Method and system for collaborative configuration of multi-element regulation resources of water, wind and light systems

Abstract

This application provides a method and system for the coordinated allocation of multi-dimensional regulatory resources in a water-wind-solar system. Based on a basic dataset of the target water-wind-solar system, a typical-extreme dual-layer scenario set is constructed. Based on this dual-layer scenario set, a flexibility gap profile of the target water-wind-solar system under each scenario is identified. Based on this profile, a resource matching degree of multi-dimensional regulatory resources is established. A resource allocation optimization model is constructed based on the flexibility gap profile and the resource matching degree, and the model is solved to obtain candidate allocation schemes. The candidate allocation schemes are then tested and verified using the typical-extreme dual-layer scenario set. Based on the verification results, the candidate allocation schemes are revised to obtain the final coordinated allocation scheme of multi-dimensional regulatory resources. This shifts the resource allocation objective from simple capacity allocation to targeted reinforcement of operational shortcomings, significantly improving the targeting and economy of resource allocation, reducing the risks of over-allocation, under-allocation, and operational infeasibility, and enhancing the feasibility of resource allocation.

Description

Technical Field

[0001] This invention relates to the field of power system planning and resource optimization and allocation technology, and in particular to a method and system for the coordinated allocation of multi-dimensional regulation resources in hydropower, wind power and solar power systems. Background Technology

[0002] With the continuous increase in the installed capacity of new energy sources such as wind power and photovoltaics, hydro-wind-solar hybrid systems are playing an increasingly important role in river basin new energy bases, grid peak shaving and supply guarantee, and clean energy consumption. Wind and solar power output is significantly affected by meteorological processes, exhibiting randomness, volatility, and anti-peak shaving characteristics. While hydropower possesses relatively strong regulation capabilities, its available regulation capacity is not constant due to constraints imposed by inflow processes, reservoir water levels, reservoir capacity boundaries, ecological flow, flood control requirements, and upstream-downstream coupling relationships. In scenarios with a high proportion of new energy integration, the system not only needs to meet conventional power balance but also simultaneously consider new energy consumption, safe supply guarantee, peak shaving reserve, continuous energy support, and grid capacity. There are clear complementary, substitutive, and constrained coupling relationships among these various regulation resources.

[0003] Currently, conventional hydropower, pumped storage, variable-speed pumped storage, electrochemical energy storage, hydrogen energy storage, and flexible retrofitting of thermal power plants can all provide regulation capabilities for hydro-wind-solar systems. Among them, electrochemical energy storage is suitable for short-term power balancing and rapid ramp-up support; pumped storage and variable-speed pumped storage are suitable for intraday peak shaving and valley filling, renewable energy consumption, and backup support; hydrogen energy storage is suitable for continuous low wind and solar power, and high load during dry periods, spanning multiple days or days to ensure power supply; conventional hydropower can rely on reservoir capacity for inter-period regulation, but is limited by water level, inflow, and dispatching procedures; flexible thermal power resources can provide basic power supply and backup support, but are limited by minimum technical output, start-up and shutdown costs, and carbon emission constraints. How to identify the truly lacking regulation capabilities in the system under different operating scenarios, and accordingly determine the configuration scale, configuration structure, and investment priority of various resources, is a key issue in the planning of hydro-wind-solar systems.

[0004] Existing technologies primarily focus on optimizing hydropower, wind power, solar power, and energy storage capacity; complementary scheduling of hydropower, wind power, and solar power; source-grid-load-storage planning; and short-term optimized scheduling of single-type energy storage or multi-energy complementarity. Some methods establish optimization models with objectives such as wind and solar curtailment rates, overall costs, power supply reliability, or output smoothness; others use clustering, typical day selection, or year-round time-series simulation to generate operating scenarios; still others introduce large models or deep learning models for wind and solar load power prediction, scenario clustering, or short-term scheduling assistance. However, what systems often lack under multi-scenario operation is not a single capacity, but rather different types of flexibility capabilities such as upscaling, downscaling, ramping, continuous energy, reserve, and cross-sectional load capacity. Existing methods typically struggle to interpret and complete low-confidence samples, abnormal operating samples, and complex extreme conditions such as continuous low wind and solar power, low water and high load, cross-sectional constraints, and resource unavailability. They also lack mechanisms to characterize the direction, magnitude, duration, frequency of occurrence, and associated constraints of flexibility gaps. Furthermore, existing large-scale model-related solutions often focus on improving prediction accuracy or assisting in short-term scheduling solutions. Directly using the output of large-scale models as the basis for resource allocation is easily affected by the quality of training data, prompt word design, and the black-box nature of the model, making it difficult to explain the reasons for scenario classification, constraint triggering, and resource reinforcement. At the same time, existing capacity allocation methods and operational verification processes are often disconnected, making it difficult to simultaneously verify operational boundaries such as reservoir capacity, energy storage state of charge, hydrogen storage status, thermal power plant deep-tunneling status, and cross-sectional load factor during the allocation phase. This can lead to problems such as constraint violations, resource over-allocation, or insufficient long-term supply capacity in actual operation. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method and system for the coordinated allocation of multi-dimensional regulatory resources in water, wind and solar systems, which can improve the pertinence and feasibility of resource allocation.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for coordinated allocation of multi-element regulatory resources in a water-wind-solar system includes: Obtain the basic dataset of the target water-wind-scenic system, and construct a typical-extreme two-layer scene set based on the basic dataset; Based on the typical-extreme dual-layer scene set, the flexibility gap profile of the target water-wind-solar system under each scene is identified. The flexibility gap profile includes upward adjustment gap, downward adjustment gap, ramping gap, continuous energy gap, backup gap and cross-sectional bearing capacity gap. Based on the aforementioned flexibility gap profile, establish a resource matching degree for diversified adjustment resources; Based on the flexibility gap profile and the resource matching degree, a resource allocation optimization model is constructed, and the resource allocation optimization model is solved to obtain candidate configuration schemes; The candidate configuration schemes are run and verified based on the typical-extreme dual-layer scenario set to obtain verification results. The candidate configuration schemes are then modified according to the verification results to obtain the final multi-dimensional adjustment resource coordination configuration scheme.

[0007] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A multi-element regulation resource coordination configuration system for a water-wind-solar system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the various steps of the aforementioned multi-element regulation resource coordination configuration method for a water-wind-solar system.

