A multi-energy resource collaborative planning and scheduling method, device, equipment and storage medium

Abstract

This invention relates to the field of energy planning technology and discloses a multi-energy resource collaborative planning and scheduling method, apparatus, equipment, and storage medium. First, this invention utilizes a Kriging estimation sample interpolation method based on least squares support vector machines to predict the time-series output of new energy sources at the target planning location, obtaining a predicted dataset of new energy time-series output at the target planning location. Second, a load forecast dataset is determined, and a multi-objective planning and scheduling model is constructed by combining the target multi-energy resource dataset and the original resource dataset. Finally, the final multi-energy resource collaborative planning and scheduling result is obtained by solving the model using the new energy time-series output forecast dataset, the historical load shedding probability dataset, and the load forecast dataset. Therefore, by implementing this invention, the power output demand of the system during specific time periods can be met, laying the foundation for establishing a new power system optimization planning and scheduling operation model with a high proportion of new energy sources.

Description

Technical Field

[0001] This invention relates to the field of energy planning technology, specifically to a method, apparatus, equipment, and storage medium for multi-energy resource collaborative planning and scheduling. Background Technology

[0002] Sustainability, resilience, and carbon neutrality are the goals for future power energy planning and development. Clean and low-carbon, safe and controllable, flexible and efficient, intelligent and user-friendly, and open and interactive are the basic characteristics of the new power system being built in my country. With the increasing severity of environmental problems and energy shortages, new energy sources, represented by solar and wind power, have received widespread attention and development. However, due to the intermittency and instability of new energy sources, their power output cannot consistently meet the balance between power supply and demand and load requirements, especially during specific periods such as peak system load and peak power generation from new energy sources. Furthermore, with the popularization of electric vehicles and other new energy vehicles, and changes in electricity consumption behavior due to users' active participation in demand response, the power system needs more dispatchable and flexible resources to maintain system stability and security.

[0003] Energy storage technology, as a key means to address the volatility of new energy sources, has developed rapidly. Energy storage systems release electricity when needed to meet load requirements. However, the introduction of new energy sources and energy storage makes system scheduling more complex, requiring coordination and optimization of multiple energy forms to ensure the reliability and stability of power supply. How to rationally plan new energy sources, energy storage, energy efficiency improvements, and demand response, and how to coordinate and optimize the scheduling of multiple energy forms to provide the required power output within a specific time period, is a challenging problem. This necessitates the development of appropriate planning methods and scheduling algorithms.

[0004] Currently, new energy resource planning and related microgrid planning have attracted the interest of many researchers worldwide. Existing literature has proposed an iterative optimization planning method based on the minimum cut-set approach, addressing various challenges including economic viability, reliability, and power generation variability. Existing literature has also studied agent-based economic dispatch of new energy. The planning and design of new energy storage mainly revolves around technical engineering and economic decisions, utilizing cost-benefit analysis and operations research methods to solve optimization dispatch and power generation expansion planning problems. While existing methods can obtain relatively accurate joint planning results that meet demand, they lack consideration for 8760 hours of load data with long-term seasonal characteristics, comprehensive consideration of multiple energy forms, meeting the power generation output demand constraints of the system during specific periods, and the potential benefits of reliable power generation output. Therefore, they have significant limitations.

[0005] In summary, the coordinated optimization planning and scheduling of multi-energy resources to meet the output demand during specific time periods is a key issue in the field of new power system planning. It plays a fundamental role in many studies and applications. However, existing planning methods do not fully consider the correlation between load and multi-energy resource power generation output, do not fully explore the flexible resources that the system can plan, have large planning errors, and have certain limitations. They cannot comprehensively describe and characterize the interaction and changing characteristics of the joint planning and scheduling of new energy and energy storage, nor can they accurately, realistically, and effectively simulate the effective interaction between new energy, energy storage, and the power grid. Summary of the Invention

[0006] In view of this, the present invention provides a multi-energy resource collaborative planning and scheduling method, device, equipment and storage medium to solve the problems that existing planning methods do not fully consider the correlation between load and multi-energy resource power generation output, do not fully explore the system's planable flexibility resources, have large planning errors and certain limitations, and cannot comprehensively describe the interaction and change characteristics of the joint planning and scheduling of new energy and energy storage, and accurately, realistically and effectively simulate the effective interaction between new energy, energy storage and the power grid.

[0007] In a first aspect, the present invention provides a multi-energy resource collaborative planning and scheduling method, the method comprising:

[0008] The process involves obtaining the initial distributed renewable energy historical output dataset, historical load shedding probability dataset, target multi-energy resource dataset, and original resource dataset for the target planning location; processing the initial distributed renewable energy historical output dataset using a Kriging estimation sample interpolation method based on least squares support vector machine to obtain the renewable energy time-series output prediction dataset for the target planning location; establishing a load prediction model based on the historical load shedding probability dataset and determining the load prediction dataset; establishing a multi-objective planning and scheduling model based on the target multi-energy resource dataset and the original resource dataset; and solving the multi-objective planning and scheduling model based on the renewable energy time-series output prediction dataset, the historical load shedding probability dataset, and the load prediction dataset to obtain the multi-energy resource collaborative planning and scheduling results.

[0009] The multi-energy resource collaborative planning and scheduling method provided by this invention first processes the initial distributed renewable energy historical output dataset using a Kriging estimation sample interpolation method improved based on least squares support vector machines. This allows for the prediction of renewable energy time-series output at the target planning location, resulting in a predicted dataset. Secondly, a load forecasting model is established based on the historical load shedding probability dataset, and the load forecasting dataset is determined. Then, a complete multi-objective planning and scheduling model is constructed by combining the target multi-energy resource dataset and the original resource dataset. Finally, the model is solved using the renewable energy time-series output prediction dataset, the historical load shedding probability dataset, and the load forecasting dataset to obtain the final multi-energy resource collaborative planning and scheduling result. Therefore, by implementing this invention, the reliable power generation capacity and flexibility of multi-energy resources are characterized. This overcomes the randomness of renewable energy generation, the uncertainty of load, and the impact of load forecasting errors, as well as the shortcomings of traditional planning methods that do not fully consider the physical constraints of various flexible resources and renewable energy and energy storage systems during actual operation. It can meet the power output demand of the system during specific periods, laying the foundation for establishing a new power system optimization planning and scheduling operation model with a high proportion of renewable energy.

[0010] In one optional implementation, the initial distributed historical output dataset of new energy sources is processed using a Kriging estimation sample interpolation method based on least squares support vector machine to obtain a time-series output prediction dataset of new energy sources at the target planning location, including:

[0011] The initial distributed renewable energy historical output dataset is normalized to obtain the target distributed renewable energy historical output dataset. Based on the target distributed renewable energy historical output dataset, the initial variable variance function is calculated. The initial variable variance function is fitted using a least squares support vector machine to obtain the target variable variance function. Based on the target variable variance function, the renewable energy time-series output prediction dataset for the target planning location is obtained through Kriging estimation sample point interpolation.

[0012] The multi-energy resource collaborative planning and scheduling method provided by this invention first calculates an initial variable variance function based on a normalized target distributed renewable energy historical output dataset. Then, it obtains the optimal target variable variance function through least squares support vector machine fitting. Finally, it uses Kriging estimation sample interpolation to predict the time-series output of renewable energy at the target planning location and obtains a predicted dataset of renewable energy time-series output at the target planning location, providing support for subsequent multi-energy resource collaborative planning and scheduling.

