Double-layer multi-time-scale rolling optimization method and system for electricity-hydrogen hybrid energy storage system

Abstract

The invention provides a double-layer multi-time-scale rolling optimization method and system for an electricity-hydrogen hybrid energy storage system, and relates to the field of optimization scheduling of the electricity-hydrogen hybrid energy storage system, and the method comprises the steps: constructing a wind-solar prediction error fuzzy set; constructing an upper-layer long-time-scale energy scheduling model, and solving to obtain long-time-scale hydrogen storage capacity and electric quantity unbalance; under the generated typical scene, constructing a short time scale power scheduling model of a lower layer, and solving to obtain the hydrogen storage capacity of the short time scale; performing rolling optimization by taking the hydrogen storage capacity of the short time scale obtained by the lower layer solution as the initial hydrogen storage capacity of the long time scale energy scheduling model to realize double-layer multi-time scale rolling optimization; according to the method, the influence of different time scales and wind and light uncertainty on optimal scheduling of the electricity-hydrogen hybrid energy storage system is considered, a double-layer scheduling model is constructed, double-layer multi-time scale rolling optimization is carried out, and a scheduling scheme considering economy and robustness is provided for weekly and day-ahead flexible scheduling of the electricity-hydrogen hybrid energy storage system.

Description

Technical Field

[0001] The present disclosure relates to the field of optimized scheduling of electric-hydrogen hybrid energy storage systems, and in particular to a double-layer multi-time-scale rolling optimization method and system for electric-hydrogen hybrid energy storage systems. Background Art

[0002] Energy storage, as a flexible resource, plays a vital role in smoothing fluctuations in renewable energy. However, most current energy storage technologies, primarily battery storage, have limited storage capacity and short durations (mostly concentrated within a day), making them less suitable for addressing long-term source-load imbalances.

[0003] With the development of electricity-to-hydrogen technology, hydrogen storage has garnered widespread attention as a new energy storage method. Hydrogen, as an ideal clean energy source, offers flexible storage capacity and the ability to store energy for extended periods (daytime). Electricity and hydrogen complement each other and play a vital role in the energy mix. Electricity-hydrogen hybrid energy storage systems enable the interconnection of electricity and hydrogen. However, the differences in scheduling between the two energy sources, as well as the uncertainty in wind and solar output brought about by high penetration of renewable energy, complicate the optimal scheduling of electric-hydrogen hybrid energy storage systems.

[0004] Currently, the operation and scheduling of electric-hydrogen hybrid energy storage systems is mostly focused on a short timescale, with a 24-hour scheduling cycle. This fails to fully leverage the advantages of hydrogen energy as a long-term energy storage system, which can be stored across multiple days. It also fails to coordinate the short-term optimization of day-ahead electricity with the long-term optimization of week-ahead hydrogen energy. Long-term hydrogen storage scheduling has not fully explored the temporal characteristics of wind and solar output forecasts, nor has it characterized the uncertainty of typical wind and solar forecast characteristics. Consequently, existing operation and scheduling technologies for electric-hydrogen hybrid energy storage systems cannot balance hydrogen storage cycle scheduling based on wind and solar forecast information with short-term day-ahead scheduling, resulting in insufficient scheduling reliability and practicality.

[0005] In addition, uncertainty optimization models include two types: stochastic optimization and robust optimization to characterize the uncertainty of renewable energy sources such as wind and solar power. However, the probability distribution of uncertain parameters is actually unknown or difficult to estimate, or a large number of scenarios need to be sampled, making it impossible to cover all the complex variation characteristics of uncertainty factors through stochastic optimization. The out-of-sample performance is poor and the solution efficiency is low. Although robust optimization has strong risk resistance, the frequency of the worst scenarios in reality is relatively low, resulting in overly conservative results. Distributed robust optimization combines the advantages of robust optimization and stochastic optimization. It does not need to assume that the wind and solar power forecast errors meet the precise probability distribution, and can solve the problem of the difficulty in obtaining accurate probability distribution parameters in stochastic optimization. At the same time, by combining probability distribution information, it improves conservatism while ensuring robustness. Therefore, for the problem of different temporal accuracy of wind and solar power output uncertainty over long time scales, the robustness of existing technologies is insufficient.

