A method, system and device for configuring reserve resources of a power system

Abstract

This invention discloses a method, system, and equipment for configuring reserve resources in a power system. The method includes: performing pre-scheduling based on a preset scenario dataset and acquiring power flow data for each scenario of the power system after pre-scheduling; dynamically partitioning the power system according to the power flow data to obtain target partitioning results; inputting the target partitioning results and the scenario dataset into a preset regional reserve resource configuration model to solve for the total upper and lower reserve demand of each sub-region with the objective of minimizing the expected total cost of the system; acquiring the day-ahead forecast data of the power system and inputting the day-ahead forecast data and the total upper and lower reserve demand into a preset day-ahead reserve resource configuration model to determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day for reserve resource configuration. Therefore, by implementing this invention, the accuracy of reserve resource configuration can be improved.

Description

Technical Field

[0001] This invention relates to the field of power systems, and more particularly to a method, system, and equipment for configuring backup resources in a power system. Background Technology

[0002] With a high proportion of renewable energy sources such as wind and solar power being integrated into the power system, their output exhibits significant intermittency and volatility. Coupled with frequent extreme weather events, this introduces multiple uncertainties to the power system's power generation, grid, and load sides. To balance power supply and demand in real time and ensure the safe and stable operation of the system, it is essential to pre-configure sufficient backup resources (such as rapidly adjustable thermal power, hydropower, pumped storage, and energy storage). The scientific and rational allocation of backup resources is a key measure to address these uncertainties, prevent large-scale power outages, and enhance grid resilience. It is also the core foundation for the economical, safe, and low-carbon operation of the new power system.

[0003] Existing technologies for allocating reserve resources often employ static rules based on a uniform proportion across the entire system (such as allocating according to a fixed proportion based on maximum load) or optimization methods focused on a single resource (such as thermal power). These methods have significant shortcomings in accuracy: First, in the spatial dimension, they fail to fully consider the actual transmission capacity of the power grid and the impact of network congestion, potentially leading to severe shortages of reserve resources in locally congested areas and over-allocation in other areas, failing to achieve precise spatial layout. Second, in the resource dimension, they lack the coordinated optimization and complementary use of various flexible resources such as hydropower, pumped storage, and energy storage, making it difficult to fully leverage their rapid response and low-carbon advantages. Third, in the decision-making dimension, they often sever the connection between medium- and long-term strategic reserves and short-term specific scheduling, failing to construct a multi-timescale coordinated decision-making framework, resulting in a disconnect between reserve allocation plans and actual operational needs, and requiring improvement in the overall economy and security of the allocation. Summary of the Invention

[0004] This invention provides a method, system, and equipment for configuring backup resources in a power system, which can improve the accuracy of backup resource configuration.

[0005] The first aspect of this invention provides a method for configuring standby resources in a power system, comprising: Pre-scheduling is performed based on a pre-set scenario dataset, and power flow data for each scenario of the power system is obtained after pre-scheduling. Based on the power flow data, the power system is dynamically partitioned using an improved semi-supervised spectral clustering algorithm to obtain the target partitioning results; The target partitioning result and the scenario dataset are input into a preset regional reserve resource configuration model. With the goal of minimizing the total expected cost of the system, the total upper and lower reserve requirements of each sub-region in the target partitioning result in the future short-to-medium-term preset period are obtained. Obtain the day-ahead forecast data of the power system, input the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource allocation model, optimize under the condition of meeting the minimum reserve constraints of each sub-region, determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, and allocate reserve resources for each generator unit in each sub-region according to the base load output point and the upper and lower reserved reserve capacity.

[0006] The power system reserve resource allocation method provided in this application can significantly improve the accuracy of reserve resource allocation. Specifically, it is reflected in the following aspects: First, the method of pre-scheduling and obtaining power flow data based on a preset scenario dataset provides an accurate data foundation reflecting multiple uncertainties for subsequent analysis, making the analysis of system congestion modes and operating states closer to actual possible scenarios and improving the reliability of the analysis from the source. Second, the method of dynamically partitioning the power network structure by improving the semi-supervised spectral clustering algorithm based on power flow data, and placing easily congested lines at the partition boundaries and ensuring that each sub-region contains a reserve power source, achieves a scientific division of the power network structure, optimizes the spatial layout of reserve resources, and improves the independence and effectiveness of reserve regulation within the partition. Third, the method of inputting the partitioning results and scenario dataset into a regional reserve resource allocation model to solve for the total reserve demand of each sub-region, through random optimization... The method comprehensively considers the probability weights of multiple uncertain scenarios in the short and medium term, and accurately determines the upper and lower limits of the required reserve capacity for each zone with the goal of minimizing the expected total cost. This avoids reserve shortages or redundancies caused by single-scenario predictions, and improves the accuracy of configuration from the time dimension and the overall planning level. Finally, based on the day-ahead forecast data and the total reserve demand, the method determines the specific base load output and reserve capacity of each unit through the day-ahead reserve resource configuration model. By using more accurate short-term forecast data, and under the condition of meeting the total zone constraints and minimum reserve constraints, the method finely allocates reserve resources to each unit, realizing the accurate implementation from the regional total to the specific unit strategy. Thus, it forms a complete technical chain from macro-situation analysis, scientific zoning, total optimization to micro-precise configuration, and systematically improves the accuracy and economy of reserve resource configuration in the power system when dealing with multiple uncertainties of source and load.

[0007] Further, the step of inputting the target partitioning result and the scenario dataset into a preset regional reserve resource allocation model, with the objective of minimizing the expected total system cost, to solve for the total upper reserve demand and lower reserve demand of each sub-region in the target partitioning result within a preset short-to-medium term, includes: Construct a stochastic optimization model for regional reserve resource allocation based on scenario-based joint pre-scheduling; Based on the target partitioning results and the scenario dataset, with the objective function being the minimum expected sum of scenario probability-weighted fixed costs, scenario operating costs, scenario reserve operating costs, and scenario risk costs, and under the operational constraints of thermal power units, hydropower units, pumped storage resources, and energy storage resources, the stochastic optimization model for the allocation of reserve resources in the region is solved to obtain the total upper reserve demand and lower reserve demand of each sub-region at each moment within the preset short-to-medium-term period.

[0008] This method, which inputs the partitioning results and scenario datasets into the regional reserve resource configuration model to solve for the total reserve demand of each sub-region, comprehensively considers the probability weights of multiple uncertain scenarios in the short and medium term through a stochastic optimization method. With the goal of minimizing the expected total cost, it accurately determines the upper and lower limits of the reserve capacity required for each partition, avoiding insufficient or redundant reserves caused by single scenario prediction. It improves the accuracy of configuration from the perspectives of time dimension and total planning.

