New energy resource collaborative scheduling method and device, terminal equipment and storage medium

Abstract

The invention discloses a new energy resource collaborative scheduling method and device, terminal equipment and a storage medium, and relates to the technical field of power systems, and the method comprises the steps: collecting multi-dimensional feature data of a target resource in a new energy system, the multi-dimensional feature data comprising adjustment feature data, carbon constraint data and spatial and temporal distribution data; generating a three-dimensional feature vector set by using the multi-dimensional feature data, and determining a to-be-classified scene which meets new energy consumption rate constraints and has a carbon cost lower than a regional mean value based on three-dimensional feature vectors; classifying the to-be-classified scene into a low-carbon consumption type scene, a flexible balance type scene and an emergency guarantee type scene, and configuring a scheduling optimization weight of each scene according to historical data; and setting a target function of the scheduling model based on the scheduling optimization weight, solving the target function to obtain an optimal scheduling strategy, and controlling the target resource to execute the optimal scheduling strategy. The method can adapt to a complex operation scene, so that the effect of resource collaborative scheduling is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, apparatus, terminal equipment and storage medium for collaborative scheduling of new energy resources. Background Technology

[0002] Currently, the volatility and uncertainty of renewable energy output significantly impact the dispatch of high-proportion renewable energy sources. Flexibility resources in renewable energy systems refer to various resources capable of rapidly responding to changes in power system demand and possessing adjustment capabilities. These resources include energy storage systems (such as battery storage and pumped hydro storage), adjustable loads (such as industrial and commercial loads whose electricity consumption time can be flexibly adjusted through demand response mechanisms), interruptible loads (loads whose power supply can be interrupted when the grid needs it), and some power sources with flexible generation capabilities (such as some gas turbine generators). The main role of flexibility resources is to balance the intermittency and uncertainty of renewable energy generation (such as solar and wind power) and improve the flexibility and reliability of the power system.

[0003] Existing methods for coordinated scheduling of new energy resources typically categorize flexible resources based on their adjustment characteristics to construct a scheduling model targeting new energy absorption rate and system operating costs. The scheduling strategy is then derived through scenario analysis and hierarchical optimization algorithms. However, simply categorizing flexible resources based on their adjustment characteristics makes it difficult to determine the synergistic effects of different resources across various dimensions, and fails to adapt to complex operating scenarios, resulting in poor coordinated resource scheduling performance. Summary of the Invention

[0004] This invention provides a method, device, terminal equipment, and storage medium for collaborative scheduling of new energy resources, which can solve the technical problem that the existing technology of simply classifying flexible resources according to their adjustment characteristics cannot adapt to complex operating scenarios, resulting in poor collaborative scheduling of resources.

[0005] This invention provides a method for coordinated scheduling of new energy resources, comprising:

[0006] Collect multi-dimensional feature data of target resources in the new energy system, wherein the multi-dimensional feature data includes regulation characteristic data, carbon constraint data and spatiotemporal distribution data;

[0007] A three-dimensional feature vector set is generated from the multi-dimensional feature data, and the scenarios to be classified are determined based on the three-dimensional feature vectors, which meet the constraints of new energy consumption rate and have carbon costs lower than the regional average.

[0008] The scenarios to be classified are divided into low-carbon consumption scenarios, flexible balance scenarios, and emergency support scenarios, and the scheduling optimization weight of each scenario is configured according to historical data.

[0009] The objective function of the scheduling model is set based on the scheduling optimization weights, the objective function is solved to obtain the optimal scheduling strategy, and the target resource is controlled to execute the optimal scheduling strategy.

[0010] Furthermore, the step of generating a three-dimensional feature vector set from the multi-dimensional feature data includes:

[0011] The multi-dimensional feature data is normalized and then mapped to a three-dimensional space to generate a three-dimensional feature vector.

[0012] Furthermore, the step of determining the scenarios to be classified based on the three-dimensional feature vectors that meet the constraints of new energy absorption rate and have carbon costs lower than the regional average includes:

[0013] Wavelet packet transform technology is used to perform multi-scale analysis on the three-dimensional feature vector to determine high-frequency detail components;

[0014] A thresholding algorithm is used to nonlinearly compress the high-frequency detail components in the three-dimensional feature vector to obtain denoised feature data.

[0015] Cluster analysis is performed on the denoised feature data to obtain the final cluster set;

[0016] From the final cluster set, identify the scenarios to be classified that meet the new energy consumption rate constraint and have carbon costs lower than the regional average.

[0017] Furthermore, the clustering analysis performed on the denoised feature data to obtain the final cluster set includes:

[0018] An adaptive resolution spatial grid is constructed based on the denoised feature data, wherein the grid density of the spatial grid is divided according to the feature value distribution;

[0019] Determine the neighborhood radius of each sample in the spatial grid, and determine the neighborhood range of each sample based on the neighborhood radius;

[0020] The samples are sorted in ascending order of reachability to generate a permutation list, wherein the reachability is the shortest distance from one sample to another within a neighborhood.