[0006] Therefore, existing technologies mostly focus on the short time scale of 24 hours a day, failing to fully utilize the advantages of long-term storage of hydrogen energy. At the same time, they fail to fully consider the time series characteristics of the prediction when describing the uncertainty of wind and solar power over a long period of time. Summary of the Invention

[0007] In order to solve the above problems, the present disclosure proposes a two-layer multi-time scale rolling optimization method and system for an electric-hydrogen hybrid energy storage system. Considering the impact of different time scales and wind and solar uncertainties on the optimal scheduling of the electric-hydrogen hybrid energy storage system, a two-layer scheduling model is constructed, and two-layer multi-time scale rolling optimization is performed to provide a scheduling solution that takes into account both economy and robustness for the flexible scheduling of the electric-hydrogen hybrid energy storage system before the week and before the day.

[0008] According to some embodiments, the present disclosure adopts the following technical solutions: A two-layer multi-time-scale rolling optimization method for electric-hydrogen hybrid energy storage system includes: Using Wasserstein distance, the fuzzy set of wind and solar prediction error is constructed; Based on the fuzzy set of wind and solar power forecast errors and aiming at maximizing power self-balancing, an upper-layer long-time scale energy dispatch model is constructed to solve the long-time scale hydrogen storage capacity and power imbalance. Based on the fuzzy set of wind and solar power prediction errors, a typical scenario is generated for the potential distribution of short-time-scale wind and solar power output. In this typical scenario, the long-time-scale hydrogen storage and power imbalance transmitted by the upper layer are received. With the goal of optimizing comprehensive economic efficiency, a short-time-scale power dispatch model is constructed at the lower layer to solve for the short-time-scale hydrogen storage. The short-time-scale hydrogen storage capacity obtained by solving the lower layer is used as the initial hydrogen storage capacity of the long-time-scale energy scheduling model, and rolling optimization is performed to achieve double-layer multi-time-scale rolling optimization.

[0009] According to some embodiments, the present disclosure adopts the following technical solutions: The electric-hydrogen hybrid energy storage system has a two-layer multi-time-scale rolling optimization system, including: The fuzzy set construction module is configured to: construct the wind and solar prediction error fuzzy set using Wasserstein distance; The energy dispatch module is configured to: construct an upper-layer long-time scale energy dispatch model based on the fuzzy set of wind and solar power forecast errors and aim at maximizing power self-balancing, and solve for the long-time scale hydrogen storage capacity and power imbalance; The power dispatch module is configured to: generate typical scenarios for the potential distribution of short-time-scale wind and solar power output based on the fuzzy set of wind and solar power forecast errors. In these typical scenarios, it receives the long-time-scale hydrogen storage capacity and power imbalance value transmitted from the upper layer, and with the goal of optimizing comprehensive economic efficiency, constructs a short-time-scale power dispatch model at the lower layer to solve for the short-time-scale hydrogen storage capacity. The rolling optimization module is configured to use the short-time-scale hydrogen storage capacity obtained by the lower layer as the initial hydrogen storage capacity of the long-time-scale energy scheduling model for rolling optimization, thereby realizing double-layer multi-time-scale rolling optimization.

[0010] According to some embodiments, the present disclosure adopts the following technical solutions: A computer program product includes a computer program, which, when executed by a processor, implements the double-layer multi-time-scale rolling optimization method for an electric-hydrogen hybrid energy storage system.

[0011] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the double-layer multi-time-scale rolling optimization method of the electric-hydrogen hybrid energy storage system is implemented.