[0009] Furthermore, the method includes: the fixed costs include the start-up and shutdown costs of thermal power units and pumped storage units, as well as the configuration costs of standby capacity for thermal power, pumped storage, hydropower and energy storage; The operating costs of the scenario include the fuel costs of thermal power units, the operating costs of pumped storage units, and carbon emission costs. The standby operating costs in the scenario include the actual standby call costs for thermal power, pumped storage, hydropower, and energy storage. The risk costs in this scenario include the cost of load shedding and the cost of wind and solar power curtailment.

[0010] Further, the step of inputting the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource allocation model, and optimizing it under the condition of satisfying the minimum reserve constraint of each sub-region, to determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, includes: Based on the day-ahead forecast data and the total reserve demand of each sub-region, a day-ahead reserve configuration model is constructed. With the objective function of minimizing the total day-ahead operating cost of the power system, the day-ahead reserve configuration model is solved under the conditions of satisfying the operating constraints of generator units, the minimum reserve constraints of each sub-region, and the zone reserve capacity constraints. This yields the base load output point, upper reserved reserve capacity, and lower reserved reserve capacity of each generator unit in the next day. The zone reserve capacity constraints are composed of the total upper reserve demand and the total lower reserve demand.

[0011] This method, which uses day-ahead forecast data and total reserve demand to determine the specific base load output and reserve capacity of each unit through a day-ahead reserve resource allocation model, utilizes more accurate short-term forecast data to finely allocate reserve resources to each unit while meeting the total regional constraints and minimum reserve constraints. This achieves accurate implementation from regional totals to specific unit strategies.

[0012] Furthermore, the method includes: The day-ahead total operating cost includes day-ahead operating cost, day-ahead reserve configuration cost, and day-ahead risk cost; wherein, the day-ahead operating cost includes at least the fuel cost of thermal power units, the day-ahead reserve configuration cost is the cost incurred from reserving reserve capacity, and the day-ahead risk cost includes loss of load cost and wind / solar curtailment cost.

[0013] Furthermore, the method includes: The operating constraints of the generator set are as follows: for any generator set that is in operation at any time, the sum of the base load output and the reserved upper reserve capacity shall not exceed the maximum technical output, and the difference between the base load output and the reserved lower reserve capacity shall not be less than the minimum technical output. The minimum reserve constraint for each sub-region is: the total upper reserve and the total lower reserve configured in the region must be no less than the preset ratio of wind and solar power output and load in the corresponding region, respectively. The zoned reserve capacity constraints are as follows: the sum of the upper reserve capacity configured for all thermal power units, pumped storage units, hydropower units, and energy storage units within a zone shall not exceed the total upper reserve demand of each zone at the corresponding time, and the sum of the lower reserve capacity configured for all thermal power units, pumped storage units, hydropower units, and energy storage units within a zone shall not be less than the total lower reserve demand of each zone at the corresponding time.

[0014] Furthermore, the step of dynamically partitioning the power system based on the power flow data using an improved semi-supervised spectral clustering algorithm to obtain the target partitioning result includes: Based on the power flow data, construct a power transfer distribution factor matrix for congested lines that takes into account uncertainties; Based on the power transfer distribution factor matrix, a similarity connection matrix is ​​constructed, and based on the similarity connection matrix, a normalized Laplace matrix is ​​constructed. The normalized Laplacian matrix is ​​subjected to eigenvalue decomposition to obtain several eigenvectors. The eigenvectors are then clustered to obtain the target partitioning result.

[0015] This method of pre-scheduling and acquiring power flow data based on a pre-set scenario dataset provides an accurate data foundation that reflects multiple uncertainties for subsequent analysis. This makes the analysis of system blocking modes and operating states closer to actual possible scenarios, thereby improving the reliability of the analysis from the source.

[0016] Further, the clustering of each of the feature vectors to obtain the target partitioning result includes: Select the k smallest non-zero eigenvalues ​​from the eigenvectors to form an eigenvector matrix, where k is the preset number of partitions; Clustering algorithms are used to cluster the row vectors of the feature vector matrix to divide the nodes of the power system into k clusters, forming the initial partitioning results; The initial partitioning result is adjusted based on the supervision condition that each cluster contains at least one generator node that can be used to provide backup, in order to optimize the spatial distribution of backup resources in the partitions, until all clusters meet the supervision condition, and the target partitioning result is obtained.

[0017] Another embodiment of the present invention provides a backup resource allocation system for a power system, comprising: The first module is used to perform pre-scheduling based on a preset scenario dataset and to obtain power flow data for each scenario of the power system after pre-scheduling. The second module, based on the power flow data, dynamically partitions the power system using an improved semi-supervised spectral clustering algorithm to obtain the target partitioning results; The third module is used to input the target partitioning result and the scenario dataset into a preset regional reserve resource configuration model, with the goal of minimizing the total expected cost of the system, and to solve for the total upper and lower reserve requirements of each sub-region in the target partitioning result in the future short-to-medium-term preset period. The fourth module is used to acquire the day-ahead forecast data of the power system, input the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource allocation model, optimize it under the condition of meeting the minimum reserve constraints of each sub-region, determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, and allocate reserve resources for each generator unit in each sub-region according to the base load output point and the upper and lower reserved reserve capacity.

[0018] Another embodiment of the present invention provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the steps of the power system backup resource configuration method as described in the present invention. Attached Figure Description

[0019] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating one embodiment of the power system backup resource allocation method provided in this application; Figure 2 This is a flowchart illustrating one embodiment of steps S201 to S203 provided in this application; Figure 3 This is a schematic diagram of the pre-scheduling results provided in this application; Figure 4 This is a schematic diagram of the partitioned spare capacity results provided in this application; Figure 5 This is a schematic diagram of the dynamic partitioning results provided in this application; Figure 6 This is a schematic diagram of the alternative configuration results provided in this application; Figure 7 This is a schematic diagram of the structure of an embodiment of the power system backup resource allocation system provided in this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0023] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0024] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0025] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0026] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0027] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0028] To ensure real-time balance between power supply and demand and guarantee the safe and stable operation of the system, sufficient reserve resources (such as rapidly adjustable thermal power, hydropower, pumped storage, and energy storage) must be pre-configured. The scientific and rational allocation of reserve resources plays a crucial role. Some technologies employ static rules based on a uniform proportion across the entire system (such as allocating reserves at a fixed proportion according to maximum load) or optimization methods focused on a single resource (such as thermal power) when allocating reserve resources. However, these methods are difficult to guarantee in terms of accuracy.

