Cross-domain water engineering group multi-target collaborative scheduling method and system

Abstract

The invention discloses a cross-domain water engineering group multi-target collaborative scheduling method and system, and the method comprises the following steps: S1, collecting the parameter information of a cross-domain water engineering group, and constructing a cross-domain water engineering group multi-target scheduling model based on the collected parameter information; s2, building a multi-target scheduling model based on a cross-domain water engineering group, and solving an optimal scheduling scheme of the water engineering group by using a multi-population cooperation search method; and S3, outputting an optimal scheduling scheme. Belongs to the field of water engineering optimization scheduling, and by adopting the cross-domain water engineering group multi-target collaborative scheduling method and system, the complex hydraulic relation is considered, and the optimal configuration of water resources under different operation conditions is ensured by comprehensively analyzing the hydraulic relation between the water engineering groups and the scheduling strategy of each water engineering.

Description

Technical Field

[0001] The present invention relates to the technical field of water project optimization scheduling, and in particular to a multi-objective collaborative scheduling method and system for a cross-domain water project group. Background Art

[0002] As an important engineering facility for the development and utilization of water resources by human society, water projects have multiple functions such as flood control, irrigation, power generation, water supply, shipping, and ecological protection. With the rapid development of the economy and society, the contradiction between water supply and demand has become increasingly prominent, and the cross-domain collaborative multi-objective scheduling of water projects has become a research hotspot. In particular, for a cross-domain water project group composed of multiple water projects, it has the characteristics of complexity, multi-objective, uncertainty, and spatiotemporal heterogeneity, so its scheduling needs to comprehensively consider the coordination and optimization of each subsystem, so the complexity and challenge of scheduling are more significant.

[0003] The multi-objective scheduling of a cross-domain water project group usually includes goals such as flood prevention and disaster reduction, water supply guarantee, hydropower generation, and ecological and environmental protection. In the scheduling process, a variety of constraints need to be considered, including water balance constraints, water level constraints, flood discharge capacity constraints, and water demand constraints. Based on the above goals and constraints, a multi-objective optimization scheduling model for a cross-domain water project group is established. Optimization theory is the basis for processing the multi-objective scheduling model of a cross-domain water project group. Existing optimization theories mainly include traditional optimization methods such as linear programming, nonlinear programming, and dynamic programming to study how to select the best decision-making plan under various constraints to achieve one or some goals. However, the above-mentioned optimization methods have the following obvious limitations in the application process: ① Computational complexity problem: When faced with multi-objective, nonlinear constraints and scheduling tasks with high uncertainty, the computational efficiency of traditional optimization methods is low, making it difficult to achieve real-time scheduling. ② Insufficient adaptability of methods: Traditional methods have great disadvantages in dealing with dynamic demands, especially fluctuations in water demand caused by climate change. They lack flexibility and cannot achieve adaptive optimization according to changes in actual demand. ③ Diversity of solutions: A single optimization method usually cannot provide a diverse set of feasible solutions, which limits the ability of managers to balance different goals and makes it difficult to achieve multifunctional coordinated operation of the scheduling system.

[0004] To solve the above problems, intelligent optimization algorithms have been further developed. They simulate evolution and group behavior in nature, conduct global search in a huge solution space, do not rely on linear or convex assumptions, can handle nonlinear and complex constraint problems, and can flexibly respond to changes in the environment and demand. They perform well in multi-objective scheduling, which has led to their widespread use in water project group scheduling.

[0005] However, there are still some deficiencies in the current research on the application of intelligent optimization algorithms in water project scheduling. For example, the current intelligent optimization algorithms lack effective processing of asymmetric constraints in the operation of water project groups. There are decision variable constraints, inequality constraints, equality constraints and boundary constraints in water project scheduling. Due to the existence of these constraints, many infeasible solution areas are generated in the search space, the feasible solution area is extremely reduced, and the interconnectedness of the feasible solution area is more serious, which makes it difficult for the algorithm to cross the infeasible area to find the optimal feasible individual that takes into account both convergence and distribution.

[0006] In summary, the existing water projects and the coordinated optimization among flood control, water supply, ecology and power generation functions lack an efficient constraint processing mechanism. Summary of the invention

[0007] The purpose of the present invention is to provide a multi-objective collaborative scheduling method and system for a cross-domain water project group to solve the above-mentioned technical problems.

[0008] To achieve the above object, the present invention provides a multi-objective coordinated scheduling method for a cross-domain water project group, comprising the following steps: S1. Collect parameter information of a cross-domain water project group, and build a multi-objective scheduling model of a cross-domain water project group based on the collected parameter information; S2, based on the construction of a multi-objective scheduling model for a cross-domain water project group, the optimal scheduling scheme for the water project group is solved by using a multi-group cooperative search method; S3. Output the optimal scheduling plan.

[0009] Preferably, in step S1, the parameter information of the cross-domain water project group includes: the total number of water projects , the total number of time periods during the scheduling period , No. The power generation coefficient of each hydropower project , No. Water project Power generation flow in the period , No. Water projects in Power generation head , Time-controlled water demand at the site , Ecological Section Control Station In time period Required ecological flow ,in , Indicates the total number of ecological section control stations; The objectives of the multi-objective scheduling model for cross-domain water project groups include maximizing the total power generation of the water project group, minimizing the water shortage rate, and minimizing the ecological flow deviation; The constraints of the multi-objective scheduling model of cross-domain water project groups include water balance constraints, hydraulic connection constraints between water projects, initial and final water level constraints, water level constraints, power generation flow constraints, output constraints and outflow constraints: The decision variable of the multi-objective scheduling model of cross-domain water project groups is the outflow flow during the water project scheduling period.

