Hydro-wind complementary energy dispatching method, device and equipment and storage medium

Abstract

The application relates to a water-wind-solar complementary energy scheduling method, device, equipment and storage medium. The method comprises the following steps: generating a scenario tree according to historical water-wind-solar resource data, and determining the probability corresponding to each scenario; determining a multi-energy complementary random optimization scheduling model based on a multi-stage random programming method, wherein the random optimization scheduling model takes the maximum total operation benefit as an objective function, and the total operation benefit is determined according to the probability of each scenario, the energy storage at the end of the scheduling period of the hydropower station, the average output of the water-wind-solar complementary power generation system and the abandoned water energy; converting the objective function into a plurality of sub-objective functions based on the augmented Lagrange method and the diagonal quadratic approximation method, wherein one sub-objective function corresponds to a multi-stage programming problem of one scenario; and solving each sub-objective function according to an optimization algorithm to obtain a target scheduling strategy. The method can effectively solve the water-wind-solar complementary optimization scheduling strategy under uncertain conditions.

Description

Technical Field

[0001] This application relates to the field of energy dispatching technology, and in particular to a method, apparatus, equipment and storage medium for hydro-wind-solar hybrid energy dispatching. Background Technology

[0002] Hydropower, wind power, and solar energy possess certain complementary output characteristics. Relying on the regulation capacity and transmission channels of existing and newly added large and medium-sized hydropower stations and pumped storage power stations, they have the foundation and advantages for integrated hydropower, wind power, and solar energy development and application. However, hydropower, wind power, and solar energy systems—that is, energy systems combining hydropower, wind power, and solar photovoltaic power generation—are extremely closely related to meteorological conditions, and due to their dependence on the variability of natural conditions, they exhibit strong uncertainty. Therefore, conducting research on the optimal scheduling of complementary hydropower, wind power, and solar energy in river basins, and exploring effective ways to achieve integrated hydropower, wind power, and solar energy resource allocation and scheduling under uncertain conditions, is conducive to improving the economic efficiency of comprehensive renewable energy development and the utilization rate of transmission channels, and realizing the efficient development and utilization of clean energy.

[0003] Therefore, optimizing the scheduling of water, wind, and solar power in a basin under uncertain conditions is a technical challenge that urgently needs to be addressed. Summary of the Invention

[0004] Therefore, it is necessary to provide a water-wind-solar hybrid energy dispatching method, apparatus, equipment, and storage medium that can accurately determine the optimal dispatching strategy for watershed water-wind-solar hybrid energy under uncertain conditions, in order to address the above-mentioned technical problems.

[0005] Firstly, this application provides a method for dispatching hydro-wind-solar hybrid energy, including:

[0006] A scene tree is generated based on historical water, wind, and light resource data, and the probability corresponding to each scene is determined.

[0007] A multi-energy complementary stochastic optimization scheduling model is determined based on a multi-stage stochastic programming method. The stochastic optimization scheduling model takes maximizing the total operating benefit as its objective function. The total operating benefit is determined based on the probability of each scenario, the energy storage at the end of the scheduling period of the hydropower station, the average output of the hydro-wind-solar complementary power generation system, and the abandoned water energy.

[0008] The objective function is solved using the augmented Lagrange method and the diagonal quadratic approximation method, and the target scheduling strategy is obtained.

[0009] In one embodiment, the constraints of the multi-energy complementary stochastic optimization scheduling model include: piecewise fitting of the hydropower station's power generation function, water balance equation, flow relationship, upper and lower limits of power station output, upper and lower limits of reservoir capacity, upper and lower limits of outflow, and unexpected constraints.

[0010] In one embodiment, the objective function is solved based on the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target scheduling strategy, including:

[0011] The augmented Lagrange method is used to relax unexpected constraints, thereby adding a penalty function to the objective function, which includes a quadratic term. The quadratic term of the penalty function is approximated using the diagonal quadratic approximation method, and the objective function is transformed into multiple sub-objective functions, where each sub-objective function corresponds to a multi-stage planning problem for a scenario. The optimization algorithm is used to solve each sub-objective function to obtain the target scheduling strategy.

[0012] In one embodiment, a scene tree is generated based on historical water, wind, and solar resource data, including:

[0013] The historical water and landscape resource sequence is obtained, which includes multiple sets of historical water and landscape resource data. Each historical water and landscape resource data is divided into multiple stages according to a preset time period, and each stage corresponds to a layer in the scene tree. The scene tree is determined based on the historical water and landscape resource data in each stage. The path from the leaf node to the root node in the scene tree represents a complete scene sequence.

[0014] In one embodiment, a scene tree is determined based on historical water, wind, and light resource data from each stage, including:

[0015] Obtain the preset scene tree structure, randomly select values ​​from the historical water, wind and light resource data of each stage and assign them to each node in the preset scene tree structure to determine the initial scene tree; iteratively update the initial scene tree based on the preset update strategy until the preset iteration conditions are met, and then determine the scene tree.

[0016] In one embodiment, determining the probability corresponding to each scenario includes:

[0017] Calculate the distance between each scene sequence in the scene tree and each historical water, wind and light resource data, and determine the distance matrix. The element in the i-th row and j-th column of the distance matrix represents the distance between the i-th scene sequence and the j-th group of historical water, wind and light resource data. For the i-th scene, the minimum element value in the i-th row of the distance matrix is ​​used as the scene distance value of the i-th scene. For each scene, the probability corresponding to the scene is determined based on the scene distance value.

[0018] In one embodiment, determining the probability corresponding to a scene based on the scene distance value includes:

[0019] The target scene distance value is determined by summing the scene distance value with a preset constant; the reciprocal of the target scene distance value is normalized to determine the probability corresponding to the scene.

[0020] Secondly, this application also provides a hydro-wind-solar hybrid energy dispatching device, comprising:

[0021] The first generation module is used to generate a scene tree based on historical water, wind and light resource data, and determine the probability of each scene.

