Nuclear power storage integrated joint robust capacity planning method and system

Abstract

The invention discloses a nuclear storage integrated joint robust capacity planning method and system, and the method comprises the steps: carrying out the clustering of a preprocessed historical data set through employing a K-medoids algorithm based on density weighting, and obtaining a combined scene set which comprises a typical scene set and an extreme weather scene set, modeling the wind and light output uncertainty based on a typical scene set and an extreme weather scene set to obtain a wind and light output uncertainty set constraint, constructing an NPSI joint robust capacity planning model based on the wind and light output uncertainty set constraint, and based on improved Camp; the NPSI combined robust capacity planning model is solved through a CG algorithm, a capacity planning result is obtained, and therefore accurate scene clustering can be achieved, the influence of extreme weather is considered, the random uncertainty and overload of new energy are considered, the system adjustment requirement can be flexibly responded, and improved Camp is used; the CG algorithm is used for solving the model, the solving efficiency is higher, the solving result is more accurate, and therefore the planning accuracy is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power systems, in particular to a nuclear pumped storage integration combined robust capacity planning method and system. BACKGROUND

[0002] Energy storage is a powerful means to deal with the instability of renewable energy output, among which pumped storage accounts for more than 76%; at the same time, nuclear power has small environmental pollution, large unit area installed capacity, low marginal cost than thermal power, nuclear fuel is theoretically non-exhaustible, and operation is reliable, which is one of the important links to solve the energy shortage, combined with the excellent regulation capacity of pumped storage, which can greatly release the potential of nuclear power construction. But nuclear power has the characteristics of high construction cost, high safety requirement, technical limitation of frequent change of power load, difficulty in fuel design and post-processing caused by frequent change of power, high risk of misoperation, etc. The early operation mode only carries basic load and does not participate in peak regulation. Nuclear power runs at rated or near rated power, and other units cooperate with load changes to adjust power. After technical iteration, the newly added nuclear power units generally have a certain peak regulation capacity, but they still cannot match thermal power units in terms of peak regulation depth and economy, and there are certain safety hazards. Nuclear pumped storage is the best choice to take advantage of their respective advantages. The combined operation of nuclear power and pumped storage can better meet the needs of the power system and ensure that nuclear power units operate more stably as base load.

[0003] Due to the fact that pumped storage power stations are mainly restricted by the water storage capacity of the upper pool and the water resource conditions they receive, in the current context of prominent peak regulation problems, the planning and construction of pumped storage power stations have also become an important part of supplementing the peak regulation capacity of the power system and ensuring the safe and stable operation of the system. For a long time, the purpose of the construction of peak regulation power sources is to optimize the operation of the power system, but in the planning and development of pumped storage power stations, due to the lack of unified planning for the construction of pumped storage power stations, combined with the current non-uniformity of the operating mode and electricity price mode of pumped storage power stations, the actual construction of pumped storage power stations has not been coordinated with the changes in the demand for grid peak regulation and frequency modulation. Therefore, according to the demand for grid peak regulation and frequency modulation, researching the pumped storage capacity planning method of nuclear pumped storage integration (NPSI) power station has important theoretical and practical significance. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a nuclear pumped storage integration combined robust capacity planning method and system, which can improve the accuracy of planning.

[0005] To solve the above technical problems, a technical solution adopted by the present application is:

[0006] A nuclear pumped storage integration combined robust capacity planning method, comprising the steps of:

[0007] obtain a historical data set, and pre-process the historical data set to obtain a pre-processed historical data set;

[0008] cluster the pre-processed historical data set using a K-center clustering algorithm based on density weighting to obtain a combined scenario set, the combined scenario set including a typical scenario set and an extreme weather scenario set;

[0009] model wind-solar power output uncertainty based on the typical scenario set and the extreme weather scenario set to obtain wind-solar power output uncertainty set constraints;

[0010] construct a nuclear-storage integrated joint robust capacity planning model based on the wind-solar power output uncertainty set constraints, and solve the nuclear-storage integrated joint robust capacity planning model based on an improved constraint and column generation algorithm to obtain a capacity planning result.

[0011] To solve the above technical problems, another technical solution adopted by the present application is:

[0012] A nuclear-storage integrated joint robust capacity planning system includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the following steps when executing the computer program:

[0013] obtain a historical data set, and pre-process the historical data set to obtain a pre-processed historical data set;

[0014] cluster the pre-processed historical data set using a K-center clustering algorithm based on density weighting to obtain a combined scenario set, the combined scenario set including a typical scenario set and an extreme weather scenario set;

[0015] model wind-solar power output uncertainty based on the typical scenario set and the extreme weather scenario set to obtain wind-solar power output uncertainty set constraints;

[0016] construct a nuclear-storage integrated joint robust capacity planning model based on the wind-solar power output uncertainty set constraints, and solve the nuclear-storage integrated joint robust capacity planning model based on an improved constraint and column generation algorithm to obtain a capacity planning result.

[0017] The beneficial effects of the present application are that: the K-medoids algorithm based on density weighting is used to cluster the pretreated historical data set, to obtain a combined scene set including a typical scene set and an extreme weather scene set, the wind and light output uncertainty is modeled based on the typical scene set and the extreme weather scene set, the wind and light output uncertainty set constraint is obtained, the NPSI joint robust capacity planning model is constructed based on the wind and light output uncertainty set constraint, and the improved C&CG algorithm is used to solve the NPSI joint robust capacity planning model, to obtain the capacity planning result, so that the K-medoids algorithm based on density weighting can realize accurate scene clustering, the influence of extreme weather is considered, the NPSI joint robust capacity planning model considers the random uncertainty of new energy and load overload, can flexibly respond to system regulation demand, the improved C&CG algorithm is used to solve the model, the solving efficiency is higher and the solving result is more accurate, so that the planning accuracy is improved. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 A step flow chart of a nuclear and storage integrated joint robust capacity planning method is provided for an embodiment of the present application.

[0019] Figure 2 A structure schematic diagram of a nuclear and storage integrated joint robust capacity planning system is provided for an embodiment of the present application.

[0020] Figure 3 A standard IEEE-24 node system source load configuration and network structure schematic diagram in the nuclear and storage integrated joint robust capacity planning method is provided for an embodiment of the present application.

[0021] Figure 4 A system annual operation cost schematic diagram of different examples in the nuclear and storage integrated joint robust capacity planning method is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0022] To explain the technical content, the achieved purposes and effects of the present application in detail, the following will be explained in combination with embodiments and the accompanying drawings.

[0023] Please refer to Figure 1 A nuclear and storage integrated joint robust capacity planning method, comprising the steps of:

[0024] Obtaining a historical data set, and pretreating the historical data set to obtain a pretreated historical data set;

[0025] Using a K-medoids algorithm based on density weighting to cluster the pretreated historical data set, to obtain a combined scene set, the combined scene set including a typical scene set and an extreme weather scene set;

[0026] modeling wind and light output uncertainty based on the typical scenario set and the extreme weather scenario set to obtain wind and light output uncertainty set constraints;

[0027] constructing a nuclear and storage integrated joint robust capacity planning model based on the wind and light output uncertainty set constraints, and solving the nuclear and storage integrated joint robust capacity planning model based on an improved constraint and column generation algorithm to obtain a capacity planning result.