[0013] In one optional implementation, a multi-objective planning and scheduling model is established based on the target multi-energy resource dataset and the original resource dataset, including:

[0014] Obtain the objective function and the set of objective constraints; determine the output demand for the target time period based on the original resource dataset; and construct a multi-objective planning and scheduling model based on the objective function and the set of objective constraints, using the output demand for the target time period and the target multi-energy resource dataset.

[0015] The multi-energy resource collaborative planning and scheduling method provided by this invention, based on the obtained objective function and objective constraint set, and combined with the target time period output demand and target multi-energy resource dataset, can construct a multi-objective planning and scheduling model that meets the target time period output demand, thus providing support for subsequent multi-energy resource collaborative planning and scheduling.

[0016] In one optional implementation, obtaining the objective function and the set of objective constraints includes:

[0017] Obtain multiple resource cost data and initial constraint sets; establish an objective function based on the multiple resource cost data; transform the initial constraint set using the Big-M method to obtain the objective constraint set.

[0018] The multi-energy resource collaborative planning and scheduling method provided by this invention can transform the nonlinear initial constraint set into linear objective constraints through the Big-M method, which can guarantee the efficiency, convergence, and global optimality of the subsequent multi-objective planning and scheduling model.

[0019] In one optional implementation, a multi-objective planning and scheduling model is constructed based on the objective function and the set of objective constraints, utilizing the target time period output demand and the target multi-energy resource dataset, including:

[0020] Based on the objective function and the set of objective constraints, an initial planning and scheduling model is constructed using the target time period output demand and the target multi-energy resource dataset. The initial planning and scheduling model is solved to obtain the solution result. It is determined whether the solution result is the optimal solution. If the solution result is the optimal solution, the initial planning and scheduling model is determined to be a multi-objective planning and scheduling model. If the solution result is not the optimal solution, the initial constraint set and multiple resource cost data are adjusted until an optimal solution exists and the multi-objective planning and scheduling model is determined.

[0021] The multi-energy resource collaborative planning and scheduling method provided by this invention can further determine the final multi-objective planning and scheduling model by examining the initial planning and scheduling model.

[0022] In one optional implementation, determining the output demand for a target time period based on the original resource dataset includes:

[0023] Based on preset conditions, the power generation cost data is calculated from the original resource dataset; using the power generation cost data as a basic reference value, the power output demand for the target period is determined using the electricity price curve or load peak-valley curve.

[0024] The multi-energy resource collaborative planning and scheduling method provided by this invention can calculate the current power generation cost data that meets the conditions using existing resources, and then determine the output demand for the target period that meets the current conditions.

[0025] Secondly, the present invention provides a multi-energy resource collaborative planning and scheduling device, the device comprising:

[0026] The acquisition module is used to acquire the target multi-energy resource dataset and the original resource dataset. The target multi-energy resource dataset includes annual 8760-hour load curve data, new energy physical and price data, energy storage physical parameters and price data, multiple electricity cost data, new energy storage incentive data, and carbon emission parameters and price data. The first determination module is used to determine the output demand for the target period based on the original resource dataset. The construction module is used to construct a multi-objective planning and scheduling model based on the output demand for the target period and the target multi-energy resource dataset. The second determination module is used to determine the multi-energy resource collaborative planning and scheduling results based on the multi-objective planning and scheduling model.

[0027] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the multi-energy resource collaborative planning and scheduling method described in the first aspect or any corresponding embodiment.

[0028] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions, which are used to cause a computer to execute the multi-energy resource collaborative planning and scheduling method described in the first aspect or any of its corresponding embodiments.

[0029] Fifthly, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute the multi-energy resource collaborative planning and scheduling method described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0030] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0031] Figure 1 This is a flowchart illustrating the multi-energy resource collaborative planning and scheduling method according to an embodiment of the present invention;

[0032] Figure 2 This is a flowchart illustrating another multi-energy resource collaborative planning and scheduling method according to an embodiment of the present invention;

[0033] Figure 3 This is a flowchart illustrating another multi-energy resource collaborative planning and scheduling method according to an embodiment of the present invention;

[0034] Figure 4 This is a flowchart illustrating a multi-energy resource collaborative optimization planning and scheduling method for meeting output demands during a specific time period, according to an embodiment of the present invention.

[0035] Figure 5 This is a schematic diagram illustrating the specific process for calculating the time-series output curve of new energy sources according to an embodiment of the present invention;

[0036] Figure 6 This is a schematic diagram of the construction process of a multi-objective planning and scheduling model according to an embodiment of the present invention;

[0037] Figure 7 This is a structural block diagram of a multi-energy resource collaborative planning and scheduling device according to an embodiment of the present invention;

[0038] Figure 8 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] This invention provides a multi-energy resource collaborative planning and scheduling method. By characterizing the reliable power generation capacity and flexibility of multiple energy resources, it overcomes the randomness of new energy power generation, the uncertainty of load, and the impact of load forecasting errors, as well as the shortcomings of traditional planning methods that do not fully consider the physical constraints of multiple flexible resources, new energy and energy storage systems during actual operation. It can meet the power output demand of the system in specific time periods and lay the foundation for establishing a new power system optimization planning and scheduling operation model with a high proportion of new energy.

[0041] According to an embodiment of the present invention, a method for collaborative planning and scheduling of multi-energy resources is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0042] This embodiment provides a multi-energy resource collaborative planning and scheduling method, which can be used in electronic devices such as computers, mobile phones, and tablets. Figure 1 This is a flowchart of a multi-energy resource collaborative planning and scheduling method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0043] Step S101: Obtain the initial distributed new energy historical output dataset, historical load failure probability dataset, target multi-energy resource dataset, and original resource dataset for the target planning location.

[0044] Specifically, the initial distributed historical output dataset of new energy sources may include time-series data of the historical output of new energy power plants in the vicinity of the target planning location or time-series meteorological data of multiple meteorological stations; the historical failure probability dataset is used to characterize the reliability index requirements of the failure probability.

[0045] Furthermore, the target multi-energy resource dataset may include annual 8760-hour load curve data, new energy physical and price data, energy storage physical parameters and price data, multiple electricity cost data, new energy storage incentive data, and carbon emission parameters and price data. Among these, the multiple electricity cost data may include energy efficiency improvement ranges and costs, and demand response electricity cost data.

[0046] Furthermore, the original resource dataset may include relevant data of existing resources, but not the aforementioned target multi-functional resource dataset.

[0047] Step S102: The initial distributed historical output dataset of new energy is processed by the Kriging estimation sample interpolation method based on least squares support vector machine to obtain the time series output prediction dataset of new energy at the target planning location.

[0048] Specifically, by using the Kriging estimation sample interpolation method based on least squares support vector machine to process the initial distributed renewable energy historical output dataset, the temporal output of renewable energy at the target planning location can be predicted, overcoming the randomness of renewable energy power generation and providing support for subsequent multi-energy resource collaborative planning and scheduling.

[0049] Step S103: Establish a load forecasting model and determine the load forecasting dataset based on the historical load loss probability dataset.