[0022] The second generation module is used to determine the multi-energy complementary stochastic optimization scheduling model based on the multi-stage stochastic programming method. The stochastic optimization scheduling model takes the maximization of total operating benefits as the objective function. The total operating benefits are determined based on the probability of each scenario, the energy storage at the end of the scheduling period of the hydropower station, the average output of the hydro-wind-solar complementary power generation system, and the abandoned water energy.

[0023] The determination module is used to solve the objective function based on the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target scheduling strategy.

[0024] In one embodiment, the constraints of the multi-energy complementary stochastic optimization scheduling model include: piecewise fitting of the hydropower station's power generation function, water balance equation, flow relationship, upper and lower limits of power station output, upper and lower limits of reservoir capacity, upper and lower limits of outflow, and unexpected constraints.

[0025] In one embodiment, the determining module is specifically used to relax unexpected constraints according to the augmented Lagrange method to add a penalty function to the objective function, the penalty function including a quadratic term; to approximate the quadratic term of the penalty function according to the diagonal quadratic approximation method, and to transform the objective function into multiple sub-objective functions, wherein each sub-objective function corresponds to a multi-stage planning problem of a scenario; and to solve each sub-objective function according to the optimization algorithm to obtain the target scheduling strategy.

[0026] In one embodiment, the first generation module is specifically used to obtain a historical water, scenery and light resource sequence, which includes multiple sets of historical water, scenery and light resource data; divide each historical water, scenery and light resource data into multiple stages according to a preset time period, with each stage corresponding to a layer in a scene tree; determine the scene tree based on each historical water, scenery and light resource data in each stage, where the path from the leaf node to the root node in the scene tree represents a complete scene sequence.

[0027] In one embodiment, the first generation module is specifically used to obtain a preset scene tree structure, randomly select values ​​from the historical water, wind and light resource data of each stage and assign values ​​to each node in the preset scene tree structure to determine the initial scene tree; iteratively update the initial scene tree based on a preset update strategy until the preset iteration conditions are met, and determine the scene tree.

[0028] In one embodiment, the first generation module is specifically used to calculate the distance between each scene sequence in the scene tree and each historical water, wind and light resource data, and determine the distance matrix. The element in the i-th row and j-th column of the distance matrix represents the distance between the i-th scene sequence and the j-th group of historical water, wind and light resource data. For the i-th scene, the minimum element value in the i-th row of the distance matrix is ​​used as the scene distance value of the i-th scene. For each scene, the probability corresponding to the scene is determined based on the scene distance value.

[0029] In one embodiment, the first generation module is specifically used to determine the target scene distance value based on the sum of the scene distance value and a preset constant; and to normalize the reciprocal of the target scene distance value to determine the probability corresponding to the scene.

[0030] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any of the methods described in the first aspect above.

[0031] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the methods described in the first aspect above.

[0032] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements any of the methods described in the first aspect above.

[0033] The aforementioned hydro-wind-solar hybrid energy dispatching method, device, equipment, and storage medium can generate a scenario tree based on historical hydro-wind-solar resource data and determine the probability corresponding to each scenario. A multi-energy complementary stochastic optimization dispatching model is determined based on a multi-stage stochastic programming method. The stochastic optimization dispatching model takes maximizing the total operating benefit as its objective function. The total operating benefit is determined based on the probability of each scenario, the energy storage at the end of the dispatching period of the hydropower station, the average output of the hydro-wind-solar hybrid power generation system, and the abandoned water energy. The objective function is solved based on the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target dispatching strategy. In this way, when determining the optimal scheduling model, the multi-stage stochastic programming method reasonably considers different scenarios and the uncertainties of each scenario. At the same time, by using the augmented Lagrange method and the diagonal quadratic approximation method, the objective function in the optimal scheduling model is transformed into multiple sub-objective functions for solving separately. This avoids the correlation between the original objective function and the scheduling strategies under different scenarios, and can effectively solve the optimal scheduling strategy of hydro-wind-solar hybrid under uncertain conditions. It also accelerates the solution speed of the stochastic optimization model of hydro-wind-solar hybrid under uncertain conditions, which can better guide the operation of hydro-wind-solar power generation system, help to formulate effective scheduling strategies and emergency plans, improve the system's adaptability and response speed, and thus ensure a continuous and stable power supply. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart illustrating a hydro-wind-solar hybrid energy dispatching method in one embodiment;

[0036] Figure 2 This is a flowchart illustrating the steps for solving the objective function in one embodiment;

[0037] Figure 3 This is a flowchart illustrating the steps of generating a scene tree in one embodiment;

[0038] Figure 4 This is a flowchart illustrating the steps of determining the scene tree based on historical water, wind, and light resource data at each stage in one embodiment.

[0039] Figure 5 This is a schematic diagram of a preset scene tree structure in one embodiment;

[0040] Figure 6 This is a flowchart illustrating the steps for determining the probability of each scenario in one embodiment.

[0041] Figure 7 This is a flowchart illustrating the probability steps for determining each scenario in another embodiment;

[0042] Figure 8 This is a flowchart illustrating the hydro-wind-solar hybrid energy dispatching method in another embodiment;

[0043] Figure 9 This is a schematic diagram showing the change of the maximum error of the unexpected constraint with the number of iterations in one embodiment.