[0028] From the above description, the beneficial effects of the present application are that the K-medoids algorithm based on density weighting is used to cluster the pretreated historical data set to obtain a combined scenario set, which includes a typical scenario set and an extreme weather scenario set, wind and light output uncertainty is modeled based on the typical scenario set and the extreme weather scenario set to obtain wind and light output uncertainty set constraints, an NPSI joint robust capacity planning model is constructed based on the wind and light output uncertainty set constraints, and the NPSI joint robust capacity planning model is solved based on an improved constraint and column generation (C&CG) algorithm to obtain a capacity planning result, so that the K-medoids algorithm based on density weighting can realize accurate scenario clustering, the influence of extreme weather is considered, the NPSI joint robust capacity planning model constructed considers the random uncertainty of new energy and load overload, can flexibly respond to system adjustment demand, the improved C&CG algorithm is used to solve the model, the solving efficiency is higher and the solving result is more accurate, and therefore the planning accuracy is improved.

[0029] Further, the pretreatment of the historical data set to obtain the pretreated historical data set comprises:

[0030] The historical data set is dimensionally reduced using principal component analysis to obtain the pretreated historical data set.

[0031] From the above description, for a system with dozens of nodes, the feature variable dimension of the historical data set can reach thousands of dimensions, and a large number of redundant features are contained therein, such as the total photovoltaic output being always 0 at night. The principal component analysis (PCA) is used to reduce the dimension of the historical data set, which can effectively improve the data processing efficiency.

[0032] Further, the clustering of the pretreated historical data set using the K-medoids algorithm based on density weighting to obtain the combined scenario set comprises:

[0033] An initial scenario set is obtained according to the pretreated historical data set;

[0034] Gaussian kernel density estimation is performed on each scenario in the initial scenario set to obtain a density value of each scenario;

[0035] determining a density-weighted distance of each scene according to the density value of the scene;

[0036] determining a number of typical scenes using a silhouette coefficient method;

[0037] performing clustering on the initial scene set based on the number of typical scenes using a K-medoids clustering algorithm to obtain a typical scene set;

[0038] performing clustering on the edge scene set using a density-weighted K-medoids clustering algorithm to obtain an extreme weather scene set.

[0039] As can be seen from the above description, in actual system operation, wind and light output and system load have strong spatio-temporal coupling effect, and there may be extreme scenes mainly composed of edge scenes in the data set. Direct use of conventional clustering methods may affect the typical scenes and clusters obtained by clustering. The K-medoids clustering algorithm based on density weighting is used to select a typical scene set, and the final clustering center is the actual scene, not the average of each scene, which greatly reduces the interference of outliers on the selection of cluster center. The silhouette coefficient method is used to determine the number of typical scenes, which requires that the difference between scenes in the same cluster is small enough, and the difference between scenes in different clusters is large enough, ensuring the accuracy of the selection of the typical scene set. In addition, the K-medoids clustering algorithm based on density weighting is used to select a representative extreme weather scene set, which not only considers the difference between the typical scene and the extreme weather scene, but also considers the difference between the existing extreme weather scene and the new extreme weather scene, ensuring that the new selected extreme weather scene can provide the maximum incremental information for the model, thereby effectively improving the reliability and accuracy of scene selection.

[0040] Further, the Gaussian kernel density estimation on each scene in the initial scene set to obtain a density value of each scene comprises:

[0041]

[0042] wherein f i represents the density value of the initial scene X i , X j , X i represent different initial scenes, δ represents the effective radius of the neighborhood, and N s represents the number of initial scenes.

[0043] As can be seen from the above description, Gaussian kernel density estimation (GKDE) can quantify the relative density or importance of each scene in the data space. By calculating the density value of each scene, it can be identified which scenes in the data set are more common or representative, which helps to more accurately process and analyze high-dimensional data and improve the accuracy and robustness of the model.

[0044] Further, the determining of the density-weighted distance of each scene according to the density value of the scene comprises:

[0045]

[0046] In the formula, d ij represents the density-weighted distance of scene x i and scene x j .

[0047] As can be seen from the above description, the density-weighted distance can more accurately reflect the similarity and difference between scenes.

[0048] Further, the constructing of the nuclear-pumped storage integrated joint robust capacity planning model based on the uncertain set constraints of the wind and light output comprises:

[0049] establishing a target function of minimizing annual investment cost and operation cost based on the uncertain set constraints of the wind and light output, and nuclear-pumped storage integrated joint operation power station pumped storage planning constraints and comprehensive scene intra-day operation constraints corresponding to the target function;

[0050] obtaining the nuclear-pumped storage integrated joint robust capacity planning model according to the target function, the nuclear-pumped storage integrated joint operation power station pumped storage planning constraints and the comprehensive scene intra-day operation constraints.

[0051] As can be seen from the above description, the nuclear-pumped storage integrated joint robust capacity planning model is constructed to optimize the coordinated operation of nuclear power units and pumped storage power stations, and to cope with the uncertainty of wind and light output. The model can ensure the optimal configuration of the output and operation strategy of nuclear power units and pumped storage power stations under different operation scenes, effectively reduce the system operation risk, improve the flexibility and reliability of the power system, optimize resource allocation, reduce the overall cost, enhance the accommodation capacity of renewable energy of the power system, achieve the balance between economy and reliability, and provide strong support for the stable operation of the power grid.

[0052] Further, the target function is specifically:

[0053]

[0054] In the formula, C total represents the total annual cost, C Inv represents the annual investment cost of the nuclear-pumped storage integrated joint operation power station, C Ope represents the annual operation cost of the nuclear-pumped storage integrated joint operation power station participating in the system operation, c C represents the unit capacity configuration cost of the nuclear-pumped storage integrated joint operation power station pumped storage, C represents the planning reservoir capacity of pumped storage of the nuclear-pumped storage integrated combined operation power station, α represents the discount rate, T P C represents the unit capacity configuration cost of pumped storage of the nuclear-pumped storage integrated combined operation power station, C represents the rated power of pumped storage of the nuclear-pumped storage integrated combined operation power station, α represents the discount rate, T life C represents the life of the nuclear-pumped storage integrated combined operation power station, π r C represents the probability of the combined scenario r, C represents the daily operation cost under the combined scenario r, i represents the node number of the planned nuclear-pumped storage integrated combined operation power station, U represents the output set of renewable energy under uncertainty, z represents a 0-1 variable of uncertain wind power output, v represents a 0-1 variable of uncertain photovoltaic output, Ξ represents the value set of the above 0-1 planning variables, p represents continuous planning variables including conventional unit output, line power, etc., and Ψ represents the value set of the above continuous planning variables.

[0055] As can be seen from the above description, since nuclear power planning is affected by various technical and non-technical factors, the planning capacity and cost of the nuclear power part are fixed by default in the NPSI combined operation power station planning, and the benefits of pumped storage are considered on this basis, and the capacity construction cost and coal saving benefits of pumped storage are modeled and analyzed, and the proposed planning model is planned for a single level year, so as to convert the equipment investment cost to a unified annual investment cost.