[0010] Preferably, the objective function expression for maximizing the total power generation of the water project group is as follows: (1) (2) In the formula, is the total power generation of the hydropower project group during the entire dispatch period, in kW·h; For the Water project Output of the time period, in kW, where , ; for The calculation period is in hours, the unit is h; The objective function expression for minimizing water shortage rate is as follows: (3) In the formula, is the water shortage rate during the entire dispatch period; Indicates the control site in the time period The flow rate, in m 3 / s; The objective function expression of the minimum ecological flow deviation is as follows: (4) In the formula, is the ecological flow deviation during the entire scheduling period; Ecological Section Control Station In time period The flow rate, in m 3 / s.

[0011] Preferably, the water balance constraint expression is as follows: (5) In the formula, , Respectively Water project Initial and final storage capacity of the time period, in m 3 ; For the Water project Inflow flow rate during the period, in m 3 / s; For the Water project Total outflow during the period, in m 3 / s; The hydraulic connection constraint expression between water projects is as follows: (6) In the formula, For the Upstream water projects connected to water projects In the The outbound flow rate during the period, in m 3 / s, and ; For the Water project The interval flow rate of the time period, in m 3 / s; For the The number of all upstream water projects directly connected to the water project; The initial and final water level constraint expressions are as follows: (7) In the formula, and All are water projects The initial water level at the beginning of the time period, in m; and They are the final water levels at the end of the time period, in m; The water level constraint expression is as follows: (8) In the formula, and Respectively Water project The minimum and maximum water levels allowed in the time period, in meters; For the Water project The water level of the time period, in m; The power generation flow constraint expression is as follows: (9) In the formula, and Respectively Water project The minimum and maximum power generation flow in the time period, in m 3 / s; The output constraint expression is as follows: (10) In the formula, and Respectively Water project The minimum and maximum power generation output of the time period, in kW; The outbound flow constraint expression is as follows: (11) In the formula, , They are the minimum outbound flow and the maximum outbound flow, in m 3 / s.

[0012] Preferably, step S2 specifically includes the following steps: S21, Initialization: Initialize the number of iterations k =1, and determine the maximum number of iterations , and set global adjustment parameters , and represents the number of elite storage is the default parameter; S22. According to the preset range of decision variables Generate initial optimization population , and calculate the fitness of individuals in the initial optimization population ; S23, initial optimization population Perform reference point based environment selection Get the first level of the Pareto frontier population individuals, and regard the first level of the Pareto frontier population individuals as the first generation of initialized elite population ; S24, implement multi-population adaptation strategy: based on the complexity of the solution model, the initial optimization population Adaptive populations are divided into three subpopulation categories: main population, support population, and auxiliary population, and environmental selection strategies are assigned to the main population, support population, and auxiliary population respectively; S25, executing the evolutionary strategy of multi-population cooperative search algorithm: executing the multi-factor set optimization strategy and the population density diffusion strategy in sequence; S26, executing the search characteristic optimization strategy: dynamically selecting according to the search characteristics of the multi-factor set optimization strategy and the population density diffusion strategy to determine the optimal population; S27, determine whether the maximum number of iterations has been reached If yes, then output the preferred population described in step S26, otherwise return to step S24.

[0013] Preferably, in step S22, the decision variable initialization expression is as follows: (12) In the formula, For the Individuals of a population The value of the decision variable; Indicates that in the interval Functions for generating uniformly distributed random numbers; is the total number of individuals in the population; is the total number of decision variables; In step S24, the optimal population expression is as follows: (13) In the formula, To find the best population; , , , , They are the main population, the first supporting population, the second supporting population, the first auxiliary population, and the second auxiliary population respectively; M is the number of targets in the multi-objective scheduling model of the cross-domain water project group; (14) (15) (16) In the formula, , , The number of individuals in the population used for optimization in three types of sub-populations: the main population, the supporting population, and the auxiliary population; is the number of subpopulations; To support populations; the degree of constraint violation for the dominant population; is an integer upward function; The main population adopts an environmental selection strategy based on the reference point, the first support population and the second support population both adopt an environmental selection strategy that considers the reference point and crowding distance constraints, and the first auxiliary population and the second auxiliary population both adopt an environmental selection strategy that does not consider constraints: (17) (18) (19) (20) (twenty one) (twenty two) In the formula, Represents the first Offspring individuals; , , , They represent the environment selection strategy based on reference point, the environment selection strategy based on crowding distance, the environment selection strategy based on reference point without considering constraints, and the environment selection strategy based on crowding distance without considering constraints; , They represent the evolutionary strategies of updating offspring individuals using multi-population cooperative search strategy and genetic method respectively; , Respectively represent The iteration and The main population obtained in the iteration; , Respectively represent The iteration and The first support population obtained in the iteration; , Respectively represent The iteration and The second support population obtained in the iteration; , Respectively represent The iteration and The first auxiliary population obtained in the iteration; , Respectively represent The iteration and The second auxiliary population obtained in the iteration.

[0014] Preferably, the multi-factor set optimization strategy described in step S25 is: introducing three types of learning parameters, namely, edge optimal solution learning factor A, mean optimal learning factor B, and random optimal learning factor C, into the convergent population, and using normal distribution to obtain an updated convergent population: (twenty three) In the formula, For the The strategy generated by the multi-population cooperative search strategy in the first iteration Individuals of a population The value of the decision variable; Indicates In the first iteration, the first Population individual d The value of the decision variable; Indicated in The marginal optimal learning factor obtained during the iteration process; Indicated in The first The first Mean learning factor for decision variables; Indicated in The first The first The random learning factor of decision variables and are the current number of iterations and the maximum number of iterations respectively; Among them, the edge optimal solution learning factor A is updated using the optimal edge optimal learning factor on each target: (twenty four) In the formula, represents the global optimal population; It is in In the iteration, by The optimal population of the generation and Elite population of the generation , using environmental non-dominated sorting according to the Pareto front level as the first level sorting function Select the best individual obtained; express The optimal individual corresponding to the target is Indicates the total number of targets; is a random integer in the target dimension range [0,1]; Indicates randomly selecting a target dimension The optimal individual for each decision variable; is a random number that follows the [0,1] distribution; The mean optimal learning factor B is updated based on the mean distribution learning strategy of the current best and worst populations: (25) In the formula, Indicates the creation of a random function that follows a normal distribution; represents the number of individuals in the global optimal population; represents the first decision variables; The random optimal learning factor C adopts the mutation perturbation strategy for updating: (26) In the formula, Represents the number of global optimal populations Randomly select an individual from the population decision variables; The expression of population density diffusion strategy is as follows: (27) In the formula, Indicated in The first generation of the population density diffusion strategy in the iterative process The individual The value of the decision variable; and Indicates the In the first iteration, the first and The first d The value of the decision variable, and represents the number of } are two integers chosen randomly.