[0044] Figure 10 This is a structural block diagram of a hydro-wind-solar hybrid energy dispatching device in one embodiment;

[0045] Figure 11 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0047] The foundation of integrated renewable energy development in river basins lies in the mutual complementarity of power source characteristics. Under the new circumstances of building a new power system, there is a consensus that the functional positioning of hydropower is gradually shifting from primarily generating electricity to emphasizing both electricity generation and capacity support. Wind and solar power generation exhibit strong randomness in output and cannot provide reliable capacity support for the power system; therefore, new energy sources are gradually shifting from supplementing incremental energy and electricity consumption to becoming the main source of incremental growth. On the other hand, the development costs of wind and solar power generation continue to decrease, and their low-cost advantage can mitigate the on-grid price of hydropower, maximizing overall benefits. Under the new circumstances of building a new power system, large-scale, high-proportion, market-oriented, and high-quality development have become new characteristics and requirements for renewable energy development. my country's major river basins, such as the Jinsha River, Lancang River, Yalong River, Dadu River, and Yellow River, are rich in hydropower resources and have a good foundation for development. At the same time, they also possess abundant new energy resources such as wind and solar power, providing a natural advantage for integrated hydropower-wind-solar power development. The integrated development of renewable energy, primarily based on water, wind, and solar power, spatially integrates the abundant hydropower resources of a river basin with the rich wind and solar resources in its surrounding areas. Through integrated and large-scale development, it achieves complementary advantages, improves the absorption and storage capacity of renewable energy, and provides the power grid with almost 100% clean electricity. This is the only way for renewable energy to achieve high-quality leapfrog development in the new era.

[0048] my country's Jinsha River, Yalong River, Dadu River, Wujiang River, Lancang River, and Yarlung Tsangpo River basins possess abundant hydropower resources and suitable wind and solar energy resources. Hydropower, wind, and solar energy exhibit certain complementary characteristics, and relying on the regulation capacity of existing and newly added large and medium-sized hydropower stations and pumped-storage power stations, as well as transmission channels, they possess the foundation and advantages for integrated hydropower, wind, and solar energy development and application. However, hydropower, wind, and solar power systems—energy systems combining hydropower, wind power, and solar photovoltaic power generation—are extremely closely related to meteorological conditions and exhibit strong uncertainty due to their dependence on the variability of natural conditions. Therefore, conducting research on the optimized scheduling of complementary hydropower, wind, and solar energy in river basins, and exploring effective ways to achieve integrated hydropower, wind, and solar energy resource allocation and scheduling under uncertain conditions, is beneficial to improving the economic efficiency and channel utilization rate of comprehensive renewable energy development, enhancing the scale, competitiveness, and development quality of hydropower, wind, solar, and energy storage development, accelerating the large-scale, high-proportion development of renewable energy, and achieving efficient development and utilization of clean energy.

[0049] Therefore, the optimal scheduling of water, wind and solar power complementarity in a basin under uncertain conditions is a technical problem that urgently needs to be solved.

[0050] In view of this, this application provides a hydro-wind-solar hybrid energy dispatching method that can effectively solve for the optimal dispatching strategy of hydro-wind-solar hybrid energy under uncertain conditions. The hydro-wind-solar hybrid energy dispatching method provided in this application can be executed by a hydro-wind-solar hybrid energy dispatching device, which can be implemented by software, hardware, or a combination of both. It can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software. In the following method embodiments, the execution subject is always described using a computer device as an example. The computer device can be a server or a desktop computer; this application does not limit the specific type of computer device.

[0051] In one exemplary embodiment, such as Figure 1 As shown, a method for scheduling hydro-wind-solar hybrid energy is provided, including the following steps 101 to 103. Wherein:

[0052] Step 101: Generate a scene tree based on historical water, wind and light resource data, and determine the probability corresponding to each scene.

[0053] Optionally, a scenario tree can be a tree-based method for describing and modeling the possible states and evolution processes of various random variables in a hydro-wind-solar energy system at different times, presenting the uncertainties faced in the scheduling of hydro-wind-solar energy in an intuitive and systematic way.

[0054] Optionally, the elements of a scenario tree can include nodes and branches. Nodes include root nodes, internal nodes, and leaf nodes. The root node typically represents the initial moment of the scheduling cycle and contains various information about the system in its initial state, such as the initial water level of the reservoir and the initial operating state of the wind and solar power stations. Internal nodes represent intermediate moments and states during the scheduling process. Each internal node corresponds to a specific time period and a certain combination of values ​​for random variables within that time period. For example, an internal node might represent the state in time period t where wind power output is in a certain range, solar power output is a certain value, and the reservoir inflow is a specific value. Leaf nodes can correspond to various possible final states at the end of the scheduling cycle and contain all relevant information about the system at the end of the entire scheduling process, such as the final reservoir water level, wind and solar power generation, and total system power generation. Branches are used to connect different nodes and can represent the process of transitioning from a certain state in one time period to another state in the next time period.

[0055] Optionally, historical water, wind, and solar resource data may include historical water resource data, historical wind energy resource data, and historical solar energy resource data. Historical water resource data may include annual hydrological and meteorological data, runoff data, and water resource volume data, such as precipitation data, evaporation data, river runoff data, surface runoff data, and groundwater runoff data. Historical wind energy resource data may include annual wind speed data, wind direction data, and wind power density data. Historical solar energy resource data may include annual solar radiation data, sunshine duration data, and solar altitude angle and azimuth angle data.

[0056] In one possible implementation, generating a scene tree based on historical water, wind, and light resource data can be achieved by dividing the entire time range into several time periods according to a preset time span. Then, in each time period, random variables are discretized according to a preset discretization method, such as the equal probability interval method. Finally, starting from the initial stage, the discretization results of each time period are used as nodes and connected through branches to form a scene tree.

[0057] For example, there are 3 discrete states in the first time period. In the second period, the state There may be two subsequent states. ,state There may be 3 subsequent states. And so on, gradually building a complete scene tree.

[0058] In one possible implementation, the probability of each scene in the scene tree can be determined based on the discretization method of the random variable and the original distribution. For example, in the equal probability interval method, the probability of each discrete interval is equal, and the probability of the scene is obtained by calculating the product of the probabilities of the discrete intervals traversed by each scene.

[0059] For example, in a two-time-phase scene tree, the first time-phase discrete state The probability is The second period starts from arrive The probability is So, what is the scenario? The probability is .

