Industrial flexible load day-ahead scheduling potential equivalence method, system, device and medium

Abstract

The application discloses an industrial flexible load day-ahead scheduling potential equivalence method, system, device and medium, comprising the following steps: acquiring interaction data of an industrial flexible load and a power grid, calculating parameters of a data-driven industrial flexible load equivalence scheduling model according to the interaction data; embedding the data-driven industrial flexible load equivalence scheduling model into a power grid day-ahead scheduling model; solving the power grid day-ahead scheduling model embedded with the data-driven industrial flexible load equivalence scheduling model to obtain an industrial flexible load scheduling plan, scheduling the industrial flexible load according to the industrial flexible load scheduling plan, and completing industrial flexible load day-ahead scheduling potential equivalence. The method, system, device and medium can accurately schedule the industrial flexible load, improve the flexible regulation capacity of the power grid, and improve the accommodation capacity of new energy.

Description

Technical Field

[0001] This invention belongs to the field of power system optimization and dispatching, and relates to a method, system, equipment and medium for valuing the day-ahead dispatching potential of flexible industrial loads. Background Technology

[0002] Day-ahead dispatching of power systems is a crucial link in maximizing energy efficiency and production benefits. Industrial flexible loads, due to their large capacity and flexible adjustment capabilities, are playing an increasingly important role in day-ahead dispatching. Traditional industrial flexible load dispatching methods often rely on user reports or manual experience to determine the load's adjustable range, lacking an accurate perception of its actual adjustable capacity. Therefore, they cannot accurately dispatch industrial flexible loads, resulting in low grid flexibility and limited capacity for absorbing renewable energy. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, system, device and medium for valuing the day-ahead scheduling potential of industrial flexible loads. This method, system, device and medium can accurately schedule industrial flexible loads, improve the power grid's flexible regulation capability and its ability to absorb new energy sources.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] In one aspect, the present invention provides a method for equivalence of industrial flexible load day-ahead scheduling potential, comprising:

[0006] Acquire interaction data between industrial flexible loads and the power grid, and calculate the parameters of a data-driven industrial flexible load equivalent scheduling model based on the interaction data;

[0007] The data-driven industrial flexible load equivalent scheduling model is embedded into the day-ahead power grid scheduling model;

[0008] Solve the day-ahead dispatch model of the power grid embedded with the data-driven industrial flexible load equivalent dispatch model to obtain the industrial flexible load dispatch plan. Perform industrial flexible load dispatch according to the industrial flexible load dispatch plan to complete the day-ahead dispatch potential equivalence of industrial flexible load.

[0009] The further improvement of the industrial flexible load day-ahead scheduling potential equivalence method described in this invention lies in:

[0010] The interaction data between the industrial flexible load and the power grid includes the declared volume and price information of the industrial flexible load in the day-ahead ancillary services market, as well as the industrial flexible load production plan.

[0011] The process of acquiring the interaction data between industrial flexible loads and the power grid, and calculating the parameters of the data-driven industrial flexible load equivalent scheduling model based on the interaction data, is as follows:

[0012] Acquire interaction data between industrial flexible loads and the power grid;

[0013] The interactive data is preprocessed;

[0014] Construct a set of input information based on the preprocessed interaction data;

[0015] Randomly generate industrial flexible load scheduling instructions, input the industrial flexible load scheduling instructions into the industrial flexible load production process optimization model to verify whether the industrial flexible load scheduling instructions can be executed, and use the verification results of different industrial flexible load scheduling instructions as labels to construct an equivalent scheduling label data set.

[0016] Construct an industrial flexible load day-ahead equivalent scheduling dataset based on the input information set and the equivalent scheduling label data set;

[0017] Based on the day-ahead equivalent scheduling dataset of industrial flexible loads, calculate the parameters of the data-driven industrial flexible load equivalent scheduling model.

[0018] The specific process of calculating the parameters of the data-driven industrial flexible load equivalent scheduling model based on the day-ahead equivalent scheduling dataset of industrial flexible load is as follows:

[0019] Based on the day-ahead equivalent scheduling dataset of industrial flexible loads, the parameters of the data-driven industrial flexible load equivalent scheduling model are calculated using an equivalent neural network.