[0015] Preferably, the preferred population expression in step S26 is as follows: (28) In the formula, For the The offspring population is selected during the iteration and is considered as the preferred population.

[0016] A multi-objective collaborative scheduling system for a cross-domain water project group is used to execute a multi-objective collaborative scheduling method for a cross-domain water project group, which includes a water project model initialization module and a water project multi-objective scheduling multi-group cooperative search method solution module, wherein the water project model initialization module is used to construct a cross-domain water project group multi-objective scheduling model based on parameter information of the cross-domain water project group, with the objectives of maximizing the total power generation of the water project group, minimizing the water shortage rate and minimizing the ecological flow deviation, with water balance constraints, hydraulic connection constraints between water projects, initial and final water level constraints, water level constraints, power generation flow constraints, output constraints and outbound flow constraints as constraints, and with the outbound flow during the water project scheduling period as a decision variable; The water project multi-objective scheduling multi-group cooperative search method solution module is used to solve the optimal scheduling plan for water project groups based on the multi-group cooperative search method.

[0017] Preferably, the water project multi-objective scheduling multi-population cooperative search method solution module includes an initialization unit, a multi-population adaptation strategy unit, a multi-population cooperative search algorithm evolution strategy unit, a search characteristic optimization strategy unit, and a water project optimal scheduling solution acquisition unit; Among them, the initialization unit is used to obtain the initialized elite population; The multi-group adaptation strategy unit is used to execute the multi-group adaptation strategy; The multi-population cooperative search algorithm evolution strategy unit is used to execute the multi-population cooperative search algorithm evolution strategy; The search feature optimization strategy unit is used to execute the search feature optimization strategy; The water project optimal scheduling scheme acquisition unit is used to output the final preferred population and use the final preferred population as the optimal scheduling scheme for the water project group.

[0018] Therefore, the present invention adopts the above-mentioned multi-objective coordinated scheduling method and system for a cross-domain water project group, which has the following beneficial effects: 1. It has the advantages of simple implementation, excellent optimization efficiency, strong global search capability and high feasibility, and is suitable for a wide range of application scenarios. When dealing with complex optimization and scheduling problems of cross-domain water conservancy projects, it can maintain high optimization efficiency in complex water conservancy systems, avoid falling into local optimal solutions, and achieve optimal allocation of water resources, providing a more effective technical means for solving multi-objective engineering optimization problems of cross-domain water conservancy projects. 2. By implementing multiple population adaptation strategies, the model operation efficiency and global optimization capability can be significantly improved, and the algorithm can be reduced to a certain extent to find the optimal solution dilemma in complex models, which improves the feasibility and practicality of optimization decision-making, and demonstrates excellent adaptability and powerful optimization performance in complex and changing application environments; 3. By executing the evolutionary strategy of multi-population cooperative search algorithm, the information of excellent solutions in the population is used to conduct global search in a huge search space, which enhances the convergence and distribution of the solution set and avoids falling into the local optimum, so as to quickly and efficiently find a better and feasible scheduling solution; 4. By introducing intelligent optimization algorithms, dynamic adjustment of scheduling schemes can be achieved under multiple objectives and multiple constraints to ensure the balance of various objective functions in the global optimal solution.

[0019] In summary, the present invention achieves a balance between global search and local development through strategies such as division of labor and cooperative search of multiple population adaptation strategies, and to a certain extent reduces the dilemma of the algorithm being unable to find the optimal solution under complex models, improves the feasibility and practicality of optimal decision-making, and demonstrates excellent adaptability and powerful optimization performance in complex and changeable application environments, achieving optimal allocation of water resources, and providing a more effective technical means for solving cross-domain multi-objective engineering optimization problems of water projects. It not only improves the accuracy and operability of the scheduling plan, but also significantly improves the adaptability when facing complex water conservancy systems, and provides solid technical support for the sustainable management of water resources.

[0020] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of a multi-objective collaborative scheduling method for a cross-domain water project group according to the present invention; Figure 2 A two-dimensional plane projection diagram of the Pareto front of the present invention and the comparative method described in the embodiment under two scheduling objectives in the scheduling engineering problem; Figure 3 A three-dimensional graph of the Pareto front of the present invention and the comparative method described in the embodiment under three scheduling objectives in the scheduling engineering problem; Figure 4 The water level variation diagram of the water project 1 described in the embodiment is selected based on the power generation target sorting of 20 solution sets; Figure 5 The water level variation diagram of the water project 2 described in the embodiment - selecting 20 solution sets based on the power generation target ranking; Figure 6 The water level variation diagram of the water project 3 described in the embodiment - selecting 20 solution sets based on the power generation target ranking; Figure 7 The water level variation diagram of the water project 4 described in the embodiment - selecting 20 solution sets based on the power generation target ranking; Figure 8 This is a water level variation diagram of the water project 5 described in the embodiment, in which 20 solution sets are selected based on the power generation target sorting. DETAILED DESCRIPTION