[0060] Step 102: Determine the multi-energy complementary stochastic optimization scheduling model based on the multi-stage stochastic programming method.

[0061] Among them, the stochastic optimization scheduling model takes the maximization of total operating benefits as the objective function. The total operating benefits are determined based on the probability of each scenario, the energy storage at the end of the scheduling period of the hydropower station, the average output of the hydro-wind-solar hybrid power generation system, and the abandoned water energy.

[0062] Optionally, the energy dispatching problem of hydropower, wind power, and solar power can be divided into multiple time stages, each facing different decision-making and uncertainty factors. For example, the dispatching cycle can be divided into multiple stages with time intervals of hours, days, or months. In each stage, decision variables such as the output of hydropower units, wind power units, and photovoltaic units need to be determined based on information such as water conditions, wind conditions, sunlight conditions, and load demand at that time.

[0063] Optionally, when determining the multi-energy complementary stochastic optimization scheduling model, factors such as the randomness of hydropower, wind power, and solar power output and the uncertainty of load demand can be considered, and random variables can be introduced to describe these uncertainties.

[0064] Optionally, the objective function of the multi-energy complementary stochastic optimization scheduling model can be expressed as follows:

[0065]

[0066] Where S represents the number of scenes, M represents the number of months, and N represents the number of hydroelectric power stations. Let be the probability of scenario s. For the average output of the hydro-wind-solar hybrid power generation system, The number of hours in the time period. For energy storage at the end of the hydropower station's operation period, For the purpose of discarding water energy, For different benefit coefficients.

[0067] It should be noted that in the average output of the hydro-wind-solar hybrid power generation system, the power output of the hydropower station is mainly analyzed, while the power output of the wind and solar power generation can be preset.

[0068] Optionally, the constraints of the multi-energy complementary stochastic optimization scheduling model include: piecewise fitting of the power generation function of the hydropower station, water balance equation, flow relationship, upper and lower limits of power station output, upper and lower limits of reservoir capacity, upper and lower limits of outflow, and unexpected constraints.

[0069] Optionally, the piecewise fitting expression for the power generation function of the hydropower station is shown below:

[0070]

[0071] in, Represents the power generation function. Where H is the total number of segments in the power generation function. The coefficients representing the power generation function are obtained by fitting historical data. These represent the reservoir capacities of the upstream and downstream reservoirs, respectively. The expression represents the power generation flow rate. It can be understood that the superscript i in the above expression represents the i-th scenario, and the subscript n represents the n-th hydropower station. This will not be elaborated further.

[0072] Alternatively, the expression for the water balance equation can be as follows:

[0073]

[0074] in, This represents the storage capacity at time t. Mn represents the reservoir capacity at time t-1, and Mn represents the total number of upstream reservoirs. This indicates the discharge flow from the upstream reservoir. This indicates that the reservoir runoff is flowing in.

[0075] Optionally, the expression for the flow relationship is as follows:

[0076]

[0077] in, Indicates the outflow from the reservoir. This indicates the discharge rate of water.

[0078] Optionally, the upper and lower limits of the power plant's output can be expressed by the following formula: ,in, and These represent the minimum and maximum output, respectively.

[0079] Optionally, the upper and lower limits of reservoir capacity can be expressed by the following formula: ,in, and These represent the minimum and maximum storage capacities, respectively.

[0080] Optionally, the upper and lower limits of the discharge flow can be expressed by the following formula: ,in, and These represent the minimum and maximum discharge flows, respectively.

[0081] Optionally, unexpected constraints can be represented by the following formula: ,in, This represents the decision vector corresponding to scene i in the scene tree. Represents a scene in the scene tree The corresponding decision vectors, scenario i and scenario i A set of scenarios belonging to shared nodes SH(t) represents the maximum number of time periods that share the same scenario tree nodes within the scheduling period.

[0082] Step 103: Solve the objective function based on the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target scheduling strategy.

[0083] Optionally, the augmented Lagrange method can be used to solve constrained optimization problems. It combines the ideas of the Lagrange multiplier method and the penalty function method. By introducing Lagrange multipliers and penalty terms, an augmented Lagrange function is constructed based on the objective function.

[0084] Optionally, the augmented Lagrangian function can be approximated twice based on the diagonal quadratic approximation method, transforming the augmented Lagrangian function into multiple sub-objective functions. Then, the multiple sub-objective functions are solved separately to obtain the target scheduling strategy.

[0085] The aforementioned hydro-wind-solar hybrid energy dispatching method can generate a scenario tree based on historical hydro-wind-solar resource data and determine the probability corresponding to each scenario. A multi-energy complementary stochastic optimization dispatching model is determined based on a multi-stage stochastic programming method. The stochastic optimization dispatching model takes maximizing the total operating benefit as its objective function. The total operating benefit is determined based on the probability of each scenario, the energy storage at the end of the dispatching period of the hydropower station, the average output of the hydro-wind-solar hybrid power generation system, and the abandoned water energy. The objective function is solved based on the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target dispatching strategy. In this way, when determining the optimal scheduling model, the multi-stage stochastic programming method reasonably considers different scenarios and the uncertainties of each scenario. At the same time, by using the augmented Lagrange method and the diagonal quadratic approximation method, the objective function in the optimal scheduling model is transformed into multiple sub-objective functions for solving separately. This avoids the correlation between the original objective function and the scheduling strategies under different scenarios, and can effectively solve the optimal scheduling strategy of hydro-wind-solar hybrid under uncertain conditions. It also accelerates the solution speed of the stochastic optimization model of hydro-wind-solar hybrid under uncertain conditions, which can better guide the operation of hydro-wind-solar power generation system, help to formulate effective scheduling strategies and emergency plans, improve the system's adaptability and response speed, and thus ensure a continuous and stable power supply.

