； The flow rate of the downstream channel should be constrained. The water distribution flow of any downstream channel should be 0.6-1.0 times its design flow rate: ； Water balance constraint: the sum of the flow rates of the lower-level channels that distribute water at the same time at any time should be equal to the flow rate of the upper-level channels ; Time constraint: the water distribution time of each irrigation group shall not exceed the maximum allowable water delivery time T of the water distribution channel: in, is the design flow of the jth downstream channel; is the water distribution time of the jth lower channel.

4. The method for dynamic water distribution wheel irrigation grouping of the smart irrigation system according to claim 3 is characterized in that: It also includes constraint processing of constraints and updating of multi-objective functions: Define the constraint violation amount for each constraint: for , the violation amount is , for , the violation amount is , for , the violation amount is , for , the violation amount is ; According to the amount of constraint violation, design the penalty term of the multi-objective function: in, 、 、 and are penalty factors corresponding to the amount of constraint violation; Update the multi-objective function according to the penalty term: in, For penalty items.

5. The method for dynamic water distribution wheel irrigation grouping of the smart irrigation system according to claim 2, characterized in that: The step S3 further comprises: S31. Generate several reference points, which are combinations of different objectives. The objectives are the 、 、 and ; S32. Use the Latin hypercube sampling method to randomly generate multiple rotation groups, use the rotation groups that satisfy the constraint violation amount as individuals in the initial population, and calculate the multi-objective function value of each individual; S33, perform crossover and mutation operations on the individuals of the population, and then convert the parent population P t and the offspring population Q t Merge into population R t , the new population size is 2n; S34, the population R t Stratify by non-dominated level, and select n individuals as the new population P based on the reference point and non-dominated level t+1 ; S35, determine whether the number of population iterations reaches the maximum number of iterations, if so, go to step S36, otherwise return to step S33; S36. Select the rotation irrigation formation with the smallest multi-objective function value in the new population obtained in the last iteration as the optimal rotation irrigation formation.

6. The method for dynamic water distribution wheel irrigation grouping of the smart irrigation system according to claim 5 is characterized in that: The total number of reference points is: Where H is the total number of reference points; p is the number of divisions for each target direction; is the total number of targets.

7. The method for dynamic water distribution wheel irrigation of the smart irrigation system according to claim 5, characterized in that: When performing a crossover operation, the expression for generating two offspring is: in, and is the first population before the crossover operation To the parent generation; and for and The offspring obtained after the crossover operation; β is the distribution factor, which controls the degree of similarity between the offspring individuals and the parent individuals; The expression for mutation operation on individuals is: Among them, δ is the mutation probability; For the Parameters related to irrigation in each individual; for The parameter value obtained after mutation; and They are The maximum and minimum values ​​of .

8. The method for dynamic water distribution wheel irrigation of the smart irrigation system according to claim 5, characterized in that: According to the reference point and non-dominated level, n individuals are selected as the new population P t+1 The methods include: When Rank1+Rank2+...+Rank s =n, put Rank1, Rank2,...,Rank s All individuals in the new population P t+1 ,Rank1,Rank2,...,Rank s All are non-dominant hierarchical stratification; When Rank1+Rank2+...+Rank s-1 = K< n, and Rank1+Rank2+...+Rank s > n, K individuals and s n-K individuals are selected from the population and placed into the new population P t+1 .

Images