[0012] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device comprises: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device implements the double-layer multi-time-scale rolling optimization method of the electric-hydrogen hybrid energy storage system.

[0013] Compared with the prior art, the present invention has the following beneficial effects: The present invention takes into account the uncertainty of wind and solar power output at multiple time scales in the electric-hydrogen hybrid energy storage system, and proposes a double-layer multi-time scale rolling optimization method. It constructs a distributed robust optimization model on a long time scale to perform energy scheduling; and performs power optimization on a short time scale, thereby improving the economy and flexibility of system scheduling.

[0014] The present invention constructs a fuzzy set based on the Wasserstein distance, sets the fuzzy set radius and confidence level corresponding to different prediction days according to the actual wind and solar power output prediction characteristics, handles the uncertainty of long-time wind and solar power output, and adopts distributed robust optimization to solve the long-time energy scheduling model, realizes the cross-day transfer of hydrogen energy, reduces the daily imbalance of electricity, and smoothes the long-time wind and solar power output fluctuations.

[0015] The double-layer multi-time-scale rolling optimization method of the present invention provides decision guidance for short-time-scale power optimization by receiving long-time-scale energy scheduling results, and feeds back the short-time-scale scheduling results to the next stage for continued rolling optimization on the long-time scale, thereby realizing the coordinated optimization of energy and power corresponding to multi-time-scale hydrogen storage and electricity storage, and achieving a balance between economy and robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings, which constitute a part of the present disclosure, are used to provide a further understanding of the present disclosure. The exemplary embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation to the present disclosure.

[0017] Figure 1 This is a structural diagram of the electric-hydrogen hybrid energy storage system of Example 1.

[0018] Figure 2 This is a double-layer multi-time-scale rolling optimization framework for the electric-hydrogen hybrid energy storage system of Example 1.

[0019] Figure 3 This is the rolling optimization model structure of Example 1. DETAILED DESCRIPTION

[0020] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0021] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present disclosure belongs.

[0022] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0023] Example 1 In one embodiment of the present disclosure, a double-layer multi-time-scale rolling optimization method for an electric-hydrogen hybrid energy storage system is provided, comprising: Using Wasserstein distance, the fuzzy set of wind and solar prediction error is constructed; Based on the fuzzy set of wind and solar power forecast errors and aiming at maximizing power self-balancing, an upper-layer long-time scale energy dispatch model is constructed to solve the long-time scale hydrogen storage capacity and power imbalance. Based on the fuzzy set of wind and solar power prediction errors, a typical scenario is generated for the potential distribution of short-time-scale wind and solar power output. In this typical scenario, the long-time-scale hydrogen storage and power imbalance transmitted by the upper layer are received. With the goal of optimizing comprehensive economic efficiency, a short-time-scale power dispatch model is constructed at the lower layer to solve for the short-time-scale hydrogen storage. The short-time-scale hydrogen storage capacity obtained by solving the lower layer is used as the initial hydrogen storage capacity of the long-time-scale energy scheduling model, and rolling optimization is performed to achieve double-layer multi-time-scale rolling optimization.

[0024] As an embodiment, the disclosed double-layer multi-time-scale rolling optimization method for the electric-hydrogen hybrid energy storage system takes into account the impact of different time scales and wind and solar uncertainties on the optimal scheduling of the electric-hydrogen hybrid energy storage system, constructs a double-layer scheduling model, and performs double-layer multi-time-scale rolling optimization. Figure 1 The flexible scheduling of the electric-hydrogen hybrid energy storage system before the week and before the day provides a scheduling solution that takes into account both economy and robustness, such as Figure 2 The specific implementation process is as follows: Step 1: Construct a long-time-scale (week-ahead) Wasserstein fuzzy set Specifically, distributional robust optimization combines the robust optimization theory with the stochastic optimization theory to establish a fuzzy set, which is used to obtain the uncertainty set of the potential true distribution, and the corresponding set contains the worst case of the true distribution.