[0086] In one exemplary embodiment, such as Figure 2 As shown, optionally, the objective function is solved based on the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target scheduling strategy, including the following steps 201 to 203. Wherein:

[0087] Step 201: Relax the unexpected constraints according to the augmented Lagrange method to add a penalty function to the objective function.

[0088] The penalty function includes quadratic terms.

[0089] Optionally, unintended constraints make decisions in different scenarios interconnected and inseparable. The objective function of step 102 can generally be represented as follows:

[0090]

[0091] in, Let represent the probability of the i-th scenario. This represents the revenue function of a hydro-wind-solar power generation system.

[0092] The constraints in step 102 are generally represented as follows:

[0093]

[0094]

[0095] in, This represents the physical constraints under scenario i, where I represents the total number of scenarios. Let i be the decision vector corresponding to scenario i. Represents a scene in the scene tree The corresponding decision vectors, scenario i and scenario i A set of scenarios that belong to shared nodes.

[0096] Optionally, the augmented Lagrangian method can be used to relax the unintended constraints and penalize violations of these constraints. The penalty for violating unintended constraints is then added to the objective function, resulting in the augmented Lagrangian objective function, as shown below:

[0097]

[0098] in, express The corresponding Lagrange multiplier vectors; and Represents the decision vector; Representing vectors with vector The inner product; Representing vectors and Euclidean distance, This represents the penalty parameter; if the original model has a solution, the augmented Lagrangian method is finitely convergent.

[0099] Step 202: Approximate the quadratic term of the penalty function using the diagonal quadratic approximation method, and convert the objective function into multiple sub-objective functions.

[0100] One sub-objective function corresponds to a multi-stage planning problem for a scenario.

[0101] Optionally, according to the diagonal quadratic approximation method, the quadratic term of the penalty function... It can be expanded to approximate as:

[0102]

[0103] Optionally, a local approximation of the inner product can be expressed as:

[0104]

[0105] in, and Represents the decision vector and The estimated value, the approximation method assumes the decision vector Reference point The neighborhood of . Substituting the inner product into the quadratic term, we get:

[0106]

[0107] in, This represents a scenario whose past and current states are the same as those of scenario i. The corresponding decision, i.e., scenario i and scenario 2. The decision sequence before time t is the same.

[0108] Alternatively, the augmented Lagrangian objective function can be written as:

[0109]

[0110] in, express The corresponding Lagrange multiplier vector.

[0111] Optionally, the augmented Lagrange objective function described above can be decomposed into sub-objective functions, which are represented as follows:

[0112]

[0113] Step 203: Solve each sub-objective function according to the optimization algorithm to obtain the target scheduling strategy.

[0114] Optionally, each sub-objective function can be solved separately according to the optimization algorithm to obtain the target scheduling strategy under each scenario. The optimization algorithm can be gradient descent, conjugate gradient, genetic algorithm, particle swarm optimization algorithm, ant colony algorithm, etc., and the embodiments of this application do not limit it.

[0115] The above-mentioned approach relaxes unexpected constraints using the augmented Lagrange method to add a penalty function to the objective function, which includes a quadratic term. The quadratic term of the penalty function is then approximated using the diagonal quadratic approximation method, transforming the objective function into multiple sub-objective functions. Each sub-objective function corresponds to a multi-stage planning problem for a specific scenario. The optimization algorithm solves for each sub-objective function to obtain the target scheduling strategy. This process transforms the objective function into multiple independent sub-objective functions, avoiding the correlation between the original objective function and the scheduling strategies under different scenarios, thus improving the solution speed of the water-wind-solar hybrid stochastic optimization model under uncertain conditions.

[0116] In one exemplary embodiment, such as Figure 3 As shown, optionally, a scene tree is generated based on historical water, wind, and light resource data, including the following steps 301 to 303. Wherein:

[0117] Step 301: Obtain the historical sequence of water, scenery and light resources.

[0118] The historical water, scenery and light resources sequence includes multiple sets of historical water, scenery and light resources data.

[0119] Optionally, there are no restrictions on the method of obtaining historical water, wind, and solar resource sequences. They can be obtained directly from a database or from departments such as meteorological departments, water conservancy departments, and hydrological monitoring stations.

[0120] Optionally, the historical water, wind, and light resource sequence is represented as follows: , , Represents the number of historical sequences, for example, At 20:00, It can represent water, wind, and solar resource data from the past 1 to 20 years.

[0121] Step 302: Divide the historical water and scenery resource data into multiple stages according to the preset time period, with each stage corresponding to a layer in the scene tree.

[0122] Optional, Depend on composition, , This indicates the number of time periods in a time series. For example, when T is 12, it can mean that the historical water, wind and solar resources data are divided into 12 months, which is the number of time periods within the scheduling cycle.

[0123] Step 303: Determine the scene tree based on the historical water, wind and light resource data in each stage.

[0124] In the scene tree, the path from the leaf node to the root node represents a complete scene sequence.

[0125] In one possible implementation, historical water, wind, and light resource data for each stage can be clustered according to a preset clustering algorithm, with each cluster representing a node; then, the scene tree can be determined based on the clustering results for each stage.

[0126] In another possible implementation, such as Figure 4 As shown, optionally, a scene tree is determined based on historical water, wind, and light resource data in each stage, including the following steps 401 to 402, wherein:

[0127] Step 401: Obtain the preset scene tree structure, randomly select values ​​from the historical water, wind and light resource data of each stage and assign them to each node in the preset scene tree structure to determine the initial scene tree.

[0128] Optionally, the water, wind, and solar energy resource scenarios are represented as follows: , , This represents the total number of scenes; similarly, a scene... Depend on composition, , This indicates the number of time periods in the time series.

[0129] Optional, such as Figure 5 As shown, this is a possible preset scene tree structure. Values ​​can be randomly selected from historical water, wind, and light resource data at various stages and assigned to each node in the preset scene tree structure. .

[0130] Optional, It is in vector form and can include Where N represents the total amount of energy resources, and similarly, Also in vector form .