[0021] The boundaries of the clusters are determined based on the permutation list and density jump points, resulting in an initial cluster set;

[0022] The adjacent clusters in the initial cluster set whose spatiotemporal feature dimension center distance is less than the cross-regional transmission delay threshold are merged into a new cluster to obtain the final cluster set.

[0023] Furthermore, configuring the scheduling optimization weights for each scenario based on historical data includes:

[0024] Extract the historical best scheduling record data of the target resource, construct a training dataset based on the historical best scheduling record data, and construct a scheduling model based on the training dataset;

[0025] The input feature data is input into the scheduling model, and the scheduling optimization weight corresponding to each input feature data is output. The input feature data includes scenario category code, real-time carbon price volatility and adjustment resource availability. The scheduling optimization weight includes new energy consumption rate weight, carbon cost weight and adjustment cost weight.

[0026] Furthermore, the objective function of setting the scheduling model based on the scheduling optimization weights includes:

[0027] Based on the aforementioned scheduling optimization weights, renewable energy absorption rate, total carbon emission cost, adjustment cost, actual renewable energy output, predicted renewable energy output, carbon emissions of the i-th type of target resource, carbon intensity coefficient of the i-th type of resource, real-time carbon trading price, total carbon emissions, and regional carbon emission quota, the objective function of the scheduling model is set.

[0028] Furthermore, the step of solving the objective function to obtain the optimal scheduling strategy and controlling the target resource to execute the optimal scheduling strategy includes:

[0029] The objective function is solved using a spatiotemporal hierarchical algorithm to obtain the optimal scheduling strategy.

[0030] The corresponding scheduling instructions are generated according to the optimal scheduling strategy, and the target resource is controlled to perform corresponding actions according to the scheduling instructions.

[0031] This invention provides a new energy resource collaborative scheduling device, comprising:

[0032] The feature data acquisition module is used to acquire multi-dimensional feature data of target resources in the new energy system, wherein the multi-dimensional feature data includes regulation characteristic data, carbon constraint data and spatiotemporal distribution data;

[0033] The module for determining the scenario to be classified is used to generate a set of three-dimensional feature vectors from the multi-dimensional feature data, and to determine the scenarios to be classified based on the three-dimensional feature vectors that meet the constraints of new energy consumption rate and have carbon costs lower than the regional average.

[0034] The scheduling optimization weight configuration module is used to classify the scenarios to be classified into low-carbon consumption scenarios, flexible balance scenarios, and emergency support scenarios, and to configure the scheduling optimization weight of each scenario based on historical data.

[0035] The scheduling optimization module is used to set the objective function of the scheduling model based on the scheduling optimization weights, solve the objective function to obtain the optimal scheduling strategy, and control the target resource to execute the optimal scheduling strategy.

[0036] 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. When the processor executes the computer program, it implements the new energy resource collaborative scheduling method as described above.

[0037] The present invention provides a computer-readable storage medium, comprising: a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the above-described new energy resource collaborative scheduling method.

[0038] The following benefits can be obtained by implementing the present invention:

[0039] This invention generates a set of three-dimensional feature vectors based on multi-dimensional feature data of target resources. Based on the three-dimensional feature vectors, it determines the scenarios that meet the constraints of new energy consumption rate and achieves accurate classification of different scenarios, including low-carbon consumption scenarios with carbon costs lower than the regional average, flexible balance scenarios, and emergency support scenarios. It can also configure the scheduling optimization weights for each scenario based on historical data, so that the scheduling model can adapt to complex operating scenarios, thereby effectively improving the effect of resource collaborative scheduling.

[0040] Furthermore, by constructing a spatial grid with adaptive resolution, this invention can accurately reflect the distribution of different new energy resources in the feature space, which helps to identify the similarities and differences between resources, thereby optimizing resource allocation and improving the utilization and absorption rate of new energy. Attached Figure Description

[0041] 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.

[0042] Figure 1 This invention provides a method for collaborative scheduling of new energy resources.

[0043] Figure 2 This is a schematic diagram of a matching rule for adjusting resource status thresholds and clustering results provided in an embodiment of the present invention;

[0044] Figure 3 This is a schematic diagram of scene classification provided by an embodiment of the present invention;

[0045] Figure 4 This is a new energy resource collaborative scheduling method provided by an embodiment of the present invention. Detailed Implementation

[0046] 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.

[0047] 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.

[0048] 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.

[0049] 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.

[0050] 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, or B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0051] 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).

[0052] 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.

[0053] See Figure 1 To address the problem that existing technologies, which simply classify flexible resources according to their adjustment characteristics, cannot adapt to complex operating scenarios and result in poor resource collaborative scheduling, an embodiment of the present invention provides a new energy resource collaborative scheduling method, comprising:

[0054] S1. Collect multi-dimensional feature data of target resources in the new energy system, including regulation characteristic data, carbon constraint data and spatiotemporal distribution data;

[0055] In this embodiment of the invention, the regulation characteristic data includes spatiotemporal complementary resource parameters, rapid regulation resource parameters, and load-side resource parameters. The spatiotemporal complementary resource parameters may include the maximum schedulable energy exchange per unit time, the charge-discharge cycle efficiency of energy storage devices, and the time delay characteristics of energy transmission across regions. The rapid regulation resource parameters may include the power regulation speed, the physical constraints of device start-up and shutdown, and the weight of the impact of unit regulation on device lifespan. The load-side resource parameters include a comprehensive score based on user response willingness and hardware capabilities, the upper limit of power that can be reduced or transferred on the load side, and the duration of load regulation sustainability.