[0050] Specifically, based on the reliability index requirements of load loss probability, a load forecasting model is obtained by modeling the load forecasting error and randomness in the form of probability distribution through a chance constraint model. This load forecasting model can determine the load forecasting dataset and overcome the impact of load forecasting error in the multi-energy resource collaborative planning and scheduling process.

[0051] In this process, the opportunity constraint, which requires the load loss probability index to be expressed in the form of a probability distribution, is transformed into a linear inequality constraint.

[0052] Various probability distribution functions, including hyperbolic distribution function, normal distribution function, and truncated normal distribution function, can be used to represent the error of load forecasting for the next day. In this embodiment, the load forecasting error is represented by a truncated normal distribution with a mean of 0 and a standard deviation of 5% of the hourly load forecast, as shown in the following relationship (1):

[0053]

[0054] Where: PD t Represents the predicted load at time t; e t E(·) represents the load forecast error at time t; E(·) represents the expected value of the random variable.

[0055] Furthermore, opportunity constraints are a method for considering uncertainty or risk in optimization problems. In opportunity-constrained optimization problems, the constraints appear in the form of certain probabilities, usually expressed as an inequality that is satisfied with a certain probability. Data-wise, a single opportunity constraint can be expressed as the following relation (2):

[0056] Pr(g(x,ξ)≤0)≥α t (2)

[0057] In the formula: g(x,ξ) represents the constraint function; x represents the decision variable; ξ represents the random variable; α t This represents the minimum probability of satisfying the constraint.

[0058] In the planning and scheduling model, the following relationship (3) shows that the power generation output of each energy source at the planned location will meet the hourly load with a predetermined probability:

[0059]

[0060] Where: LOLP t This indicates the reliability index requirements that need to be met based on the probability of load failure. This represents the amount of electricity purchased from the power grid at time t. This indicates the planned photovoltaic power output. Indicates energy storage output. This indicates the contribution of energy efficiency improvement measures. This represents the power output contribution in demand response. By setting this value appropriately as a trade-off between economy and reliability, the above relationship (3) ensures sufficient power generation output in real-time scheduling.

[0061] Furthermore, the chance constraint represented by the above relation (3) is transformed into an equivalent deterministic inequality in the solution process, as shown in the following relation (4):

[0062]

[0063] In the formula: It represents 100*(1-LOLP) t ) th Percentiles of the standard normal distribution.

[0064] Step S104: Establish a multi-objective planning and scheduling model based on the target multi-energy resource dataset and the original resource dataset.

[0065] Specifically, a corresponding multi-objective planning and scheduling model can be constructed based on the target multi-energy resource dataset and the original resource dataset.

[0066] Step S105: Based on the new energy time-series output prediction dataset, historical load failure probability dataset, and load prediction dataset, solve the multi-objective planning and scheduling model to obtain the multi-energy resource collaborative planning and scheduling results.

[0067] Specifically, by inputting the new energy time-series output prediction dataset, the historical load failure probability dataset, and the load prediction dataset into the multi-objective programming scheduling model for solving, the optimal solution that meets various constraints and multiple objectives, such as the output demand for a specific period, can be obtained.

[0068] Furthermore, the optimal solution for multiple objectives can be used to obtain the results of multi-energy resource collaborative planning and scheduling.

[0069] The multi-energy resource collaborative planning and scheduling method provided in this embodiment first processes the initial distributed renewable energy historical output dataset using a Kriging estimation sample interpolation method based on least squares support vector machines. This allows for the prediction of renewable energy time-series output at the target planning location, resulting in a predicted dataset. Secondly, a load forecasting model is established based on the historical load shedding probability dataset, and the load forecasting dataset is determined. Then, a complete multi-objective planning and scheduling model is constructed by combining the target multi-energy resource dataset and the original resource dataset. Finally, the model is solved using the renewable energy time-series output prediction dataset, the historical load shedding probability dataset, and the load forecasting dataset to obtain the final multi-energy resource collaborative planning and scheduling result. Therefore, by implementing this invention, the reliable power generation capacity and flexibility of multi-energy resources are characterized. This overcomes the randomness of renewable energy generation, the uncertainty of load, and the impact of load forecasting errors, as well as the shortcomings of traditional planning methods that do not fully consider the physical constraints of various flexible resources and renewable energy and energy storage systems during actual operation. It can meet the power output demand of the system during specific periods, laying the foundation for establishing a new power system optimization planning and scheduling operation model with a high proportion of renewable energy.

[0070] This embodiment provides a multi-energy resource collaborative planning and scheduling method, which can be used in electronic devices such as computers, mobile phones, and tablets. Figure 2 This is a flowchart of a multi-energy resource collaborative planning and scheduling method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0071] Step S201: Obtain the initial distributed renewable energy historical output dataset, historical load failure probability dataset, target multi-energy resource dataset, and original resource dataset for the target planning location. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0072] Step S202: The initial distributed historical output dataset of new energy is processed by the Kriging estimation sample interpolation method based on least squares support vector machine to obtain the time series output prediction dataset of new energy at the target planning location.

[0073] Specifically, step S202 includes:

[0074] Step S2021: Normalize the initial distributed renewable energy historical output dataset to obtain the target distributed renewable energy historical output dataset.

[0075] Specifically, normalization is performed using the following relation (5):

[0076]

[0077] In the formula: This represents the power generation output of the l-th renewable energy power station in the vicinity of the target planning location at time i; The output data is normalized.

[0078] Step S2022: Calculate the initial variable variance function based on the target distributed new energy historical output dataset.

[0079] Specifically, based on the target distributed renewable energy historical output dataset and the spatial distance between power plants, the experimental variance function, i.e., the initial variance function γ, of the Kriging estimation sample interpolation method can be calculated. * (h ij ).

[0080] Step S2023: The initial variable variance function is fitted using a least squares support vector machine to obtain the target variable variance function.

[0081] Specifically, the core idea of ​​the Kriging algorithm is to obtain the numerical values ​​of unsampled locations by linearly weighting the combinations of the numbers from the observation points. For an unsampled point Z(x0), its predicted value Z0 is... * (x0) can be expressed as shown in the following relation (6):

[0082]

[0083] In the formula: λ i Z(x) represents the i-th observation point. i The weighting factor of Z(x0); n represents the number of observations involved in the prediction. Weight λ i The calculation of is the core of the Kriging algorithm, and it is usually obtained through the following system of linear equations as shown in equation (7):

[0084]

[0085] In the formula: γ(h) ij ) represents the variable variance function, which describes x i With x j The spatial dependency between them. ij and h i0 They are x i and x j , and x i The distance or spatial offset between x0 and x0; μ represents the Lagrange multiplier used to satisfy the constraint that the sum of weights is 1.

[0086] Wherein, weight λ i :

[0087]

[0088] Furthermore, a key problem in the Kriging algorithm is determining the spatial variation of variables, and then estimating unknown values ​​based on known samples. This variation is the variable variance function γ(h) ij ).

[0089] Let P be the training sample of historical time series power output of new energy sources, and P = (x i ,y i ), i = 1, 2, ..., m. Where m represents the number of training samples; x i ,y i Let represent the output matrix of the training set. To estimate the unknown function, a linear regression function is constructed, as shown in equation (9):

[0090]

[0091] In the formula: ω represents the mapping function used to map training sample data to a high-dimensional space; b represents the corresponding weight vector; and b represents the bias vector.