[0131] Optional, such as Figure 5 As shown, it can be understood that at t=1, the value of the root node in the scene tree is... , ,..., They are all the same, at time t=2, and ,..., They are the same. and ,..., They are the same.

[0132] Step 402: Iteratively update the initial scene tree based on the preset update strategy until the preset iteration conditions are met, and then determine the scene tree.

[0133] Optionally, the preset update strategy may include a greedy algorithm or a genetic algorithm. The target scene is determined according to the greedy algorithm or the genetic algorithm, and the nodes in the scene tree are replaced according to the target scene. The target scene may be the scene closest to the historical water, wind and light resource sequence. The embodiments of this application do not limit the process of determining the target scene according to the preset update strategy.

[0134] Optionally, the distance between the scene sequence and the historical water, wind and light resource sequence can be represented by Euclidean distance or Manhattan distance, and this application embodiment does not limit this.

[0135] For example, the distance between a scene sequence and a historical water, wind, and light resource sequence calculated using Euclidean distance can be expressed by the following formula:

[0136]

[0137] in, This represents the preset importance coefficient for a given time period.

[0138] Optionally, the deviation between the scene sequence and the historical waterscape resource sequence. This can be expressed by the following formula:

[0139]

[0140]

[0141]

[0142]

[0143] in, This is an array determined by sorting the scene sequence according to the distance between it and the historical water and landscape resource sequence. Indicates the historical sequence of water, scenery, and natural resources. This indicates the number of iterations. As the number of iterations gradually increases, the deviation between the scene sequence and the historical water and landscape resource sequence gradually decreases. It can be seen that as the number of iterations increases from 0 to... As the deviation gradually decreases, the values ​​of the scene tree gradually converge to the historical water and landscape resource sequence. For example, the parameter values ​​of the above formula are respectively... .

[0144] Optionally, preset iteration conditions include the number of iterations reaching a preset number, or the average distance between all scene sequences in the scene tree and the historical actual sequence being less than a set threshold.

[0145] The above-mentioned acquisition of historical water, wind and solar resources sequence includes multiple sets of historical water, wind and solar resources data; each historical water, wind and solar resources data is divided into multiple stages according to a preset time period, and each stage corresponds to a layer in a scene tree; the scene tree is determined based on the historical water, wind and solar resources data in each stage. The path from the leaf node to the root node in the scene tree represents a complete scene sequence. In this way, a scene tree that approximates the actual historical water, wind and solar resources data can be obtained, and the uncertainty of water, wind and solar energy can be quantified through this scene tree.

[0146] In one exemplary embodiment, such as Figure 6 As shown, optionally, the probability corresponding to each scenario is determined, including the following steps 601 to 602. Wherein:

[0147] Step 601: Calculate the distance between each scene sequence in the scene tree and each historical water, wind and light resource data, and determine the distance matrix.

[0148] In the distance matrix, the element in the i-th row and j-th column represents the distance between the i-th scene sequence and the j-th group of historical water, wind and light resource data.

[0149] Optionally, after determining the scene tree, a distance matrix can be determined based on each scene sequence in the scene tree and each historical water, wind and light resource data. The distance is calculated in the same way as in step 402 above, and will not be described again in this embodiment.

[0150] Step 602: For the i-th scene, the minimum element value of the i-th row of the distance matrix is ​​used as the scene distance value of the i-th scene.

[0151] Optionally, the scene distance value of the i-th scene can be represented by the minimum distance between the scene sequence and the historical water, wind, and light resource sequence, i.e. That is, the smallest element in the i-th row of the distance matrix is ​​taken as the scene distance value of the scene.

[0152] Step 603: For each scenario, determine the probability corresponding to the scenario based on the scenario distance value.

[0153] Optionally, when determining the probability of each scenario, the scenario distance value can be determined first, and then the probability of the scenario can be determined based on the scenario distance value.

[0154] In one possible implementation, the probability of a scene can be calculated by normalizing the value based on the reciprocal of the scene distance value, and the normalized result can be used as the probability of that scene.

[0155] In another possible implementation, optional, such as Figure 7As shown, the probability corresponding to a scene is determined based on the scene distance value, including the following steps 701 to 702. Wherein:

[0156] Step 701: Determine the target scene distance value based on the sum of the scene distance value and a preset constant.

[0157] Step 702: Normalize the reciprocal of the target scene distance value to determine the probability corresponding to the scene.

[0158] Optionally, the scene distance value corresponding to the scene can be added to a preset constant to determine the target scene distance value. Then, normalization is performed based on the reciprocal of the target scene distance value to determine the probability corresponding to the scene, which can be expressed by the following formula:

[0159]

[0160] Where A is a preset constant. This represents the probability of the i-th scenario. Adding the scenario distance value to a preset constant ensures that the probability of a "super scenario" that is very close to the historical sequence is not too high.

[0161] The distance matrix is ​​determined by calculating the distance between each scene sequence in the scene tree and each historical water, wind and light resource data. For the i-th scene, the minimum element value of the i-th row of the distance matrix is ​​used as the scene distance value of the i-th scene. For each scene, the probability corresponding to the scene is determined based on the scene distance value. In this way, the similarity between the scene and the historical situation can be considered intuitively, highlighting scenes with a high degree of matching with historical data, and the calculation process is relatively simple.

[0162] As an optional implementation method, such as Figure 8 As shown in the embodiments of this application, the hydro-wind-solar hybrid energy dispatching method may include the following specific steps:

[0163] Step 801: Obtain the historical sequence of water, scenery and light resources.

[0164] The historical water, scenery and light resources sequence includes multiple sets of historical water, scenery and light resources data.

[0165] Step 802: Divide the historical water and scenery resource data into multiple stages according to the preset time period, with each stage corresponding to a layer in the scene tree.