[0056] Further, the improved constraint and column generation algorithm is used to solve the nuclear-pumped storage integrated robust capacity planning model, and the capacity planning result includes:

[0057] The constraint and column generation algorithm is used to solve the objective function in the pre-scheduling stage, and the pre-scheduling stage objective function is obtained;

[0058] The pre-scheduling stage objective function is transformed to obtain a transformed pre-scheduling stage objective function, and the transformed pre-scheduling stage objective function is a max-min problem;

[0059] The min problem in the transformed pre-scheduling stage objective function is dually transformed based on the auxiliary variable in the re-scheduling stage to obtain a re-scheduling stage objective function, and the re-scheduling stage objective function is a max problem;

[0060] The re-scheduling stage objective function is solved to obtain a capacity planning result.

[0061] From the above description, the improved C&CG algorithm is used to solve the integrated nuclear and storage robust capacity planning model, aiming to efficiently solve complex optimization problems and obtain reliable capacity planning results. In the pre-scheduling stage, the C&CG algorithm is used to solve the objective function, which can effectively handle the complex constraints and variables of large-scale problems and gradually generate the optimal solution. The pre-scheduling stage objective function is converted into a max-min problem, further clarifying the robustness of the optimization objective, i.e., the optimal performance of the system is still guaranteed under the most unfavorable circumstances. In the rescheduling stage, the auxiliary variable is used to transform the min problem into a max problem, simplifying the solving process and improving the calculation efficiency. Finally, the capacity planning results obtained based on the rescheduling stage objective function can balance the system's economy and reliability, while effectively dealing with the uncertainty of wind and solar power output, providing a scientific basis for the efficient operation and capacity configuration of the integrated nuclear and storage system.

[0062] Further, the pre-scheduling stage objective function is specifically:

[0063]

[0064] In the formula, I i,t represents the running state symbol of thermal power units, η represents the auxiliary variable, which aims to separate the main problem and the sub-problem, and Constraints under the base scenario represents the constraint conditions under the base scenario;

[0065] The transformed pre-scheduling stage objective function is specifically:

[0066]

[0067] In the formula, represents the actual output of wind power under uncertain scenarios, represents the actual output of photovoltaic under uncertain scenarios, represents the abandoned power of wind power, represents the abandoned power of photovoltaic, ΔD d,t represents the load shedding power, U wind represents the wind power output uncertainty set constraint, U pv represents the photovoltaic output uncertainty set constraint, U represents the wind and solar power output uncertainty set, and respectively represent 0-1 variables in the wind power output uncertainty set, and respectively represent 0-1 variables in the photovoltaic output uncertainty set;

[0068] The rescheduling stage objective function is specifically:

[0069]

[0070] In the formula, u represents the planning variable under the uncertain scene after the dual transformation, and θ, β and χ respectively represent the dual variables of the constraints, and A, B, C and D respectively represent the matrices corresponding to the constraints.

[0071] As can be known from the above description, the improved C&CG algorithm is used for flexible conversion of 0-1 variables, effectively solves the difficulty caused by the pumped storage operation mode in the two-stage NPSI robust capacity planning framework, ensures the solving accuracy, improves the solving speed, and has excellent solving efficiency.

[0072] Referring to Figure 2 , another embodiment of the present application provides a nuclear and storage integrated joint robust capacity planning system, which comprises a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor implements each step in the above nuclear and storage integrated joint robust capacity planning method when executing the computer program.

[0073] The above nuclear and storage integrated joint robust capacity planning method and system can be applied to the NPSI joint robust capacity planning scene, and the following will be described through a specific embodiment:

[0074] Referring to Figure 1 , Figure 3 and Figure 4 , the embodiment one of the present application is:

[0075] A nuclear and storage integrated joint robust capacity planning method comprises the following steps:

[0076] S1, obtaining a historical data set and preprocessing the historical data set to obtain a preprocessed historical data set, specifically comprising S11-S12:

[0077] S11, obtaining a historical data set. The historical data set comprises historical wind power, photovoltaic output and load data of each node of the system.

[0078] In an optional embodiment, in order to fully depict the time and space distribution characteristics of new energy and load, the wind power, photovoltaic output and load data of each node of the system in 8760 hours of the planning year are obtained to obtain the historical data set X.

[0079] The scene length N T is generally determined by the characteristics of wind, light and load, and 24 hours are selected as a single scene length, i.e., N T = 24, in order to highlight the intraday fluctuation. The historical data set X contains N s scenes. Taking the annual operation historical data set as an example, N s = 365.

[0080] S12, dimensionality reduction is performed on the historical data set using principal component analysis to obtain a preprocessed historical data set, and specifically includes:

[0081] (1) The historical data set is subjected to z-score standardization to obtain a standardized data set, specifically:

[0082]

[0083] In the formula, X st represents the standardized data set, σ f represents the variance of the feature F f . represents the variance of the feature .

[0084] (2) The covariance matrix of the standardized data set is calculated, specifically:

[0085]

[0086] In the formula, ∑ cov represents the covariance matrix, N sl represents the total number of features.

[0087] (3) The covariance matrix is subjected to eigenvalue decomposition to obtain a matrix composed of eigenvalues from large to small and corresponding eigenvectors :

[0088]

[0089] The purpose of dimensionality reduction is to compress the number of feature variables of the data set while trying to ensure the difference between different scenarios. When the first K feature variables are selected to represent the original data, the data variance ratio retained is:

[0090]

[0091] In the formula, Λ K represents the data variance ratio.

[0092] (4) According to the expected data variance ratio to be retained, the feature matrix used to represent the original data is uniquely determined r , and finally the dimensionality-reduced and compressed scenario set, i.e., the preprocessed historical data set X , is obtained as:

[0093]

[0094] Through the above preprocessing, the computational complexity and computation time are greatly reduced without affecting the clustering and screening effect.

[0095] S2, clustering the preprocessed historical data set using a K-center clustering algorithm based on density weighting to obtain a combined scene set, the combined scene set including a typical scene set and an extreme weather scene set, specifically including S21-S26:

[0096] S21, obtaining an initial scene set according to the preprocessed historical data set.

[0097] S22, performing Gaussian kernel density estimation on each scene in the initial scene set to obtain a density value of each scene, specifically:

[0098]

[0099] In the formula, f i represents the density value of the initial scene X i , X j , X i represents different initial scenes, δ represents the effective radius of the neighborhood, and N s represents the number of initial scenes.

[0100] S23, determining the density-weighted distance of each scene according to the density value of each scene, specifically:

[0101]

[0102] In the formula, d ij represents the density-weighted distance of the scene x i to the scene x j .

[0103] Unlike the existing method of using a truncated distance to perform density estimation, Gaussian kernel density estimation is used in the present application to assign density to discrete points. Compared with the truncated distance density estimation method, GKDE does not depend on a specific distribution assumption, and thus performs well when dealing with complex or unknown distribution data, and provides a smoother and more accurate probability density function estimation. At the same time, due to the characteristics of the Gaussian kernel function, such as infinite support area, smoothness and insensitivity to extreme values, the Gaussian kernel function still has excellent performance when performing density estimation on large-scale data and multi-density distribution data.