[0029] See Figure 1 To improve the accuracy of backup resource allocation, an embodiment of the present invention provides a backup resource allocation method for a power system, including steps S101 to S104: Step S101: Perform pre-scheduling based on the preset scenario dataset and obtain power flow data for each scenario of the power system after pre-scheduling; In some embodiments, it is first necessary to prepare a "typical set of scenarios to be regulated" that covers multiple uncertainties in the future (such as load fluctuations, changes in renewable energy output, extreme weather risks, etc.). Each scenario s in this set corresponds to a given probability of occurrence. Subsequently, for each scenario *s* in this scenario set, an independent day-ahead safe and economical dispatch is performed, without considering wildfire disasters and reserve configurations, and with operational economics as the sole optimization objective. Each dispatch is solved using a complete power system optimization model that includes generator operation constraints and network power flow constraints (usually using a DC power flow model). After solving, detailed operating status of the power system within a complete dispatch cycle (e.g., the next 24 hours, with time intervals *t*) under this scenario can be obtained, with the most critical result being the power flow data.

[0030] It should be noted that the power flow data includes line power flow information and node injected power information. Line power flow information refers to the active power flowing through each line l at each time t and in each scenario s. By comparing this power value with the line's maximum transmission power... By comparing the data and combining it with a predefined over-capacity ratio threshold, it is possible to determine and record whether the line is blocked in a given scenario. Node injection power information refers to the net injection power of each node j at each time t and in each scenario s. .

[0031] Step S102: Based on the power flow data, the power system is dynamically partitioned using an improved semi-supervised spectral clustering algorithm to obtain the target partitioning result; Please refer to Figure 2 In some embodiments, step S102 includes steps S201 to S203: Step S201: Based on the power flow data, construct a power transfer distribution factor matrix for congested lines that takes into account uncertainties; In some embodiments, step S201 includes: using the probability of each scenario in the scenario dataset as weights, and combining the congestion status and limit violation degree of the lines in the pre-scheduled power flow data under each scenario, a weighted calculation is performed to obtain a matrix H that comprehensively represents the influence of nodes on congested lines, whose elements... This represents the power transfer distribution factor of node i to line l. Specifically, firstly, based on the acquired power flow data, it is determined whether each line l in each scenario s is blocked: when the power flowing through line l at time t... Exceeding its transmission limit And reach the preset over-limit capacity ratio When this occurs, it is marked as blocked, and a 0-1 variable is defined. (Blocking is 1, otherwise it's 0). Then, combine the probabilities of each scenario. The power transfer distribution factor matrix H of the blocked line is obtained by weighting the severity of line over-limit and the inherent electrical relationship between the nodes and the line.

[0032] In some embodiments, the expression for the power transfer distribution factor matrix H of the congested line is: ; ; ; ; ; ; In the formula, Let be the power transfer distribution factor from node i to line l. Let be the power transmission distribution coefficient matrix of the i-th node to line l. It is the proportion of capacity exceeding the limit. It is the probability that line l will exceed the limit in scenario s. It is the probability of scenario s occurring. It is a 0-1 variable (takes 1 only when line l is blocked in scenario s). Let t be the power flowing through line k in scenario s at time s. It is the upper limit of the line transmission power. It is the total number of nodes. It is the total number of blocked lines. The network impedance matrix, To remove the admittance matrix of the row and column containing the balance node, Let be the net injection power of node j in scenario s at time t.

[0033] Step S202: Based on the power transfer distribution factor matrix, construct a similarity connection matrix, and based on the similarity connection matrix, construct a normalized Laplace matrix; In some embodiments, firstly, an adaptively scaled Gaussian function is used to calculate the similarity between any two nodes i and j based on the power transfer distribution factor matrix H, where the scale parameter... The similarity connection matrix is ​​determined based on the distance from node i to its k-th nearest neighbor. The calculation formula is: In the formula, Let be the scale parameter, representing the distance from node i to its k-th nearest neighbor. This represents the i-th row of the H matrix. Then, based on the similarity matrix... Construct the normalized Laplace matrix L, which is the normalization operator used in spectral clustering to extract graph structure features. The calculation formula is: The Laplace matrix is ​​defined as L = D – W, ​​where D is a diagonal matrix composed of the diagonal elements of the weighted adjacency matrix. The normalized Laplace matrix has better result consistency, and thus a normalized Laplace matrix can be constructed, expressed as: .

[0034] It should be noted that the similarity calculation is based on a key assumption: if the influence patterns of two nodes on critical congested lines in the system (i.e., the corresponding row vectors in matrix H) are highly similar, then they should belong to the same "region" in an electrical sense.

[0035] Step S203: Perform eigenvalue decomposition on the normalized Laplacian matrix to obtain several eigenvectors, and cluster each eigenvector to obtain the target partitioning result.

[0036] In some embodiments, eigenvalue decomposition is performed on the normalized Laplacian matrix, specifically by performing eigenvalue decomposition on the normalized Laplacian matrix L to solve for its eigenvalues ​​and corresponding eigenvectors. According to spectral theory, the eigenvalues ​​of the normalized Laplacian matrix L can be regarded as a measure of the ease of graph partitioning.

[0037] In some embodiments, clustering the feature vectors to obtain the target partitioning result includes: selecting the feature vectors corresponding to the k smallest non-zero feature values ​​in the feature vectors to form a feature vector matrix, where k is a preset number of partitions; using a clustering algorithm to cluster the row vectors of the feature vector matrix to divide the nodes of the power system into k clusters, forming an initial partitioning result; adjusting the initial partitioning result with the supervision condition that each cluster contains at least one generator node that can be used to provide backup, to optimize the spatial distribution of backup resources in the partitions, until all clusters satisfy the supervision condition, and obtaining the target partitioning result. Specifically, firstly, in order to divide the system nodes into k regions (clusters), it is necessary to select the feature vectors corresponding to the k smallest non-zero feature values ​​of matrix L, and use these k feature vectors as column vectors to form an n-dimensional feature vector matrix (n is the total number of system nodes). Each row of this matrix is ​​a k-dimensional vector, corresponding to a node in the original network. These row vectors re-express the similarity relationship of nodes in the new feature space. Subsequently, the K-means clustering algorithm is used with a preset number of partitions k as the number of cluster centers. Through iterative calculation, all row vectors (i.e., all nodes) are divided into k different clusters, thus forming an initial partitioning result. However, this initial result is based solely on network topology and electrical similarity and may not meet the engineering requirements for backup resource allocation. Therefore, supervisory information must be introduced to adjust it, checking each cluster generated by the initial clustering. If a cluster is found to contain no generator nodes, the composition of the cluster needs to be adjusted to reassign some or all of the nodes in the cluster to neighboring clusters that contain generator nodes, or to merge adjacent clusters, until all k clusters meet the supervisory condition of "containing at least one usable generator node".

[0038] It should be noted that this monitoring information is defined as: each partition (cluster) must contain at least one generator node that can be used to provide backup.

[0039] This method of pre-scheduling and acquiring power flow data based on a pre-set scenario dataset provides an accurate data foundation that reflects multiple uncertainties for subsequent analysis. This makes the analysis of system blocking modes and operating states closer to actual possible scenarios, thereby improving the reliability of the analysis from the source.