[0092] Currently, there is no very good method for selecting the variance function model in general interpolation. A reasonable variance function model is usually determined by comparing different variance function models. However, this method is both time-consuming (requiring multiple kriging interpolation calculations) and, due to the very limited variety of variance function models, makes it difficult for the variance function to describe the spatial distribution characteristics of real data.

[0093] To overcome existing problems, Least Squares Support Vector Machine (LS-SVM) is introduced to fit the experimental variance function, thus avoiding the subjectivity of selecting the variance function model. Based on Least Squares Support Vector Machine, there is no need to determine the type of the basic variance function model; instead, the experimental variance function can be fitted directly based on its own distribution plot.

[0094] The least squares support vector machine is based on the structural risk minimization criterion. When training on the training sample set, its objective function and equality constraints are shown in the following equations (10) and (11):

[0095]

[0096] In the formula: J represents the penalty function; e i y represents the bias of the training sample data; b represents a constant; e represents the output error; C represents the regularization parameter, which can adjust the penalty for the error term; y i This represents the elements of the output vector.

[0097] Solve the above linear equations and construct the Lagrange function, as shown in the following relation (12):

[0098]

[0099] Where: β i This represents the Lagrange multiplier.

[0100] The partial derivatives of the Lagrangian function with respect to each variable are calculated. By applying the KKT (Karush-Kuhn-Tucker) conditions, the variables ω and e are eliminated, and the prediction model is obtained as shown in the following relation (13):

[0101]

[0102] In the formula: x represents the support vector; x i K(x, x) represents the i support vector input elements; i ) represents the corresponding kernel function, and its expression is shown in the following relation (14):

[0103]

[0104] In the formula: σ represents the variance of the output elements of the support vector.

[0105] Therefore, the experimental variogram γ was fitted using a least-squares support vector machine. * (h ij The optimal variance function, i.e., the initial variance function γ(h), can be obtained. ij ).

[0106] Step S2024: Based on the target variable variance function, the new energy time series output prediction dataset for the target planning location is obtained by processing the sample interpolation method of Kriging estimation.

[0107] Specifically, according to γ(h) ij Calculate the output parameter λ of each new energy power station. i ,i=1…n, and then the estimated value of the new energy output at the planned target location can be calculated.

[0108] Finally, based on the historical normalized time-series power output data of new energy sources in the vicinity of the planned location, the normalized time-series power output curve of the new energy source at the proposed location is calculated, and the corresponding new energy time-series power output prediction dataset is obtained.

[0109] Step S203: Establish a load forecasting model and determine the load forecasting dataset based on the historical load loss probability dataset. For details, please refer to [link to relevant documentation]. Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0110] Step S204: Establish a multi-objective planning and scheduling model based on the target multi-energy resource dataset and the original resource dataset. For details, please refer to [link to details]. Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0111] Step S205 involves solving the multi-objective planning and scheduling model based on the new energy time-series output prediction dataset, the historical load shedding probability dataset, and the load prediction dataset, to obtain the multi-energy resource collaborative planning and scheduling results. For details, please refer to [link to relevant documentation]. Figure 1 Step S105 of the illustrated embodiment will not be described again here.

[0112] The multi-energy resource collaborative planning and scheduling method provided in this embodiment first calculates the initial variable variance function based on the normalized target distributed renewable energy historical output dataset, and then obtains the optimal target variable variance function through least squares support vector machine fitting. Finally, the kriging estimation sample point interpolation method is used to predict the renewable energy time-series output at the target planning location and obtain the renewable energy time-series output prediction dataset for the target planning location. Further, a load prediction model is established based on the historical load shedding probability dataset and the load prediction dataset is determined. Then, a complete multi-objective planning and scheduling model is constructed by combining the target multi-energy resource dataset and the original resource dataset. Finally, the model is solved using the renewable energy time-series output prediction dataset, the historical load shedding probability dataset, and the load prediction dataset to obtain the final multi-energy resource collaborative planning and scheduling result. This method overcomes the randomness of renewable energy generation, the uncertainty of load, and the impact of load prediction errors, as well as the shortcomings of traditional planning methods that do not fully consider the physical constraints of various flexible resources and renewable energy and energy storage systems during actual operation. It can meet the power output demand of the system in specific time periods and lays the foundation for establishing a new power system optimization planning and scheduling operation model with a high proportion of renewable energy.

[0113] This embodiment provides a multi-energy resource collaborative planning and scheduling method, which can be used in electronic devices such as computers, mobile phones, and tablets. Figure 3 This is a flowchart of a multi-energy resource collaborative planning and scheduling method according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0114] Step S301: Obtain the initial distributed renewable energy historical output dataset, historical load failure probability dataset, target multi-energy resource dataset, and original resource dataset for the target planning location. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0115] Step S302 involves processing the initial distributed historical output dataset of new energy sources using a Kriging estimation sample interpolation method based on least squares support vector machines to obtain the time-series output prediction dataset of new energy sources at the target planning location. For details, please refer to [link to relevant documentation]. Figure 2 Step S202 of the illustrated embodiment will not be described again here.

[0116] Step S303: Establish a load forecasting model and determine the load forecasting dataset based on the historical load loss probability dataset. For details, please refer to [link to relevant documentation]. Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0117] Step S304: Establish a multi-objective planning and scheduling model based on the target multi-energy resource dataset and the original resource dataset.

[0118] Specifically, step S304 includes:

[0119] Step S3041: Obtain the objective function and the set of objective constraints.

[0120] In some optional implementations, step S3041 above includes:

[0121] Step a1: Obtain multiple resource cost data and an initial set of constraints.

[0122] Step a2: Establish the objective function based on multiple resource cost data.

[0123] Step a3: Use the Big-M method to transform the initial constraint set to obtain the target constraint set.

[0124] The Big-M method represents a technique for solving linear programming problems. It finds initial basic feasible solutions by introducing artificial variables into the constraints. It is applicable to linear programming problems where the constraints are equality or greater than or equal to the constraints.

[0125] Furthermore, multiple resource cost data can include the cost of purchasing electricity from the grid, new energy sources, energy efficiency improvements, demand response, energy storage investment costs, equipment maintenance and operation costs, etc.

[0126] Furthermore, the initial set of constraints may include: electricity and capacity-related electricity cost constraints; physical and operational constraints of existing new energy sources; physical and operational constraints of planned new energy sources and energy storage; financial constraints on investment costs; constraints on resilience requirements, environmental requirements, and economic expectations; power generation output constraints for specific time periods; and incentive constraints for new energy source confidence generation capacity, etc.

[0127] If the initial constraint set is nonlinear, it can be transformed into a linear target constraint set using the Big-M method.

[0128] Specifically, model objectives can be determined based on multiple resource cost data, including: new energy power generation resources, energy efficiency improvement, demand response, optimal energy storage capacity planning and scheduling, and minimizing total costs (investment costs, operating costs, electricity costs, carbon emission costs, confidence generation capacity incentives, etc.).