[0056] Carbon constraint data can include resource carbon intensity coefficients, regional carbon quota remaining amounts, and real-time carbon trading prices. The resource carbon intensity coefficient can include the carbon intensity of thermal power units calculated based on fuel type and unit efficiency, the carbon intensity of new energy units (default value is 0), and inter-regional power transmission calculated based on the average carbon intensity of the receiving-end power grid. For example, the carbon intensity coefficient C of a certain thermal power unit... coal Calculated using the following formula:

[0057]

[0058] Among them, F coal For coal consumption, EF CO2 P is the carbon emission factor for coal combustion. max This is the maximum output of the generator unit.

[0059] Spatiotemporal distribution data can be obtained through Geographic Information System (GIS) and Energy Management System (EMS), including the distribution of new energy power plants, the spatiotemporal distribution of loads, and the utilization rate of inter-regional power transmission channels. The distribution of new energy power plants includes latitude and longitude coordinates, installed capacity, and predicted output curves (15-minute granularity); the spatiotemporal distribution of loads includes regional load density, peak and valley load distribution, and spatial clustering results of adjustable loads; the utilization rate of inter-regional power transmission channels includes historical maximum transmission power and current channel congestion rate.

[0060] S2. Generate a set of three-dimensional feature vectors from multi-dimensional feature data, and determine the scenarios to be classified based on the three-dimensional feature vectors that meet the constraints of new energy consumption rate and have carbon costs lower than the regional average.

[0061] S3. Classify the scenarios to be classified into low-carbon consumption scenarios, flexible balance scenarios, and emergency support scenarios, and configure the scheduling optimization weight for each scenario based on historical data.

[0062] In this embodiment of the invention, resource state thresholds and matching rules with clustering results are set. The resource state thresholds may include a photovoltaic / wind power output threshold, an energy storage state of charge threshold, and a carbon price fluctuation threshold, and their expressions are as follows:

[0063] P RES =[0.3P 预测max 0.7P 预测max ];

[0064] Among them, P RES P is the threshold value for photovoltaic / wind power output. 预测max For the maximum predicted output of new energy sources, when the real-time output is >0.7P. 预测max The time was determined to be a "peak period", with real-time output <0.3P. 预测max The time is the "low period"; the energy storage charge status threshold is set as follows: low power warning is <20%, high power adjustable zone is ∈[40%,80%], and full charge idle zone is >90%; the carbon price fluctuation threshold is set as follows: daily carbon price fluctuation rate <5% is defined as a low carbon price stable period, and daily carbon price fluctuation rate >10% is defined as a high carbon price sensitive period.

[0065] Please see Figure 2 In this embodiment of the invention, different scenarios can be divided according to the set adjustment resource status threshold and the parameters of each cluster in the clustering results.

[0066] Please see Figure 2 For example, if a cluster has an adjustment capacity index of 0.75, a carbon cost of 70% of the regional average, a spatiotemporal synergy of 0.68, and its current photovoltaic output reaches 85% of Pmax, then a low-carbon consumption scenario is triggered.

[0067] To eliminate uncertainty in boundary conditions, a fuzzy membership function is used for correction: First, the adjustment capability index membership is defined:

[0068]

[0069] Where, μ flex (x) represents the fuzzy membership degree of the regulation capability index x.

[0070] S4. Based on the scheduling optimization weight, set the objective function of the scheduling model, solve the objective function to obtain the optimal scheduling strategy, and control the target resource to execute the optimal scheduling strategy.

[0071] This invention generates a set of three-dimensional feature vectors based on multi-dimensional feature data of target resources. Based on the three-dimensional feature vectors, it determines the scenarios that meet the constraints of new energy consumption rate and achieves accurate classification of different scenarios, including low-carbon consumption scenarios with carbon costs lower than the regional average, flexible balance scenarios, and emergency support scenarios. It can also configure the scheduling optimization weights for each scenario based on historical data, so that the scheduling model can adapt to complex operating scenarios, thereby effectively improving the effect of resource collaborative scheduling.

[0072] In one embodiment, step S2, generating a three-dimensional feature vector set from multi-dimensional feature data, includes:

[0073] After normalizing the multi-dimensional feature data, it is mapped to a three-dimensional space to generate a three-dimensional feature vector.

[0074] In the embodiments of the invention, the sampling frequency of the above multi-source data is unified to a 15-minute granularity, and missing values ​​are filled by linear interpolation; resources are divided into regional nodes according to the power grid topology to ensure consistent data spatial ownership; data points exceeding the mean ± 3 times the standard deviation are removed based on the 3σ principle.