[0022] In the description of the present invention, it should be noted that the terms "upper", "lower", "inside", "outside", etc. indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, or the positions or positional relationships in which the invented product is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In the description of the present invention, it should also be noted that, unless otherwise clearly specified and limited, the terms "setting", "installation", and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0023] The embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0024] like Figure 1 As shown, a multi-objective coordinated scheduling method for a cross-domain water project group includes the following steps: S1. Collect parameter information of a cross-domain water project group, and build a multi-objective scheduling model of a cross-domain water project group based on the collected parameter information; In step S1, the parameter information of the cross-domain water project group includes: the total number of water projects , the total number of time periods during the scheduling period , No. The power generation coefficient of each hydropower project , No. Water project Power generation flow in the period , No. Water projects in Power generation head , Time-controlled water demand at the site , Ecological Section Control Station In time period Required ecological flow ,in , Indicates the total number of ecological section control stations; The objectives of the multi-objective scheduling model for cross-domain water project groups include maximizing the total power generation of the water project group, minimizing the water shortage rate, and minimizing the ecological flow deviation; The constraints of the multi-objective scheduling model of cross-domain water project groups include water balance constraints, hydraulic connection constraints between water projects, initial and final water level constraints, water level constraints, power generation flow constraints, output constraints and outflow constraints: The decision variable of the multi-objective scheduling model of cross-domain water project groups is the outflow flow during the water project scheduling period.

[0025] For most water projects, maximizing power generation benefits is generally an important goal of their optimized scheduling. As the proportion of hydropower in the power system increases, the demand for power output during the dry season also increases. Therefore, for the power generation target, the objective function expression for maximizing the total power generation of the water project group is as follows: (1) (2) In the formula, is the total power generation of the hydropower project group during the entire dispatch period, in kW·h; For the Water project Output of the time period, in kW, where , ; for The calculation period is in hours, and the unit is h. From the perspective of the supply side (water project group) and the demand side, the minimum water shortage rate is selected to minimize the degree of water shortage in water supply. The objective function expression of the minimum water shortage rate is as follows: (3) In the formula, is the water shortage rate during the entire dispatch period; Indicates the control site in the time period The flow rate, in m 3 / s; Ecological flow is the flow required by the river ecosystem to perform different ecological functions. There is a critical point for ecological flow. Once the flow exceeds the critical point, whether it is too high or too low, it will affect the ecological balance. The objective function expression for the minimum deviation of ecological flow is as follows: (4) In the formula, is the ecological flow deviation during the entire scheduling period; Ecological Section Control Station In time period The flow rate, in m 3 / s.

[0026] Preferably, the water balance constraint expression is as follows: (5) In the formula, , Respectively Water project Initial and final storage capacity of the time period, in m 3 ; For the Water project Inflow flow rate during the period, in m 3 / s; For the Water project Total outflow during the period, in m 3 / s; The hydraulic connection constraint expression between water projects is as follows: (6) In the formula, For the Upstream water projects connected to water projects In the The outbound flow rate during the period, in m 3 / s, and ; For the Water project The interval flow rate of the time period, in m 3 / s; For the The number of all upstream water projects directly connected to the water project; The initial and final water level constraint expressions are as follows: (7) In the formula, and All are water projects The initial water level at the beginning of the time period, in m; and They are the final water levels at the end of the time period, in m; The water level constraint expression is as follows: (8) In the formula, and Respectively Water project The minimum and maximum water levels allowed in the time period, in meters; For the Water project The water level of the time period, in m; The power generation flow constraint expression is as follows: (9) In the formula, and Respectively Water project The minimum and maximum power generation flow in the time period, in m 3 / s; The output constraint expression is as follows: (10) In the formula, and Respectively Water project The minimum and maximum power generation output of the time period, in kW; The outbound flow constraint expression is as follows: (11) In the formula, , They are the minimum outbound flow and the maximum outbound flow, in m 3 / s.

[0027] S2, based on the construction of a multi-objective scheduling model for a cross-domain water project group, the optimal scheduling scheme for the water project group is solved by using a multi-group cooperative search method; Step S2 specifically includes the following steps: S21. Initialization: Initialize the number of iterations and determine the maximum number of iterations , and set global adjustment parameters , and represents the number of elite storage is the default parameter; S22. According to the preset range of decision variables Generate initial optimization population , and calculate the fitness of individuals in the initial optimization population ; In step S22, the decision variable initialization expression is as follows: (12) In the formula, For the Individuals of a population The value of the decision variable; Indicates that in the interval Functions for generating uniformly distributed random numbers; is the total number of individuals in the population; is the total number of decision variables; S23, initial optimization population Perform reference point based environment selection Get the first level of the Pareto frontier population individuals, and regard the first level of the Pareto frontier population individuals as the first generation of initialized elite population ; S24, executing multiple population adaptation strategies: based on the complexity of the solution model, the initial optimal population is adaptively divided into three sub-population categories: main population, support population and auxiliary population, and environmental selection strategies are assigned to the main population, support population and auxiliary population respectively; In step S24, the optimal population expression is as follows: (13) In the formula, To find the best population; , , , , They are the main population, the first supporting population, the second supporting population, the first auxiliary population, and the second auxiliary population respectively; M is the number of targets in the multi-objective scheduling model of the cross-domain water project group; (14) (15) (16) In the formula, , , The number of individuals in the population used for optimization in three types of sub-populations: the main population, the supporting population, and the auxiliary population; is the number of subpopulations; To support populations; is the constraint violation degree of the main population. If the constraint violation degree is 0, it means that the solution does not violate the model constraints. is an integer upward function; The main population adopts an environmental selection strategy based on the reference point, the first support population and the second support population both adopt an environmental selection strategy that considers the reference point and crowding distance constraints, and the first auxiliary population and the second auxiliary population both adopt an environmental selection strategy that does not consider constraints: (17) (18) (19) (20) (twenty one) (twenty two) In the formula, Represents the first Offspring individuals; , , , They represent the environment selection strategy based on reference point, the environment selection strategy based on crowding distance, the environment selection strategy based on reference point without considering constraints, and the environment selection strategy based on crowding distance without considering constraints; , They represent the evolutionary strategies of updating offspring individuals using multi-population cooperative search strategy and genetic method respectively; , Respectively represent The iteration and The main population obtained in the iteration; , Respectively represent The iteration and The first support population obtained in the iteration; , Respectively represent The iteration and The second support population obtained in the iteration; , Respectively represent The iteration and The first auxiliary population obtained in the iteration; , Respectively represent The iteration and The second auxiliary population obtained in the iteration.