[0129] In this embodiment, the objective function is shown in the following relation (15):

[0130] Minimize:BillCost+IncestmentCost+DRCost+EECost+LostLoadCost

[0131] +WeightedEmssionCost-Revenue (15)

[0133] Where BillCost represents the cost of purchasing electricity from the grid, as shown in the following equation (16):

[0134]

[0135] In the formula: m represents the m-th month; M represents the total number of months, and we take M = 12; These represent the power of the peak load (peak), second peak load (mpeak), second valley load (opeak), and valley load (speak) in the m-th month, respectively. These represent the electricity demand prices (yuan / kW) for the corresponding time periods (or moments) of the peak load, second peak load, second valley load, and valley load in month m; These represent the electricity consumption during the corresponding time periods of the peak load (peak), second peak load (mpeak), second valley load (opeak), and valley load (speak) in the m-th month, respectively; These represent the electricity prices (yuan / kWh) for the corresponding time periods of peak load (peak), second peak load (mpeak), second valley load (opeak), and valley load (speak) in month m, respectively; MeterCharge represents the fixed cost of the electricity meter.

[0136] Furthermore, InvestmentCost represents the investment cost of distributed energy resources, as shown in the following equation (17):

[0137]

[0138] In the formula: Let C represent the photovoltaic system capacity variable; e represents the e-th energy efficiency improvement measure; E represents the total number of energy efficiency improvement measures; C pv This indicates the investment cost per unit of photovoltaic capacity; Indicates the planned energy storage capacity; C batt,p This indicates the investment cost per unit of energy storage capacity; Indicates planned energy storage capacity; C batt,e This indicates the investment cost per unit of energy storage capacity; This represents the installed capacity of the e-th energy efficiency improvement strategy; This represents the investment cost of the e-th energy efficiency improvement strategy.

[0139] Furthermore, DRCost represents the initial investment cost of the demand response, as shown in the following equation (18):

[0140]

[0141] In the formula: t represents the t-th time period; T represents the total number of time periods, taken as T = 8760; DR t C represents the demand response capacity during time period t; dr This indicates the cost of demand response investment.

[0142] Furthermore, EECost represents the initial investment cost for improving energy efficiency, as shown in the following equation (19):

[0143]

[0144] In the formula: NE represents the installed capacity of energy efficiency improvement strategy e; NE represents the total number of energy efficiency improvement strategies; Price_EE e This represents the investment cost of the e-th unit of the energy efficiency improvement strategy.

[0145] Furthermore, LostLoadCost represents the load loss cost, as shown in the following equation (20):

[0146]

[0147] In the formula: ll t It indicates lost load or emergency demand response; VOLL indicates the cost per unit of lost load or emergency demand response.

[0148] Furthermore, WeightedEmssionCost represents the emission cost, as shown in the following equation (21):

[0149]

[0150] In the formula: This indicates the amount of electricity purchased from the power grid during a given time period; w E This represents the carbon emission cost weighting coefficient; Φ represents the unit carbon emission price of the power grid during time period t; p This represents the carbon emission constant per kilowatt-hour of the power grid.

[0151] Furthermore, Revenue represents the revenue from power generation capacity, as shown in the following relationship (22):

[0152] Revenue = P r *Ccp (twenty two)

[0153] In the formula: P r Indicates the confidenced generation capacity supplied to the grid; C cp This indicates the price of the confidence generation capacity.

[0154] Furthermore, the demand constraints for capacity and electricity after new energy sources and energy storage in multi-energy resource planning are shown in the following equations (23) to (30), which are used to calculate the cost of purchasing electricity from the grid. Industrial and commercial electricity charges are calculated monthly according to the electricity charge structure, mainly including electricity charges and capacity charges. The planned new energy sources and energy storage change the net load shape by reducing peak load and reducing overall electricity consumption, thereby reducing demand charging and energy costs.

[0155]

[0156]

[0157] In the formula: Represents the set of peak load periods in the power grid; Represents the set of secondary peak load periods in the power grid; Represents the set of periods of sub-valley load in the power grid; This represents the set of periods of peak load on the power grid.

[0158] Furthermore, the total investment cost constraint is shown in the following equation (31):

[0159]

[0160] In the formula: MF represents the upper limit of the total initial investment cost.

[0161] Furthermore, the demand response planning and scheduling constraints are shown in the following equations (32) to (34):

[0162]

[0163] In the formula: DRH represents the upper limit of annual demand hours; Indicates the planned capacity for demand response; An indicator variable representing whether time period t participates in demand response ( Indicate participation, (Indicating non-participation); This indicates the maximum power of the demand response.

[0164] Furthermore, the output constraint of photovoltaic power generation is shown in the following relationship (35):

[0165]

[0166] In the formula: This represents the photovoltaic power generation coefficient for time period t, which is the new energy time-series output prediction data calculated in step S202. This indicates the planned photovoltaic output during time period t.

[0167] Furthermore, the maximum total photovoltaic power generation constraint is shown in the following equation (36):

[0168]

[0169] In the formula: SOLL represents the maximum photovoltaic output load ratio; D t This represents the total load during time period t.

[0170] Furthermore, the energy storage-related constraints are shown in the following equations (37) to (45). Here, the Big-M method is used to transform the semilinear constraints generated by the coupling of energy storage planning and operation into linear constraints, as shown in the following equations (43) and (44), in order to increase the model's solution efficiency, convergence, and ensure the global optimality of the solution results.

[0171]

[0172]

[0173] In the formula: This indicates the state of charge of the energy storage battery during time period t; This represents the charging power of the energy storage battery during time period t; CC represents the discharge power of the energy storage battery during time period t; batt This indicates the maximum number of cycles per year for energy storage batteries; An indicator variable representing whether the energy storage battery is in a charging state during time period t. Indicates charging. (Indicates no charging); An indicator variable representing whether the energy storage battery is in a discharged state during time period t. Indicates discharge. (Indicates no discharge); M big It represents a very large number.

[0174] Furthermore, the system power generation load constraint is shown in the following relationship (46):

[0175]

[0176] In the formula: This represents the planned photovoltaic output during time period t; This represents the discharge power of the energy storage battery during time period t; This indicates the output of the energy efficiency improvement strategy during time period t; This represents the output of the demand response during time period t.

[0177] Furthermore, during specific time periods GH and RH, the multi-energy system needs to meet the contractually agreed power generation output to obtain capacity revenue, as shown in equations (47) and (48). During the GH period, the multi-energy system must continuously generate electricity to ensure energy balance in the system, and during the RH period, the multi-energy system must be positive, meaning that the energy storage system cannot use the grid for charging during this period.

[0178]

[0179] Furthermore, the relevant constraints of the energy efficiency improvement plan are shown in the following equation (49):

[0180]

[0181] In the formula: This represents the energy-saving contribution per unit of time period t for the energy efficiency improvement strategy e.

[0182] Step S3042: Determine the output demand for the target time period based on the original resource dataset.

[0183] In some optional implementations, step S3042 above includes:

[0184] Step b1: Calculate power generation cost data based on preset conditions and the original resource dataset.

[0185] The preset condition is that no new renewable energy power generation resources or new energy storage equipment are considered.

[0186] Specifically, without considering new renewable energy power generation resources and new energy storage equipment, the power generation cost under the current conditions can be calculated based on the original resource dataset as a benchmark reference value.

[0187] Step b2: Using power generation cost data as a basic reference value, determine the power output demand for the target period using the electricity price curve or load peak-valley curve.

[0188] Among them, the electricity price curve is used to reflect the price changes of electricity resources; the load peak-valley curve is used to characterize the electricity load of the power system within a certain period of time. Through the load peak-valley curve, we can understand the operating status of the power system and provide support for power resource scheduling, planning and operation.