[0075] Normalization was performed using a combination of Min-Max linear normalization and Z-score standardization to eliminate dimensional differences. Power parameters (such as energy transfer capacity and maximum adjustable power) were normalized to the [0,1] interval using Min-Max normalization. Efficiency parameters (such as charge / discharge efficiency) retained their original percentage values. Carbon intensity coefficient and carbon price were standardized using Z-score. Geographic coordinates were converted to polar coordinates relative to the regional center. Time series data (such as power output curves) were extracted using Fourier transform to obtain the dominant frequency component as a feature.

[0076] In this embodiment of the invention, the expression for the three-dimensional feature vector set V is as follows:

[0077] V = {(A i C i ,S i )|i=1,2,…,M}

[0078] Among them, A i For the assessment of adjustment ability, C i S is a carbon cost influencing factor. i For spatiotemporal correlation, M represents the total number of flexible resources. The regulation capacity score can be calculated by weighting parameters such as power, efficiency, and response speed based on the regulation capacity dimension. Weights can be dynamically allocated according to resource type, such as emphasizing ramp rate for energy storage and elasticity index for load-side resources. Carbon cost impact factors can be generated by aggregating carbon intensity, carbon quota, and carbon price data based on the carbon attribute dimension. The spatiotemporal correlation matrix can be calculated by combining spatiotemporal characteristics, resource geographical location, and time distribution.

[0079] The embodiments of the present invention map multi-dimensional feature data into a three-dimensional space after normalization, which can effectively simplify the analysis process, reduce computational complexity, and at the same time retain the main features and structural information of the data.

[0080] In one embodiment, step S2, determining the scenarios to be classified based on the three-dimensional feature vector, which meet the new energy absorption rate constraint and have carbon costs lower than the regional average, includes:

[0081] S21. Wavelet packet transform technology is used to perform multi-scale analysis on the three-dimensional feature vectors to determine the high-frequency detail components.

[0082] In this embodiment of the invention, to address the characteristic noise caused by the fluctuation of new energy output and the randomness of load, wavelet packet transform can be used to perform multi-scale analysis on the three-dimensional feature vector set, including:

[0083] First, the Daubechies 6 (db6) wavelet basis function is selected to balance time-frequency localization characteristics and computational efficiency. Based on the characteristics of the new energy output cycle, a three-level decomposition is set (corresponding to 15-minute, 1-hour, and 4-hour time scales). Each feature dimension (regulation capability, carbon properties, and spatiotemporal characteristics) is decomposed independently to obtain the low-frequency approximate component (A3) and the high-frequency detail component (D1-D3).

[0084] S22. A threshold algorithm is used to perform nonlinear compression on the high-frequency detail components in the three-dimensional feature vector to obtain the denoised feature data.

[0085] In this embodiment of the invention, an improved SUREShrink threshold algorithm is used to filter out high-frequency noise while retaining key fluctuation features; wherein the SURE threshold λ j Calculation formula: Where σ j N represents the standard deviation of subband noise. j The number of data points in the sub-band; the threshold is dynamically adjusted based on the sensitivity of new energy consumption to obtain the adjusted threshold λ′. j Calculation formula: λ′j =λ j ×(1+α·absorption rate deviation), where λ j The SURE threshold is used to determine noise components, α is the sensitivity coefficient, and the attenuation rate deviation is the percentage difference between the current predicted value and the target value. Nonlinear compression is applied to high-frequency detail components: D′ j (t)=sign(D j (t))·(|D j (t)|-λ′ j ), of which (D j (t) and D′ j (t) represents the subband signal before and after denoising in time period t.

[0086] S23. Perform cluster analysis on the denoised feature data to obtain the final cluster set;

[0087] In this embodiment of the invention, each cluster in the final cluster set corresponds to a scenario.

[0088] S24. Determine the scenarios to be classified from the final cluster set that meet the new energy consumption rate constraint and whose carbon cost is lower than the regional average.

[0089] In this embodiment of the invention, the new energy consumption rate constraint η c for:

[0090]

[0091] Carbon cost constraint C CO2,c The expression is as follows:

[0092]

[0093] Among them, P RES,c P represents the actual power absorbed by new energy sources within the cluster. curtail,c P represents the power of wind and solar power curtailed within the cluster. load，c E represents the total load power within the region. i Let CI represent the carbon emissions of the i-th target resource. i Let λ be the carbon intensity coefficient of the i-th type of target resource. carbon For real-time carbon trading prices, Q quota E represents the total regional carbon allowance. total This represents the region's total actual carbon emissions. This serves as the upper limit for the region's carbon cost budget.

[0094] In this embodiment of the invention, feature enhancement can also be performed on the determined scene to be classified, which requires calculating the adjustment capability index: I flex =ω1·A avg +ω2·R maxWhere ω1=0.6 and ω2=0.4 correspond to the weights of average adjustment capability and maximum climbing rate, respectively, A avg For average regulating capacity, R max This represents the maximum gradient. The spatiotemporal coordination also needs to be calculated. Where d ij D represents the geographical distance between resources. trands The maximum effective distance for cross-regional transmission is given by n, where n is the total number of scenarios to be classified.