[0040] Step S103: Input the target partitioning result and the scenario dataset into the preset regional reserve resource configuration model, with the goal of minimizing the total expected cost of the system, and solve for the total upper and lower reserve requirements of each sub-region in the target partitioning result in the future short and medium term preset period. In some embodiments, step S103 includes: constructing a stochastic optimization model for regional reserve resource allocation based on scenario joint pre-scheduling; and, based on the target partitioning results and the scenario dataset, using the objective function of minimizing the expected sum of scenario probability-weighted fixed costs, scenario operating costs, scenario reserve operating costs, and scenario risk costs, and under the condition of satisfying the operating constraints of thermal power units, hydropower units, pumped storage resources, and energy storage resources, solving the stochastic optimization model for regional reserve resource allocation to obtain the total upper reserve demand and the total lower reserve demand of each sub-region at each moment within the short-to-medium-term preset period.

[0041] In some embodiments, a stochastic optimization model for regional reserve resource allocation based on scenario-based joint pre-scheduling is first constructed. This model uses a short-to-medium term of one week or more for scheduling, and its core lies in employing stochastic programming methods to handle the multiple uncertainties in the scenario dataset. The objective function of the model is to minimize the probability of each scenario. The objective function is the sum of the expected weighted pre-week operating costs, risk costs, and reserve configuration costs of the power system, expressed mathematically as follows: In the formula, S represents the total number of scenes. For scenario probabilities, For fixed costs that change only over time and do not change with the scenario, For the scene operation cost, To reserve operating costs for the scenario Costs associated with specific scenarios.

[0042] In some embodiments, the method includes: the fixed costs include the start-up and shutdown costs of thermal power units and pumped storage units, as well as the standby capacity configuration costs of thermal power, pumped storage, hydropower, and energy storage; the scenario operating costs include the fuel costs of thermal power units, the operating costs of pumped storage units, and the carbon emission costs; the scenario standby operating costs include the actual standby call costs of thermal power, pumped storage, hydropower, and energy storage; and the scenario risk costs include the cost of load shedding and the cost of wind and solar curtailment.

[0043] In some embodiments, fixed costs Costs that are independent of the specific application scenario are mainly divided into start-up and shutdown costs for thermal power / pumped storage power plants. Configuration cost of backup power capacity Two parts, the expression is: ; ; ; In the formula, and These are the start-up and shutdown costs for thermal power and pumped storage power, respectively. , , and The standby capacity configuration costs for thermal power, pumped storage, hydropower, and energy storage are expressed as follows: ; ; ; ; ; ; In the formula, The number of thermal power units. The number of hydroelectric generating units. The number of energy storage units; and It is the start-stop state variable of unit i at time t, which is set to 1 at the start-stop time and 0 at other times; and These represent the upper and lower standby configuration power of hydropower unit i at time t. and These are the start-up and shutdown costs for thermal power units and pumped-storage units, respectively. / , / , / , / These are the unit capacity configuration costs for upper / lower standby power for thermal power, pumped storage, hydropower, and energy storage, respectively.

[0044] In some embodiments, the scenario operating cost In terms of power generation, the output cost of hydropower units is very small and can be ignored. The initial dispatch output of energy storage resources is much smaller compared to thermal power and pumped storage, and the cost of standby deployment of energy storage devices is relatively high. Therefore, the scenario operation cost of energy storage devices is also included in the standby cost; pure operation costs are not considered. Thus, the scenario operation cost mainly consists of the fuel cost of thermal power units. Pumped storage operation costs and carbon emission costs It consists of three parts, and the expressions are as follows: ; ; ; ; In the formula, , , It is the coal consumption coefficient of thermal power unit i. It's the price of coal. and These are the unit operating costs for power generation and water pumping of the pumped-storage unit, respectively. The price for carbon emission trading. This represents the carbon emission coefficient of thermal power units. It is the carbon quota coefficient. It refers to the number of loads that involve carbon emissions. It is the power of load i involving carbon emissions in scenario s at time t.

[0045] It is worth noting that the coal consumption cost uses a quadratic function, which will consume a significant amount of time when the problem size is large. Therefore, the original model can be modified... To linearize the coal consumption function, it is divided into piecewise linearization segments, resulting in m segments: The expression is: ; ; In the formula, Let be the slope of the l-th segment after piecewise linearization of thermal power unit i at time t under scenario s. Let l be the segmented output of thermal power unit i after piecewise linearization in scenario s at time t.

[0046] In some embodiments, scenario standby operating costs It is mainly divided into thermal power, hydropower, pumped storage, and energy storage. For each time period and each scenario, the actual reserve demand is generally different, and the expressions are as follows: ; ; ; ; ; In the formula, and These represent the upper and lower standby outputs of hydropower unit i at time t in scenario s. / , / , / , / These are the standby unit power dispatch costs for thermal power, pumped storage, hydropower, and energy storage, respectively.

[0047] In some embodiments, scenario risk cost Depending on the scenario, it is mainly divided into loss-of-load cost and wind / solar curtailment cost, expressed as: ; In the formula, The number of load nodes, Let i be the expected load loss value of load i at time t in scenario s. The cost of penalty for unit loss of load. and These are the numbers of wind power and solar power stations, respectively. and These represent the curtailment power of wind and solar power plants at time t in scenario s, respectively. and These represent the unit penalty cost for abandoning wind and solar power.

[0048] In some embodiments, the model needs to be solved under detailed operational constraints of various resources. These constraints together constitute the feasible domain for joint pre-scheduling of multiple scenarios, which mainly includes the operational constraints of thermal power units, hydropower units, pumped storage resources and energy storage resources.

[0049] In some embodiments, thermal power unit constraints include output constraints, upper and lower reserve constraints, ramping constraints, minimum start-up and shutdown time constraints (start-stop constraints), etc. The thermal power output constraints and upper limit reserve selection constraints considering the initial scheduling of resource output are as follows: ; ; ; ; ; In the formula, Let i be the actual output of thermal power unit i in scenario s at time t. Let i be the initial power output of thermal power unit i at time t. and These represent the upper and lower standby outputs of thermal power unit i at time t in scenario s. and These represent the upper and lower reserve configuration power of thermal power unit i at time t, respectively. and These are the minimum and maximum technical output of thermal power unit i, respectively (applicable to both the initial dispatch output and the actual output after standby configuration).

[0050] The ramping constraint for thermal power units is: ; In the formula, This represents the upper limit of the ramping power of thermal power unit i.

[0051] The start-up and shutdown of the generating units are not scenario-specific; that is, the start-up and shutdown status of the units remains consistent across all scenarios. The minimum start-up and shutdown time constraints (start-up and shutdown constraints) for thermal power units are: ; ; ; ; ; ; In the formula, It is the state variable of unit i at time t, with 1 for start-up and 0 for shutdown; and It is the number of hours that unit i has been continuously started or shut down at time t; and These are the minimum downtime and minimum start-up time for unit i.