[0104] S24, determining the number of typical scenes using the contour coefficient method.

[0105] S25, clustering the initial scene set based on the number of typical scenes using a K-center clustering algorithm to obtain a typical scene set.

[0106] Wherein, when clustering, the cluster center of each cluster is selected as a typical scene, and the typical scene set is specifically:

[0107]

[0108] where X typ denotes the typical scenario set, denotes the load demand in the typical scenario, denotes the wind power output in the typical scenario, denotes the photovoltaic power output in the typical scenario, N n denotes the number of system nodes, denotes the number of typical scenarios.

[0109] S26, filtering an edge scenario set from the typical scenario set, and clustering the edge scenario set using a density-weighted K-center clustering algorithm to obtain an extreme weather scenario set.

[0110] Specifically, weights of scenarios in different clusters are obtained, and the edge scenario set is filtered from the typical scenario set through a preset threshold ε and the weights, such as scenarios with weights greater than the preset threshold ε are removed.

[0111] wherein the extreme weather scenario set is specifically:

[0112]

[0113] where X ext denotes the extreme weather scenario set, denotes the number of extreme weather scenarios, denotes the load demand in the extreme weather scenario, denotes the wind power output in the extreme weather scenario, denotes the photovoltaic power output in the extreme weather scenario.

[0114] The combined scenario set is specifically:

[0115]

[0116] where X c denotes the combined scenario set, and the number of combined scenarios

[0117] In an optional implementation, the method further comprises:

[0118] For each scenario, a weight is assigned according to a scenario occurrence frequency.

[0119] Specifically, the weight of a representative scenario (a typical scenario or an extreme weather scenario) is defined as the number of cluster center points contained in the representative scenario, that is, the weight of a typical scenario is the number of elements in the cluster of the typical scenario The weight of an extreme weather scenario is the number of elements in the cluster of the extreme weather scenario |Ω q |. The combined scenario weight vector W is represented as:

[0120]

[0121] wherein, denotes the weight of the typical scenario, denotes the weight of the extreme weather scenario. The sum of each element in the weight vector is the total number of scenarios in the planning period, specifically:

[0122]

[0123] By using the BDW-KM algorithm for clustering, the difference between different clusters is obvious, while the scenarios within the cluster have good similarity, which ensures the flexibility and adequacy of subsequent planning.

[0124] S3, based on the typical scenario set and the extreme weather scenario set, the wind and light output uncertainty is modeled, and the wind and light output uncertainty set constraint is obtained, specifically:

[0125]

[0126] wherein, U wind denotes the wind power output uncertainty set constraint, U pv denotes the photovoltaic output uncertainty set constraint, denotes the real output of wind power under uncertain scenarios, denotes the output of the wind farm, and denote 0-1 variables in the wind power output uncertainty set, and denote the upward and downward fluctuation range of wind power uncertainty output, denotes the time uncertainty budget of the wind power uncertainty set, denotes the spatial uncertainty budget of the wind power uncertainty set, z i,t denotes the wind power output coefficient under uncertain scenarios, denotes the predicted wind power output coefficient, denotes the average wind power output fluctuation coefficient under uncertain scenarios, denotes the smooth uncertainty budget of the wind power uncertainty set, w(i) denotes the wind farm node set, and t denotes the operation period, denotes the real output of photovoltaic under uncertain scenarios, denotes the output of the photovoltaic power station, and denote 0-1 variables in the photovoltaic output uncertainty set, and denote the upward and downward fluctuation range of photovoltaic uncertainty output, a time uncertainty budget representing the photovoltaic uncertainty set, a spatial uncertainty budget representing the photovoltaic uncertainty set, i,t a photovoltaic output coefficient under uncertainty, a photovoltaic output coefficient under prediction, a photovoltaic output fluctuation average coefficient under uncertainty, a smoothing uncertainty budget representing the photovoltaic uncertainty set, pv(i) represents a photovoltaic power station node set, and φ represents a wind power prediction error coefficient, a photovoltaic prediction error coefficient.

[0127] Here, the budget uncertainty set is used to model the wind and light uncertainty, which can accurately reflect the wind and light uncertainty.

[0128] S4, based on the wind and light output uncertainty set constraint, a nuclear and storage integrated joint robust capacity planning model is constructed, and the nuclear and storage integrated joint robust capacity planning model is solved based on an improved constraint and column generation algorithm to obtain a capacity planning result, specifically including S41-S46:

[0129] S41, based on the wind and light output uncertainty set constraint, a target function of minimizing annual investment cost and operation cost is established, and a nuclear and storage integrated joint operation power station pumped storage planning constraint and a comprehensive scene daily operation constraint corresponding to the target function are established.

[0130] The target function is specifically:

[0131]

[0132] In the formula, C total represents the total annual cost, C Inv represents the annual investment cost of the nuclear and storage integrated joint operation power station, C Ope represents the annual operation cost of the nuclear and storage integrated joint operation power station participating in system operation, c C represents the unit capacity configuration cost of the nuclear and storage integrated joint operation power station pumped storage, represents the planned reservoir capacity of the nuclear and storage integrated joint operation power station pumped storage, c P represents the unit unit configuration cost of the nuclear and storage integrated joint operation power station pumped storage, represents the rated power of the nuclear and storage integrated joint operation power station pumped storage, α represents the discount rate, T life represents the life of the nuclear and storage integrated joint operation power station, π r represents the probability of the combined scene r, represents the daily operation cost under the combination scenario r, i represents the node number of the planned NPSI power station, U represents the output set of renewable energy under uncertain scenarios, z represents the 0-1 variable of uncertain output of wind power, v represents the 0-1 variable of uncertain output of photovoltaic, Ξ represents the value set of the above 0-1 planning variables, p represents the continuous planning variable including the output of conventional units, line power and the like, and Ψ represents the value set of the above continuous planning variable. The objective function is a min-max-min problem.

[0133] The daily operation cost includes the operation cost of the nuclear and storage integrated combined operation power station, the operation cost of the conventional unit and the penalty cost of abandoned wind and light, and is specifically:

[0134]

[0135] In the formula, represents the daily operation cost, represents the output of the unit in the nuclear and storage integrated combined operation power station, represents the output of the conventional thermal power unit, I i,t represents the running state symbol of the thermal power unit, represents the abandoned power of wind power, represents the abandoned power of photovoltaic, ΔD d,t represents the cut-off power of load, C gen represents the operation cost of the conventional thermal power unit, C int represents the operation cost of the nuclear and storage integrated combined operation power station, C wind represents the operation cost of the wind power field, C pv represents the operation cost of the photovoltaic power station, C D represents the penalty cost of cut-off load.