[0020] The constraints of the day-ahead optimization dispatch model of the power grid include power balance constraints, reserve capacity constraints, and cross-sectional capacity constraints.

[0021] The constraints of the current power grid optimization scheduling model are:

[0022]

[0023]

[0024]

[0025]

[0026]

[0027] in, To select the z-th adjustment scheme for the k-th industrial flexible load, and These are the actual power consumption and the baseline power consumption of the k-th industrial flexible load in the t-th time period, respectively. and These are the parameters for the industrial flexible load equivalent scheduling model.

[0028] The day-ahead scheduling model of the power grid, which is embedded with a data-driven industrial flexible load equivalent scheduling model, is solved by a solver.

[0029] In a second aspect, the present invention provides an equivalent system for industrial flexible load day-ahead scheduling potential, comprising:

[0030] The first calculation module is used to acquire the interaction data between industrial flexible loads and the power grid, and calculate the parameters of the data-driven industrial flexible load equivalent scheduling model based on the interaction data.

[0031] An embedding module is used to embed the data-driven industrial flexible load equivalent scheduling model into the day-ahead power grid scheduling model;

[0032] The execution module is used to solve the day-ahead dispatch model of the power grid embedded with the data-driven industrial flexible load equivalent dispatch model, obtain the industrial flexible load dispatch plan, perform industrial flexible load dispatch according to the industrial flexible load dispatch plan, and complete the day-ahead dispatch potential equivalent of industrial flexible load.

[0033] In three aspects, the present invention provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the industrial flexible load day-ahead scheduling potential equivalence method.

[0034] In four aspects, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method for equivalence of the day-ahead scheduling potential of industrial flexible loads.

[0035] The present invention has the following beneficial effects:

[0036] In practical operation, the method, system, equipment, and medium for valuing the potential of day-ahead scheduling of industrial flexible loads described in this invention calculate the parameters of a data-driven equivalent scheduling model of industrial flexible loads based on the interaction data between the industrial flexible loads and the power grid. This fully leverages the interaction data between the industrial flexible loads and the power grid to solve the day-ahead scheduling model of the power grid embedded with the data-driven equivalent scheduling model of industrial flexible loads. By comprehensively considering data analysis, machine learning, and optimization techniques, it achieves refined management and scheduling of loads, thereby improving the power grid's flexible adjustment capability and its ability to absorb new energy sources. Attached Figure Description

[0037] The accompanying drawings, which form part of this specification, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0038] Figure 1 This is a flowchart of the method of the present invention;

[0039] Figure 2 Example diagram of an equivalent model for flexible industrial loads;

[0040] Figure 3 A graph showing the response cost-response power relationship for an equivalent model of flexible industrial loads;

[0041] Figure 4 This is a system structure diagram of the present invention. Detailed Implementation

[0042] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0043] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0044] The present invention will now be described in further detail with reference to the accompanying drawings:

[0045] Example 1

[0046] refer to Figure 1 The method for equivalence of day-ahead scheduling potential for industrial flexible loads as described in this invention includes the following steps:

[0047] 1) Considering the complementary characteristics of flexible industrial loads, aggregate flexible loads for small and medium-sized industries;

[0048] The specific operation of step 1) is as follows:

[0049] 11) Collect load data from smart meters, sensors and monitoring equipment in industrial enterprises in real time, clean and process the collected load data to ensure the accuracy and availability of the data, and integrate and standardize the load data from different enterprises for subsequent analysis and aggregation.

[0050] 12) For the characteristics of flexible industrial load regulation, such as regulation capacity, regulation delay, and regulation cost, the Max-min normalization method is used to normalize them. A larger normalized result indicates better performance of the flexible industrial load in this aspect. The normalization formula is:

[0051]

[0052] q qi For the normalized industrial flexible load characteristics, q i For the original industrial flexible load characteristics, min 1≤j≤n {q i} represents the lower limit of the flexible load characteristics of industry, max 1≤j≤n {q i} represents the upper limit of the industrial flexible load characteristics.