[0008] The beneficial effects of this invention are as follows: A typical-extreme dual-layer scenario set is constructed based on the basic dataset of the target water-wind-solar system. Based on this scenario set, a flexibility gap profile of the target water-wind-solar system under each scenario is identified. A resource matching degree for multi-dimensional regulatory resources is established based on the flexibility gap profile and resource matching degree. A resource allocation optimization model is constructed based on the flexibility gap profile and resource matching degree, and the model is solved to obtain candidate configuration schemes. The candidate configuration schemes are then tested and verified using the typical-extreme dual-layer scenario set to obtain verification results. Based on these results, the candidate configuration schemes are revised to obtain the final multi-dimensional regulatory resource coordination configuration scheme. Thus, by using the typical-extreme dual-layer scenario set, the flexibility requirements under both normal operation and complex extreme conditions can be captured, avoiding the limitations of traditional methods. The method relies solely on configuration blind spots caused by typical daily or average scenarios. It introduces a flexibility gap profiling mechanism to quantify the system's actual shortage capacity under different scenarios from six dimensions: upscaling, downscaling, ramping, continuous energy, reserve, and cross-sectional capacity. This shifts the resource allocation objective from simple capacity allocation to targeted reinforcement of operational shortcomings, significantly improving the pertinence and economy of resource allocation. Furthermore, by establishing resource matching degree and constructing an optimization model coupled with operational verification, boundary constraints can be verified simultaneously during the configuration phase, avoiding a disconnect between planning and operation. The final solution can take into account the needs of multiple time scales, such as short-term power support, intraday peak shaving, and long-term supply assurance. Moreover, through interpretable gap identification and verification feedback, it reduces the risks of over-allocation, under-allocation, and actual operational infeasibility of resource allocation, thereby improving the feasibility of resource allocation. Attached Figure Description

[0009] Figure 1 This is a flowchart of a multi-element regulation resource collaborative configuration method for a water-wind-solar system according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a multi-element regulation resource collaborative configuration system for a water, wind, and solar system according to an embodiment of the present invention. Detailed Implementation

[0010] Definitions:

[0011] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0012] Existing technologies for hydropower, wind power, and solar power system planning research primarily focus on capacity optimization and scheduling, targeting wind and solar curtailment rates, cost, or reliability. They employ clustering, typical day or time-series simulations to generate scenarios and introduce large-scale models to assist in prediction and scheduling. However, real-world systems lack various flexibility capabilities such as upsizing, ramp-up, and sustained energy, rather than just single-capacity adjustments. Existing methods struggle to identify and complete low-confidence and complex extreme conditions, lacking a clear representation of the direction, magnitude, frequency, and associated constraints of flexibility gaps. Large-scale model outputs are affected by data quality and black-box nature, making it difficult to provide interpretable data for resource allocation. Furthermore, capacity allocation and operational verification are disconnected, failing to simultaneously verify constraints such as reservoir water levels, energy storage charge, thermal power plant deep regulation, and cross-sectional load during the planning phase. This easily leads to problems such as constraint violations, over-allocation, or insufficient long-term supply during actual operation.

[0013] To address at least the aforementioned issues, a typical-extreme dual-layer scenario set is constructed based on the basic dataset of the target water-wind-solar system. Based on this set, a flexibility gap profile of the target water-wind-solar system under each scenario is identified. A resource matching degree for multi-dimensional regulatory resources is established based on the flexibility gap profile and resource matching degree. A resource allocation optimization model is constructed based on the flexibility gap profile and resource matching degree, and the model is solved to obtain candidate configuration schemes. These schemes are then validated using the typical-extreme dual-layer scenario set, and the validation results are used to revise the schemes, resulting in the final multi-dimensional regulatory resource coordination configuration scheme. In this way, by using a typical-extreme dual-layer scenario set, the flexibility requirements under normal operation and complex extreme conditions can be captured. This avoids the configuration blind spots caused by traditional methods that rely solely on typical daily or average scenarios. By introducing a flexibility gap profiling mechanism, the system's actual shortage capacity under different scenarios is quantified from six dimensions: upscaling, downscaling, ramping, continuous energy, reserve, and cross-sectional capacity. This shifts the resource allocation objective from simple capacity allocation to targeted reinforcement of operational shortcomings, significantly improving the targeting and economy of resource allocation. Furthermore, by establishing resource matching degree and constructing an optimization model coupled with operation verification, boundary constraints can be verified simultaneously during the configuration phase, avoiding a disconnect between planning and operation. The final solution can take into account the needs of multiple time scales, such as short-term power support, intraday peak shaving, and long-term supply assurance. Moreover, through interpretable gap identification and verification feedback, the risks of over-allocation, under-allocation, and actual operational infeasibility of resource allocation are reduced, improving the feasibility of resource allocation.

[0014] The following details a method for the coordinated allocation of multi-element regulatory resources in a water-wind-solar system according to the present invention. Please refer to [link / reference]. Figure 1 The method 100 includes steps 101 to 105: Step 101: Obtain the basic dataset of the target water-wind-scenic system, and construct a typical-extreme dual-layer scene set based on the basic dataset.

[0015] Step 102: Based on the typical-extreme dual-layer scenario set, identify the flexibility gap profile of the target water-wind-solar system under each scenario. The flexibility gap profile includes upward adjustment gap, downward adjustment gap, ramping gap, continuous energy gap, backup gap and cross-sectional bearing capacity gap.

[0016] Step 103: Establish the resource matching degree of multi-adjustment resources based on the aforementioned flexibility gap profile.

[0017] Step 104: Construct a resource configuration optimization model based on the flexibility gap profile and the resource matching degree, and solve the resource configuration optimization model to obtain candidate configuration schemes.

[0018] Step 105: Perform runtime verification on the candidate configuration scheme based on the typical-extreme dual-layer scenario set to obtain the verification result, and modify the candidate configuration scheme according to the verification result to obtain the final multi-dimensional adjustment resource coordination configuration scheme.

[0019] In one embodiment of the present invention, step 101 includes steps 1011 to 1016: Step 1011: Obtain the basic dataset of the target water-wind-sun system.

[0020] Specifically, multi-source heterogeneous data of the target water, wind and solar system are collected. The multi-source heterogeneous data includes new energy and load data, hydropower and regulation resource data, as well as power grid, meteorological and regulatory data. The multi-source heterogeneous data is processed by unifying the time scale, imputing missing values, removing outliers and normalizing the dimensions to obtain the basic dataset.