[0166] Step 803: Obtain the preset scene tree structure, randomly select values ​​from the historical water, wind and light resource data of each stage and assign them to each node in the preset scene tree structure to determine the initial scene tree;

[0167] Step 804: Iteratively update the initial scene tree based on the preset update strategy until the preset iteration conditions are met, and then determine the scene tree.

[0168] In the scene tree, the path from the leaf node to the root node represents a complete scene sequence.

[0169] Step 805: Calculate the distance between each scene sequence in the scene tree and each historical water, wind and light resource data, and determine the distance matrix.

[0170] In the distance matrix, the element in the i-th row and j-th column represents the distance between the i-th scene sequence and the j-th group of historical water, wind and light resource data.

[0171] Step 806: For the i-th scene, the minimum element value of the i-th row of the distance matrix is ​​used as the scene distance value of the i-th scene.

[0172] Step 807: For each scene, determine the target scene distance value based on the sum of the scene distance value and a preset constant.

[0173] Step 808: For each scenario, normalize the reciprocal of the target scenario distance value to determine the probability corresponding to the scenario.

[0174] Step 809: Determine the multi-energy complementary stochastic optimization scheduling model based on the multi-stage stochastic programming method.

[0175] Among them, the stochastic optimization scheduling model takes the maximization of total operating benefits as the objective function. The total operating benefits are determined based on the probability of each scenario, the energy storage at the end of the scheduling period of the hydropower station, the average output of the hydro-wind-solar hybrid power generation system, and the abandoned water energy. The constraints of the multi-energy complementary stochastic optimization scheduling model include: piecewise fitting of the power generation function of the hydropower station, water balance equation, flow relationship, upper and lower limits of power station output, upper and lower limits of reservoir capacity, upper and lower limits of outflow, and unexpected constraints.

[0176] Step 810: Relax the unexpected constraints according to the augmented Lagrange method to add a penalty function to the objective function.

[0177] The penalty function includes quadratic terms.

[0178] Step 811: Approximate the quadratic term of the penalty function using the diagonal quadratic approximation method, and transform the objective function into multiple sub-objective functions.

[0179] One sub-objective function corresponds to a multi-stage planning problem for a scenario.

[0180] Step 812: Solve each sub-objective function according to the optimization algorithm to obtain the target scheduling strategy.

[0181] For example, taking a cascade hydropower station and surrounding wind and solar power stations in a certain river basin as an example, a stochastic optimization scheduling model for hydropower-wind-solar complementary power generation is established, and a diagonal quadratic approximation algorithm is used to solve the model, with the penalty parameters set as follows: , , and The performance of the corresponding diagonal quadratic approximation algorithm is shown in Table 1. The maximum error of the unexpected constraint changes with the number of iterations as follows: Figure 9 As shown, 901 is the penalty parameter. The curve showing the change over time, with 902 representing the penalty parameter. The curve showing the change over time, with 903 representing the penalty parameter. The curves showing the changes over time indicate that as the number of iterations increases, the maximum error of the unexpected constraint decreases significantly. The method provided in this application can effectively solve the optimal scheduling strategy for water-wind-solar hybrid systems under uncertain conditions.

[0182] Table 1

[0183]

[0184] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0185] Based on the same inventive concept, this application also provides a hydro-wind-solar complementary energy dispatching device for implementing the above-mentioned hydro-wind-solar complementary energy dispatching method. The solution provided by this device is similar to the solution described in the above-described method. Therefore, the specific limitations of one or more hydro-wind-solar complementary energy dispatching device embodiments provided below can be found in the limitations of the hydro-wind-solar complementary energy dispatching method above, and will not be repeated here.

[0186] In one exemplary embodiment, such as Figure 10 As shown, a hydro-wind-solar hybrid energy dispatching device 1000 is provided, comprising: a first generation module 1001, a second generation module 1002, and a determination module 1003, wherein:

[0187] The first generation module 1001 is used to generate a scene tree based on historical water, wind and light resource data, and determine the probability corresponding to each scene;

[0188] The second generation module 1002 is used to determine the multi-energy complementary stochastic optimization scheduling model based on the multi-stage stochastic programming method. The stochastic optimization scheduling model takes the maximization of total operating benefits as the objective function. The total operating benefits are determined based on the probability of each scenario, the energy storage at the end of the scheduling period of the hydropower station, the average output of the hydro-wind-solar complementary power generation system, and the abandoned water energy.

[0189] The determination module 1003 is used to solve the objective function based on the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target scheduling strategy.

[0190] In one embodiment, the constraints of the multi-energy complementary stochastic optimization scheduling model include: piecewise fitting of the hydropower station's power generation function, water balance equation, flow relationship, upper and lower limits of power station output, upper and lower limits of reservoir capacity, upper and lower limits of outflow, and unexpected constraints.

[0191] In one embodiment, the determining module 1003 is specifically used to relax unexpected constraints according to the augmented Lagrange method to add a penalty function to the objective function, the penalty function including a quadratic term; to approximate the quadratic term of the penalty function according to the diagonal quadratic approximation method, and to convert the objective function into multiple sub-objective functions, wherein each sub-objective function corresponds to a multi-stage planning problem of a scenario; and to solve each sub-objective function according to the optimization algorithm to obtain the target scheduling strategy.

[0192] In one embodiment, the first generation module 1001 is specifically used to obtain a historical water, scenery and light resource sequence, which includes multiple sets of historical water, scenery and light resource data; divide each historical water, scenery and light resource data into multiple stages according to a preset time period, with each stage corresponding to a layer in a scene tree; determine the scene tree based on each historical water, scenery and light resource data in each stage, where the path from the leaf node to the root node in the scene tree represents a complete scene sequence.

[0193] In one embodiment, the first generation module 1001 is specifically used to obtain a preset scene tree structure, randomly select values ​​from the historical water, wind and light resource data of each stage and assign values ​​to each node in the preset scene tree structure to determine the initial scene tree; iteratively update the initial scene tree based on a preset update strategy until the preset iteration conditions are met, and determine the scene tree.