[0053] 13) For the normalized industrial flexible load adjustment characteristics, quantify them into Boolean values, where 1 indicates the presence of a certain characteristic and 0 indicates the absence of a certain characteristic. Then, based on the characteristic values ​​of each industrial flexible load, construct the industrial flexible load complementary coordination matrix Q. k for:

[0054]

[0055] Among them, the industrial flexible load complementary coordination matrix Q k Each column in the matrix corresponds to an industrial flexible load, such as energy storage, electrolytic aluminum, and industrial furnaces. The industrial flexible load complementarity coordination matrix Q... k Each row in the table represents the quantified, multi-dimensional adjustment characteristics of the industrial flexible load, such as adjustment capacity, adjustment delay, and adjustment cost.

[0056] 14) Based on the flexible load complementarity coordination matrix Q of each industry k To aggregate flexible industrial loads, machine learning algorithms such as K-means clustering and nearest neighbor algorithm can be used.

[0057] 2) Obtain interaction data between industrial flexible loads and the power grid, and construct a day-ahead equivalent scheduling dataset for industrial flexible loads;

[0058] The specific operation of step 2) is as follows:

[0059] 21) Collect interaction data between industrial flexible loads and the power grid through power companies, energy trading markets or related institutions. The interaction data includes the declared volume and declared price information of industrial flexible loads in the day-ahead ancillary service (peak shaving) market, and the production plan of industrial flexible loads.

[0060] 22) Preprocess the interactive data, including data cleaning, outlier removal, and missing data filling.

[0061] 23) Based on the preprocessed interactive data, construct the input information set, in which the interactive data is integrated and divided according to the preset time interval and time window. The appropriate time granularity, such as hourly, daily or weekly, can be selected according to the needs and actual situation.

[0062] 24) Considering the load demand of the power grid, the price signals of the energy market, and the flexibility of industrial loads, different flexible industrial load dispatching instructions are randomly generated. The flexible industrial load dispatching instructions are input into the flexible industrial load production process optimization model to verify whether the flexible industrial load dispatching instructions can be executed. The verification results are used as labels to construct an equivalent dispatching label data set. Based on the equivalent dispatching label data set and the input information set, an industrial flexible load day-ahead equivalent dispatching dataset is constructed.

[0063] 3) Establish a data-driven industrial flexible load equivalent scheduling model and determine the parameters of the data-driven industrial flexible load equivalent scheduling model.

[0064] The specific operation of step 3) is as follows:

[0065] 31) Considering day-ahead scheduling operations, interactive data structures, and production regulation characteristics, establish an equivalent scheduling model for flexible industrial loads, similar to generating units or energy storage systems:

[0066]

[0067] in, M represents the load power of the industrial flexible load k. k For the corresponding coefficient matrix, N k These are the reported parameters to be determined for the equivalent model, corresponding to power, ramp, and energy boundaries.

[0068] 32) Based on the equivalent scheduling tag data set, generate an initial equivalent power set containing all executable industrial flexible load scheduling instructions, and use a neural network to predict the adjustment direction and distance of the power set in order to eliminate power points in the power set that cannot be scheduled. The neural network is an estimating neural network.

[0069]

[0070] Among them, Ω k Let be the power set, β be the scaling factor, and z be the displacement factor.

[0071] 33) Based on the adjusted distance of the power set, an offline neural network model is trained to determine the parameters of the data-driven industrial flexible load equivalent scheduling model. The trained neural network model is used as the equivalent neural network. Specifically, the day-ahead equivalent scheduling dataset of industrial flexible load obtained in step 2) is input, and the regularization term of the loss function of the equivalent neural network is adjusted according to the distance output set of the estimation neural network. Through stochastic gradient descent, the weight parameters of the equivalent neural network are continuously adjusted, and the output is the parameters of the optimized equivalent model.

[0072]

[0073] Where, θ t Here are the network weight parameters to be updated, where δ is the learning rate and ε is the smoothing coefficient. and is the mean of the first and second moments of the gradient.

[0074] 34) Use the PyTorch deep learning library to build and train a deep learning model. Through multiple rounds of training, obtain converged results. Divide the input information set in step 2) into a test set and substitute it into the equivalent neural network. Perform inverse normalization on the output of the equivalent neural network to obtain the parameters of the data-driven industrial flexible load equivalent scheduling model.