[0021] In one optional implementation, time-series data such as wind and solar power output, load, inflow, water level, energy storage SOC (state of charge of batteries), hydrogen storage capacity, and cross-sectional tidal flow are time-aligned, and data on negative power output, water levels exceeding limits, abnormal flow, SOC exceeding limits, and cross-sectional data that are significantly exceeding limits are corrected. Furthermore, high-frequency data are smoothed, downsampled, or windowed.

[0022] The new energy and load data include wind power data, photovoltaic data, and load data; the wind power data includes wind power installed capacity, historical output, predicted output, available power, curtailment records, and prediction errors; the photovoltaic data includes photovoltaic installed capacity, historical output, predicted output, available power, irradiance, temperature, and curtailment records; the load data includes historical load, predicted load, peak-to-valley difference, maximum ramp rate, and load pressure indicators such as high temperature and cold wave.

[0023] The hydropower and regulation resource data includes hydropower and reservoir data, as well as regulation resource data. The hydropower and reservoir data includes hydropower installed capacity, unit output upper and lower limits, inflow, outflow, reservoir water level, reservoir capacity, water level-storage capacity relationship curve, ecological flow, flood control restrictions, and dispatching procedures. The regulation resource data includes the power capacity, energy capacity, efficiency, response time, operating boundary, availability status, and investment cost of pumped storage, variable speed pumped storage, electrochemical energy storage, hydrogen energy storage, and thermal power flexibility resources.

[0024] The power grid, meteorological, and regulatory data include power grid data, meteorological data, and dispatching and maintenance data; the power grid data includes key section transmission limits, section load rates, transmission capacity constraints, and section over-limit records; the meteorological data includes wind speed, irradiance, rainfall, temperature, humidity, and extreme weather indicators; the dispatching and maintenance data includes load plans, power adjustment instructions, water level control requirements, standby requirements, unit maintenance plans, and operation and dispatching rules.

[0025] Step 1012: Construct scene feature vectors based on the aforementioned basic dataset.

[0026] Specifically, scene feature vectors are constructed based on the basic dataset according to a preset time window.

[0027] In one alternative implementation, the preset time window is 24 hours, 48 ​​hours, or 168 hours.

[0028] Specifically, a scene feature vector is constructed for the i-th sample in the basic dataset according to a preset time window, as follows: ; In the formula, Let i represent the scene feature vector constructed for the i-th sample. Indicates wind power output rate. Indicates photovoltaic power output rate. Indicates the load factor. Indicates the level of water abundance or scarcity. Indicates the initial storage capacity ratio of the reservoir. This indicates the maximum ramp rate of net load. Indicates reserve margin, Indicates the cross-sectional load rate. This indicates the adjustment of resource availability.

[0029] The calculation methods for each scene feature are as follows: ; In the formula, Indicates the preset time window. express Actual wind power output at all times Indicates wind power installed capacity. express Real-time photovoltaic output Indicates photovoltaic installed capacity. express System load at all times This indicates the maximum load within the statistical period.

[0030] Water abundance / sparseness level The water level is determined based on the average inflow within the window and the percentile of the same period over many years. For example, water levels below the 25th percentile are recorded as low water, water levels above the 75th percentile are recorded as high water, and the rest are recorded as normal water.

[0031] ; ; In the formula, This indicates the reservoir capacity at the start of the window. This indicates the reservoir capacity corresponding to the normal water level or the available regulating capacity. express Net load at all times express Net load at all times and These represent the average available standby capacity and the average required standby capacity within the window, respectively. Indicates the critical section at Current flow or transmission power at any given moment Indicates the cross-sectional transmission limit. This indicates that the resource capacity can be adjusted within the window. This indicates the rated capacity of the corresponding resource.

[0032] Step 1013: Standardize the scene feature vector to obtain a standardized scene feature vector. .

[0033] The first after standardization The first sample The eigenvalues ​​are as follows: ; In the formula, Represents the standardized first The first sample 1 eigenvalue, Indicates the first The first sample 1 eigenvalue, Indicates the first The first sample The mean of the eigenvalues, Indicates the first The first sample The standard deviation of each eigenvalue This represents the smallest positive number that prevents the denominator from being zero.

[0034] Step 1014: Cluster the scene features based on the standardized scene feature vectors to obtain the initial clustered scene.

[0035] In one alternative implementation, clustering is performed based on the standardized scene feature vector using methods such as K-means (K-means clustering algorithm), K-medoids (K-medoids clustering algorithm), hierarchical clustering, Gaussian mixture model, or spectral clustering to obtain an initial clustered scene.

[0036] Specifically, taking K-means as an example, its clustering objective function can be expressed as: ; In the formula, This represents the sum of squared errors from all samples to their respective cluster centers. Indicates the number of clusters. Indicates the first A sample set of clusters, Indicates the first There are 10 clusters. The number of clusters is... The results can be jointly determined by the contour coefficient, elbow rule, Davies-Bouldin index, and semantic analysis of scene distribution features by the large model. The large model does not directly replace the clustering algorithm here, but is used to assist in judging different... Does the scene division under the value conform to the operating rules of the water, wind and light system?

[0037] Step 1015: Based on the initial clustering scene identification, typical scenes and extreme scenes are obtained.

[0038] Step 1016: Obtain a typical-extreme dual-layer scene set based on the typical scenario and the extreme scenario, specifically as follows: ; In the formula, This represents a typical-extreme two-layer scene set. This represents a set of typical scenarios obtained through clustering and semantic correction of a large model. This represents the set of extreme scenarios identified through preset extreme operating trigger conditions, large model semantic correction, and physical constraint verification. Indicates the number of typical scenarios. This indicates the number of extreme scenarios. Each scenario includes a scenario number, scenario label, scenario level, scenario weight, key operational features, confidence level, triggering conditions, associated constraints, and possible gap types.

[0039] As described above, by constructing and standardizing scene feature vectors, the scale differences between data of different dimensions such as wind power, photovoltaics, and runoff are eliminated, making the clustering results more objective in reflecting the system's operating mode. Typical scenarios and extreme scenarios are constructed in two layers, which can not only cover the mainstream operating mode under normal scheduling, but also explicitly include low-probability, high-risk composite conditions such as continuous low wind and low light, and high load during the planning stage. This avoids the overwhelming of extreme samples by traditional methods due to peak shaving or averaging. The two-layer structure enables the resource allocation scheme to not only meet the efficiency and economic requirements for most periods, but also to have reliable supply guarantee and safety over-limit prevention capabilities under extreme conditions.

[0040] In one embodiment of the present invention, step 1015 includes steps 10151 to 10152: Step 10151: In the initial clustering scenario, the samples that are located between two or more cluster centers and have a small difference in distance from the nearest cluster center are marked as low-confidence samples, and the remaining samples are taken as typical scenarios.