[0028] S25. Execute the evolutionary strategy of multi-population cooperative search algorithm: sequentially execute the multi-factor set optimization strategy (to fully consider and measure the high-quality information in the population optimization process) and the population density diffusion strategy (taking into account the population clustering search phenomenon caused by the rapid convergence of the population in the optimization process, the population density diffusion strategy is introduced. Through the density diffusion strategy based on population density, the population density is reduced, the population exploration range is increased, and the organic combination of convergence efficiency and population diversity is achieved); use the information of excellent solutions in the population to conduct a global search in a huge search space, enhance the convergence and distribution of the solution set, avoid falling into the local optimum, and thus quickly and efficiently find a better and feasible solution set.

[0029] The multi-factor set optimization strategy described in step S25 is: introducing three types of learning parameters, namely, edge optimal solution learning factor A, mean optimal learning factor B, and random optimal learning factor C, into the convergent population, and using normal distribution to obtain an updated convergent population: (twenty three) In the formula, For the In the first iteration, the first Individuals of a population The value of the decision variable; Indicates In the first iteration, the first Population individual d The value of the decision variable; Indicated in The marginal optimal learning factor obtained during the iteration process; Indicated in The first The first Mean learning factor for decision variables; Indicated in The first The first Machine learning factors for decision variables; and are the current number of iterations and the maximum number of iterations respectively; Among them, the edge optimal solution learning factor A is updated by using the optimal edge optimal learning factor on each target. The edge optimal solution learning factor A enables the population to spread the search area at a faster speed in the early stage of the search: (twenty four) In the formula, represents the global optimal population; It is in In the iteration, by The optimal population of the generation and Elite population of the generation , using environmental non-dominated sorting according to the Pareto front level as the first level sorting function Select the best individual obtained; express The optimal individual corresponding to the target is Indicates the total number of targets; is a random integer in the target dimension range [0,1]; Indicates randomly selecting a target dimension The optimal individual for each decision variable; is a random number that follows the [0,1] distribution; The mean optimal learning factor B is updated based on the mean distribution learning strategy of the current best and worst populations, which can effectively guide the population to jump out of the local optimal pseudo-Pareto optimal solution: (25) In the formula, Indicates the creation of a random function that follows a normal distribution; represents the number of individuals in the global optimal population; represents the first decision variables; The random optimal learning factor C is updated using a mutation perturbation strategy to increase local exploration capabilities: (26) In the formula, Represents the number of global optimal populations Randomly select an individual from the population decision variables; The expression of population density diffusion strategy is as follows: (27) In the formula, Indicated in The first generation of the population density diffusion strategy in the iterative process The individual The value of the decision variable; and Respectively represent In the first iteration, the first and The first d The value of the decision variable, and represents the number of } are two integers chosen randomly.

[0030] S26, executing the search characteristic optimization strategy: dynamically selecting according to the search characteristics of the multi-factor set optimization strategy and the population density diffusion strategy to determine the optimal population; The preferred population expression in step S26 is as follows: (28) In the formula, For the The offspring population is selected during the iteration and is considered as the preferred population.

[0031] S27, determine whether the maximum number of iterations has been reached If yes, then output the preferred population described in step S26, otherwise return to step S24.

[0032] S3. Output the optimal scheduling plan.

[0033] A multi-objective collaborative scheduling system for a cross-domain water project group is used to execute a multi-objective collaborative scheduling method for a cross-domain water project group, which includes a water project model initialization module and a water project multi-objective scheduling multi-population cooperative search method solution module, wherein the water project model initialization module is used to construct a cross-domain water project group multi-objective scheduling model based on parameter information of the cross-domain water project group, with the goals of maximizing the total power generation of the water project group, minimizing the water shortage rate and minimizing the ecological flow deviation, with water balance constraints, hydraulic connection constraints between water projects, initial and final water level constraints, water level constraints, power generation flow constraints, output constraints and outbound flow constraints as constraints, and with the outbound flow during the water project scheduling period as a decision variable; the water project multi-objective scheduling multi-population cooperative search method solution module is used to solve the optimal scheduling plan for the water project group based on the multi-population cooperative search method.

[0034] Specifically, the multi-population cooperative search method solution module for multi-objective scheduling of water projects includes an initialization unit, a multi-population adaptation strategy unit, a multi-population cooperative search algorithm evolution strategy unit, a search feature optimization strategy unit, and a water project optimal scheduling solution acquisition unit; wherein the initialization unit is used to obtain an initialized elite population; the multi-population adaptation strategy unit is used to execute the multi-population adaptation strategy; the multi-population cooperative search algorithm evolution strategy unit is used to execute the multi-population cooperative search algorithm evolution strategy; the search feature optimization strategy unit is used to execute the search feature optimization strategy; and the water project optimal scheduling solution acquisition unit is used to output the final optimized population, and use the final optimized population as the optimal scheduling solution for the water project group.