[0136] The operation cost of the conventional thermal power unit is specifically:

[0137]

[0138] In the formula, N g represents the set of conventional thermal power units, represents the start-up and shutdown cost of the thermal power unit at the i node in the t period, represents the operation cost of the thermal power unit at the i node in the t period, su i , sd i represents the unit start-up and shutdown cost of the unit, ru i,t , rd i,t represents the start-up and shutdown state of the unit in the t period, a i , b i and c i respectively represent the fuel cost coefficient of the thermal power unit, P i,tPn(t) represents the output of the thermal power unit in the basic scenario.

[0139] The operation cost of the nuclear and pumped storage integrated combined operation power station is specifically:

[0140] C int = C nu + C ps ;

[0141] In the formula, C nu represents the peak shaving cost of the nuclear power unit, C ps represents the peak shaving cost of the pumped storage unit.

[0142] The peak shaving cost of the nuclear power unit is included in the operation cost, which can be represented as:

[0143]

[0144] In the formula, T represents the total period, N nu represents the total number of nuclear power units, represents the operation cost of the nuclear power unit at the i node in the t period, represents the peak shaving cost of the nuclear power unit at the i node in the t period, represents the power generation cost coefficient of the nuclear power unit i, represents the output of the nuclear power unit in the basic scenario, C N represents the peak shaving cost coefficient of the nuclear power unit, represents the maximum output of the nuclear power unit.

[0145] The peak shaving cost coefficient of the nuclear power unit is specifically:

[0146] C N = C F + σ N C S ;

[0147] In the formula, C F represents the peak shaving fuel cost of the nuclear power unit, σ N represents the nuclear safety value coefficient, C S represents the peak shaving safety cost of the nuclear power unit.

[0148] The pumped storage unit participates in power system dispatching through mutual conversion between the two working states of power generation and pumping, and its operation cost mainly includes unit start-up cost, and shutdown cost can be ignored. Therefore, the peak shaving cost of the pumped storage unit is specifically:

[0149]

[0150] In the formula, N c represents the total number of pumped storage units, is a 0-1 state variable, which is equal to 1 when the unit i switches from the shutdown state to the power generation state at time period t, and is equal to 0 otherwise, is the cost of starting the generator of the pumped storage unit i once, is a 0-1 state variable, which is equal to 1 when the unit i switches from the shutdown state to the pumping state at time period t, and is equal to 0 otherwise, is the cost of starting the motor of the pumped storage unit i once.

[0151] The operation cost of the wind farm is specifically:

[0152]

[0153] wherein N w represents a set of wind farms, k w represents a wind curtailment penalty coefficient.

[0154] The operation cost of the photovoltaic power station is specifically:

[0155]

[0156] wherein N pv represents a set of photovoltaic power stations, k pv represents a light curtailment penalty coefficient.

[0157] The load shedding penalty cost is specifically:

[0158]

[0159] wherein N d represents a set of load nodes, k D represents a load shedding penalty coefficient.

[0160] The pumped storage integrated joint operation power station pumped storage planning constraints include pumped storage integrated joint operation power station investment constraints, pumped storage integrated joint operation power station pumped storage reservoir planning constraints, pumped storage integrated joint operation power station pumped storage unit planning constraints, and pumped storage integrated joint operation power station pumped storage continuous full-load hours constraints, and the like, and are specifically:

[0161]

[0162] wherein Π max represents an upper limit of the pumped storage integrated joint operation power station pumped storage part investment, x i represents a 0-1 variable of the pumped storage integrated joint operation power station equipment construction, represents a lower limit of the pumped storage integrated joint operation power station pumped storage part reservoir capacity, represents an upper limit of the pumped storage integrated joint operation power station pumped storage part reservoir capacity, represents a lower limit of the pumped storage integrated joint operation power station pumped storage unit power, represents the upper limit of the power of the pumped storage unit of the nuclear-pumped storage integrated combined operation power station, represents the minimum value of the continuous full-load hours of the pumped storage power station, represents the maximum value of the continuous full-load hours of the pumped storage power station.

[0163] The intra-day operation constraints of the comprehensive scenario include operation constraints in a basic scenario and operation constraints in an uncertain scenario. The operation constraints in the basic scenario include node power balance constraints, line power flow constraints, wind power plant output constraints, photovoltaic power station output constraints, conventional thermal power unit operation constraints, output constraints and reserve constraints, minimum start-up / shut-down time constraints and ramping constraints, and operation constraints of the nuclear-pumped storage integrated combined operation power station. The operation constraints in the uncertain scenario include node power balance constraints, line power flow constraints, rescheduling constraints of conventional units, flexible scheduling constraints of the nuclear-pumped storage integrated combined operation power station, uncertain wind power output constraints, and uncertain photovoltaic power output constraints.

[0164] The node power balance constraints are specifically as follows:

[0165]

[0166] In the formula, i represents a node where different units are located, d represents a node where different loads are located, and l represents different transmission lines. N int , and N l respectively represent a set of nuclear-pumped storage integrated units and a set of transmission lines in the power system. PL l,t represents transmission power of the transmission line in the power system. D d,t represents a power prediction value of a user load.

[0167] The line power flow constraints are specifically as follows:

[0168]

[0169] In the formula, θ m,t represents a phase angle of a first end node of the transmission line l at time t, θ n,t represents a phase angle of a last end node of the transmission line l at time t, x l represents an electric reactance value of the transmission line l, θ ref represents a reference node phase angle value, represents the maximum power allowed to flow through the transmission line l.

[0170] The wind power plant output constraints are specifically as follows:

[0171]

[0172] In the formula, P P (t) represents the predicted value of the output power of the wind turbine at time t.

[0173] The output constraint of the photovoltaic power station is specifically:

[0174]

[0175] In the formula, P (t) represents the predicted value of the output power of the wind turbine at time t.

[0176] The operation constraint, output constraint and standby constraint of the conventional thermal power unit are specifically:

[0177]

[0178]

[0179] In the formula, Pmin (t) represents the lower limit of the output power of the unit i, Pmax (t) represents the upper limit of the output power of the unit i, Pmax (t) represents the maximum value of the upper and lower adjustable standby amount of the unit i.

[0180] The minimum start / stop time constraint and the ramp constraint are specifically:

[0181]

[0182] In the formula, Pmin (t) represents the start-up limit of the unit, Pmax (t) represents the shutdown limit of the unit, UR (t) represents the start / stop state of the unit at time t, n DR (t) represents the upper ramp limit, n DR (t) represents the lower ramp limit.

[0183] The operation constraint of the nuclear and pumped storage integrated combined operation power station includes the constraint of the nuclear power and pumped storage part, decouples the combined output of the NPSI power station, and is specifically:

[0184]

[0185] In the formula, P (t) represents the output of the pumped storage unit in the basic scenario.