[0041] Specifically, let and Samples The distance to the nearest and second nearest cluster centers, and the low-confidence sample criterion, can be expressed as: ; In the formula, This indicates the low confidence threshold. This represents a very small positive number to prevent the denominator from being zero. When the difference between the nearest and second-nearest distances is small, it indicates that the sample is located on the cluster boundary, and the initial clustering results are unstable.

[0042] Step 10152: Use the trained large model to determine whether the low-confidence sample is a typical scenario or an extreme scenario candidate sample. If it is an extreme scenario candidate sample, determine whether the extreme scenario candidate sample meets the preset extreme operation triggering conditions. If it does, perform physical constraint verification on the extreme scenario candidate sample. If the verification passes, the extreme scenario candidate sample is used as an extreme scenario.

[0043] Specifically, for low-confidence samples, their original operational characteristics, representative scenarios of the nearest and second-nearest clusters, and relevant meteorological, maintenance, and scheduling rule information are input into the trained large model. The trained large model then determines whether the sample is closer to the current cluster, the second-nearest cluster, or should be considered an extreme scenario candidate. If it is an extreme scenario candidate, it is determined whether the extreme scenario candidate meets the preset extreme operation triggering conditions. The preset extreme operation triggering conditions include continuous low wind and low light, low water and high load, high water and high new energy output combined with limited cross-section, unavailable regulation resources, low reservoir water level for supply, and high ramp load, etc. Taking a single sample as an example, its extreme scenario trigger function can be expressed as: ; In the formula, Indicates the first Each sample triggers a preset extreme running condition. Indicates the first The sample did not trigger the preset extreme operation trigger condition. This indicates an indicator function that returns 1 if the condition is true and 0 otherwise. Indicates the duration of continuous low wind and low light. These represent the corresponding trigger thresholds.

[0044] If an extreme scenario candidate sample meets the preset extreme operation triggering conditions, it is determined whether the scenario label of the extreme scenario candidate sample contradicts the physical constraint verification. If there is a contradiction, the scenario label is corrected; if it cannot be corrected, the sample is removed. If there is no contradiction, the extreme scenario candidate sample is used as an extreme scenario. The physical constraint verification is specifically as follows: ; ; ; ; In the formula, This represents the total power generation capacity of power sources including wind power, solar power, hydropower, and thermal power. This indicates the discharge or generation power of regulating resources such as energy storage, pumped hydro storage, or hydrogen storage. This indicates the charging, pumping, or electrical power consumption of energy storage, pumped hydro storage, or hydrogen production equipment. This indicates the permissible power balance deviation threshold. Indicates the reservoir capacity or equivalent water storage. and These represent the minimum and maximum allowed storage capacities, respectively. Indicates the state of charge of electrochemical energy storage. and These represent the minimum and maximum permissible states of charge, respectively. This indicates the actual load factor or transmission power of a key transmission section. This indicates the upper limit of the cross-section allowable.

[0045] For example, if the scene is labeled "section-restricted", but the scene meets the requirements in all time periods. ,in If the threshold for cross-section restriction warning is not met, it indicates that the label is inconsistent with the cross-section verification result; if the scenario label is "reservoir low water level supply guarantee", but... ,in If the low water level warning threshold is not met, it indicates that the label is inconsistent with the reservoir's capacity status; if the scenario label is "continuous low wind and low solar power," but the wind power output and solar power output are not in a continuous low water level warning state, then the label is incorrect. If a certain time period is lower than the corresponding threshold, it indicates that the label is inconsistent with the characteristics of new energy output.

[0046] In one alternative implementation, the method further includes training the large model to obtain a trained large model.

[0047] Specifically, after clustering historical operational data, scheduling experts or rule engines manually annotate some typical clustering results to form training samples. Each training sample includes input information such as cluster center, representative sample, sample distribution characteristics, meteorological text (obtained from meteorological data, extreme weather identifiers and weather process records), maintenance plan, equipment outage information, scheduling rules (scheduling and maintenance data) and cross-sectional constraints (power grid data), as well as corresponding scene semantic labels and explanatory text.

[0048] The training samples are as follows: ; In the formula, Indicates the first Structured input for clustering scenarios Semantic tags indicating manual confirmation This represents the explanatory text corresponding to the semantic tag.

[0049] The input is as follows: ; In the formula, Indicates the cluster center. This indicates that the representative sample is... Indicates the characteristics of sample distribution. This text displays operational information such as weather, maintenance, and equipment status. This section describes the scheduling rules and cross-sectional constraints.

[0050] The large model is trained using a fine-tuning method based on instructions, enabling the model to learn the following mapping relationship to obtain the trained large model: ; Given a structured description of a clustering scenario, the model outputs semantic labels and explanatory text for that clustering scenario.

[0051] As described above, by filtering low-confidence samples through cluster distance differences, it is possible to automatically locate complex time-series segments that are at the boundaries of multiple operating modes or difficult to classify, avoiding misjudgments caused by direct hard clustering. Introducing a large model to perform semantic-level judgment on these boundary samples can help identify potential extreme evolution trends by leveraging its powerful contextual understanding capabilities. However, it does not directly accept the model output, but further verifies it through preset extreme operating trigger conditions and physical constraint checks to ensure that the samples marked as extreme scenarios are physically feasible and conform to the actual operating boundaries of the system. This significantly improves the accuracy and interpretability of extreme scenario identification, ultimately enabling typical scenarios to cover the main operating rules, and extreme scenarios to capture rare working conditions with real supply pressure, providing reliable and targeted scenario inputs for subsequent flexibility gap profiling.

[0052] In one embodiment of the present invention, step 102 includes steps 1021 to 1024: Step 1021: Calculate the net load ramp-up requirement, reserve margin, and cross-sectional utilization of the target water-wind-solar system for each scenario and time period based on the typical-extreme dual-layer scenario set.

[0053] Specifically, the net load ramp-up requirement is as follows: ; ; In the formula, Representing a scene Time period The system net load is below, Representing a scene Time period The system load under the following conditions and Representing the scene respectively Time period The actual output of wind and solar power under the current conditions Representing a scene Time period The hydroelectric power must provide power. Representing a scene Time period The net load ramp-up demand, Representing a scene Time period The system net load is as follows.

[0054] The specific reserve margin is as follows: ; In the formula, Representing a scene Time period The reserve margin below, Representing a scene Time period The available spare capacity below Representing a scene Time period The required reserve capacity.