[0035] Example The joint dispatching in the dry season needs to meet the water supply flow of the downstream section of the water project group as the goal, taking into account the multi-objective needs of ecology, power generation, etc., to maximize the benefits of comprehensive utilization of water resources. Taking the water project group consisting of five water projects in a certain basin as the research object, the three-objective hybrid water project group optimization dispatching model with the maximum total power generation, the minimum water shortage rate and the minimum ecological flow deviation of the constructed water project group (where the ecological flow demand takes the downstream site as the control section) is used to study the water dispatching in the dry season using the present invention, the third-generation non-dominated sorting genetic algorithm (NSGAIII) and the adaptive third-generation non-dominated sorting genetic algorithm (ANSGAIII).

[0036] Based on the long series of dry season runoff data of the control station, a frequency analysis was conducted, and a 75% water frequency scenario was set. When the water storage rate of the water project group was 100% at the beginning of the scheduling period, the scheduling was carried out on a ten-day scale from November to March during the dry season. There are 15 scheduling periods, each with an interval of 10 days. The constraint variables include water level, outflow flow, power generation flow, final water level equation, output and other constraints.

[0037] In order to compare and analyze the solution stability, solution efficiency and solution optimization breadth of the present invention, NSGAIII and ANSGAIII in the actual multi-objective scheduling problem of water projects, 20 independent calculations were performed on the multi-objective scheduling problem of water project groups, and the Pareto fronts obtained by the above three algorithms were used to sort the non-dominated solutions to obtain the ideal Pareto front.

[0038] The Pareto front obtained by each method and the ideal Pareto front are normalized and the hypervolume (HV) index is calculated to evaluate the convergence and distribution of the solution set. In particular, for problems with unknown Pareto fronts, the larger the HV, the better the distribution of the scheduling solution set obtained by this method, and the corresponding non-dominated solution set is closer to the real Pareto front. Therefore, the HV index can be used to compare the differences between the present invention and NSGAIII and ANSGAIII in practical applications.

[0039] Result analysis: First, by calculating and analyzing the average HV index of the three methods, as shown in Table 1, it can be seen that the average HV index of the present invention in the actual scheduling scenario is better than that of the comparison algorithm. Therefore, the stability and optimization breadth of the present invention have significant advantages over NSGAIII and ANSGAIII in the multi-objective optimization scheduling problem, indicating that the present invention has the best convergence and distribution when solving the multi-objective optimization problem of water engineering group operation, and the performance is very excellent. Therefore, the present invention has stronger exploration ability, search efficiency and solution efficiency for solving the multi-objective optimization problem of water engineering.

[0040] Table 1 Evaluation index values ​​obtained by different algorithms ;

[0041] At the same time, in order to compare the differences in the objective functions of the present invention, NSGAIII and ANSGAIII in the obtained solution sets, three-dimensional graphic visualization and two-dimensional plane projection visualization are used to evaluate the performance of different algorithms from three aspects: the dominance relationship of the Pareto front, the convergence of the solution set and the distribution of the solution set. Convergence and distribution are measured based on the range of the target value and the population distribution on the coordinate axis, and the Pareto front dominance relationship is measured based on the positional relationship between the solution sets. A wider and more uniform solution set and a better Pareto front dominance relationship indicate better convergence performance. Therefore, the scheme corresponding to the median HV value is selected from 20 independent calculations of each method and plotted as shown below. Figure 2 and Figure 3 From the Pareto frontier graph shown, it can be seen that: in the actual scheduling case study, compared with the other two algorithms, the Pareto frontier represented by the present invention is closer to the ideal Pareto frontier than the Pareto frontier represented by NSGAIII and ANSGAIII, and the Pareto frontiers obtained by the two algorithms NSGAIII and ANSGAIII are dominated by the results obtained by the present invention, which can be explained that the convergence of the present invention is better than that of NSGAIII and ANSGAIII, proving that it has better convergence ability in actual scheduling problems, and the solutions are more diverse and more representative. Therefore, the present invention performs well in terms of convergence and distribution. At the same time, the point distribution of the present invention on the power generation target is significantly better than that of other algorithms, indicating that the distribution of the present invention is stronger than that of other algorithms.

[0042] The solution set corresponding to the median HV is selected from the 20 results obtained by the present invention to analyze the water level change process, and the 20 solutions (NO.1-NO.20) with the largest total power generation in the solution set are selected and arranged in ascending order of total power generation. The water level change process of the above solutions on the water project is plotted as follows: Figure 4-Figure 8 As shown by Figure 4-Figure 8It can be seen that in the present invention, under the condition of 75% water inflow frequency in the dry season (slightly dry year), when the water project group starts to adjust from the normal water storage level and is constrained to the flood limit water level at the end of the scheduling period, all the water shortage rates obtained by the present invention are 0, indicating that even in a slightly dry year, the water storage capacity of the water project group at the end of the flood season can guarantee the pressure saline flow, meet the water supply demand of downstream sites, and provide a variety of effective alternatives for scheduling decisions to make the best decision plan. In order to meet the downstream water supply demand and increase the power generation efficiency, the water level of each water project, except for water project 1 and water project 2, shows a significant downward trend in the entire operation. As the most upstream water projects, the outflow of water project 1 (NO.1) and water project 2 (NO.2) is directly related to the inflow of downstream water projects 4 (NO.4) and water project 5 (NO.5). In order to ensure the water inflow of water projects 4 (NO.4) and water projects 5 (NO.5), the operating water level of water project 1 (NO.1) dropped significantly in the initial operation period, so the outflow was significantly higher than other periods; but in order to meet the water level constraint at the end of the dispatch period, the outflow was reduced in the later period so that the water level was stored back to the flood season limit water level or the initial water level. On the basis of ensuring the lower limit of the different pressure-salinity flow of the downstream section, under the 75% water inflow frequency, the water level change process of the water project under each scheme is basically consistent in trend. Since water project 5 (NO.5) is located at the most downstream of the basin, its water level operation is affected by the upstream water inflow, and the water level under each solution scheme is quite different. In addition, as a daily regulating water project, its reservoir capacity is small and its water storage capacity is limited, so the fluctuation range is large, but in most periods it can maintain a high operating head and increase the power generation efficiency. In terms of the speed of water level drop, the water level of water project 1 (NO.1) dropped fastest in November, water project 2 (NO.2) and water project 3 (NO.3) in November-December, water project 4 (NO.4) in December-February, and water project 5 (NO.5) in March. The water level of upstream water projects (water project 1 (NO.1), water project 2 (NO.2) and water project 3 (NO.3)) dropped faster in the first half of the dry season, while the water level of downstream water projects (water project 4 (NO.4) and water project 5 (NO.5)) dropped faster in the second half of the dispatching period. All water projects can meet the water level constraints before the flood season, do not face the pressure of water level drop, and can maximize the power generation benefits while completing the water level drop task.