[0052] In some embodiments, the constraints of hydropower units include output constraints, upstream and downstream reserve selection constraints, reservoir capacity constraints, and flow constraints. Daily initial reservoir capacity pre-scheduling constraints are applied for the day before, the day of, and several days after extreme weather events. The output constraints and upstream and downstream reserve selection constraints of hydropower units are similar to those of thermal power units, and their expressions are as follows: ; ; ; ; ; In the formula, Let i be the initial power output of hydropower unit i at time t. The maximum technical output of hydropower unit i (applicable to both initial dispatch output and actual output after standby configuration).

[0053] The storage capacity constraint and flow constraint are: ; ; ; ; ; ; In the formula, Let i be the energy conversion coefficient of the hydropower unit. Let be the power generation flow rate of hydropower unit i at time t in scenario s. and These are the lower and upper limits of the power generation flow of hydropower unit i, respectively. The reservoir capacity of hydropower unit i in scenario s at time t. The inflow rate of hydropower unit i at time t. The amount of water discarded by hydropower unit i at time t in scenario s. and These are the lower and upper limits of the reservoir capacity for hydropower unit i, respectively, and T is the scheduling period. and These are the lower and upper limits of the water discharge for hydropower unit i, respectively.

[0054] The reservoir capacity of hydropower units needs to meet the constraint of consistency between the beginning and end of the scheduling cycle. This constraint is not required during daily scheduling. However, a stricter capacity range constraint needs to be added to the initial reservoir capacity at the beginning and end of the day before, during, and several days after extreme weather events. This can be expressed as: ; In the formula, and These are the minimum and maximum daily initial reservoir capacity of the hydropower station where hydropower unit i is located.

[0055] In some embodiments, the energy storage device has fast and accurate response capabilities, can reduce the capacity of the power system's spinning reserve and the carbon emissions from thermal power plants, and can serve as both an energy reserve and a capacity reserve. The constraints of the energy storage device are as follows: ; ; ; ; ; ; ; ; ; ; ; ; In the formula, It refers to the state of charge of energy storage device i at time t in scenario s. It is the dissipation coefficient. It refers to charge / discharge efficiency. It is the actual charging power of energy storage device i at time t in scenario s. It is the actual discharge power of energy storage device i at time t under scenario s. and These are the upper and lower limits of the state of charge of energy storage device i, respectively. and These are the upper limits of the charging and discharging power of energy storage device i (applicable to both the initial dispatch output and the actual output after standby configuration). and These are the initial scheduled charging and discharging power of energy storage device i at time t. and These represent the upper and lower standby outputs of energy storage device i at time t in scenario s. and These are the upper and lower standby configuration capacities of energy storage device i at time t, respectively.

[0056] The capacity of energy storage devices does not need to remain consistent throughout the day, but can be configured more precisely the day before, the day of, or several days after extreme weather events, or at other times. This can be represented as: ; In the formula, and These represent the minimum and maximum capacity limits of energy storage device i at the beginning of the date on which capacity needs to be configured, respectively.

[0057] A pumped-storage hydroelectric power station utilizes electricity generated during periods of low electricity load to pump water to an upper reservoir, and then releases the water to a lower reservoir to generate electricity during periods of high electricity load. It is also known as a pumped-storage hydroelectric power station. The pumped-storage model consists of two levels: the pumped-storage power station and the pumped-storage units. The pumped-storage units are further divided into constant-speed units and variable-speed units based on whether their pumping power is adjustable. They have two distinct states: pumping and generating. For constant-speed units, the pumping power is fixed because the unit's start-up and shutdown are unaffected by the environment. Therefore, whether a unit operates at a given time does not change with the environment. Thus, the pumping power and state of constant-speed units are unaffected by the environment, and the constraints are as follows: ; ; ; ; In the formula, This refers to the rated pumping power of the constant-speed unit; Let be the operating state variable of constant speed unit i at time t, where 1 represents running and 0 represents stopping; The actual pumping power of the constant-speed unit i at time t; The actual pumping power of variable speed unit i in scenario s at time t; Let be the operating state variable of variable speed unit i at time t; and These are the upper and lower limits of the pumping power of the variable speed unit, respectively. Let i be the actual power generation of unit i at time t in scenario s; Let be the operating state variable of unit i at time t; and These are the upper and lower limits of the unit's power generation capacity.

[0058] Pumped storage power stations have upper and lower reservoirs. The expression for reservoir capacity state transition is: ; ; In the formula, The unit scheduling interval time, and These represent the upper and lower reservoir capacities of the pumped storage power station at time t under scenario s. and These are the conversion coefficients between pumped storage unit pumping power, power generation, and flow rate. and These refer to the number of constant-speed and variable-speed generator sets, respectively. and The expression is: ; In the formula, and These refer to the pumping and power generation efficiency of the unit, respectively. , g, and h represent the density of water, gravitational acceleration, and the average head difference between the upper and lower reservoirs, respectively.

[0059] The capacities of the upper and lower reservoirs of a pumped storage power station, and the maximum change in the capacity of the upper reservoir during the first and last periods of the scheduling cycle, need to be constrained. The expression is as follows: ; ; ; In the formula, and These represent the upper and lower limits of the reservoir's capacity, respectively. and These represent the upper and lower limits of the reservoir's capacity. and These represent the maximum permissible reservoir capacity variation values ​​for the first and last periods within the reservoir's scheduling cycle.

[0060] The upstream and downstream reservoir capacities of pumped storage power stations need to be configured more precisely one day before, on, or several days after extreme weather events, or even longer. This can be expressed as: ; ; In the formula, and These represent the maximum and minimum capacity limits of the pumped storage power station at the beginning of the date on which capacity needs to be configured.

[0061] The reserve capacity of pumped storage power stations is configured only for the generating capacity portion, expressed as: ; ; ; ; In the formula, The initial power output of unit i at time t. and These represent the upper and lower standby outputs of unit i at time t in scenario s, respectively. and These represent the upper and lower standby capacity configurations for unit i at time t.

[0062] In some embodiments, a DC power flow model is used in the transmission network, and the power balance equation is: ; In the formula, It is the load power demand of load i at time t in scenario s. and These represent the original outputs of wind power and photovoltaic generator unit i at time t in scenario s, respectively. Let i be the actual output of hydropower unit i in scenario s at time t.

[0063] In some embodiments, regional reserve capacity demand is determined by combining the zoning results and the joint pre-scheduling results of the previous week's scenarios. The actual generation-side resources provided for reserve include thermal power, hydropower, pumped storage, and energy storage. The expression for short-to-medium-term regional reserve capacity demand is as follows: ; ; In the formula, j is the partition number. and These represent the total upper and lower reserve capacity of region j at time t, respectively.