[0055] The specific utilization of the cross-section is as follows: ; In the formula, Representing a scene Time period The key section load rate, Representing a scene Time period The key section of the power flow or transmission power is below. This indicates the maximum allowable transmission power of the cross-section. When Approaching or exceeding This indicates that the scenario faces risks such as limited cross-section or strained transmission capacity.

[0056] Step 1022: Calculate the upward adjustment gap, downward adjustment gap, ramping gap, continuous energy gap, reserve gap, and cross-section bearing capacity gap based on the net load ramping demand, the reserve margin, and the cross-section utilization status.

[0057] Specifically, the upward adjustment gap is as follows: ; In the formula, Representing a scene Time period The upward gap below, Representing a scene Time period The available power generation capacity below This indicates that the ability to increase is available.

[0058] The specific downward adjustment gap is as follows: ; In the formula, Representing a scene Time period The downward gap below, Representing a scene Time period The required output power is as follows. Representing a scene Time period The ability to downregulate or absorb is now available.

[0059] The specific climbing gap is as follows: ; In the formula, Representing a scene Time period The gap in the slope below, Representing a scene Time period The available climbing ability.

[0060] The persistent energy gap specifically refers to: ; In the formula, Representing a scene The persistent energy gap below Representing a scene The set of consecutive gap periods below, Indicates a time interval.

[0061] The specific backup gap is: ; In the formula, Representing a scene Time period The backup gap below, when When <0, it indicates insufficient available reserve capacity, and the reserve shortfall is calculated based on the proportion of insufficient reserve margin; when When the value is greater than 0, it means that the available reserve capacity meets the required reserve capacity, and the reserve gap is 0.

[0062] The specific bearing capacity gap in the cross-section is as follows: ; In the formula, Representing a scene The cross-sectional bearing gap below.

[0063] Step 1023: Obtain the set of constraints related to the gap in each scenario.

[0064] Among them, the set of constraints related to the gap in each scenario includes reservoir water level, reservoir capacity, SOC, hydrogen storage capacity, thermal power plant ramp-up and cross-sectional limits.

[0065] Step 1024: Based on the upward adjustment gap, the downward adjustment gap, the ramping gap, the continuous energy gap, the reserve gap, the cross-sectional bearing capacity gap, and the constraint set, a flexibility gap profile is obtained, specifically as follows: ; In the formula, Representing a scene The following is a portrait of the flexibility gap. Representing a scene The upward gap below, Representing a scene The downward gap below, Representing a scene The gap in the slope below, Representing a scene The set of constraints associated with the gap. A flexibility gap profile can describe the gap type, direction, magnitude, duration, frequency of occurrence, associated constraints, and impact indicators.

[0066] As described above, six types of gaps are calculated based on physical quantities such as net load ramp-up demand, reserve margin, and cross-sectional utilization. This gives the quantification of flexibility gaps a clear operational physical meaning. By introducing a set of constraints related to the gaps, each type of gap is not only labeled with its amplitude and time period, but also associated with the constraint boundaries that trigger the gap. This clearly answers when, what constraints, and what kind of flexibility capabilities the system lacks, providing precise demand positioning for subsequent resource matching and avoiding blind allocation. At the same time, the dual-layer scenario set covers both normal and extreme operating conditions, enabling the profile to identify efficiency shortcomings under typical operating conditions and identify supply guarantee and safety over-limit risks under extreme operating conditions, ensuring that resource allocation decisions are both targeted and comprehensive.

[0067] In one embodiment of the present invention, step 103 includes steps 1031 to 1032: Step 1031: Establish a resource capability vector for multi-dimensional adjustment resources. The resource capability vector includes resource upswing, downswing, ramp-up, continuous energy, reserve, cross-section mitigation capability, and unit configuration or operating cost.

[0068] The aforementioned diversified regulatory resources include conventional hydropower, pumped storage, variable-speed pumped storage, electrochemical energy storage, hydrogen energy storage, and thermal power flexibility resources.

[0069] The resource capability vector is specifically: ; In the formula, The capability vector representing resource r. These represent the upward adjustment, downward adjustment, ramp-up, sustained energy, reserve, and fault mitigation capability of resource r, respectively. This represents the unit configuration or operating cost of resource r.

[0070] Step 1032: Calculate the resource matching degree of the multi-faceted adjustment resources based on the flexibility gap profile and the resource capability vector, specifically as follows: ; In the formula, This represents the resource matching degree of resource r to scenario s. Represents the normalization function. These represent the weighting coefficients, , , , , , These represent the normalized upward adjustment gap, downward adjustment gap, ramp-up gap, continuous energy gap, reserve gap, and cross-sectional bearing capacity gap, respectively.

[0071] As described above, each type of regulatory resource is abstracted into a multi-dimensional capability vector that includes upward adjustment, downward adjustment, ramp-up, continuous energy, reserve, cross-section mitigation capacity, and cost. This enables a structured correspondence between resource characteristics and flexibility gap profiles. By calculating the matching degree between the gap profile and the resource capability vector, it is possible to identify the contribution of each type of resource to a specific gap in a specific scenario. This reveals the complementary and substitutive relationships of different resources such as hydropower, pumped storage, electrochemical energy storage, and hydrogen energy storage in various dimensions, providing a quantitative basis for optimal allocation.

[0072] In one alternative implementation, it further includes: The resource reinforcement direction is determined based on the flexibility gap profile, the resource matching degree, and the resource capability vector. The resource reinforcement direction characterizes the priority relationship of different adjustment resources in addressing major flexibility gaps, and serves as the basis for determining the candidate resource set, configuration scale boundary, and investment priority when subsequently constructing a resource allocation optimization model.

[0073] Specifically, for each scenario Select resource matching degree Resources with higher levels of contribution and a positive marginal contribution to the main gap should be used as supplementary resources: ; In the formula, Representing a scene The following are the priority recommendations for reinforcement resources. For example, when the energy shortage is mainly a short-term ramp-up shortage, electrochemical energy storage or variable-speed pumped hydro storage should be prioritized as reinforcement directions; when the energy shortage is mainly a continuous energy shortage or a cross-day supply guarantee shortage, hydrogen energy storage, pumped hydro storage, or thermal power flexibility resources should be prioritized as reinforcement directions; when the energy shortage is mainly a down-adjustment consumption scenario or a large-scale new energy generation scenario, pumped hydro storage, variable-speed pumped hydro storage, electrochemical energy storage, or hydrogen production load should be prioritized as reinforcement directions; when the energy shortage is mainly due to insufficient cross-sectional capacity, local consumption-type energy storage, pumped hydro storage, or regulation resources in the relevant cross-sectional constraint area should be prioritized as reinforcement directions.