[0095] Enhanced features of the scene to be classified flex S syn It can be directly used as input data for subsequent scheduling models.

[0096] The embodiments of the present invention, through wavelet packet transform and threshold algorithm denoising, can more accurately identify and quantify the regulation capacity and spatiotemporal characteristics of new energy resources (such as wind energy and solar energy), which is conducive to improving the new energy consumption rate.

[0097] In one embodiment, step S23, performing cluster analysis on the denoised feature data to obtain the final cluster set, includes:

[0098] S23. Construct an adaptive resolution spatial grid based on the denoised feature data, wherein the grid density of the spatial grid is divided according to the feature value distribution;

[0099] In this embodiment of the invention, the expression for the mesh granularity is as follows:

[0100]

[0101] Where Δx is the grid granularity, A i To adjust the capability dimension value, N is the number of samples.

[0102] The embodiments of the present invention can further calculate the grid cell density:

[0103]

[0104] Where, ρ ijk Let C be the mesh cell density, Δx, Δy, and Δz be the mesh granularity in each dimension, and C be the granularity in each dimension. ijk This represents the number of samples within a grid unit.

[0105] S231. Determine the neighborhood radius of each sample in the spatial grid, and determine the neighborhood range of each sample based on the neighborhood radius;

[0106] In this embodiment of the invention, the neighborhood radius is the radius that contains at least 5 samples.

[0107] S232. Sort the samples in ascending order of reachability distance to generate a permutation list, where reachability distance is the shortest distance from one sample to another within a neighborhood.

[0108] S233. Determine the cluster boundaries based on the permutation list and density jump points to obtain the initial cluster set;

[0109] In this embodiment of the invention, cluster generation detection is defined as a density jump point, which is a position where the density of sample points changes significantly, for example, when the density difference of sample points is higher than a preset ratio.

[0110] S234. Merge adjacent clusters in the initial cluster set whose spatiotemporal feature dimension center distance is less than the cross-regional transmission delay threshold into a new cluster to obtain the final cluster set.

[0111] In this embodiment of the invention, the spatiotemporal feature dimension center distance refers to the distance between the two cluster centers in the spatiotemporal feature dimension, and the cross-regional transmission delay threshold refers to the maximum delay time that may exist in power transmission when transmitting power between different regions due to factors such as distance, transmission line capacity and transmission speed.

[0112] The embodiments of the present invention construct an adaptive resolution spatial grid, which can accurately reflect the distribution of different new energy resources in the feature space, helping to identify the similarities and differences between resources, thereby optimizing resource allocation and improving the utilization and absorption rate of new energy.

[0113] In one embodiment, step S3, configuring the scheduling optimization weight for each scenario based on historical data, includes:

[0114] S31. Extract historical optimal scheduling record data of the target resource, construct a training dataset based on the historical optimal scheduling record data, and construct a scheduling model based on the training dataset;

[0115] In this embodiment of the invention, historical optimal scheduling record data can be extracted from the EMS system, and the scheduling model can be constructed through supervised learning using this historical optimal scheduling record data as a training set.

[0116] S32. Input the input feature data into the scheduling model and output the scheduling optimization weight corresponding to each input feature data. The input feature data includes scenario category code, real-time carbon price volatility and adjustment resource availability. The scheduling optimization weight includes new energy consumption rate weight, carbon cost weight and adjustment cost weight.

[0117] In this embodiment of the invention, the weights of new energy absorption rate, carbon cost, and adjustment cost are the scheduling optimization weights corresponding to low-carbon absorption scenario, flexible balance scenario, and emergency support scenario, respectively.

[0118] In this embodiment of the invention, the scheduling model outputs the optimal weight combination (w1, w2, w3), where w1 is the weight of new energy absorption rate, w2 is the weight of carbon cost, and w3 is the weight of adjustment cost.

[0119] In this embodiment of the invention, by inputting the input feature data into the scheduling model, the corresponding scheduling optimization weights are output. Each scheduling optimization weight actually corresponds to a scenario, which can accurately determine the allocation of weights and help improve the accuracy of the model.

[0120] Please see Figure 3 In one embodiment, each scenario can also be weighted. For example, the total weight of a low-carbon consumption scenario can be divided into a new energy consumption rate weight, a carbon cost weight, and a regulation cost weight.

[0121] In this embodiment of the invention, the template weights can be adjusted based on the latest operating data, for example, by adjusting them according to carbon price sensitivity. When the real-time carbon price λ carbon Exceeding the monthly average At the same time, increase the weight of carbon costs.

[0122] In one embodiment, step S4, setting the objective function of the scheduling model based on the scheduling optimization weights, includes:

[0123] The objective function of the scheduling model is set based on the scheduling optimization weight, renewable energy absorption rate, total carbon emission cost, adjustment cost, actual renewable energy output, predicted renewable energy output, carbon emissions of the i-th type of target resource, carbon intensity coefficient of the i-th type of resource, real-time carbon trading price, total carbon emissions, and regional carbon emission quota.