[0025] The Wasserstein distance can be expressed as: (1) Where, is the infimum function; and They represent the actual prediction error and empirical prediction error of wind and solar power output respectively; Represents the empirical distribution and the true distribution The joint distribution of The entire collection of; Represents the norm function.

[0026] To facilitate model solving, this embodiment uses the 1-norm, and the Wasserstein distance can be rewritten into the desired form: (2) Where, is the base distance of the Wasserstein distance expressed in 1-norm.

[0027] This embodiment introduces Wasserstein distance into fuzzy set construction and obtains empirical distribution from historical forecast data. , the fuzzy set can be regarded as the true distribution Distribution around experience The radius is The sphere contains the true distribution of all possible wind and solar output forecast errors, then the Wasserstein fuzzy set Expressed as: (3) Where, Support set All probability distributions in the support set satisfy: (4) Where, is an auxiliary variable, and They represent the upper and lower bounds of the wind and solar prediction error respectively. The specific formula is: (5) Where, is the future forecast value of the distribution, Take a value for the confidence level.

[0028] Different prediction error intervals can be obtained by setting different confidence levels.

[0029] For example, different values ​​are taken for the fuzzy set radius of the first 3 days, 4th, 5th day, 6th day and 7th day of the long time scale. According to the different prediction accuracy of the wind and solar power corresponding days, the confidence level is also different. Thus, by controlling the radius and confidence level of the fuzzy set in different days, the potential real distribution of wind and solar power output on a long time scale is obtained. and ; Potential real distribution of wind and solar power output and In the period Sum the time period The potential real distributed power generation of internal wind turbines and photovoltaics is the corresponding and .

[0030] Step 2: Construct an upper-level long-time scale energy scheduling model The long-time scale energy scheduling model aims to minimize the imbalanced electricity over a long time scale, that is, to maximize the self-balance of electricity. The constraints include hydrogen energy balance constraints, hydrogen storage device model, power-to-hydrogen (P2H) device and fuel cell (FC) model.

[0031] (1) Objective function Electric-hydrogen hybrid energy storage needs to be Sub-long time scale energy scheduling, The objective function of the suboptimal scheduling model is: (6) Where, Indicates the unbalanced power within the scheduling period; and It is the time period Potential real distributed power generation of photovoltaic and wind turbines; and It is the time period Internal FC power generation and P2H device power consumption; It is the time period Internal electrical load; Indicates the total duration of the long-time scheduling cycle.

[0032] (2) Hydrogen energy balance constraints The long-term hydrogen energy balance constraint is: (7) Where, and It is the time period Hydrogen production of internal P2H device and hydrogen consumption of FC; and It is the time period The amount of hydrogen charged and discharged in the internal hydrogen storage device; It is the time period Internal hydrogen loading.

[0033] (3) Hydrogen storage device model During operation, the hydrogen storage device needs to charge and discharge hydrogen according to the working characteristics of the equipment. At the same time, the timing characteristics of hydrogen storage need to be considered to establish an operation model of the hydrogen storage device.

[0034] (8) Where, for The hydrogen storage capacity of the hydrogen storage tank during the time period; and is the amount of hydrogen charged and discharged, and Represents the efficiency of hydrogen charging and discharging; and They are the upper and lower limits of the total hydrogen storage capacity of the hydrogen energy storage device; and is a 0-1 variable. Taking the hydrogen release state as an example, when When it is 1, it means the hydrogen storage device releases hydrogen; and express The upper and lower limits of hydrogen charging and discharging within a time period. The hydrogen storage capacity of the hydrogen storage device at the beginning and end of the scheduling cycle is equal.