[0186] To ensure the safety of the nuclear power unit participating in the peak shaving operation, the range of 50% to 100% of the rated power is defined as the safety peak shaving depth range of the nuclear power, and the nuclear power peak shaving related constraints are set in this range to meet the peak shaving flexibility requirements, so as to ensure the safety of the nuclear power unit participating in the peak shaving operation, including setting multiple peak shaving depths in the rated power range and allowing the nuclear power unit to realize the transition between low power and full power at three speed gears in different peak shaving depths, improving the peak shaving capacity of the nuclear power to meet the peak shaving flexibility requirements of the system in different scenarios, wherein the nuclear power unit output constraint is specifically:

[0187]

[0188] In the formula, Pn represents the peak shaving power of the nuclear power unit at the nth peak shaving depth, Pmin represents the minimum output of the nuclear power unit, n represents the peak shaving depth identifier, represents the operation at the nth peak shaving depth, s nu Pn represents the peak shaving depth of the nuclear power unit, Pn represents the low power stage power of the nuclear power unit at the nth peak shaving depth, Pn represents the power of the transition state between high and low power of the nuclear power unit, j represents the peak shaving speed identifier, and takes values 1-3 to represent different peak shaving speeds, q t l represents the full power operation identifier at t, n,t d represents the low power operation identifier at t, n,j,t q represents the transition operation identifier at t.

[0189] The power state constraint of the nuclear power unit is specifically:

[0190]

[0191] The minimum rated / low power operation time constraint is specifically:

[0192]

[0193] In the formula, and both represent the operation identifier, and the constraint on the identifier is to ensure the minimum rated / low power operation time of the nuclear power by cyclically constraining t N from t to q t-1 l represents the full power operation identifier at t-1, n,t-1 d represents the low power operation identifier at t-1, Tl represents the minimum duration of the full power operation state, Td represents the minimum duration of the low power operation state.

[0194] The operation identifier coupling constraint is specifically:

[0195]

[0196] where q t+1 denotes the full power operation indicator at time t+1, l n,t+1 denotes the low power operation indicator at time t+1.

[0197] The pumped storage operation mode is constrained by the operation cycle, the upper and lower reservoir capacity of the pumped storage, and the daily operation reservoir capacity constraint based on the daily operation scenario established in the charging / discharging right. The upper reservoir capacity is limited, and the lower reservoir capacity is relatively large, so the upper reservoir capacity constraint is strict. The pumped storage output constraint is specifically:

[0198]

[0199] where denotes the power generation at time t, denotes the pumping power at time t.

[0200] The reservoir capacity and power balance constraint is specifically:

[0201]

[0202] where denotes the upper reservoir water storage at time t, denotes the upper reservoir water storage at time t-1, Δt denotes denotes the state variable of pumping at time t, taking value 0 or 1, η p denotes the pumping efficiency, denotes the state variable of power generation at time t, taking value 0 or 1, η t denotes the power generation efficiency.

[0203] The upper reservoir capacity upper and lower limit constraint is specifically:

[0204]

[0205] where denotes the upper limit value of the upper reservoir capacity, denotes the upper limit value of the upper reservoir capacity.

[0206] The daily operation reservoir capacity balance constraint is specifically:

[0207]

[0208] where ΔU max denotes the maximum daily change of the upper reservoir capacity, denotes the upper reservoir water storage at time T, denotes the upper reservoir water storage at the initial time.

[0209] The unit start-stop variable constraint is specifically:

[0210]

[0211] The upper and lower limit of the power generation constraint is specifically:

[0212]

[0213] In the formula, represents the lower limit of the power generation value, represents the upper limit of the power generation value.

[0214] The pumping power constraint is specifically:

[0215]

[0216] In the formula, represents the constant pumping power.

[0217] S42, according to the objective function, the nuclear storage integrated joint operation power station pumping planning constraint and the comprehensive scene daily operation constraint, a nuclear storage integrated joint robust capacity planning model is obtained.

[0218] S43, using the constraint and column generation algorithm to solve the objective function in the pre-scheduling stage, and obtaining a pre-scheduling stage objective function.

[0219] The pre-scheduling stage objective function is specifically:

[0220]

[0221] In the formula, η represents an auxiliary variable, which is used to separate the main problem and the sub-problem, and Constraints under the base scenario represents the constraint condition under the base scenario. The objective of the pre-scheduling stage is to minimize the expected total cost constraint, including the start-stop constraint of the conventional unit, the output constraint and the C&CG constraint returned in the sub-problem.

[0222] S44, the pre-scheduling stage objective function is transformed to obtain a transformed pre-scheduling stage objective function, and the transformed pre-scheduling stage objective function is a max-min problem.

[0223] The transformed pre-scheduling stage objective function is specifically:

[0224]

[0225] In the formula, U represents the wind and light output uncertainty set.

[0226] S45, performing dual transformation on the min problem in the converted pre-scheduling stage objective function based on the auxiliary variable in the re-scheduling stage to obtain a re-scheduling stage objective function, the re-scheduling stage objective function being a max problem.

[0227] The re-scheduling stage objective function is specifically:

[0228] max u∈U，θ,β,χ [θ(A-Bσ-Cχ)+Dβ];

[0229] In the formula, u represents a planning variable in an uncertain scenario after dual transformation, θ, β and χ respectively represent dual variables of constraints, and A, B, C and D respectively represent matrices corresponding to the constraints.

[0230] In an optional implementation, the method further includes:

[0231] generating a C&CG cut set constraint, specifically:

[0232]

[0233] In the formula, all C&CG cut plane obtained so far represents a cut plane constraint generated in the solving process by performing dual transformation on a sub-problem when using a C&CG algorithm, and k represents an iteration number.

[0234] S46, performing solving based on the re-scheduling stage objective function to obtain a capacity planning result.

[0235] Specifically, because the original problem has many variables and complex constraints, a detailed solving algorithm in a matrix form is given, and the matrix form of the original problem is as shown in the following formula:

[0236]

[0237] In the formula, y represents a continuous variable vector in the first stage (pre-scheduling stage), including energy storage planning capacity, p u represents an uncertain variable vector, v represents a binary variable vector in the second stage (re-scheduling stage), including 0-1 variables of the charging and discharging state of the nuclear and pumped storage integrated joint operation power station, p represents a continuous variable vector in the second stage problem, including conventional unit output, NPSI nuclear power output, line power flow, new energy output, NPSI pumped storage charging and discharging power and the like, x represents a 0-1 variable in the first stage, which is a construction decision variable of the pumped storage part of the nuclear and pumped storage integrated joint operation power station and a start-stop state variable of the conventional unit, and b, c, d, f, A, B, C, D, E, F and G respectively represent respective corresponding coefficient matrices.

[0238] Since the inner problem of the original problem contains 0-1 variables, it is a mixed integer linear programming problem and cannot be directly transformed by duality. In order to solve this problem, an auxiliary variable is introduced, and the specific improvement method is as follows: first, 0-1 auxiliary variables are introduced in the original problem The 0-1 variable v in the original problem is defined as a continuous variable, and the constraint v is equal to

[0239] The above formula is transformed into the following optimization problem:

[0240]

[0241] In the above formula, μ and λ represent the dual variables of the constraints, respectively. In the above formula, the outer layer is a max problem under an uncertain scenario, the middle layer is an auxiliary variable As the min problem of the optimization variable, the middle layer constraint is shown in (b) in the above formula; the inner layer is converted into a linear min problem, and the constraint is shown in (a) in the above formula.

[0242] The inner problem min of the above formula is transformed by duality to obtain:

[0243]

[0244] According to the minimax inequality, the following relationship is obtained:

[0245]

[0246] The optimal value of in the above formula is determined by enumerating the combination of 0 and 1 variables, as shown in the following formula:

[0247]

[0248] In the formula, Ω = {δ≤λ, δ≤0}.