[0124] In this embodiment of the invention, the expression of the objective function is as follows:

[0125]

[0126] Where w1, w2, and w3 are the weights for renewable energy consumption rate, carbon cost, and carbon cost, respectively, and η RES For the renewable energy consumption rate, C CO2 For the total cost of carbon emissions, C flex To adjust costs; P RES,actual and P RES,forecast For both actual and projected contributions to new energy sources; E i Let CI represent the carbon emissions of the i-th target resource. i Let λ be the carbon intensity coefficient of the i-th type of target resource. carbon For real-time carbon trading prices, E total Q represents the region's total actual carbon emissions. quota denoted as the total regional carbon quota, and n as the total number of scenarios to be classified.

[0127] In this embodiment of the invention, the constraints on the objective function are as follows:

[0128]

[0129] |P t -P t-1 |≤R ramp Δt

[0130]

[0131] P RES,forecast ·(1-∈)≤P RES,actual ≤P RES,forecast ·(1+∈)

[0132] Among them, SOC t+1 and SOC t η represents the state of charge of energy storage at time periods t and t+1, respectively. ch and η dis The energy storage charge / discharge efficiency, P, is given by time period t. ch,t and P dis,t These represent the charging / discharging power for time period t, where Δt is the time interval, and E is the power of the charge / discharge. rated P represents the rated capacity of the energy storage. t and P t-1 Power output for time periods t and t+1, respectively, R ramp E represents the maximum ramp rate of the equipment. total ,t represents the total carbon emissions in time period t, Q quota ∈ represents the total regional carbon quota; ∈ represents the prediction error tolerance.

[0133] In one embodiment, step S4, solving the objective function to obtain the optimal scheduling policy and controlling the target resource to execute the optimal scheduling policy, includes:

[0134] S41. The objective function is solved using a spatiotemporal hierarchical algorithm to obtain the optimal scheduling strategy;

[0135] S42. Generate corresponding scheduling instructions based on the optimal scheduling strategy, and control the target resources to perform corresponding actions according to the scheduling instructions.

[0136] In this embodiment of the invention, the spatiotemporal layering algorithm can decompose the scheduling problem into a day-ahead layer (24-hour granularity), an intraday layer (1-hour granularity), and a real-time layer according to the time scale. The day-ahead layer optimizes the inter-regional power transmission plan and energy storage charging and discharging baseline based on the wind and solar power output prediction curves, solves the global resource coordination problem, and outputs the power exchange benchmark value for each region. The intraday layer corrects the energy storage SOC trajectory and fast unit combination according to the latest prediction, allowing the inter-regional power baseline to be adjusted within ±20% to balance the prediction error. The real-time layer adjusts the interruptible load command and energy storage charging and discharging power for second-level fluctuations, and adds a ramp rate penalty term to suppress frequent adjustments.

[0137] In terms of spatial dimension, the spatiotemporal hierarchical algorithm divides the power grid into multiple autonomous regions. Each region independently solves its local optimization problem, and then uses the Lagrange multiplier method to coordinate boundary power deviations to ensure global optimality. Simultaneously, based on resource dynamic characteristics, the algorithm categorizes the regions into slow-dynamic and fast-dynamic groups. The slow-dynamic group participates in day-ahead optimization, focusing on cross-time period energy transfer; the fast-dynamic group also participates in day-ahead optimization, focusing on cross-time period energy transfer.

[0138] At the solver level, the daytime layer uses mixed integer programming (MIP) to ensure global optimality, while the real-time layer uses quadratic programming (QP) to ensure computational speed. By decoupling time and space, the complexity of high-dimensional problems is reduced, making it suitable for the fluctuating scenarios of new energy timeliness.

[0139] In this embodiment of the invention, the scheduling strategy generated by the spatiotemporal layering algorithm is a dynamic set of instructions that coordinates multiple time scales and spatial levels. For example, the day-ahead layer determines the baseline power exchange between regions and plans the charging and discharging periods of energy storage throughout the day; the intraday layer adjusts the start-up and shutdown status and output range of gas turbine units based on the latest forecasts, dynamically updates the charging and discharging power of energy storage, and allows a deviation of ±20% from the day-ahead plan; and the real-time layer issues fine-tuning of the charging and discharging power of energy storage.

[0140] Implementing the embodiments of the present invention has the following beneficial effects:

[0141] This invention generates a set of three-dimensional feature vectors based on multi-dimensional feature data of target resources. Based on the three-dimensional feature vectors, it determines the scenarios that meet the constraints of new energy consumption rate and achieves accurate classification of different scenarios, including low-carbon consumption scenarios with carbon costs lower than the regional average, flexible balance scenarios, and emergency support scenarios. It can also configure the scheduling optimization weights for each scenario based on historical data, so that the scheduling model can adapt to complex operating scenarios, thereby effectively improving the effect of resource collaborative scheduling.

[0142] Furthermore, by constructing a spatial grid with adaptive resolution, the embodiments of the present invention can accurately reflect the distribution of different new energy resources in the feature space, which helps to identify the similarities and differences between resources, thereby optimizing resource allocation and improving the utilization and absorption rate of new energy.