[0189] Specifically, using power generation cost data as a basic reference value, users can provide power generation output requirements for specific periods through electricity price curves or power companies can provide them through load peak-valley curves.

[0190] Step S3043: Based on the objective function and the set of objective constraints, a multi-objective planning and scheduling model is constructed using the target time period output demand and the target multi-energy resource dataset.

[0191] In some optional implementations, step S3043 above includes:

[0192] Step c1: Based on the objective function and the set of objective constraints, construct an initial planning and scheduling model using the target time period output demand and the target multi-energy resource dataset.

[0193] Step c2: Solve the initial planning and scheduling model to obtain the solution results.

[0194] Step c3: Determine whether the solution is the optimal solution.

[0195] Step c4: When the solution is the optimal solution, determine the initial planning and scheduling model as a multi-objective planning and scheduling model.

[0196] Step c5: If the solution is not optimal, adjust the initial constraint set and multiple resource cost data until an optimal solution exists and the multi-objective planning scheduling model is determined.

[0197] Specifically, under the training conditions of the objective function and the set of objective constraints, and with the output demand during the target time period as the objective condition, the model can be trained by combining the target multi-energy resource dataset to obtain a well-trained multi-initial planning and scheduling model.

[0198] Furthermore, tools such as MATLAB can be used to solve the model and obtain the solution results.

[0199] Furthermore, check whether the solution obtained is the globally optimal feasible solution.

[0200] Furthermore, if the current solution is the globally optimal feasible solution, it means that the total investment and operating costs are at their lowest at this time, while also meeting the power generation output demand for a specific period. In other words, the current initial planning and scheduling model is the final initial planning and scheduling model, which can provide reliable power generation capacity for the power grid during a specific period.

[0201] Furthermore, if the current solution is not the globally optimal feasible solution, the input parameters such as investment cost and maximum photovoltaic power output constraint are adjusted and the solution is recalculated until a multi-objective programming scheduling model that meets the conditions is obtained.

[0202] Step S305: Based on the new energy time-series output prediction dataset, historical load shedding probability dataset, and load prediction dataset, solve the multi-objective planning and scheduling model to obtain the multi-energy resource collaborative planning and scheduling results. For details, please refer to [link to relevant documentation]. Figure 1 Step S105 of the illustrated embodiment will not be described again here.

[0203] The multi-energy resource collaborative planning and scheduling method provided in this embodiment first processes the initial distributed renewable energy historical output dataset using a Kriging estimation sample interpolation method based on least squares support vector machines. This allows for the prediction of renewable energy time-series output at the target planning location, resulting in a predicted renewable energy time-series output dataset for that location. Secondly, a load forecasting model is established based on the historical load shedding probability dataset, and the load forecasting dataset is determined. Then, the Big-M method transforms the nonlinear initial constraint set into linear target constraints, ensuring the efficiency, convergence, and global optimality of the subsequent multi-objective planning and scheduling model. Simultaneously, existing resources can be used to calculate the current generation cost data that meets the conditions, thereby determining the output demand for the target time period that satisfies the current conditions. Furthermore, based on the obtained objective function and objective constraint set, and combined with the target period output demand and target multi-energy resource dataset, a multi-objective planning and scheduling model that meets the target period output demand can be constructed. This model overcomes the randomness of new energy power generation, the uncertainty of load, and the impact of load forecasting errors, as well as the shortcomings of traditional planning methods that do not fully consider the physical constraints of various flexible resources and the actual operation of new energy and energy storage systems. It can meet the system's output demand for specific periods and lays the foundation for establishing a new power system optimization planning and scheduling operation model with a high proportion of new energy.

[0204] In one example, a multi-energy resource collaborative optimization planning and scheduling method is provided to meet the output demand during a specific time period, such as... Figure 4 As shown, it includes:

[0205] Step S11: Normalize the historical time-series power output data of distributed new energy sources in the vicinity of the planned location, and use the Kriging estimation sample interpolation method based on least squares support vector machine to predict the time-series power output curve of new energy sources at the target planned location.

[0206] Among them, the power output curve of new energy at the planned site can be calculated by analyzing the relationship with the historical power output of existing new energy power plants in the nearby area, or it can be calculated based on the historical meteorological time series data detected by nearby meteorological stations. The first step is to normalize the historical time series data.

[0207] like Figure 5 As shown: The method of predicting the time-series power output curve of new energy sources at the target planning location using the Kriging estimation sample interpolation method based on least squares support vector machine consists of the following five steps:

[0208] Sub-step S21: Normalize the historical output time series data of new energy power plants in the vicinity of the planned location or the historical time series meteorological data of meteorological stations;

[0209] Sub-step S22: Based on the power output data of new energy power stations in the vicinity of the planned location and the spatial distance between the stations, calculate the experimental variance function γ of the Kriging sample point interpolation method. * (h ij );

[0210] Sub-step S23: Use a least squares support vector machine (LS-SVM) to fit the experimental variogram γ. * (h ij );

[0211] Sub-step S24: Obtain the optimal γ(h) ij );

[0212] Sub-step S25: According to γ(h) ij Calculate the output parameter λ of each new energy power station. i i = 1…n, and then the estimated value Z of the new energy output at the planned target location can be calculated. * (x0).

[0213] Finally, the normalized time-series power output curve of the new energy source at the proposed location was calculated based on the historical normalized time-series power output data of the new energy sources in the vicinity of the planned location.

[0214] Step S12: Based on the load loss probability reliability index requirements, the load prediction error and randomness are modeled in the form of probability distribution using a chance constraint model, and the chance constraints of the load loss probability index requirements expressed in the form of probability distribution are transformed into linear inequality constraints.

[0215] Step S13: Based on 8760 hours of time series data, and taking into account load forecasting errors, new energy sources, energy efficiency improvement, demand response, energy storage, financial and environmental constraints, construct an objective function and constraint conditions, and a multi-objective planning and scheduling model that satisfies environmental requirements and optimizes total cost.

[0216] Specifically, based on 8,760 hours of time-series data per year, it can effectively take into account the seasonality, holidays, monthly and weekly periodicity of new energy output and load, and at the same time, it can more accurately model the time coupling constraints of energy storage, and explore the long-term scheduling potential of energy storage across days, weeks and seasons.

[0217] For example, firstly, the annual 8760-hour load curve data, new energy physical and price data, energy storage physical parameters and price data, energy efficiency improvement range and costs, demand response electricity cost data, new energy storage incentive policy data, and carbon emission parameters and price data are read into the computer program. Without considering new new energy power generation resources and new energy storage equipment, the current power generation cost is calculated as a benchmark reference value. Users provide specific time-period power generation output requirements based on the electricity price curve, or power companies provide them based on the load peak-valley curve; a complete multi-objective programming scheduling model is constructed and solved to obtain the planning results and scheduling operation strategy that meet the power generation output requirements for specific time periods. Constraints define the continuous power output demand of the system for specific time periods, ensuring that the planning and scheduling results can provide reliable power output to the grid during the specified time periods, reducing the impact of the volatility and uncertainty of new energy on the stability and security of the grid, while providing flexibility resources to the grid. Figure 6 As shown, the specific steps include:

[0218] Sub-step 1: Initialize the data source file location or database read related parameters.