[0194] In one embodiment, the first generation module 1001 is specifically used to calculate the distance between each scene sequence in the scene tree and each historical water, wind and light resource data, and determine the distance matrix. The element in the i-th row and j-th column of the distance matrix represents the distance between the i-th scene sequence and the j-th group of historical water, wind and light resource data. For the i-th scene, the minimum element value in the i-th row of the distance matrix is ​​used as the scene distance value of the i-th scene. For each scene, the probability corresponding to the scene is determined based on the scene distance value.

[0195] In one embodiment, the first generation module 1001 is specifically used to determine the target scene distance value based on the sum of the scene distance value and a preset constant; and to normalize the reciprocal of the target scene distance value to determine the probability corresponding to the scene.

[0196] Each module in the aforementioned hydro-wind-solar hybrid energy dispatching method device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0197] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 11 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network. When the computer program is executed by the processor, it implements a hydro-wind-solar hybrid energy dispatching method.

[0198] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0199] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps described in any of the above method embodiments.

[0200] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps described in any of the above method embodiments.

[0201] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps described in any of the above method embodiments.

[0202] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0203] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0204] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for dispatching hydro-wind-solar hybrid energy, characterized in that, The method includes: Obtain a historical sequence of water, scenery and light resources, which includes multiple sets of historical water, scenery and light resource data; The historical water, wind and light resources data are divided into multiple stages according to a preset time period, and each stage corresponds to a layer in the scene tree. Obtain a preset scene tree structure, and randomly select values ​​from the historical water, wind and light resource data of each stage to assign values ​​to each node in the preset scene tree structure to determine the initial scene tree; The initial scene tree is iteratively updated based on a preset update strategy until the preset iteration conditions are met. The scene tree is then determined, and the probability corresponding to each scene is determined. The preset update strategy includes: determining the target scene according to a greedy algorithm or a genetic algorithm, and replacing the nodes in the scene tree according to the target scene, wherein the target scene is the scene closest to the historical water and scenery resource sequence; the preset iteration conditions include the number of iterations reaching a preset number, or the average distance between all scene sequences in the scene tree and the historical water and scenery resource sequence being less than a set threshold. A multi-energy complementary stochastic optimization scheduling model is determined based on a multi-stage stochastic programming method. The stochastic optimization scheduling model takes maximizing the total operating benefit as its objective function. The total operating benefit is determined based on the probability of each scenario, the energy storage at the end of the scheduling period of the hydropower station, the average output of the hydro-wind-solar complementary power generation system, and the abandoned water energy. The objective function is solved using the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target scheduling strategy.

2. The method according to claim 1, characterized in that, The constraints of the multi-energy complementary stochastic optimization scheduling model include: piecewise fitting of the hydropower station's power generation function, water balance equation, flow relationship, upper and lower limits of power station output, upper and lower limits of reservoir capacity, upper and lower limits of outflow, and unexpected constraints.

3. The method according to claim 2, characterized in that, The objective function is solved using the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target scheduling strategy, including: The unintended constraints are relaxed according to the augmented Lagrange method to add a penalty function to the objective function, the penalty function comprising quadratic terms; The quadratic term of the penalty function is approximated using the diagonal quadratic approximation method, and the objective function is transformed into multiple sub-objective functions, wherein each sub-objective function corresponds to a multi-stage planning problem for a scenario. The target scheduling strategy is obtained by solving the sub-objective functions according to the optimization algorithm.

4. The method according to claim 1, characterized in that, Determining the probability corresponding to each scenario includes: Calculate the distance between each scene sequence in the scene tree and each historical water, wind and light resource data, and determine the distance matrix. The element in the i-th row and j-th column of the distance matrix represents the distance between the i-th scene sequence and the j-th group of historical water, wind and light resource data. For the i-th scene, the minimum element value of the i-th row of the distance matrix is ​​used as the scene distance value of the i-th scene; For each scenario, the probability corresponding to the scenario is determined based on the scenario distance value.

5. The method according to claim 4, characterized in that, The step of determining the probability corresponding to the scene based on the scene distance value includes: The target scene distance value is determined based on the sum of the scene distance value and a preset constant; The reciprocal of the distance value to the target scene is normalized to determine the probability corresponding to the scene.

6. A hydro-wind-solar hybrid energy dispatching device, characterized in that, The device includes: The first generation module is used to obtain a historical water, wind and light resource sequence, which includes multiple sets of historical water, wind and light resource data. The historical water, wind and light resources data are divided into multiple stages according to a preset time period, and each stage corresponds to a layer in the scene tree. Obtain a preset scene tree structure, and randomly select values ​​from the historical water, wind and light resource data of each stage to assign values ​​to each node in the preset scene tree structure to determine the initial scene tree; The initial scene tree is iteratively updated based on a preset update strategy until the preset iteration conditions are met. The scene tree is then determined, and the probability corresponding to each scene is determined. The preset update strategy includes: determining the target scene according to a greedy algorithm or a genetic algorithm, and replacing the nodes in the scene tree according to the target scene, wherein the target scene is the scene closest to the historical water and scenery resource sequence; the preset iteration conditions include the number of iterations reaching a preset number, or the average distance between all scene sequences in the scene tree and the historical water and scenery resource sequence being less than a set threshold. The second generation module is used to determine a multi-energy complementary stochastic optimization scheduling model based on a multi-stage stochastic programming method. The stochastic optimization scheduling model takes the maximization of total operating benefits as its objective function. The total operating benefits are determined based on the probability of each scenario, the energy storage at the end of the scheduling period of the hydropower station, the average output of the hydro-wind-solar complementary power generation system, and the abandoned water energy. The determination module is used to solve the objective function based on the augmented Lagrange method and the diagonal quadratic approximation method to obtain the target scheduling strategy.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images