[0035] (4) P2H device and FC model The constraints of the long-time optimization model also include P2H device and FC constraints, which are expressed as (9) (10) Where, is the hydrogen production of the P2H device using electricity to produce hydrogen, are the minimum and maximum values ​​of hydrogen production, respectively; is the hydrogen consumption of FC using hydrogen for power generation, are the minimum and maximum values ​​of hydrogen consumption, respectively; and It is the time period Internal FC power generation and P2H device power consumption; It is the time period Internal electrical load; and It is the electricity-hydrogen energy conversion efficiency of the P2H device and FC.

[0036] Furthermore, for the long-time scale energy scheduling model, the objective function can be rewritten as: (11) Where, is the decision variable, that is, and ; is the coefficient matrix corresponding to the decision variables in the objective function; represents the expected function; is the supremum function; is the uncertainty variable, that is, and The objective function corresponds to the fuzzy set Under the worst conditions, the inner layer expectation function is minimized and the unbalanced electricity is minimized to the greatest extent on a long time scale.

[0037] Among them, the uncertain variables The inner function changes to: (12) Where, is the supremum function; is the infimum function; is the decision variable, that is, and ; represents the expected function; is the fuzzy set radius; is the uncertainty variable, that is, and ; is the dual variable; It is Historical forecast samples; is the number of error samples; Represents auxiliary variables.

[0038] The corresponding inner layer distribution robust optimization function model, i.e. (12), can be solved by dual transformation to obtain the final mixed integer linear programming form, which is expressed as: (13) Where, Corresponding to the constraints in the long-time scale energy scheduling model; is an abstract function; is the dual variable; It is Historical forecast samples; and are the maximum and minimum values ​​in the predicted sample set; is the number of error samples; Represents auxiliary variables; is the decision variable, that is, and ; is the uncertainty variable, that is, and .

[0039] Step 3: Generate typical scenarios of short-term wind and solar power output In layman's terms, the scenario here refers to the potential real distribution of wind and solar power output in the previous 24 hours obtained by the upper-level model, that is, the 24-hour wind and solar power. A large number of wind and solar power output arrays are obtained through Latin hypercube sampling, and then K-means clustering is performed to obtain several typical 24-hour wind and solar power arrays. This typical wind and solar power output array is a typical scenario, corresponding to a scenario generation probability. This process is the typical scenario generation process.

[0040] Specifically, the error distribution of random variables significantly impacts the rationality of random scenario generation. Kernel density estimation is used to verify that the error in wind and solar output meets the actual distribution. Latin Hypercube (LHS) sampling uses stratified sampling to ensure uniform and comprehensive sample coverage across the variable distribution. Compared to the Monte Carlo method, this method can mitigate sample accumulation and uneven distribution, achieving higher sampling accuracy within a smaller sampling scale. The sampling method essentially performs an inverse function transformation on each sample value based on the probability density function, without changing the original density function, to obtain the final sampling result.

[0041] When generating typical wind and solar power output scenarios, Latin hypercube sampling is used to obtain a large number of wind and solar power output scenarios, and kernel density estimation is used to evaluate and verify the rationality of these scenarios. Specifically, kernel density estimation is used to estimate the error distribution of wind and solar power output, thereby ensuring that the large number of scenarios generated by Latin hypercube sampling are statistically closer to the actual distribution. In other words, kernel density estimation provides a way to test and adjust the scenarios obtained by Latin hypercube sampling.

[0042] After obtaining a large number of scenarios through LHS sampling, improved K-means clustering is used to obtain typical wind and solar output scenarios. The main clustering steps are as follows: Step 1: Let the number of clusters be , randomly select a scene as a cluster center .

[0043] Step 2: Calculate the remaining scenes and The distance between them (using Euclidean distance), the scene with the farthest distance is selected as the second cluster center Then calculate the sum of the distances between the remaining scenes and the two cluster centers after removing the two cluster centers, and find the scene with the largest distance sum as the third cluster center , and so on, we get Initial cluster centers.

[0044] Step 3: Calculate the distance between the remaining scenes and each cluster center, classify the remaining scenes into the cluster center closest to them, and re-calculate the cluster center of each category.