[0219] Sub-step two: Reading and preprocessing the raw data, inputting specific time periods and power generation demand information. This includes 8760-hour load curve data, new energy physical parameters and pricing data, energy storage physical parameters and pricing data, electricity cost data, new energy storage incentive policy data, and carbon emission parameters. The cost required to meet the current load is calculated based on existing resources, and the existing resource scheduling results are used as reference data.

[0220] Sub-step 3: Construct a multi-objective planning and scheduling model based on the input data and initial parameters, and find the global optimal solution. The Big-M method is used to transform the nonlinear constraints in constructing the planning and scheduling co-optimization model into linear constraints to ensure model efficiency, convergence, and the global optimality of the results.

[0221] Sub-step four: Check if a globally optimal feasible solution exists. If no feasible solution is found, adjust the input parameters such as investment cost and maximum photovoltaic power output constraint, and solve again. If a globally optimal feasible solution is found, the total investment and operating cost is minimized, while simultaneously meeting the power output demand for a specific period, thus providing reliable power generation capacity to the grid during that specific period.

[0222] The multi-objective programming scheduling model aims to obtain the optimal multi-energy resource planning and scheduling operation strategy. The objective is to minimize total cost, including the cost of purchasing electricity from the grid, renewable energy sources, energy efficiency improvements, demand response, energy storage investment costs, equipment maintenance and operation costs, etc. The model considers the following objectives and constraints, detailed in step S3041 above.

[0223] Step S14: Input the normalized new energy time-series output curve predicted at the planned location in Step S11, the historical load shedding probability in Step S12, and the load prediction time-series data into the planning and scheduling model in Step S13 to solve for the optimal solution that meets various constraints and multiple objectives, such as the output demand in a specific time period.

[0224] This example provides a multi-energy resource collaborative optimization planning and scheduling method that meets the output demand for a specific time period. It fully considers the environment, meteorology, and uncertainties of existing new energy output in the planning area, avoiding prediction biases caused by data from a single observation or measurement point. Considering the uncertainty of load demand, the method introduces the probability of system load failure, ensuring that the planning and scheduling results meet system reliability requirements. By modeling the 8706-hour time period, it can cover the seasonal, monthly, and weekly periodic characteristics of multi-energy resources. Through joint planning and scheduling of multi-energy resources, it provides a reliable representation of power generation capacity and flexibility, and unifies the planning and scheduling of various energy forms such as new energy, energy storage, demand response, and energy efficiency improvement. This overcomes the influence of the randomness and uncertainty of new energy power generation, as well as the shortcomings of traditional planning methods that do not fully consider various flexible resources and the physical constraints of actual operation between new energy and energy storage systems. Therefore, this example characterizes the reliable power generation capacity and flexibility of multi-energy resource collaboration, overcoming the randomness of new energy power generation, load uncertainty and load forecasting errors, as well as the shortcomings of traditional planning methods that do not fully consider the physical constraints of multiple flexible resources, new energy and energy storage systems during actual operation. It can meet the power output demand of the system in specific periods and lay the foundation for establishing a new power system optimization planning and dispatch operation model with a high proportion of new energy.

[0225] This embodiment also provides a multi-energy resource collaborative planning and scheduling device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0226] This embodiment provides a multi-energy resource collaborative planning and scheduling device, such as... Figure 7 As shown, it includes:

[0227] The acquisition module 701 is used to acquire the initial distributed new energy historical output dataset, historical load failure probability dataset, target multi-energy resource dataset, and original resource dataset of the target planning location.

[0228] The processing module 702 is used to process the initial distributed historical output dataset of new energy sources through the Kriging estimation sample interpolation method based on least squares support vector machine to obtain the time series output prediction dataset of new energy sources at the target planning location.

[0229] The Establishment and Determination module 703 is used to establish a load forecasting model and determine the load forecasting dataset based on the historical load loss probability dataset.

[0230] Module 704 is established to build a multi-objective planning and scheduling model based on the target multi-energy resource dataset and the original resource dataset.

[0231] The solver module 705 is used to solve the multi-objective planning and scheduling model based on the new energy time-series output prediction dataset, the historical load failure probability dataset, and the load prediction dataset, and obtain the multi-energy resource collaborative planning and scheduling results.

[0232] In some alternative implementations, the processing module 702 includes:

[0233] The first processing submodule is used to normalize the initial distributed renewable energy historical output dataset to obtain the target distributed renewable energy historical output dataset.

[0234] The calculation submodule is used to calculate the initial variable variance function based on the target distributed new energy historical output dataset.

[0235] The second processing submodule is used to fit the initial variable variance function using a least squares support vector machine to obtain the target variable variance function.

[0236] The third processing submodule is used to obtain the new energy time-series output prediction dataset for the target planning location by processing it with the Kriging estimation sample interpolation method based on the target variable variance function.

[0237] In some alternative implementations, the establishment module 704 includes:

[0238] The `get` submodule is used to obtain the objective function and the set of objective constraints.

[0239] The first determination submodule is used to determine the output demand for the target time period based on the original resource dataset.

[0240] A submodule is constructed to build a multi-objective planning and scheduling model based on the objective function and the set of objective constraints, using the output demand during the target time period and the target multi-energy resource dataset.

[0241] In some optional implementations, the acquisition submodule includes:

[0242] The acquisition unit is used to acquire multiple resource cost data and initial constraint sets.

[0243] Establish a unit to create an objective function based on multiple resource cost data.

[0244] The transformation unit is used to transform the initial constraint set using the Big-M method to obtain the target constraint set.

[0245] In some alternative implementations, the construction submodule includes:

[0246] The building unit is used to construct an initial planning and scheduling model based on the objective function and the set of objective constraints, using the target time period output demand and the target multi-energy resource dataset.

[0247] The solution unit is used to solve the initial planning and scheduling model and obtain the solution results.

[0248] The first determining unit is used to determine the initial planning and scheduling model as a multi-objective planning and scheduling model when the solution result is the optimal solution.

[0249] The second determining unit is used to adjust the initial constraint set and multiple resource cost data when the solution is not the optimal solution, until an optimal solution exists and the multi-objective planning scheduling model is determined.

[0250] In some alternative implementations, the first determining submodule includes:

[0251] The calculation unit is used to calculate power generation cost data based on preset conditions and the original resource dataset.

[0252] The third determining unit is used to determine the output demand for the target period based on the power generation cost data as a reference value and by using the electricity price curve or load peak-valley curve.

[0253] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0254] In this embodiment, the multi-energy resource collaborative planning and scheduling device is presented in the form of functional units. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0255] This invention also provides a computer device having the above-described features. Figure 7 The multi-energy resource collaborative planning and scheduling device shown is shown.

[0256] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 8As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 8 Take a processor 10 as an example.