[0143] like Figure 4 As shown, based on the above method embodiments, corresponding apparatus embodiments are provided;

[0144] An embodiment of the present invention provides a new energy resource collaborative scheduling device, comprising:

[0145] The feature data acquisition module 10 is used to acquire multi-dimensional feature data of target resources in the new energy system. The multi-dimensional feature data includes regulation characteristic data, carbon constraint data and spatiotemporal distribution data.

[0146] The scenario determination module 20 is used to generate a set of three-dimensional feature vectors from multi-dimensional feature data, and to determine the scenarios to be classified based on the three-dimensional feature vectors that meet the constraints of new energy consumption rate and have carbon costs lower than the regional average.

[0147] The scheduling optimization weight configuration module 30 is used to classify the scenarios to be classified into low-carbon consumption scenarios, flexible balance scenarios and emergency support scenarios, and to configure the scheduling optimization weight of each scenario based on historical data.

[0148] The scheduling optimization module 40 is used to set the objective function of the scheduling model based on the scheduling optimization weight, solve the objective function to obtain the optimal scheduling strategy, and control the target resource to execute the optimal scheduling strategy.

[0149] In one embodiment, generating a set of three-dimensional feature vectors from multi-dimensional feature data includes:

[0150] After normalizing the multi-dimensional feature data, it is mapped to a three-dimensional space to generate a three-dimensional feature vector.

[0151] In one embodiment, scenarios to be classified that meet the new energy absorption rate constraint and have carbon costs lower than the regional average are determined based on three-dimensional feature vectors, including:

[0152] Wavelet packet transform technology is used to perform multi-scale analysis on three-dimensional feature vectors to determine high-frequency detail components;

[0153] A thresholding algorithm is used to nonlinearly compress the high-frequency detail components in the three-dimensional feature vector to obtain denoised feature data;

[0154] Cluster analysis is performed on the denoised feature data to obtain the final cluster set;

[0155] From the final cluster set, identify the scenarios to be classified that meet the constraints of new energy consumption rate and have carbon costs lower than the regional average.

[0156] In one embodiment, cluster analysis is performed on the denoised feature data to obtain a final cluster set, including:

[0157] An adaptive resolution spatial grid is constructed based on the denoised feature data, wherein the grid density of the spatial grid is divided according to the feature value distribution.

[0158] Determine the neighborhood radius of each sample in the spatial grid, and determine the neighborhood range of each sample based on the neighborhood radius;

[0159] The samples are sorted in ascending order of reachability to generate a permutation list, where reachability is the shortest distance between one sample and another within a neighborhood.

[0160] The boundaries of the clusters are determined based on the permutation list and density jump points, resulting in the initial cluster set;

[0161] The adjacent clusters in the initial cluster set whose spatiotemporal feature dimension center distance is less than the cross-regional transmission delay threshold are merged into a new cluster to obtain the final cluster set.

[0162] In one embodiment, the scheduling optimization weights for each scenario are configured based on historical data, including:

[0163] Extract historical best scheduling records of the target resource, construct a training dataset based on the historical best scheduling records, and construct a scheduling model based on the training dataset.

[0164] The input feature data is input into the scheduling model, and the scheduling optimization weight corresponding to each input feature data is output. The input feature data includes scenario category coding, real-time carbon price volatility and adjustment resource availability. The scheduling optimization weight includes new energy consumption rate weight, carbon cost weight and carbon cost weight.

[0165] In one embodiment, the objective function of the scheduling model is set based on the scheduling optimization weights, including:

[0166] The objective function of the scheduling model is set based on the scheduling optimization weight, renewable energy absorption rate, total carbon emission cost, adjustment cost, actual renewable energy output, predicted renewable energy output, carbon emissions of the i-th type of target resource, carbon intensity coefficient of the i-th type of resource, real-time carbon trading price, total carbon emissions, and regional carbon emission quota.

[0167] In one embodiment, solving the objective function to obtain the optimal scheduling policy and controlling the target resource to execute the optimal scheduling policy includes:

[0168] The objective function is solved using a spatiotemporal layering algorithm to obtain the optimal scheduling strategy;

[0169] Generate corresponding scheduling instructions based on the optimal scheduling strategy, and control the target resources to perform corresponding actions based on the scheduling instructions.

[0170] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the new energy resource collaborative scheduling method provided by any of the above-described method embodiments of the present invention.

[0171] 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.

[0172] Based on the above embodiments of the new energy resource collaborative scheduling 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 new energy resource collaborative scheduling method of any embodiment of the present invention.

[0173] For example, in this embodiment, the computer program can be divided into one or more modules, one or more modules are stored in memory and executed by a processor to complete the present invention. One or more module elements can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a terminal device.

[0174] Terminal devices can be computing devices such as desktop computers, laptops, handheld computers, and cloud servers. Terminal devices may include, but are not limited to, processors and memory.

[0175] 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 through various interfaces and lines.

[0176] 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 new energy resource collaborative scheduling method described in any of the above-described method embodiments of the present invention.