[0045] Step 4: Define the sum of the squares of the distances between each scene and its cluster center as the clustering error , when the clustering error of the cluster center of the 2nd iteration is When the difference is less than a certain convergence accuracy, stop the iteration and output the cluster center of the last iteration, otherwise return to Step 3.

[0046] Step 4: Build the lower-level short-time-scale power scheduling model Specifically, the lower-level short-time-scale power dispatch takes the comprehensive economic optimization as the optimization goal, and the constraints include the electric-hydrogen power balance constraint, the power storage device model, the power purchase and sales model, the P2H device and FC model, and the short-time-scale supplementary constraint of the hydrogen storage device.

[0047] (1) Objective function The short-time-scale power dispatch objective function includes the cost of purchasing and selling electricity, the cost of curtailing wind and solar power, the cost of energy storage, the operating cost of P2H devices and FCs, the cost of carbon emissions, and the penalty cost of power imbalance, which can be expressed as follows: (14) Respectively expressed as: (15) Where, and 、 and Respectively The power and time-of-use electricity price of electricity purchased from and sold to the external power grid at any given time; and Indicates the probability of different wind and solar power output scenarios; and 、 and They represent the penalty cost coefficients for wind and solar power abandonment, Moment Scene of curtailed wind and solar power; and is the energy storage cost coefficient of hydrogen and electricity storage devices; and 、 and yes The hydrogen storage device's charging and discharging power, and the electricity storage device's full power at all times; and 、 and is the charging and discharging efficiency of hydrogen and electricity storage devices; and are the operating cost coefficients of the P2H unit and FC, respectively; and yes The operating power of the P2H device and FC at the moment; is the carbon emission coefficient corresponding to the unit electricity purchase; is the environmental penalty cost coefficient per unit of carbon dioxide emissions; Penalty price for system power imbalance; The unbalanced electricity volume of the day is obtained by optimizing the upper layer for a long time scale.

[0048] (2) Electric-hydrogen power balance constraints For the short-time scale 24h day-ahead power dispatch, the electric-hydrogen power balance constraint is expressed as: (16) Where, and yes Electric hydrogen load power at each moment; and yes The hydrogen production power of the P2H device and the hydrogen consumption power of the FC at this moment.

[0049] (3) Energy storage device model The power storage device is represented by a battery in this paper. The battery stores energy by charging when the power supply is sufficient and discharges when the power supply is insufficient. Its output constraint is expressed as: (17) Where, yes The storage capacity of the power storage device at all times; and It is the upper and lower limits of the storage capacity of the power storage device; and is a 0-1 variable, Taking the discharge state as an example, when it is 1, it means the battery is discharged; and They represent the maximum charging and discharging power of the battery respectively.

[0050] (4) Electricity purchasing and sales model (18) Where, and It is a 0-1 variable. Taking the electricity purchasing status as an example, 1 means the system purchases electricity from the external power grid; and It is the upper limit of the power that the system can purchase from and sell to the external power grid.

[0051] (5) P2H device and FC model (19) (20) Where, and yes The maximum hydrogen production power of the P2H device and the maximum power generation power of the FC at that moment; and is the electricity-to-hydrogen conversion efficiency of the P2H device and FC; and Indicates the maximum climbing power of the P2H device and FC.

[0052] (6) Supplementary constraints on short-time scales of hydrogen storage devices In the short-time scale power scheduling, the hydrogen storage device receives the daily hydrogen storage capacity transmitted by the upper long-time scale and relaxes it. The short-time scale model of the hydrogen storage device is expressed as: (twenty one) Where, yes The hydrogen storage capacity of the hydrogen storage device at any moment; and It is a 0-1 variable. Taking the hydrogen release state as an example, When it is 1, it means the hydrogen storage device releases hydrogen; and They represent the maximum power of hydrogen charging and discharging of the hydrogen storage device respectively; Indicates the initial hydrogen storage capacity of the next cycle of short-term scheduling of the hydrogen storage device, which is the 25th hour in this embodiment; Represents the long time scale Suboptimal scheduling of the initial hydrogen storage capacity on the second day; and It is the relaxation coefficient for the upward and downward adjustment of the hydrogen storage capacity of the hydrogen storage device.