[0249] By substituting the following formula into the above formula, the min problem of the inner layer can be eliminated. Then, the outer max problem and the inner max problem are combined, and finally, the objective function of the subproblem is converted to the following formula:

[0250]

[0251] Which contains a bilinear term, which is linearized by the big M method. Through the above processing, the subproblem is converted into a form that can be solved, and the solution of the 0-1 variable subproblem considering energy storage operation is realized. Finally, by solving the subproblem, the worst scenario p of new energy output in the uncertain scenario is obtained worst By adding new variables, C&CG cut set constraints are generated, as shown in the following formula, and are added to the main problem for iterative solution.

[0252]

[0253] From the perspective of energy consumption, the frequent occurrence of continuous high temperatures and cold waves in recent years will stimulate a continuous increase in cooling and heating loads. From the perspective of energy production, extreme weather will affect the normal operation of various power generation facilities and reduce the power supply capacity. From a conventional perspective, there are significant differences in user electricity consumption characteristics and renewable energy power supply characteristics on typical days in different seasons. Therefore, considering both typical scenarios and extreme weather, this invention uses the BDW-KM clustering method and statistical analysis method to perform a series of operations such as clustering, reduction, and weighting on historical scenarios. By comprehensively considering the characteristics of electricity load and renewable energy output, a new type of power system NPSI comprehensive scenario set is constructed. The clustering results are shown in Table 1.

[0254] Table 1 Extracts NPSI Comprehensive Application Scenarios Analysis

[0255]

[0256] Numerical simulations were conducted using the NPSI joint robust capacity planning model constructed based on this invention in a comprehensive scenario driven by historical data. The method described above was tested using an IEEE-24 standard node system to establish the NPSI joint robust capacity planning model. The simulations included four types of power source layouts (thermal power, wind power, photovoltaic) and two types of energy device layouts (nuclear power, pumped storage). The source-load configuration and network structure of the standard IEEE-24 node system used are as follows: Figure 3 As shown in Table 2, four different types of simulation cases are configured simultaneously. The power system collaborative planning simulation operation is then carried out according to the different simulation case configuration requirements.

[0257] Table 2 Example Settings

[0258]

[0259] To comprehensively evaluate the effectiveness of NPSI joint capacity planning under different conditions, four sets of simulation examples were set up, with the following characteristics and objectives: Simulation example 1 serves as a standard simulation example, considering extreme weather and performing NPSI joint capacity planning to measure the reliability and flexibility of the system under adverse weather conditions; Simulation example 2 performs NPSI joint capacity planning under typical scenarios without considering extreme weather, aiming to evaluate the system performance under normal weather conditions; Simulation example 3 considers extreme weather but does not perform NPSI joint capacity planning, to evaluate the performance of fixed-capacity NPSI under adverse weather conditions; Simulation example 4 considers extreme weather and performs NPSI joint capacity planning under a high-proportion renewable energy system, to evaluate the system reliability and adaptability under high-penetration renewable energy access conditions.

[0260] Table 3. Programming Results of the Case Study

[0261]

[0262]

[0263] A comprehensive evaluation of the NPSI pumped storage planning was conducted through four sets of calculation examples, yielding key indicators such as the NPSI pumped storage installed capacity, annualized planning cost, and annual system operating cost for each example, as shown in Table 3. Example 1, as a standard example, demonstrates the comprehensive performance considering extreme weather and pumped storage planning. Example 2, without considering extreme weather, shows a significant reduction in the NPSI pumped storage capacity to 48.54 MW, with corresponding reductions in the annualized planning cost and annual system operating cost. This indicates that excluding extreme weather has a significant impact on costs, and that planning the system based solely on typical scenarios differs considerably from actual conditions. Planning methods that consider extreme weather can further reflect the rationality of the planning results under real-world conditions. Example 3, without NPSI pumped storage planning, has a fixed pumped storage capacity of 200MW. While the planned annualized cost is high, the system's annual operating cost is similar to Example 1. This indicates that although a fixed-capacity NPSI can maintain certain operational performance under extreme weather conditions, the higher planned annualized cost results in significantly lower overall benefits compared to Example 1, which incorporates NPSI joint robust capacity planning. This reflects how NPSI joint robust capacity planning achieves rational utilization of power generation resources and effectively avoids unnecessary investment. Example 4, under a high-proportion renewable energy system, significantly increases the planned NPSI pumped storage capacity to 341.56MW. While the planned annualized cost slightly increases, the system's annual operating cost significantly decreases. This demonstrates that increasing the NPSI capacity can substantially reduce system operating costs, reflecting the high requirements for system reliability and adaptability imposed by high-penetration renewable energy integration. Through comparative analysis, it can be seen that factors such as extreme weather, NPSI pumped storage planning, and the scale of renewable energy installations have a significant impact on NPSI planning, providing important reference for practical engineering decisions.

[0264] At the same time, the operating costs of four different sets of calculation examples were evaluated, such as Figure 4 As shown, Figure 4 The annual operating costs of different calculation examples were compared. Examples 1 and 2 show that, without considering extreme weather conditions, the system may rely more on high-carbon-emission energy sources, leading to increased costs. Example 3 demonstrates that a fixed-capacity NPSI performs better in integrating new energy sources while maintaining stable costs for thermal power and carbon emissions. Example 4 reflects that a high proportion of new energy integration significantly reduces the costs of thermal power and carbon emissions, but the cost of new energy curtailment increases, possibly due to absorption issues caused by the volatility of new energy sources. Overall, the changes in different cost items in each example reasonably reflect the system operating characteristics under their respective settings, providing a multi-dimensional reference for actual power system planning.

[0265] Please refer toFigure 2 Embodiment two of the present invention is as follows:

[0266] A nuclear-energy integrated robust capacity planning 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 nuclear-energy integrated robust capacity planning method in Embodiment 1.

[0267] In summary, this invention provides a joint robust capacity planning method and system for nuclear power and energy storage (NPSI). It uses a density-weighted K-medoids algorithm to cluster preprocessed historical datasets, obtaining a combined scenario set, including a typical scenario set and an extreme weather scenario set. Based on these sets, the uncertainty of wind and solar power output is modeled, resulting in a set of uncertainties. A joint robust capacity planning model for NPSI is constructed based on these constraints, and the improved C&CG algorithm is used to solve the model, yielding the capacity planning results. This method utilizes the density-weighted K-medoids algorithm to achieve accurate scenario clustering, considers the impact of extreme weather, and constructs a joint robust capacity planning model that takes into account the stochastic uncertainty and load overload of new energy sources, enabling flexible response to system adjustment needs. An improved C&CG algorithm is used to solve the model, resulting in higher efficiency and more accurate solutions, thus improving the accuracy of the planning. Furthermore, a density-weighted K-medoids algorithm is used to select a set of typical scenarios, ensuring that the cluster centers are actual scenarios rather than the average of all scenarios. This significantly reduces the interference of outliers in cluster center selection. The silhouette coefficient method is used to determine the number of typical scenarios, with the core requirement being that the differences between scenarios within the same cluster are sufficiently small, and the differences between scenarios in different clusters are sufficiently large, ensuring the accuracy of the typical scenario set selection. Then, a density-weighted K-medoids algorithm is used to select a representative set of extreme weather scenarios, considering not only their differences from typical scenarios but also their differences from existing extreme weather data. This ensures that the newly selected extreme weather scenarios can provide the model with the maximum incremental information, effectively improving the reliability and accuracy of scenario selection.