[0064] It should be noted that the upper reserve is mainly for dealing with load surges, while the lower reserve is mainly for dealing with renewable energy redundancy.

[0065] It's important to note that initial resource output scheduling doesn't consider specific scenarios; each scenario starts consistently. However, different scenarios generate different reserve demands, meaning the actual output of resources considering reserve requirements varies across scenarios. This is the significance of discussing reserve configurations across scenarios. The maximum reserve configuration across all scenarios represents the minimum reserve that a resource must reserve, thus generating a reserve resource assessment result. It's worth noting that pre-weekly scheduling forecasts are typically not highly accurate. Therefore, multiple possible scenarios requiring adjustment must be considered. For this reason, the reserve capacity reserved before the week only uses the total capacity of each region as the result in the next day-ahead scheduling, without focusing on the specific capacity configuration of each unit.

[0066] This method, which inputs the partitioning results and scenario datasets into the regional reserve resource configuration model to solve for the total reserve demand of each sub-region, comprehensively considers the probability weights of multiple uncertain scenarios in the short and medium term through a stochastic optimization method. With the goal of minimizing the expected total cost, it accurately determines the upper and lower limits of the reserve capacity required for each partition, avoiding insufficient or redundant reserves caused by single scenario prediction. It improves the accuracy of configuration from the perspectives of time dimension and total planning.

[0067] Step S104: Obtain the day-ahead forecast data of the power system, input the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource configuration model, optimize under the condition of meeting the minimum reserve constraints of each sub-region, determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, and configure the reserve resources of each generator unit in each sub-region according to the base load output point and the upper and lower reserved reserve capacity.

[0068] In some embodiments, the step of inputting the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource configuration model, and optimizing it under the condition of satisfying the minimum reserve constraints of each sub-region, to determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, includes: constructing a day-ahead reserve configuration model based on the day-ahead forecast data and the total reserve requirements of each sub-region; solving the day-ahead reserve configuration model with the objective function of minimizing the total day-ahead operating cost of the power system, under the condition of satisfying the operating constraints of generator units, the minimum reserve constraints of each sub-region, and the zoned reserve capacity constraints, to obtain the base load output point, upper reserved reserve capacity, and lower reserved reserve capacity of each generator unit in the next day, wherein the zoned reserve capacity constraint is composed of the total upper reserve requirement and the total lower reserve requirement.

[0069] In some embodiments, the day-ahead reserve configuration model is a refined scheduling model for the next day (24 hours) based on more accurate day-ahead forecast data. This model uses the total upper and lower reserve demand of each sub-region on the target day, obtained in step three (short-term optimization), as a rigid framework, and minimizes the total day-ahead operating cost of the power system as its objective function. The total day-ahead operating cost mainly consists of three parts: day-ahead operating cost (mainly including fuel costs of thermal power units, etc.), day-ahead reserve configuration cost (costs incurred for reserving reserve capacity for units), and day-ahead risk cost (including the cost of load shedding and wind / solar curtailment that may result from forecast deviations).

[0070] In some embodiments, the method includes: the day-ahead total operating cost includes day-ahead operating cost, day-ahead standby configuration cost, and day-ahead risk cost; wherein, the day-ahead operating cost includes at least the unit's operating cost. The aforementioned day-ahead standby configuration cost is the cost incurred from reserving standby capacity. The aforementioned day-ahead risk costs include loss of load costs and wind / solar curtailment costs. .

[0071] In some embodiments, the mathematical expression of the day-ahead standby configuration model is: Its specific expression is consistent with that of short-to-medium term regional optimization, the only difference being that the day-ahead reserve allocation model here is a simplified form of the regional reserve resource allocation model under a single forecast scenario.

[0072] In some embodiments, the solution of the model must satisfy the following key constraints to ensure that the scheme has both safety, reliability and economy: Operating constraints of generator units: including specific operating limitations of all participating thermal power, hydropower, pumped storage and energy storage units, such as ramp rate, minimum start-up and shutdown time, reservoir capacity balance, and dynamic state of charge of energy storage, to ensure the engineering feasibility of the scheme. This constraint has been elaborated in detail in step S102, so it will not be repeated here. Furthermore, since the day-ahead reserve configuration model is a simplified form of the regional reserve resource configuration model under a single prediction scenario, the regional reserve resource configuration model needs to consider multi-scenario coupling, while the day-ahead reserve configuration model does not need to consider, i.e., there is no scenario coupling constraint, but there are partition minimum reserve constraints, partition reserve capacity constraints obtained from short-term and medium-term pre-scheduling, and reserve capacity configuration variables. The expressions for the partition minimum reserve constraints and the partition reserve capacity constraints obtained from short-term and medium-term pre-scheduling are: ; ; In the formula, , , , The configurations for thermal power, pumped storage, hydropower, and energy storage are as follows: upper-level backup and lower-level backup. and These represent the minimum required reserve ratios for wind, solar, and load within region j, respectively. This constraint ensures consistency between long-term uncertain planning and short-term deterministic scheduling, guaranteeing that the day-ahead reserve resources are sufficient to cover the system risks assessed in the medium and short term. Ultimately, solving this optimization model yields the refined baseload output points for each unit. With reserved spare capacity This completes the closed loop of reserve resources from regional total planning to specific allocation at the unit level.

[0073] In some embodiments, the standby capacity configuration variable specifically refers to: for any activated, configurable standby unit i in the model at any time t, its scheduling decision variable includes the baseload output point. and the reserve capacity reserved for it They must all meet the unit's own upper and lower output limits, namely: In the formula, For the day-ahead fixed output of unit i, Reserved for unit i and The corresponding upper and lower limits of unit output. These are the start-up and shutdown variables for the day-ahead unit combinations.

[0074] It should be noted that this step uses the day-ahead dispatch results as the fixed output. Based on the load and wind and solar forecast deviations, the effectiveness of the proposed model is verified using intraday data. The upper and lower limits of unit output adjustment are the day-ahead reserve configuration. Similarly, the dispatch is carried out with economic efficiency as the objective function, and wind and solar curtailment and load shedding are taken into account as important means of balancing electricity in economic costs.

[0075] It should be noted that the total partitioned reserve demand output by S103, in the form of a "partitioned reserve capacity constraint," becomes a key constraint in the S104 optimization problem. This means that S104 refines costs only after satisfying the "reserve strategic reserve" plan established by S103. Therefore, the objective of S103 (minimizing the expected total cost) indirectly guides and constrains the optimization direction of S104 through its output total reserve demand.

[0076] This method, which uses day-ahead forecast data and total reserve demand to determine the specific base load output and reserve capacity of each unit through a day-ahead reserve resource allocation model, utilizes more accurate short-term forecast data to finely allocate reserve resources to each unit while meeting the total regional constraints and minimum reserve constraints. This achieves accurate implementation from regional totals to specific unit strategies.