[0053] Furthermore, the short-time-scale power scheduling model is a single-level linear programming problem, which can be solved directly using commercial solvers such as Gurobi. Its compact form is: (twenty two) Where, is the decision variable in the short-time-scale power scheduling stage; and is the constant matrix of the objective function; and is the constant matrix of the constraints.

[0054] Furthermore, the lower power model is solved to obtain the short-time scale optimization scheduling strategy, and the short-time As the first The initial hydrogen storage capacity of the secondary long-time scale energy scheduling is passed back to the upper energy scheduling model for the following Figure 3 Scrolling optimization shown.

[0055] Example 2 In one embodiment of the present disclosure, a dual-layer multi-time-scale rolling optimization system for an electric-hydrogen hybrid energy storage system is provided, comprising: The fuzzy set construction module is configured to: construct the wind and solar prediction error fuzzy set using Wasserstein distance; The energy dispatch module is configured to: construct an upper-layer long-time scale energy dispatch model based on the fuzzy set of wind and solar power forecast errors and aim at maximizing power self-balancing, and solve for the long-time scale hydrogen storage capacity and power imbalance; The power dispatch module is configured to: generate typical scenarios for the potential distribution of short-time-scale wind and solar power output based on the fuzzy set of wind and solar power forecast errors. In these typical scenarios, it receives the long-time-scale hydrogen storage capacity and power imbalance value transmitted from the upper layer, and with the goal of optimizing comprehensive economic efficiency, constructs a short-time-scale power dispatch model at the lower layer to solve for the short-time-scale hydrogen storage capacity. The rolling optimization module is configured to use the short-time-scale hydrogen storage capacity obtained by the lower layer as the initial hydrogen storage capacity of the long-time-scale energy scheduling model for rolling optimization, thereby realizing double-layer multi-time-scale rolling optimization.

[0056] Example 3 In one embodiment of the present disclosure, a computer program product is provided, including a computer program, which, when executed by a processor, implements the double-layer multi-time-scale rolling optimization method for the electric-hydrogen hybrid energy storage system.

[0057] Example 4 In one embodiment of the present disclosure, a non-transitory computer-readable storage medium is provided, which is used to store computer instructions. When the computer instructions are executed by a processor, the double-layer multi-time-scale rolling optimization method of the electric-hydrogen hybrid energy storage system is implemented.

[0058] Example 5 In one embodiment of the present disclosure, an electronic device is provided, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the double-layer multi-time-scale rolling optimization method for the electric-hydrogen hybrid energy storage system.

[0059] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0060] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0061] Although the above describes the specific implementation methods of the present disclosure in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present disclosure. Those skilled in the art should understand that on the basis of the technical solution of the present disclosure, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present disclosure.

Claims

1. A double-layer multi-time scale rolling optimization method for an electric-hydrogen hybrid energy storage system, characterized in that: include: Using Wasserstein distance, the fuzzy set of wind power prediction error is constructed; Based on the fuzzy set of wind and solar forecast errors and with the goal of maximizing the self-balancing of electricity, an upper-level long-time scale energy dispatch model is constructed to solve the long-time scale hydrogen storage capacity and electricity imbalance. Based on the fuzzy set of wind and solar prediction errors, a typical scenario is generated for the potential distribution of short-time-scale wind and solar output. In the typical scenario, the long-time-scale hydrogen storage and power imbalance transmitted by the upper layer are received. With the goal of comprehensive economic optimization, the short-time-scale power dispatch model of the lower layer is constructed to solve the short-time-scale hydrogen storage. The short-time scale hydrogen storage capacity obtained by solving the lower layer is used as the initial hydrogen storage capacity of the long-time scale energy scheduling model for rolling optimization to achieve double-layer multi-time scale rolling optimization.