[0043] In summary, it is proved that the method of the present invention enables a group of water projects to consider multiple conflicting objectives at the same time, such as maximizing hydropower generation revenue or improving ecological flow guarantee rate while ensuring water supply stability. In addition, the multi-objective optimization method of the present invention can provide managers with a series of feasible solutions to help find the best balance between different objectives.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A multi-objective coordinated scheduling method for a cross-domain water project group, characterized by: The following steps are involved: S1. Collect parameter information of a cross-domain water project group, and build a multi-objective scheduling model of a cross-domain water project group based on the collected parameter information; S2, based on the construction of a multi-objective scheduling model for a cross-domain water project group, the optimal scheduling scheme for the water project group is solved by using a multi-group cooperative search method; S3. Output the optimal scheduling plan.

2. A multi-objective coordinated scheduling method for a cross-domain water project group according to claim 1, characterized in that: In step S1, the parameter information of the cross-domain water project group includes: the total number of water projects , the total number of time periods during the scheduling period , No. The power generation coefficient of each hydropower project , No. Water project Power generation flow in the period , No. Water projects in Power generation head , Time-controlled water demand at the site , Ecological Section Control Station In time period Required ecological flow ,in , Indicates the total number of ecological section control stations; The objectives of the multi-objective scheduling model for cross-domain water project groups include maximizing the total power generation of the water project group, minimizing the water shortage rate, and minimizing the ecological flow deviation; The constraints of the multi-objective scheduling model of cross-domain water project groups include water balance constraints, hydraulic connection constraints between water projects, initial and final water level constraints, water level constraints, power generation flow constraints, output constraints and outflow constraints: The decision variable of the multi-objective scheduling model of cross-domain water project groups is the outflow flow during the water project scheduling period.

3. A multi-objective coordinated scheduling method for a cross-domain water project group according to claim 2, characterized in that: The objective function expression for maximizing the total power generation of the hydropower project group is as follows: (1) (2) In the formula, is the total power generation of the hydropower project group during the entire dispatch period, in kW·h; For the Water project Output of the time period, in kW, where , ; for The calculation period is in hours, the unit is h; The objective function expression of minimizing water shortage rate is as follows: (3) In the formula, is the water shortage rate during the entire dispatch period; Indicates the control site in the time period The flow rate, in m 3 / s; The objective function expression of the minimum ecological flow deviation is as follows: (4) In the formula, is the ecological flow deviation during the entire scheduling period; Ecological Section Control Station In time period The flow rate, in m 3 / s.

4. A multi-objective coordinated scheduling method for a cross-domain water project group according to claim 3, characterized in that: The water balance constraint expression is as follows: (5) In the formula, , Respectively Water project Initial and final storage capacity of the time period, in m 3 ; For the Water project Inflow flow rate during the period, in m 3 / s; For the Water project Total outflow during the period, in m 3 / s; The hydraulic connection constraint expressions between water projects are as follows: (6) In the formula, For the Upstream water projects connected to water projects In the The outbound flow rate during the period, in m 3 / s, and ; For the Water project The interval flow rate of the time period, in m 3 / s; For the The number of all upstream water projects directly connected to the water project; The initial and final water level constraint expressions are as follows: (7) In the formula, and All are water projects The initial water level at the beginning of the time period, in m; and They are the final water levels at the end of the time period, in m; The water level constraint expression is as follows: (8) In the formula, and Respectively Water project The minimum and maximum water levels allowed in the time period, in meters; For the Water project The water level of the time period, in m; The power generation flow constraint expression is as follows: (9) In the formula, and Respectively Water project The minimum and maximum power generation flow in the time period, in m 3 / s; The output constraint expression is as follows: (10) In the formula, and Respectively Water project The minimum and maximum power generation output of the time period, in kW; The outbound flow constraint expression is as follows: (11) In the formula, , They are the minimum outbound flow and the maximum outbound flow, in m 3 / s.

5. A multi-objective coordinated scheduling method for a cross-domain water project group according to claim 4, characterized in that: Step S2 specifically includes the following steps: S21, Initialization: Initialize the number of iterations k =1, and determine the maximum number of iterations , and set global adjustment parameters , and represents the number of elite storage is the default parameter; S22. According to the preset range of decision variables Generate initial optimization population , and calculate the fitness of individuals in the initial optimization population ; S23, initial optimization population Perform reference point based environment selection Get the first level of the Pareto frontier population individuals, and regard the first level of the Pareto frontier population individuals as the first generation of initialized elite population ; S24, implement multi-population adaptation strategy: based on the complexity of the solution model, the initial optimization population Adaptive populations are divided into three subpopulation categories: main population, support population, and auxiliary population, and environmental selection strategies are assigned to the main population, support population, and auxiliary population respectively; S25, executing the evolutionary strategy of multi-population cooperative search algorithm: executing the multi-factor set optimization strategy and the population density diffusion strategy in sequence; S26, executing the search characteristic optimization strategy: dynamically selecting according to the search characteristics of the multi-factor set optimization strategy and the population density diffusion strategy to determine the optimal population; S27, determine whether the maximum number of iterations has been reached If yes, then output the preferred population described in step S26, otherwise return to step S24.