[0074] In one embodiment of the present invention, step 104 includes steps 1041 to 1042: Step 1041: Based on the flexibility gap profile, the resource matching degree, and the resource reinforcement direction, determine the candidate resource set, configuration scale boundary, initial reference scale, and investment priority of the diversified adjustment resources.

[0075] Step 1042: Using the configuration scale of various adjustment resources in the candidate configuration resource set as decision variables, construct an objective function and constraints that consider investment costs, gap reduction, new energy consumption, security supply and resource matching degree based on the resource reinforcement direction and investment priority, and obtain a resource allocation optimization model.

[0076] The objective function is specifically: ; In the formula, J represents the objective function. This indicates the configuration size of resource r. These represent the weighting coefficients, Representing a scene The weight, Indicates configuration scheme In the scene The risk of unmet needs or supply shortages. Indicates configuration scheme In the scene Losses due to wind curtailment, solar curtailment, hydropower curtailment, or limited energy utilization. Indicates configuration scheme In the scene Penalties are imposed for exceeding the limits of the cross-section or violating operational constraints. This objective function is used to comprehensively weigh the resource investment costs and the effectiveness of improving operational shortcomings.

[0077] The constraints include resource allocation boundary constraints, investment budget constraints, power balance constraints, energy storage status constraints, reservoir water level and capacity constraints, and cross-sectional transmission constraints.

[0078] The resource allocation boundary constraints are specifically as follows: ; In the formula, This indicates the lower limit of the configuration size of resource r. This indicates the upper limit of the configuration size of resource r.

[0079] The specific investment budget constraints are as follows: ; In the formula, This indicates the upper limit of the investment budget.

[0080] The power balance constraint is specifically as follows: ; In the formula, Representing a scene Time period The total power generation capacity of wind power, photovoltaic power, hydropower, thermal power, and other power sources is as follows: Representing a scene Time period The discharge or power generation capacity of energy storage, pumped hydro storage, or hydrogen energy storage, etc., can be regulated. Representing a scene Time period The charging, pumping, or electrical power of energy storage, pumped storage, or hydrogen production equipment.

[0081] The energy storage state constraints are specifically as follows: ; In the formula, Representing a scene Time period The electrochemical energy storage state of charge.

[0082] The specific constraints on reservoir water level and storage capacity are as follows: ; In the formula, Representing a scene Time period The reservoir capacity or equivalent water storage capacity.

[0083] The cross-sectional transmission constraint is specifically as follows: ; In the formula, Representing a scene Time period The actual load rate or transmission power of the key transmission sections.

[0084] Constraints are used to ensure the feasibility and operational safety of the configuration scheme. Specifically, resource allocation boundary constraints limit the constructable or modifiable scale of various regulatory resources; investment budget constraints ensure the configuration scheme meets investment boundaries; power balance constraints ensure supply and demand balance in different time periods; energy storage status constraints prevent overcharging, over-discharging, or lifespan damage of energy storage; reservoir water level and capacity constraints ensure reservoir operation does not exceed ecological, flood control, and dispatching boundaries; and cross-sectional transmission constraints ensure the transmission of new energy and the safe operation of the power grid.

[0085] As described above, based on the flexibility gap profile, resource matching degree, and resource reinforcement direction, the candidate resource set, configuration scale boundary, initial reference scale, and investment priority of diversified adjustment resources are determined. This ensures that the candidate resource set, configuration scale boundary, initial reference scale, and investment priority of diversified adjustment resources are based on the quantitative matching results of gap direction, amplitude, and resource capability vector under each scenario. This ensures that new resources accurately target the weakest link in the system. The configuration scale of various adjustment resources in the candidate resource set is used as the decision variable. Based on the resource reinforcement direction and investment priority, an objective function is constructed that covers investment cost, gap reduction, new energy consumption, security supply, and resource matching degree. This ensures that the optimization process takes into account economy, security, and functionality.

[0086] In one embodiment of the present invention, step 104 further includes steps 1043 to 1046: Step 1043: Solve the resource configuration optimization model to obtain the initial candidate configuration scheme.

[0087] Specifically, linear programming, mixed-integer programming, multi-objective optimization algorithms, or heuristic optimization algorithms are used to solve the resource allocation optimization model to obtain initial candidate allocation schemes. In the formula, Let k represent the k-th initial candidate configuration scheme, k=1,…,n, where n represents the total number of initial candidate configuration schemes.

[0088] Step 1044: Calculate the comprehensive evaluation value for each initial candidate configuration scheme, specifically as follows: ; In the formula, This represents the comprehensive evaluation value of the k-th initial candidate configuration scheme. These represent the evaluation weights, This represents the gap reduction effect of the k-th initial candidate configuration scheme. This represents the effect of improving the renewable energy consumption of the k-th initial candidate configuration scheme. This indicates the safety supply or cross-sectional load-bearing capacity improvement effect of the k-th initial candidate configuration scheme. This represents the cost of the k-th initial candidate configuration.

[0089] Step 1045: Sort the initial candidate configuration schemes in descending order of the comprehensive evaluation value to obtain the sorted initial candidate configuration schemes.

[0090] Step 1046: Select the first preset number of initial candidate configuration schemes from the sorted initial candidate configuration schemes as candidate configuration schemes.

[0091] In one alternative implementation, the preset number can be 1 or 3-5.

[0092] As described above, by generating multiple initial candidate configuration schemes and calculating comprehensive evaluation values, and retaining the schemes ranked higher, the risk of a single optimal solution losing its feasibility in actual engineering due to fluctuations in boundary conditions or uncertainty in parameters is effectively avoided.

[0093] In one embodiment of the present invention, step 105 includes steps 1051 to 1053: Step 1051, construct the joint runtime state space corresponding to the candidate configuration scheme, specifically as follows: ; In the formula, Indicates candidate configuration schemes In the scene Time period The joint operating state space under, Indicates in the scene Time period Hydrogen storage under the following conditions: Indicates in the scene Time period The thermal power plant is in a deeply adjustable or adjustable state.

[0094] Step 1052: Substitute the candidate configuration scheme into the typical-extreme two-layer scenario set, calculate the operation index of the candidate configuration scheme in the corresponding scenario on a rolling basis according to the joint operation state space, and calculate the constraint violation degree, resource utilization rate and marginal adjustment contribution of the candidate configuration scheme in the corresponding scenario according to the operation index.