[0077] This application proposes a power system zonal reserve optimization method considering multiple uncertainties. Based on the pre-scheduling results of uncertain scenarios, an improved semi-supervised spectral clustering zoning algorithm is proposed, using the power transfer distribution factor of congested lines as a similarity measure. Based on the zoning results, a stochastic optimization model for regional reserve resource allocation based on scenario joint pre-scheduling is proposed, with the objective function being to minimize the expected sum of power system operating costs, risk costs, and reserve configuration costs weighted by scenario probabilities, to solve for the regional reserve capacity allocation requirements under possible short- and medium-term scenarios. Based on day-ahead forecast data, a day-ahead reserve configuration model considering zonal minimum reserve constraints and short- and medium-term pre-scheduling reserve capacity allocation is proposed, and the effectiveness of the proposed method is verified through intraday actual operation.

[0078] The power system reserve resource allocation method provided in this application can significantly improve the accuracy of reserve resource allocation. Specifically, it is reflected in the following aspects: First, the method of pre-scheduling and obtaining power flow data based on a preset scenario dataset provides an accurate data foundation reflecting multiple uncertainties for subsequent analysis, making the analysis of system congestion modes and operating states closer to actual possible scenarios and improving the reliability of the analysis from the source. Second, the method of dynamically partitioning the power network structure by improving the semi-supervised spectral clustering algorithm based on power flow data, and placing easily congested lines at the partition boundaries and ensuring that each sub-region contains a reserve power source, achieves a scientific division of the power network structure, optimizes the spatial layout of reserve resources, and improves the independence and effectiveness of reserve regulation within the partition. Third, the method of inputting the partitioning results and scenario dataset into a regional reserve resource allocation model to solve for the total reserve demand of each sub-region, through random optimization... The method comprehensively considers the probability weights of multiple uncertain scenarios in the short and medium term, and accurately determines the upper and lower limits of the required reserve capacity for each zone with the goal of minimizing the expected total cost. This avoids reserve shortages or redundancies caused by single-scenario predictions, and improves the accuracy of configuration from the time dimension and the overall planning level. Finally, based on the day-ahead forecast data and the total reserve demand, the method determines the specific base load output and reserve capacity of each unit through the day-ahead reserve resource configuration model. By using more accurate short-term forecast data, and under the condition of meeting the total zone constraints and minimum reserve constraints, the method finely allocates reserve resources to each unit, realizing the accurate implementation from the regional total to the specific unit strategy. Thus, it forms a complete technical chain from macro-situation analysis, scientific zoning, total optimization to micro-precise configuration, and systematically improves the accuracy and economy of reserve resource configuration in the power system when dealing with multiple uncertainties of source and load.

[0079] To verify the effectiveness of the proposed method, simplified power grid data was selected. The proposed model is a mixed-integer linear programming problem. The model was built on the MATLAB 2022a platform and solved using the Gurobi 10.0.3 solver called by YALMIP. A one-week scheduling period was used, with a 1-hour scheduling interval. Wind, solar, and load data were simulated based on actual power grid operation data.

[0080] The first part of the pre-scheduling results is shown below. Figure 3 As shown, thermal power accounts for a large proportion of all power output, exceeding 50%. Since photovoltaic power output is higher at noon, the output of renewable energy also shows a pattern of high output at noon and low output in the morning and evening. Hydropower accounts for a relatively large proportion during the first three days, which is related to the changes in hydropower reservoir capacity caused by the weather during that week.

[0081] The results of partitioned spare capacity are as follows Figure 4 As shown, during the partitioning process, the entire region is dynamically divided into four sub-regions, and the pre-scheduled reserve capacity configuration is allocated to each region. The scenario in the diagram with fixed reserve requirements traditionally allocates 5% of the total load and 3% of the wind and solar power output as reserve requirements. From... Figure 4 a and Figure 4 As shown in b, the reserve requirement does not need to reach the traditional configuration level for a significant portion of the time, only reaching it when the midday photovoltaic output is highly uncertain. The lower reserve follows the same trend as the upper reserve. It is important to note that the upper / lower reserve here is based on the initial resource scheduling output; therefore, it can only mitigate uncertainty. The solver selectively adjusts the initial resource output to obtain the optimal solution. Thus, the upper / lower reserve here only reflects the system's response to uncertainty and is different from the traditional upper / lower reserve. Figure 5 The results of the dynamic partitioning are shown (in 1 hour). Figure 5 a) 50h ( Figure 5 (b) Taking this as an example, the different areas are divided and change over time according to the different power flows of the lines.

[0082] The alternative configuration result for this method is as follows: Figure 6 As shown in Table 1, the required costs are as follows. Due to day-ahead wind and solar load forecasting errors, the cost of setting up load shedding risk accounts for a certain proportion of the total cost. Therefore, by effectively reducing the amount of load shedding, the total cost can be effectively controlled, thereby achieving the best economic effect. The table shows that the proposed method requires certain reserve costs and unit start-up, operation, and fuel costs, but the amount of load shedding is relatively small, effectively reducing the load shedding risk caused by day-ahead wind and solar load forecasting errors. This method can effectively utilize more day-ahead configured reserve capacity within the day, thereby reducing the risk of load shedding and improving the economy and reliability of the power system.

[0083] Table 1 Cost of Alternate Optimization Results like Figure 7 As shown, based on the above method embodiments, corresponding apparatus embodiments are provided; An embodiment of the present invention provides a backup resource allocation system for a power system, comprising: The first module 100 is used to perform pre-scheduling based on a preset scenario dataset and to obtain power flow data for each scenario of the power system after pre-scheduling. The second module 200 dynamically partitions the power system based on the power flow data using an improved semi-supervised spectral clustering algorithm to obtain the target partitioning result. The third module 300 is used to input the target partitioning result and the scenario dataset into a preset regional reserve resource configuration model, with the goal of minimizing the total expected cost of the system, and to solve for the total upper and lower reserve requirements of each sub-region in the target partitioning result in the future short-to-medium-term preset period. The fourth module 400 is used to acquire the day-ahead forecast data of the power system, input the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource configuration model, optimize it under the condition of meeting the minimum reserve constraints of each sub-region, determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, and configure the reserve resources of each generator unit in each sub-region according to the base load output point and the upper and lower reserved reserve capacity.

[0084] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the power system backup resource configuration method provided by any of the above-described method embodiments of the present invention.

[0085] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0086] Based on the above-described embodiments of the power system backup resource allocation method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the power system backup resource allocation method of any embodiment of the present invention.

[0087] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0088] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0089] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0090] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the power system backup resource configuration method described in any of the above-described method embodiments of the present invention.