[0257] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0258] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0259] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0260] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0261] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0262] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0263] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0264] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A multi-energy resource collaborative planning and scheduling method, characterized in that, The method includes: The process involves acquiring an initial distributed renewable energy historical output dataset, a historical load shedding probability dataset, a target multi-energy resource dataset, and a raw resource dataset for the target planning location. The initial distributed renewable energy historical output dataset includes multiple historical output time-series data of renewable energy power plants in the vicinity of the target planning location or historical time-series meteorological data from multiple meteorological stations. The target multi-energy resource dataset includes annual 8760-hour load curve data, renewable energy physical and price data, energy storage physical parameters and price data, multiple electricity cost data, renewable energy storage incentive data, and carbon emission parameters and price data. The historical load shedding probability dataset is used to characterize the reliability index requirements for load shedding probability. The raw resource dataset includes relevant data of existing resources but does not include the target multi-energy resource dataset. The initial distributed historical output dataset of new energy sources is processed by the Kriging estimation sample interpolation method based on least squares support vector machine to obtain the time series output prediction dataset of new energy sources at the target planning location. A load prediction model is established based on the historical load loss probability dataset, and the load prediction dataset is determined. A multi-objective planning and scheduling model is established based on the target multi-energy resource dataset and the original resource dataset. Based on the new energy time-series output prediction dataset, the historical load failure probability dataset, and the load prediction dataset, the multi-objective planning and scheduling model is solved to obtain the multi-energy resource collaborative planning and scheduling results. Specifically, the initial distributed historical output dataset of new energy sources is processed using a Kriging estimation sample interpolation method based on least squares support vector machine to obtain the time-series output prediction dataset of new energy sources at the target planning location, including: The initial distributed renewable energy historical output dataset is normalized to obtain the target distributed renewable energy historical output dataset. Based on the target distributed new energy historical output dataset, an initial variable variance function is calculated, wherein, based on the target distributed new energy historical output dataset and the spatial distance between power stations, an experimental variable variance function of the Kriging estimation sample interpolation method is calculated, and the experimental variable variance function is the initial variable variance function. The initial variance function is fitted using a least squares support vector machine to obtain the target variance function. The least squares support vector machine directly fits the experimental variance function based on its own distribution map. Based on the target variance function, the new energy time-series output prediction dataset for the target planning location is obtained by processing it with the Kriging estimation sample interpolation method. The process of establishing a load forecasting model based on the historical load loss probability dataset includes: converting the opportunity constraints, which are expressed in the form of a probability distribution, into linear inequality constraints; and, based on the linear inequality constraints, modeling the load forecasting error and randomness in the form of a probability distribution using an opportunity constraint model to obtain the load forecasting model. The linear inequality constraints are expressed as follows: In the formula: express Electricity purchased from the power grid at all times This indicates the planned photovoltaic power output. Indicates energy storage output. This indicates the contribution of energy efficiency improvement measures. Indicates the contribution of demand response efforts; express Percentiles of the standard normal distribution; express Load forecasting error at any given time; It represents the mathematical expectation of a random variable; express Forecasted load at any given time.

2. The method according to claim 1, characterized in that, A multi-objective planning and scheduling model is established based on the target multi-energy resource dataset and the original resource dataset, including: Obtain the objective function and the set of objective constraints; Determine the output demand for the target time period based on the original resource dataset; Based on the objective function and the set of objective constraints, the multi-objective planning and scheduling model is constructed using the target time period output demand and the target multi-energy resource dataset.

3. The method according to claim 2, characterized in that, Obtain the objective function and the set of objective constraints, including: Obtain multiple resource cost data and initial constraint sets; Based on the aforementioned resource cost data, the objective function is established; The initial constraint set is transformed using the Big-M method to obtain the target constraint set.

4. The method according to claim 3, characterized in that, Based on the objective function and the set of objective constraints, and utilizing the target time period output demand and the target multi-energy resource dataset, the multi-objective planning and scheduling model is constructed, including: Based on the objective function and the set of objective constraints, an initial planning and scheduling model is constructed using the target time period output demand and the target multi-energy resource dataset. Solve the initial planning and scheduling model to obtain the solution results; Determine whether the solution result is the optimal solution; When the solution result is the optimal solution, the initial planning and scheduling model is determined to be the multi-objective planning and scheduling model. If the solution result is not the optimal solution, adjust the initial constraint set and the multiple resource cost data until an optimal solution exists and determine the multi-objective planning and scheduling model.

5. The method according to claim 2, characterized in that, Determine the output demand for the target time period based on the original resource dataset, including: Based on preset conditions, calculate power generation cost data according to the original resource dataset; Based on the power generation cost data as a reference value, the power output demand for the target period is determined using the electricity price curve or the load peak-valley curve.

6. A multi-energy resource collaborative planning and scheduling device, characterized in that, The device includes: The acquisition module is used to acquire the initial distributed renewable energy historical output dataset, the historical load failure probability dataset, the target multi-energy resource dataset, and the original resource dataset for the target planning location. The initial distributed renewable energy historical output dataset includes multiple historical output time-series data of renewable energy power plants in the vicinity of the target planning location or historical time-series meteorological data from multiple meteorological stations. The target multi-energy resource dataset includes annual 8760-hour load curve data, renewable energy physical and price data, energy storage physical parameters and price data, multiple electricity cost data, renewable energy storage incentive data, and carbon emission parameters and price data. The historical load failure probability dataset is used to characterize the reliability index of the load failure probability. The original resource dataset includes relevant data of currently existing resources, but does not include the target multi-energy resource dataset. The processing module is used to process the initial distributed historical output dataset of new energy sources through the Kriging estimation sample interpolation method based on least squares support vector machine to obtain the time series output prediction dataset of new energy sources at the target planning location. The module for establishing and determining load prediction models is used to establish load prediction models and determine load prediction datasets based on the historical load loss probability dataset. A module is established to build a multi-objective planning and scheduling model based on the target multi-energy resource dataset and the original resource dataset. The solution module is used to solve the multi-objective planning and scheduling model based on the new energy time-series output prediction dataset, the historical load failure probability dataset, and the load prediction dataset, so as to obtain the multi-energy resource collaborative planning and scheduling results. The processing module includes: The first processing submodule is used to normalize the initial distributed new energy historical output dataset to obtain the target distributed new energy historical output dataset. The calculation submodule is used to calculate the initial variable variance function based on the target distributed new energy historical output dataset, wherein the experimental variable variance function of the Kriging estimation sample interpolation method is calculated based on the target distributed new energy historical output dataset and the spatial distance between power stations, and the experimental variable variance function is the initial variable variance function. The second processing submodule is used to fit the initial variance function with a least squares support vector machine to obtain the target variance function. The least squares support vector machine directly fits the experimental variance function based on its own distribution map. The third processing submodule is used to obtain the new energy time-series output prediction dataset of the target planning location based on the target variable variance function and through Kriging estimation sample interpolation method. The establishment module is specifically used to: convert the opportunity constraints, which are expressed in the form of probability distribution, into linear inequality constraints; and based on the linear inequality constraints, model the load forecasting error and randomness in the form of probability distribution using the opportunity constraint model to obtain the load forecasting model. The linear inequality constraints are expressed as follows: In the formula: express Electricity purchased from the power grid at all times This indicates the planned photovoltaic power output. Indicates energy storage output. This indicates the contribution of energy efficiency improvement measures. Indicates the contribution of demand response efforts; express Percentiles of the standard normal distribution; express Load forecasting error at any given time; It represents the mathematical expectation of a random variable; express Forecasted load at any given time.

7. A computer device, characterized in that, include: A memory and a processor are interconnected, the memory stores computer instructions, and the processor executes the computer instructions to perform the multi-energy resource collaborative planning and scheduling method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the multi-energy resource collaborative planning and scheduling method according to any one of claims 1 to 5.

9. A computer program product, characterized in that, It includes computer instructions for causing a computer to execute the multi-energy resource collaborative planning and scheduling method according to any one of claims 1 to 5.

Images