[0095] The operational indicators include supply shortage, wind, solar and hydropower curtailment, cross-section exceeding limits, reservoir exceeding limits, SOC exceeding limits, and insufficient hydrogen storage.

[0096] The specific degree of constraint violation is as follows: ; In the formula, This indicates the degree of constraint violation of the candidate configuration scheme in the corresponding scenario s. These represent the constraint violation weights.

[0097] The resource utilization rate is specifically: ; In the formula, Representing resources In candidate configuration schemes utilization rate Representing resources In the scene Time period The actual power of the call, Representing resources In candidate configuration schemes The configuration scale below, Indicates the number of verification periods.

[0098] The specific marginal adjustment contribution is as follows: ; In the formula, Representing resources The marginal adjustment contribution, This function represents the combined effects of reducing the supply gap, increasing the absorption of new energy sources, reducing the risk of supply shortages, or decreasing the number of sections exceeding the limit. Representing resources The capacity perturbation.

[0099] Step 1053: Obtain the verification result based on the constraint violation degree, the resource utilization rate, and the marginal adjustment contribution.

[0100] As described above, by constructing a joint operation state space, the time-varying states of various resources, such as hydropower reservoir capacity, energy storage charge, hydrogen storage status, thermal power deep regulation, and cross-sectional load rate, are uniformly incorporated into the verification framework. This avoids the one-sidedness of traditional planning that only performs power balance verification. By substituting candidate configuration schemes into a two-layer scenario set and calculating constraint violation, it is possible to accurately locate the time periods and resource types where the configuration scheme may exceed the limits in actual operation. Resource utilization assessment can identify inefficient configurations and prevent redundant investment. Marginal adjustment contribution measures the actual efficiency improvement of new resources under critical operating conditions, providing direction for scheme fine-tuning. The verification results formed by the combination of these three factors not only quantify the feasibility and safety margin of the scheme but also provide feedback on adjustment priorities, effectively improving the engineering reliability and economy of the configuration scheme.

[0101] In one embodiment of the present invention, step 105 further includes step 1054: Step 1054: Based on the constraint violation degree, the resource utilization rate, and the marginal adjustment contribution, the resource allocation scale in the candidate allocation scheme is modified to obtain the final multi-dimensional adjustment resource coordination allocation scheme, specifically as follows: ; In the formula, Indicates the first Resources after round correction Configuration scale, Indicates the first Wheel Resources Configuration scale, The higher the value, the more beneficial it is to reduce the gap or improve constraints by increasing the resource. Representing resources Utilization rate, if A lower value indicates that the resource may be over-allocated. Representation and Resources The degree of violation of relevant constraints , , These represent the adjusted weights for marginal contribution, resource utilization rate, and constraint violation degree, respectively. This indicates that the revised configuration scale will be projected onto the budget, construction scale, and technical feasibility domain. Inside.

[0102] In one optional implementation, the adjustment weights of each item can be adjusted according to a preset adjustment rule during actual implementation. The preset adjustment rule is as follows: if a certain type of gap is still large and the corresponding resource has a high marginal contribution, the allocation scale of that type of resource is increased; if the utilization rate of a certain type of resource is low or the marginal contribution is low, its allocation scale is reduced; if there is a violation of constraints on reservoirs, energy storage, hydrogen storage, thermal power, or cross-sectional areas, the resource allocation scale is reduced; if the constraint violation is still not eliminated, the allocation scale of the relevant resources is readjusted according to the type of constraint violation, resource matching degree, and marginal adjustment contribution.

[0103] when At higher levels, the correction formula tends to increase resources. Configuration scale; when At lower levels, the correction formula tends to reduce resources. Over-configuration; when A higher value indicates that the current configuration may lead to a constraint violation, and the correction formula tends to reduce the size of the resource allocation.

[0104] Specifically, when the revised candidate configuration scheme meets the preset requirements for safety, economy and operational feasibility, the final multi-dimensional regulatory resource coordination configuration scheme is output, including: the capacity scale of various regulatory resources, the multi-dimensional regulatory resource configuration structure, investment priority, applicable scenarios, corresponding gap types and reinforcement directions.

[0105] As described above, directly using the constraint violation rate, resource utilization rate, and marginal adjustment contribution generated during the verification stage as the basis for correction can systematically eliminate the risk of going out of bounds and redundant investment caused by the simplification of the optimization model or the compression of the scenario. This ensures that the final solution meets the operational constraints such as storage capacity, state of charge, and cross-sectional load in all verification scenarios, while improving the overall resource utilization efficiency.

[0106] Please refer to Figure 2 The present invention also provides a multi-dimensional regulation resource collaborative configuration system 200 for a water-wind-solar system, including a memory 201, a processor 202, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the various steps of the multi-dimensional regulation resource collaborative configuration method for a water-wind-solar system as described above.

[0107] In summary, the multi-dimensional regulatory resource coordination configuration method and system for water-wind-solar systems described above in this invention constructs a typical-extreme dual-layer scenario set based on the basic dataset of the target water-wind-solar system. Based on this set, it identifies the flexibility gap profile of the target system under each scenario. It then establishes the resource matching degree of multi-dimensional regulatory resources based on the flexibility gap profile. A resource configuration optimization model is constructed based on the flexibility gap profile and the resource matching degree, and the model is solved to obtain candidate configuration schemes. The candidate configuration schemes are then run and verified based on the typical-extreme dual-layer scenario set to obtain verification results. Based on these results, the candidate configuration schemes are corrected to obtain the final multi-dimensional regulatory resource coordination configuration scheme. This approach, through the typical-extreme dual-layer scenario set, captures the flexibility requirements under both normal operation and complex extreme conditions, avoiding the configuration blind spots caused by traditional methods that rely solely on typical daily or average scenarios. It introduces a flexibility gap profile mechanism, quantifying the system's actual short-term performance under different scenarios from six dimensions: up-adjustment, down-adjustment, ramp-up, continuous energy, reserve, and cross-sectional load capacity. The lack of capacity has shifted resource allocation objectives from simple capacity allocation to targeted reinforcement of operational shortcomings, significantly improving the relevance and economy of resource allocation. Furthermore, establishing resource matching degrees and constructing an optimization model coupled with operational verification allows for simultaneous verification of boundary constraints during the allocation phase, avoiding a disconnect between planning and operation. The final solution can accommodate multiple time-scale needs, including short-term power support, intraday peak shaving, and long-term supply assurance. Through interpretable gap identification and verification feedback, the risks of over-allocation, under-allocation, and operational infeasibility in resource allocation are reduced, improving the feasibility of resource allocation. In addition, each type of regulatory resource is abstracted into a multi-dimensional capability vector including upward adjustment, downward adjustment, ramp-up, continuous energy, reserve, cross-section mitigation capacity, and cost. This creates a structured correspondence between resource characteristics and flexibility gap profiles. By calculating the matching degree between gap profiles and resource capability vectors, the contribution of each type of resource to specific gaps in specific scenarios can be identified, revealing the complementary and substitutive relationships of different resources such as hydropower, pumped storage, electrochemical energy storage, and hydrogen energy storage in various dimensions, providing quantitative basis for optimal allocation.