[0075] 4) Establish and solve the day-ahead dispatch model of the power grid that incorporates the data-driven industrial flexible load equivalent dispatch model.

[0076] The specific process of step 4) is as follows:

[0077] 41) Construct a day-ahead dispatch model for the power grid that includes power balance constraints, reserve capacity constraints, and cross-sectional capacity constraints.

[0078] 42) Embed the data-driven industrial flexible load equivalent scheduling model into the day-ahead power grid scheduling model. The constraints of this day-ahead power grid scheduling model are:

[0079]

[0080]

[0081]

[0082]

[0083]

[0084] in, This indicates that the k-th industrial flexible load selects the z-th adjustment scheme. These represent the actual / baseline power consumption of the k-th industrial flexible load in the t-th time period, respectively. These are the parameters for the industrial flexible load equivalent model.

[0085] 44) Use a solver to solve the day-ahead dispatch model of the power grid embedded with the data-driven industrial flexible load equivalent dispatch model to obtain the industrial flexible load dispatch plan, and perform industrial flexible load dispatch according to the industrial flexible load dispatch plan. The commercial solver can be cplex or gurobi.

[0086] It should be noted that this invention, by analyzing and utilizing the interaction data between industrial flexible loads and the power grid, aggregates industrial flexible loads under the regional power grid into dispatchable objects similar to generating units or energy storage for use by the upper-level power grid. This allows for more accurate prediction and assessment of the potential and flexibility of industrial loads. Furthermore, this invention combines data analysis, machine learning, and optimization techniques to establish an equivalent dispatch model for industrial flexible loads, achieving refined management and dispatch of loads. Simultaneously, it considers the actual conditions and energy supply strategies within the industrial flexible loads, ensuring that dispatch instructions issued by the upper-level power grid can be effectively executed by industrial users.

[0087] Example 2

[0088] In a provincial power grid, the aggregated industrial flexible load includes two types of self-provided generators with rated capacities of 350MW and 125MW, respectively. The maximum loads of electric arc furnaces and electrolytic cells (electrolytic aluminum) are 20.72MW and 735MW, respectively. Considering this invention, its equivalent dispatch potential model is as follows: Figure 2 As shown, the response cost-response power relationship of the equivalent model is as follows: Figure 3 As shown.

[0089] As can be seen, compared with the method of directly superimposing equipment parameters to equalize scheduling potential, this invention can more accurately describe the adjustment capability boundary of industrial flexible loads, ensuring that the power grid does not issue scheduling instructions that industrial flexible loads cannot execute. Figure 3 It can be seen that the equivalent model can achieve equal response power within a certain cost range, thereby improving the grid's flexible adjustment capability and its ability to absorb new energy sources while ensuring its own benefits.

[0090] Example 3

[0091] refer to Figure 4 The industrial flexible load day-ahead scheduling potential equivalent system of the present invention includes:

[0092] The first calculation module is used to acquire the interaction data between industrial flexible loads and the power grid, and calculate the parameters of the data-driven industrial flexible load equivalent scheduling model based on the interaction data.

[0093] An embedding module is used to embed the data-driven industrial flexible load equivalent scheduling model into the day-ahead power grid scheduling model;

[0094] The execution module is used to solve the day-ahead dispatch model of the power grid embedded with the data-driven industrial flexible load equivalent dispatch model, obtain the industrial flexible load dispatch plan, perform industrial flexible load dispatch according to the industrial flexible load dispatch plan, and complete the day-ahead dispatch potential equivalent of industrial flexible load.

[0095] In this embodiment, the calculation module includes:

[0096] The acquisition module is used to acquire interaction data between industrial flexible loads and the power grid;

[0097] The preprocessing module is used to preprocess the interactive data;

[0098] A module is created to construct a set of input information based on the preprocessed interactive data.