[0091] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0092] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for allocating reserve resources in a power system, characterized in that, include: Pre-scheduling is performed based on a pre-set scenario dataset, and power flow data for each scenario of the power system is obtained after pre-scheduling. Based on the power flow data, the power system is dynamically partitioned using an improved semi-supervised spectral clustering algorithm to obtain the target partitioning results; The target partitioning result and the scenario dataset are input into a preset regional reserve resource configuration model. With the goal of minimizing the total expected cost of the system, the total upper and lower reserve requirements of each sub-region in the target partitioning result in the future short-to-medium-term preset period are obtained. Obtain the day-ahead forecast data of the power system, input the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource allocation model, optimize under the condition of meeting the minimum reserve constraints of each sub-region, determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, and allocate reserve resources for each generator unit in each sub-region according to the base load output point and the upper and lower reserved reserve capacity.

2. The method for allocating reserve resources in a power system according to claim 1, characterized in that, The step of inputting the target partitioning results and the scenario dataset into a preset regional reserve resource allocation model, with the objective of minimizing the expected total system cost, and solving for the total upper and lower reserve requirements of each sub-region in the target partitioning results within a preset short-to-medium term, includes: Construct a stochastic optimization model for regional reserve resource allocation based on scenario-based joint pre-scheduling; Based on the target partitioning results and the scenario dataset, with the objective function being the minimum expected sum of scenario probability-weighted fixed costs, scenario operating costs, scenario reserve operating costs, and scenario risk costs, and under the operational constraints of thermal power units, hydropower units, pumped storage resources, and energy storage resources, the stochastic optimization model for the allocation of reserve resources in the region is solved to obtain the total upper reserve demand and lower reserve demand of each sub-region at each moment within the preset short-to-medium-term period.

3. The method for allocating standby resources in a power system according to claim 2, characterized in that, include: The fixed costs include the start-up and shutdown costs of thermal power units and pumped storage units, as well as the configuration costs of standby capacity for thermal power, pumped storage, hydropower, and energy storage. The operating costs of the scenario include the fuel costs of thermal power units, the operating costs of pumped storage units, and carbon emission costs. The standby operating costs in the scenario include the actual standby call costs for thermal power, pumped storage, hydropower, and energy storage. The risk costs in this scenario include the cost of load shedding and the cost of wind and solar power curtailment.

4. The method for allocating reserve resources in a power system according to claim 1, characterized in that, The step of inputting the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource allocation model, and optimizing it under the condition of satisfying the minimum reserve constraint of each sub-region, to determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, includes: Based on the day-ahead forecast data and the total reserve demand of each sub-region, a day-ahead reserve configuration model is constructed. With the objective function of minimizing the total day-ahead operating cost of the power system, the day-ahead reserve configuration model is solved under the conditions of satisfying the operating constraints of generator units, the minimum reserve constraints of each sub-region, and the zone reserve capacity constraints. This yields the base load output point, upper reserved reserve capacity, and lower reserved reserve capacity of each generator unit in the next day. The zone reserve capacity constraints are composed of the total upper reserve demand and the total lower reserve demand.

5. The method for allocating standby resources in a power system according to claim 4, characterized in that, include: The day-ahead total operating cost includes day-ahead operating cost, day-ahead reserve configuration cost, and day-ahead risk cost; wherein, the day-ahead operating cost includes at least the fuel cost of thermal power units, the day-ahead reserve configuration cost is the cost incurred from reserving reserve capacity, and the day-ahead risk cost includes loss of load cost and wind / solar curtailment cost.

6. The method for allocating standby resources in a power system according to claim 4, characterized in that, include: The operating constraints of the generator set are as follows: for any generator set that is in operation at any time, the sum of the base load output and the reserved upper reserve capacity shall not exceed the maximum technical output, and the difference between the base load output and the reserved lower reserve capacity shall not be less than the minimum technical output. The minimum reserve constraint for each sub-region is: the total upper reserve and the total lower reserve configured in the region must be no less than the preset ratio of wind and solar power output and load in the corresponding region, respectively. The zoned reserve capacity constraints are as follows: the sum of the upper reserve capacity configured for all thermal power units, pumped storage units, hydropower units, and energy storage units within a zone shall not exceed the total upper reserve demand of each zone at the corresponding time, and the sum of the lower reserve capacity configured for all thermal power units, pumped storage units, hydropower units, and energy storage units within a zone shall not be less than the total lower reserve demand of each zone at the corresponding time.

7. The method for allocating reserve resources in a power system according to claim 1, characterized in that, The step of dynamically partitioning the power system based on the power flow data using an improved semi-supervised spectral clustering algorithm to obtain the target partitioning results includes: Based on the power flow data, construct a power transfer distribution factor matrix for congested lines that takes into account uncertainties; Based on the power transfer distribution factor matrix, a similarity connection matrix is ​​constructed, and based on the similarity connection matrix, a normalized Laplace matrix is ​​constructed. The normalized Laplacian matrix is ​​subjected to eigenvalue decomposition to obtain several eigenvectors. The eigenvectors are then clustered to obtain the target partitioning result.

8. The method for allocating standby resources in a power system according to claim 2, characterized in that, The clustering of each feature vector to obtain the target partitioning result includes: Select the k smallest non-zero eigenvalues ​​from the eigenvectors to form an eigenvector matrix, where k is the preset number of partitions; Clustering algorithms are used to cluster the row vectors of the feature vector matrix to divide the nodes of the power system into k clusters, forming the initial partitioning results; The initial partitioning result is adjusted to optimize the spatial distribution of backup resources in each partition, with the condition that each cluster contains at least one generator node that can be used to provide backup, until all clusters meet the supervision condition, thus obtaining the target partitioning result.

9. A backup resource allocation system for a power system, characterized in that, include: The first module is used to perform pre-scheduling based on a preset scenario dataset and to obtain power flow data for each scenario of the power system after pre-scheduling. The second module, based on the power flow data, dynamically partitions the power system using an improved semi-supervised spectral clustering algorithm to obtain the target partitioning results; The third module is used to input the target partitioning result and the scenario dataset into a preset regional reserve resource configuration model, with the goal of minimizing the total expected cost of the system, and to solve for the total upper and lower reserve requirements of each sub-region in the target partitioning result in the future short-to-medium-term preset period. The fourth module is used to acquire the day-ahead forecast data of the power system, input the day-ahead forecast data and the total upper and lower reserve requirements into a preset day-ahead reserve resource allocation model, optimize it under the condition of meeting the minimum reserve constraints of each sub-region, determine the base load output point and upper and lower reserved reserve capacity of each generator unit in the next day, and allocate reserve resources for each generator unit in each sub-region according to the base load output point and the upper and lower reserved reserve capacity.

10. A terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the power system standby resource allocation method as described in any one of claims 1-8.

Images