[0108] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for the coordinated allocation of multi-element regulatory resources in a water-wind-solar system, characterized in that, include: Obtain the basic dataset of the target water-wind-scenic system, and construct a typical-extreme two-layer scene set based on the basic dataset; Based on the typical-extreme dual-layer scene set, the flexibility gap profile of the target water-wind-solar system under each scene is identified. The flexibility gap profile includes upward adjustment gap, downward adjustment gap, ramping gap, continuous energy gap, backup gap and cross-sectional bearing capacity gap. Based on the aforementioned flexibility gap profile, establish a resource matching degree for diversified adjustment resources; Based on the flexibility gap profile and the resource matching degree, a resource allocation optimization model is constructed, and the resource allocation optimization model is solved to obtain candidate configuration schemes; The candidate configuration schemes are run and verified based on the typical-extreme dual-layer scenario set to obtain verification results. The candidate configuration schemes are then modified according to the verification results to obtain the final multi-dimensional adjustment resource coordination configuration scheme.

2. The method according to claim 1, characterized in that, The typical-extreme two-layer scenario set constructed based on the aforementioned basic dataset includes: Construct scene feature vectors based on the aforementioned basic dataset; The scene feature vector is standardized to obtain a standardized scene feature vector; Clustering is performed based on the standardized scene feature vectors to obtain the initial clustered scene; Based on the initial clustering scene identification, typical and extreme scenes are obtained; Based on the typical scenarios and the extreme scenarios, a typical-extreme dual-layer scenario set is obtained.

3. The method according to claim 2, characterized in that, Based on the initial clustering scene identification, typical and extreme scenes are obtained, including: In the initial clustering scenario, samples that are located between two or more cluster centers and have a small difference in distance from the nearest cluster center are marked as low-confidence samples, and the remaining samples are used as typical scenarios. The trained large model is used to determine whether the low-confidence sample is a typical scenario or an extreme scenario candidate sample. If it is an extreme scenario candidate sample, it is determined whether the extreme scenario candidate sample meets the preset extreme operation triggering conditions. If it does, the extreme scenario candidate sample is physically constrained and verified. If the verification passes, the extreme scenario candidate sample is used as an extreme scenario.

4. The method according to claim 1, characterized in that, Based on the typical-extreme dual-layer scene set, the flexibility gap profile of the target water-wind-light system in each scene is identified as follows: Based on the typical-extreme dual-layer scenario set, calculate the net load ramp-up demand, reserve margin and cross-sectional utilization of the target water-wind-solar system under each scenario and time period; Calculate the upward adjustment gap, downward adjustment gap, ramping gap, continuous energy gap, reserve gap, and cross-section bearing capacity gap based on the net load ramping demand, the reserve margin, and the cross-section utilization status. Obtain the set of constraints related to the gap in each scenario; A flexibility gap profile is obtained based on the upward adjustment gap, the downward adjustment gap, the ramping gap, the continuous energy gap, the backup gap, the cross-sectional bearing capacity gap, and the constraint set.

5. The method according to claim 1, characterized in that, Based on the aforementioned flexibility gap profile, the resource matching degree for establishing diversified adjustment resources includes: Establish a resource capability vector for multi-dimensional adjustment resources. The resource capability vector includes resource up-adjustment, down-adjustment, ramp-up, continuous energy, reserve, cross-section mitigation capability, and unit configuration or operating cost. The resource matching degree of multi-dimensional adjustment resources is calculated based on the flexibility gap profile and the resource capability vector.

6. The method according to claim 1, characterized in that, Constructing a resource allocation optimization model based on the aforementioned flexibility gap profile and resource matching degree includes: Based on the flexibility gap profile, the resource matching degree, and the resource reinforcement direction, determine the candidate resource set, configuration scale boundary, initial reference scale, and investment priority of the diversified adjustment resources; Using the configuration scale of various adjustment resources in the candidate configuration resource set as decision variables, and based on the resource reinforcement direction and investment priority, an objective function and constraints considering investment cost, gap reduction, new energy consumption, security supply and resource matching degree are constructed to obtain a resource allocation optimization model.

7. The method according to claim 1, characterized in that, Solving the resource allocation optimization model yields candidate allocation schemes, including: The resource allocation optimization model is solved to obtain initial candidate allocation schemes; Calculate the comprehensive evaluation value for each initial candidate configuration scheme; The initial candidate configuration schemes are sorted in descending order of the comprehensive evaluation value to obtain the sorted initial candidate configuration schemes. Select the first preset number of initial candidate configuration schemes from the sorted initial candidate configuration schemes as candidate configuration schemes.

8. The method according to claim 1, characterized in that, The candidate configuration schemes were validated based on the typical-extreme two-layer scenario set, and the validation results include: Construct a joint runtime state space corresponding to the candidate configuration schemes; Substitute the candidate configuration schemes into the typical-extreme two-layer scenario set, calculate the operational indicators of the candidate configuration schemes in the corresponding scenarios on a rolling basis according to the joint operational state space, and calculate the constraint violation degree, resource utilization rate and marginal adjustment contribution of the candidate configuration schemes in the corresponding scenarios based on the operational indicators. The verification result is obtained based on the constraint violation degree, the resource utilization rate, and the marginal adjustment contribution.

9. The method according to claim 8, characterized in that, Based on the verification results, the candidate configuration schemes are revised to obtain the final multi-dimensional adjustment resource coordination configuration scheme, including: The resource allocation scale in the candidate allocation scheme is modified based on the constraint violation degree, the resource utilization rate, and the marginal adjustment contribution to obtain the final multi-dimensional adjustment resource coordination allocation scheme.

10. A multi-element regulation resource coordination and allocation system for a water-wind-solar system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements each step of the multi-element regulation resource collaborative configuration method for a water-wind-solar system as described in any one of claims 1 to 9.

Images