[0099] The first construction module is used to randomly generate industrial flexible load scheduling instructions, input the industrial flexible load scheduling instructions into the industrial flexible load production process optimization model to verify whether the industrial flexible load scheduling instructions can be executed, and use the verification results of different industrial flexible load scheduling instructions as labels to construct an equivalent scheduling label data set.

[0100] The second construction module is used to construct an industrial flexible load day-ahead equivalent scheduling dataset based on the input information set and the equivalent scheduling label data set.

[0101] The second calculation module is used to calculate the parameters of the data-driven industrial flexible load equivalent scheduling model based on the day-ahead equivalent scheduling dataset of the industrial flexible load.

[0102] Example 4

[0103] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of a method for equivalence of industrial flexible load day-ahead scheduling potential. The memory may include main memory, such as high-speed random access memory (RAM), and may also include non-volatile memory, such as at least one disk storage device. The processor, network interface, and memory are interconnected via an internal bus, which may be an industry-standard architecture bus, a peripheral component interconnection standard bus, an extended industry-standard architecture bus, etc. The bus may be categorized as an address bus, a data bus, a control bus, etc. The memory stores the program; specifically, the program may include program code, which includes computer operation instructions. The memory may include main memory and non-volatile memory, and provides instructions and data to the processor.

[0104] Example 5

[0105] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the industrial flexible load day-ahead scheduling potential equivalence method. Specifically, the computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. The volatile memory may include random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include read-only memory (ROM), hard disk, flash memory, optical disk, magnetic disk, etc.

[0106] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0107] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0108] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0109] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for valuing the potential of flexible day-ahead load scheduling in industry, characterized in that, include: Acquire interaction data between industrial flexible loads and the power grid; The interactive data is preprocessed; Construct a set of input information based on the preprocessed interaction data; Randomly generate industrial flexible load scheduling instructions, input the industrial flexible load scheduling instructions into the industrial flexible load production process optimization model to verify whether the industrial flexible load scheduling instructions can be executed, and use the verification results of different industrial flexible load scheduling instructions as labels to construct an equivalent scheduling label data set. Construct an industrial flexible load day-ahead equivalent scheduling dataset based on the input information set and the equivalent scheduling label data set; Based on the day-ahead equivalent scheduling dataset of industrial flexible loads, calculate the parameters of the data-driven industrial flexible load equivalent scheduling model; The data-driven industrial flexible load equivalent scheduling model is embedded into the day-ahead power grid scheduling model; Solve the day-ahead dispatch model of the power grid with an embedded data-driven industrial flexible load equivalent dispatch model to obtain the industrial flexible load dispatch plan. Perform industrial flexible load dispatch according to the industrial flexible load dispatch plan to complete the day-ahead dispatch potential equivalence of industrial flexible load. Considering day-ahead scheduling operations, interactive data structures, and production regulation characteristics, establish an equivalent scheduling model for flexible industrial loads, similar to generating units or energy storage systems: in, For the load power of the industrial flexible load k, For the corresponding coefficient matrix, These are the reported parameters to be determined for the equivalent model, corresponding to power, ramp, and energy boundaries; Based on the equivalent scheduling tag data set, an initial equivalent power set containing all executable industrial flexible load scheduling instructions is generated. A neural network is used to predict the adjustment direction and distance of the power set in order to eliminate power points in the power set that cannot be scheduled. The neural network is an estimating neural network. in, For power collection, This is the scaling factor. It is the displacement coefficient; Based on the distance adjustment of the power set, an offline neural network model is trained to determine the parameters of the data-driven industrial flexible load equivalent scheduling model. This trained neural network model is used as an equivalent neural network, input into the day-ahead equivalent scheduling dataset of industrial flexible loads. The regularization term of the loss function of the equivalent neural network is adjusted according to the distance output set of the estimating neural network. Through stochastic gradient descent, the weight parameters of the equivalent neural network are continuously adjusted, and the output is the parameters of the optimized equivalent model. in, For the network weight parameters to be updated, For learning rate, For smoothing coefficients, and The first and second moments of the gradient are the mean values. The PyTorch deep learning library is used to build and train a deep learning model. Through multiple rounds of training, convergence results are obtained. The input information set is divided into a test set and substituted into an equivalent neural network. The output of the equivalent neural network is denormalized to obtain the parameters of the data-driven industrial flexible load equivalent scheduling model.