2. The double-layer multi-time scale rolling optimization method for the electric-hydrogen hybrid energy storage system according to claim 1, characterized in that: The wind and solar power prediction error fuzzy set is used to set the fuzzy set radius and confidence level corresponding to different prediction days according to the actual wind and solar power output prediction characteristics, so as to deal with the uncertainty of wind and solar power output on a long time scale.

3. The double-layer multi-time scale rolling optimization method for the electric-hydrogen hybrid energy storage system according to claim 1, characterized in that: The objective is to maximize the self-balancing of electricity, which can be expressed as the objective function: In the formula, Indicates the unbalanced electricity in the dispatch cycle; and It is the time period Potential real distributed power generation of photovoltaic and wind turbines; and It is the time period Internal FC power generation and P2H device power consumption; It is the time period Internal electrical load; Indicates the total duration of the long-time scheduling cycle.

4. The double-layer multi-time scale rolling optimization method for the electric-hydrogen hybrid energy storage system according to claim 1, characterized in that: The typical scene generation adopts Latin hypercube scene generation and K-means clustering to obtain a typical scene of wind and solar power output in a short time scale.

5. The double-layer multi-time scale rolling optimization method for the electric-hydrogen hybrid energy storage system according to claim 1, characterized in that: The above-mentioned objective is to optimize the comprehensive economic performance, which is represented by an objective function consisting of the cost of purchasing and selling electricity, the cost of curtailing wind and solar power, the cost of energy storage, the operating cost of P2H devices and FC, the cost of carbon emission and the penalty cost of electricity imbalance.

6. The double-layer multi-time scale rolling optimization method for the electric-hydrogen hybrid energy storage system according to claim 1, characterized in that: The long-time scale energy scheduling model also includes constraints, specifically hydrogen energy balance constraints, hydrogen storage device models, power-to-hydrogen devices and fuel cell models; The short-time scale power dispatch model also includes constraints, specifically, electric-hydrogen power balance constraints, power storage device constraints, power purchase and sales constraints, P2H device and FC operation constraints, and short-time scale supplementary constraints for hydrogen storage devices.

7. The double-layer multi-time scale rolling optimization system of the electric-hydrogen hybrid energy storage system is characterized by: include: The fuzzy set construction module is configured to: construct the fuzzy set of wind power prediction error using Wasserstein distance; The energy dispatch module is configured as follows: based on the fuzzy set of wind and solar forecast errors, with the goal of maximizing the self-balancing of electricity, an upper-layer long-time scale energy dispatch model is constructed to solve the long-time scale hydrogen storage capacity and electricity imbalance; The power dispatch module is configured to: generate typical scenarios for the potential distribution of short-time-scale wind and solar power output based on the fuzzy set of wind and solar power prediction errors; in typical scenarios, receive the long-time-scale hydrogen storage and power imbalance transmitted by the upper layer, and build a short-time-scale power dispatch model for the lower layer with the goal of comprehensive economic optimization to solve for the short-time-scale hydrogen storage; The rolling optimization module is configured to use the short-time scale hydrogen storage capacity obtained by the lower layer solution as the initial hydrogen storage capacity of the long-time scale energy scheduling model, and perform rolling optimization to achieve double-layer multi-time scale rolling optimization.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the double-layer multi-time scale rolling optimization method for the electric-hydrogen hybrid energy storage system according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, the double-layer multi-time scale rolling optimization method of the electric-hydrogen hybrid energy storage system as described in any one of claims 1-6 is implemented.

10. An electronic device, characterized in that: include: A processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes the double-layer multi-time scale rolling optimization method for the electric-hydrogen hybrid energy storage system as described in any one of claims 1-6.

Images