6. A multi-objective coordinated scheduling method for a cross-domain water project group according to claim 5, characterized in that: In step S22, the decision variable initialization expression is as follows: (12) In the formula, For the Individuals of a population The value of the decision variable; Indicates that in the interval Functions for generating uniformly distributed random numbers; is the total number of individuals in the population; is the total number of decision variables; In step S24, the optimal population expression is as follows: (13) In the formula, To find the best population; , , , , They are the main population, the first supporting population, the second supporting population, the first auxiliary population, and the second auxiliary population respectively; M is the number of targets in the multi-objective scheduling model of the cross-domain water project group; (14) (15) (16) In the formula, , , The number of individuals in the population used for optimization in three types of sub-populations: the main population, the supporting population, and the auxiliary population; is the number of subpopulations; To support populations; the degree of constraint violation for the dominant population; is an integer upward function; The main population adopts an environmental selection strategy based on the reference point, the first support population and the second support population both adopt an environmental selection strategy that considers the reference point and crowding distance constraints, and the first auxiliary population and the second auxiliary population both adopt an environmental selection strategy that does not consider constraints: (17) (18) (19) (20) (21) (22) In the formula, Represents the first Offspring individuals; , , , They represent the environment selection strategy based on reference point, the environment selection strategy based on crowding distance, the environment selection strategy based on reference point without considering constraints, and the environment selection strategy based on crowding distance without considering constraints; , They represent the evolutionary strategies of updating offspring individuals using multi-population cooperative search strategy and genetic method respectively; , Respectively represent The iteration and The main population obtained in the iteration; , Respectively represent The iteration and The first support population obtained in the iteration; , Respectively represent The iteration and The second support population obtained in the iteration; , Respectively represent The iteration and The first auxiliary population obtained in the iteration; , Respectively represent The iteration and The second auxiliary population obtained in the iteration.

7. A multi-objective coordinated scheduling method for a cross-domain water project group according to claim 6, characterized in that: The multi-factor set optimization strategy described in step S25 is: introducing three types of learning parameters, namely, edge optimal solution learning factor A, mean optimal learning factor B, and random optimal learning factor C, into the convergent population, and using normal distribution to obtain an updated convergent population: (23) In the formula, For the The first generation of Population strategy d The value of the decision variable; Indicates In the first iteration, the first Population individual d The value of the decision variable; Indicated in The marginal optimal learning factor obtained during the iteration process; Indicated in The first The first Mean learning factor for decision variables; Indicated in The first The first random learning factors for decision variables; and is the current number of iterations and the maximum number of iterations; Among them, the edge optimal solution learning factor A is updated using the optimal edge optimal learning factor on each target: (24) In the formula, represents the global optimal population; It is in In the iteration, by The optimal population of the generation and Elite population of the generation , using environmental non-dominated sorting according to the Pareto front level as the first level sorting function Select the best individual obtained; express The optimal individual corresponding to the target is Indicates the total number of targets; The target dimension range A random integer within ; Indicates randomly selecting a target dimension The optimal individual for each decision variable; is a random number that follows the [0,1] distribution; The mean optimal learning factor B is updated based on the mean distribution learning strategy of the current best and worst populations: (25) In the formula, Indicates the creation of a random function that follows a normal distribution; represents the number of individuals in the global optimal population; represents the first decision variables; The random optimal learning factor C adopts the mutation perturbation strategy for updating: (26) In the formula, Indicates the number of global optimal populations Randomly select an individual from the population decision variables; The expression of population density diffusion strategy is as follows: (27) In the formula, Indicated in The first generation of the population density diffusion strategy in the iterative process The individual The value of the decision variable; and Respectively represent In the first iteration, the first and The first d The value of the decision variable, and represents the number of } are two integers chosen randomly.

8. A multi-objective coordinated scheduling method for a cross-domain water project group according to claim 7, characterized in that: The preferred population expression in step S26 is as follows: (28) In the formula, For the The offspring population is selected during the iteration and is considered as the preferred population.

9. A multi-objective coordinated dispatching system for a cross-domain water project group, characterized by: A method for executing a multi-objective collaborative scheduling method for a cross-domain water project group as described in any one of claims 1 to 8, comprising a water project model initialization module and a water project multi-objective scheduling multi-group cooperative search method solving module, wherein the water project model initialization module is used to construct a cross-domain water project group multi-objective scheduling model based on parameter information of the cross-domain water project group, with the objectives of maximizing the total power generation of the water project group, minimizing the water shortage rate and minimizing the ecological flow deviation, with water balance constraints, hydraulic connection constraints between water projects, initial and final water level constraints, water level constraints, power generation flow constraints, output constraints and outbound flow constraints as constraints, and with the outbound flow during the water project scheduling period as a decision variable; The water project multi-objective scheduling multi-group cooperative search method solution module is used to solve the optimal scheduling plan for water project groups based on the multi-group cooperative search method.

10. A multi-objective coordinated dispatching system for a cross-domain water project group according to claim 9, characterized in that: The water project multi-objective scheduling multi-population cooperative search method solution module includes an initialization unit, a multi-population adaptation strategy unit, a multi-population cooperative search algorithm evolution strategy unit, a search feature optimization strategy unit, and a water project optimal scheduling solution acquisition unit; Among them, the initialization unit is used to obtain the initialized elite population; The multi-group adaptation strategy unit is used to execute the multi-group adaptation strategy; The multi-population cooperative search algorithm evolution strategy unit is used to execute the multi-population cooperative search algorithm evolution strategy; The search feature optimization strategy unit is used to execute the search feature optimization strategy; The water project optimal scheduling scheme acquisition unit is used to output the final preferred population and use the final preferred population as the optimal scheduling scheme for the water project group.

Images