[0268] 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's 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 nuclear energy storage integrated combined robust capacity planning method, characterized in that, The method comprises the steps of: obtaining a historical data set and preprocessing the historical data set to obtain a preprocessed historical data set; performing clustering on the preprocessed historical data set using a K-center clustering algorithm based on density weighting to obtain a combined scenario set, the combined scenario set comprising a typical scenario set and an extreme weather scenario set; modeling wind and light output uncertainty based on the typical scenario set and the extreme weather scenario set to obtain wind and light output uncertainty set constraints; constructing a nuclear and storage integrated joint robust capacity planning model based on the wind and light output uncertainty set constraints, and solving the nuclear and storage integrated joint robust capacity planning model based on an improved constraint and column generation algorithm to obtain a capacity planning result.

2. The nuclear energy storage integrated combined robust capacity planning method of claim 1, wherein, The preprocessing of the historical data set to obtain a preprocessed historical data set comprises: performing dimensionality reduction on the historical data set using principal component analysis to obtain a preprocessed historical data set.

3. The nuclear energy and storage integrated robust capacity planning method of claim 1, wherein, The clustering of the preprocessed historical data set using a K-center clustering algorithm based on density weighting to obtain a combined scenario set comprises: obtaining an initial scenario set from the preprocessed historical data set; performing Gaussian kernel density estimation on each scenario in the initial scenario set to obtain a density value of each scenario; determining a density-weighted distance of each scenario based on the density value of each scenario; determining the number of typical scenarios using a silhouette coefficient method; performing clustering on the initial scenario set using a K-center clustering algorithm based on the number of typical scenarios and the density-weighted distance to obtain a typical scenario set; screening an edge scenario set from the typical scenario set and performing clustering on the edge scenario set using a K-center clustering algorithm based on density weighting to obtain an extreme weather scenario set.

4. The nuclear energy and storage integrated combined robust capacity planning method of claim 3, wherein, The Gaussian kernel density estimation on each scenario in the initial scenario set to obtain a density value of each scenario comprises: where f i represents the density value of the initial scene X i , X j , X i represents different initial scenes, δ represents the effective radius of the neighborhood, and N s represents the number of initial scenes.

5. The nuclear energy and storage integrated combined robust capacity planning method of claim 4, wherein, The determination of the density-weighted distance of each scenario based on the density value of each scenario comprises: where d ij represents the density-weighted distance of scene x i to scene x j .

6. The nuclear energy and storage integrated combined robust capacity planning method of claim 1, wherein, The construction of a nuclear and storage integrated joint robust capacity planning model based on the wind and light output uncertainty set constraints comprises: establishing a target function for minimizing annual investment and operation costs, and nuclear and storage integrated joint operation power station pumping and storage planning constraints and comprehensive scenario intra-day operation constraints corresponding to the target function based on the wind and light output uncertainty set constraints; obtaining a nuclear and storage integrated joint robust capacity planning model based on the target function, the nuclear and storage integrated joint operation power station pumping and storage planning constraints, and the comprehensive scenario intra-day operation constraints.

7. The nuclear energy and storage integrated combined robust capacity planning method of claim 6, wherein, The target function is specifically: minC total = C Inv + C Ope ; where C total denotes the total annual cost, C Inv denotes the annual investment cost of the nuclear-pumped storage integrated joint operation power station, C Ope denotes the annual operation cost of the nuclear-pumped storage integrated joint operation power station participating in system operation, c C denotes the unit reservoir configuration cost of pumped storage of the nuclear-pumped storage integrated joint operation power station, denotes the planned reservoir capacity of pumped storage of the nuclear-pumped storage integrated joint operation power station, c P denotes the unit unit configuration cost of pumped storage of the nuclear-pumped storage integrated joint operation power station, denotes the rated power of pumped storage of the nuclear-pumped storage integrated joint operation power station, a denotes the discount rate, T life denotes the life of the nuclear-pumped storage integrated joint operation power station, p r denotes the probability of the combined scenario r, denotes the daily operation cost under the combined scenario r, i denotes the node number of the planned nuclear-pumped storage integrated joint operation power station, U denotes the output set of renewable energy under uncertainty, z denotes a 0-1 variable of uncertain wind power output, v denotes a 0-1 variable of uncertain photovoltaic output, Ξ denotes a value set of the above 0-1 planning variables, p denotes continuous planning variables including conventional unit output, line power, etc., and Ψ denotes a value set of the above continuous planning variables.

8. The nuclear energy and storage integrated combined robust capacity planning method of claim 7, wherein, The solving of the nuclear and storage integrated joint robust capacity planning model based on the improved constraint and column generation algorithm to obtain a capacity planning result comprises: solving the target function using a constraint and column generation algorithm in a pre-scheduling stage to obtain a pre-scheduling stage target function; transforming the pre-scheduling stage target function to obtain a transformed pre-scheduling stage target function, the transformed pre-scheduling stage target function being a max-min problem; In the rescheduling stage, a min problem in the converted prescheduling stage objective function is dually transformed based on an auxiliary variable to obtain a rescheduling stage objective function, which is a max problem; The rescheduling stage objective function is solved to obtain a capacity planning result.

9. The nuclear energy and storage integrated combined robust capacity planning method of claim 8, wherein, The prescheduling stage objective function is specifically: In the formula, I i,t denotes the operation state symbol of the thermal power unit, η denotes an auxiliary variable, which is used to separate the main problem and the sub-problem, and Constraints under the base scenario denotes the constraint condition under the base scenario. The converted prescheduling stage objective function is specifically: wherein represents the real output of wind power in the uncertain scenario, represents the real output of photovoltaic in the uncertain scenario, represents the curtailment power of wind power, represents the curtailment power of photovoltaic, ΔD d,t represents the outage power of load, U wind represents the wind power output uncertainty set constraint, U pv represents the photovoltaic output uncertainty set constraint, U represents the wind and photovoltaic output uncertainty set, and respectively represent 0-1 variables in the wind power output uncertainty set, and respectively represent 0-1 variables in the photovoltaic output uncertainty set; The rescheduling stage objective function is specifically: max u∈U，θ，β，χ [θ(A-Bσ-Cχ)+Dβ] In the formula, u represents a planning variable in an uncertain scenario after dual transformation, θ, β and χ respectively represent dual variables of constraints, and A, B, C and D respectively represent matrices corresponding to the constraints.

10. A nuclear and storage integrated combined robust capacity planning system comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein, The processor implements each step in the nuclear and storage integrated joint robust capacity planning method in any one of claims 1 to 9 when executing the computer program.

Images