2. The method for valuing the potential of day-ahead flexible load scheduling in industry according to claim 1, characterized in that, The interaction data between the industrial flexible load and the power grid includes the declared volume and price information of the industrial flexible load in the day-ahead ancillary services market, as well as the industrial flexible load production plan.

3. The method for valuing the potential of day-ahead flexible load scheduling in industry according to claim 1, characterized in that, The constraints of the day-ahead dispatch model of the power grid include power balance constraints, reserve capacity constraints, and cross-sectional capacity constraints.

4. The method for valuing the potential for day-ahead flexible load scheduling in industry according to claim 3, characterized in that, The constraints of the day-ahead dispatch model for the power grid are: in, For the first The first industrial flexible load option Various adjustment schemes, The first The industrial flexible load in the first Actual power consumption and baseline power consumption for each time period These are the parameters for the industrial flexible load equivalent scheduling model.

5. The method for valuing the potential for day-ahead flexible load scheduling in industry according to claim 1, characterized in that, The day-ahead dispatch model of the power grid is solved by a solver that incorporates a data-driven industrial flexible load equivalent dispatch model.

6. An equivalent system for the potential of flexible day-ahead load scheduling in industry, characterized in that, include: The first computing module is used to acquire interaction data between industrial flexible loads and the power grid; The interactive data is preprocessed; Construct a set of input information based on the preprocessed interaction data; Randomly generated industrial flexible load scheduling instructions are input into an industrial flexible load production process optimization model to verify whether the instructions can be executed. The verification results of different instructions are used as labels to construct an equivalent scheduling label dataset. An industrial flexible load day-ahead equivalent scheduling dataset is constructed based on the input information set and the equivalent scheduling label dataset. The parameters of the data-driven industrial flexible load equivalent scheduling model are calculated based on the dataset. An embedding module is used to embed the data-driven industrial flexible load equivalent scheduling model into the day-ahead power grid scheduling model; The execution module is used to solve the day-ahead dispatch model of the power grid embedded with the data-driven industrial flexible load equivalent dispatch model, obtain the industrial flexible load dispatch plan, perform industrial flexible load dispatch according to the industrial flexible load dispatch plan, and complete the day-ahead dispatch potential equivalence of industrial flexible load. Considering day-ahead scheduling operations, interactive data structures, and production regulation characteristics, establish an equivalent scheduling model for flexible industrial loads, similar to generating units or energy storage systems: in, For the load power of the industrial flexible load k, For the corresponding coefficient matrix, These are the reported parameters to be determined for the equivalent model, corresponding to power, ramp, and energy boundaries; Based on the equivalent scheduling tag data set, an initial equivalent power set containing all executable industrial flexible load scheduling instructions is generated. A neural network is used to predict the adjustment direction and distance of the power set in order to eliminate power points in the power set that cannot be scheduled. The neural network is an estimating neural network. in, For power collection, This is the scaling factor. It is the displacement coefficient; Based on the distance adjustment of the power set, an offline neural network model is trained to determine the parameters of the data-driven industrial flexible load equivalent scheduling model. This trained neural network model is used as an equivalent neural network, input into the day-ahead equivalent scheduling dataset of industrial flexible loads. The regularization term of the loss function of the equivalent neural network is adjusted according to the distance output set of the estimating neural network. Through stochastic gradient descent, the weight parameters of the equivalent neural network are continuously adjusted, and the output is the parameters of the optimized equivalent model. in, For the network weight parameters to be updated, For learning rate, For smoothing coefficients, and The first and second moments of the gradient are the mean values. The PyTorch deep learning library is used to build and train a deep learning model. Through multiple rounds of training, convergence results are obtained. The input information set is divided into a test set and substituted into an equivalent neural network. The output of the equivalent neural network is denormalized to obtain the parameters of the data-driven industrial flexible load equivalent scheduling model.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the industrial flexible load day-ahead scheduling potential equivalence method as described in any one of claims 1-5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the industrial flexible load day-ahead scheduling potential equivalence method as described in any one of claims 1-5.

Images