[0177] 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, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0178] 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 collaborative scheduling of new energy resources, characterized in that, include: Collect multi-dimensional feature data of target resources in the new energy system, wherein the multi-dimensional feature data includes regulation characteristic data, carbon constraint data and spatiotemporal distribution data; A three-dimensional feature vector set is generated from the multi-dimensional feature data, and the scenarios to be classified are determined based on the three-dimensional feature vectors, which meet the constraints of new energy consumption rate and have carbon costs lower than the regional average. The scenarios to be classified are divided into low-carbon consumption scenarios, flexible balance scenarios, and emergency support scenarios, and the scheduling optimization weight of each scenario is configured according to historical data. The objective function of the scheduling model is set based on the scheduling optimization weights, the objective function is solved to obtain the optimal scheduling strategy, and the target resource is controlled to execute the optimal scheduling strategy.

2. The new energy resource collaborative scheduling method as described in claim 1, characterized in that, The step of generating a three-dimensional feature vector set from the multi-dimensional feature data includes: The multi-dimensional feature data is normalized and then mapped to a three-dimensional space to generate a three-dimensional feature vector.

3. The new energy resource collaborative scheduling method as described in claim 1, characterized in that, The process of determining the scenarios to be classified based on the three-dimensional feature vectors that meet the constraints of new energy absorption rate and have carbon costs lower than the regional average includes: Wavelet packet transform technology is used to perform multi-scale analysis on the three-dimensional feature vector to determine high-frequency detail components; A thresholding algorithm is used to nonlinearly compress the high-frequency detail components in the three-dimensional feature vector to obtain denoised feature data. Cluster analysis is performed on the denoised feature data to obtain the final cluster set; From the final cluster set, identify the scenarios to be classified that meet the new energy consumption rate constraint and have carbon costs lower than the regional average.

4. The new energy resource collaborative scheduling method as described in claim 3, characterized in that, The clustering analysis performed on the denoised feature data yields a final cluster set, including: An adaptive resolution spatial grid is constructed based on the denoised feature data, wherein the grid density of the spatial grid is divided according to the feature value distribution; Determine the neighborhood radius of each sample in the spatial grid, and determine the neighborhood range of each sample based on the neighborhood radius; The samples are sorted in ascending order of reachability to generate a permutation list, wherein the reachability is the shortest distance from one sample to another within a neighborhood. The boundaries of the clusters are determined based on the permutation list and density jump points, resulting in an initial cluster set; The adjacent clusters in the initial cluster set whose spatiotemporal feature dimension center distance is less than the cross-regional transmission delay threshold are merged into a new cluster to obtain the final cluster set.

5. The new energy resource collaborative scheduling method as described in claim 1, characterized in that, The configuration of scheduling optimization weights for each scenario based on historical data includes: Extract the historical best scheduling record data of the target resource, construct a training dataset based on the historical best scheduling record data, and construct a scheduling model based on the training dataset; The input feature data is input into the scheduling model, and the scheduling optimization weight corresponding to each input feature data is output. The input feature data includes scenario category code, real-time carbon price volatility and adjustment resource availability. The scheduling optimization weight includes new energy consumption rate weight, carbon cost weight and carbon cost weight.

6. The new energy resource collaborative scheduling method as described in claim 1, characterized in that, The objective function of setting the scheduling model based on the scheduling optimization weights includes: Based on the aforementioned scheduling optimization weights, renewable energy absorption rate, total carbon emission cost, adjustment cost, actual renewable energy output, predicted renewable energy output, carbon emissions of the i-th type of target resource, carbon intensity coefficient of the i-th type of resource, real-time carbon trading price, total carbon emissions, and regional carbon emission quota, the objective function of the scheduling model is set.

7. The new energy resource collaborative scheduling method as described in claim 1, characterized in that, The process of solving the objective function to obtain the optimal scheduling strategy and controlling the target resource to execute the optimal scheduling strategy includes: The objective function is solved using a spatiotemporal hierarchical algorithm to obtain the optimal scheduling strategy. The corresponding scheduling instructions are generated according to the optimal scheduling strategy, and the target resource is controlled to perform corresponding actions according to the scheduling instructions.

8. A new energy resource collaborative scheduling device, characterized in that, include: The feature data acquisition module is used to acquire multi-dimensional feature data of target resources in the new energy system, wherein the multi-dimensional feature data includes regulation characteristic data, carbon constraint data and spatiotemporal distribution data; The module for determining the scenario to be classified is used to generate a set of three-dimensional feature vectors from the multi-dimensional feature data, and to determine the scenarios to be classified based on the three-dimensional feature vectors that meet the constraints of new energy consumption rate and have carbon costs lower than the regional average. The scheduling optimization weight configuration module is used to classify the scenarios to be classified into low-carbon consumption scenarios, flexible balance scenarios, and emergency support scenarios, and to configure the scheduling optimization weight of each scenario based on historical data. The scheduling optimization module is used to set the objective function of the scheduling model based on the scheduling optimization weights, solve the objective function to obtain the optimal scheduling strategy, and control the target resource to execute the optimal scheduling strategy.

9. A terminal device, characterized in that, The method 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 new energy resource collaborative scheduling method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, include: A stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the new energy resource collaborative scheduling method as described in any one of claims 1-7.

Images