A power transmission network planning method considering wind and light uncertainty and data center virtual energy storage characteristics

Abstract

The application discloses a power transmission network planning method considering wind and light uncertainty and data center virtual energy storage characteristics, comprising the following steps: (1) obtaining basic parameters required for power transmission network optimization planning; (2) establishing a data center virtual charging and discharging model according to the data center virtual energy storage characteristics; (3) constructing an upper-layer network frame planning model; (4) introducing a conditional risk value to measure the risk loss caused by wind and light uncertainty; (5) constructing a lower-layer actual energy storage planning model; (6) solving the model to obtain an optimal solution of the power transmission network frame planning. The application considers that the data center is used as a virtual energy storage resource to participate in the power transmission network planning, actual energy storage investment can be effectively reduced on the basis of reducing the system net load peak-valley difference, the economy of the power transmission network planning is improved, and meanwhile, the application quantifies the influence of different risk aversion degrees on power transmission line construction and actual energy storage capacity configuration.

Description

Technical Field

[0001] This invention relates to the field of power system planning technology, specifically to a power transmission network planning method that considers the uncertainties of wind and solar power and the characteristics of virtual energy storage in data centers. Background Technology

[0002] The output of wind and solar power is random and fluctuating, and large-scale wind and solar grid connection poses challenges to power system planning and operation, especially in addressing transmission congestion and excessive peak-shaving pressure on the grid. Optimizing the planning of transmission line construction and actual energy storage system capacity can help alleviate insufficient peak-shaving caused by the mismatch between wind / solar power and load, and improve the economy and security of grid planning. However, the investment and operating costs of actual energy storage are high, and with the continuous increase in the current wind and solar grid connection capacity, higher requirements are placed on the capacity and charging / discharging strategies of actual energy storage, leading to further increases in actual energy storage costs and making large-scale promotion and application difficult. If the concept of actual energy storage is expanded to include loads with flexible and adjustable demand-side characteristics as virtual energy storage and its participation in transmission grid planning, the economy of transmission grid planning can be further improved.

[0003] Meanwhile, as digital technology continues to permeate all sectors of the economy and society, data centers, as core facilities of digital "new infrastructure," are experiencing continuous growth in scale and energy consumption. The high electricity costs resulting from their massive power consumption will constrain the sustainable development of data centers. On the other hand, data loads have spatiotemporal transfer characteristics, making them a demand-side flexible resource capable of participating in grid response. If the virtual energy storage characteristics of data centers are considered in the planning of transmission grids with a high proportion of wind and solar energy, it can not only reduce actual energy storage investment and improve the economy and flexibility of transmission grid planning, but also lower the energy costs of data centers. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a power grid planning method that considers the uncertainties of wind and solar power and the characteristics of virtual energy storage in data centers. This method reduces actual energy storage investment and improves the economic efficiency of power grid planning by minimizing the net load peak-to-valley difference.

[0005] The objective of this invention is achieved through the following technical solution: The present invention provides a power grid planning method that considers the uncertainties of wind and solar power and the virtual energy storage characteristics of data centers, comprising the following steps: Step S1: Obtain the basic parameters required for power transmission network optimization planning; Step S2: Based on the virtual energy storage characteristics of the data center, establish a virtual charging and discharging model for the data center; Step S3: Construct an upper-level grid planning model that minimizes the investment cost of transmission lines and maximizes the available transfer capacity (ATC) to decide on the construction of transmission lines; Step S4: Introduce conditional value at risk to measure the risk loss caused by the uncertainty of wind and solar power, and establish a mathematical model; Step S5: Construct a lower-level actual energy storage planning model with the actual energy storage investment cost, the compensation cost for the data center, and the risk loss cost as the minimum, optimize the actual energy storage configuration, and improve the investment decision of the upper-level transmission line; Step S6: The improved nested hybrid particle swarm optimization algorithm is used to iteratively solve the bi-level programming model to obtain the optimal solution for the power transmission network architecture planning.

[0006] Furthermore, in step S1, the basic parameters required for power grid optimization planning include: wind farm access location and predicted output curve, photovoltaic power station access location and predicted output curve, actual energy storage access location, conventional power plant access location and rated power, data center access location, data center server parameters, load values ​​of each node, and basic line parameter information. Furthermore, in step S2, the virtual charging and discharging model of the data center is as follows: (1) In the formula, T represents the total time period divided into planning cycles; Let t be the power of the data center at time t; , The charging and discharging status of the virtual energy storage in the data center at time t; , These represent the discharge and charging power of the virtual energy storage in the data center at time t, respectively. The amount of interactive data load that needs to be processed in the data center at time t; This represents the original batch data load in the data center at time t. The data load of batch processing data transferred into the data center at time t; This represents the batch processing data load transferred out of the data center at time t. The operating power of a single server in the data center when idle; This refers to the operating power of a single server in a data center under full load. Data center power efficiency is defined as the ratio of total data center energy consumption to IT facility energy consumption. Let t be the rate at which the data center server processes the data load. Let t be the number of servers running at time t.

[0007] Furthermore, in step S3, a superstructure-grid planning model is constructed to minimize the investment cost of transmission lines and maximize the available transmission capacity, and the construction of transmission lines is decided upon. The objective function is: (2) (3) In the formula, N B For the set of candidate routes; N u Re represents the scene set; Re represents the set of nodes in the receiving region. For the upper-level network planning model, represents whether or not to construct the candidate line between nodes ij; The investment cost of constructing new lines between nodes ij; The active power of the regular load at node i in scenario u, excluding the data center; The active power of the data center at node i in scenario u; The active power of node i in scenario u under normal load conditions, excluding the data center; This represents the active power of node i in the data center of scenario u under the base state.

[0008] Furthermore, step S3 also includes the constraints for constructing the upper-level grid planning model, including constraints on the number of new lines, line transmission capacity, power balance, node voltage, and conventional power plant operation.

[0009] Furthermore, in step S4, the conditional value of risk is calculated at a confidence level. Below, CVaR is the conditional mean when the risk loss exceeds VaR. VaR reflects the maximum potential loss of the grid portfolio at a given confidence level. Considering the uncertainties of wind farm and photovoltaic power plant output, CVaR is used to measure the risk loss of wind and solar curtailment caused by wind and solar forecast errors. The loss function is: (9) In the formula, x represents the decision variables for power grid planning, including whether the lines are built and the actual energy storage capacity; y represents uncertainties, including the output of wind farms and photovoltaic power plants. , These represent the absolute prediction errors for wind power and solar power in scenario u at time t, respectively. Let be the total absolute prediction error for wind power and solar power under scenario u at time t; The negative spinning reserve capacity of the system under scenario u at time t represents the minimum output capacity of the controllable power generation equipment; The unit penalty cost for curtailing wind and solar power.

[0010] Furthermore, in step S5, a lower-level actual energy storage planning model is constructed by minimizing the actual energy storage investment cost, the compensation cost for the data center, and the risk loss cost. This optimizes the actual energy storage configuration and improves the investment decision for the upper-level transmission lines. The objective function is: (13) In the formula, This represents the actual investment cost of energy storage. To compensate for the costs of data centers; L is the risk aversion coefficient, representing the degree of investor's aversion to risk. The larger the L value, the more risk factors the investor considers; For the CVaR of the power grid portfolio; The actual energy storage investment cost is: (14) In the formula, This refers to the investment cost per unit capacity of an actual energy storage system. The planned capacity of the actual energy storage system deployed at node i; The discount rate; This refers to the actual service life of the energy storage system. The compensation cost given to the data center is: (15) In the formula, To compensate the data center for its unit cost; Let t be the data center power of node i in scenario u at time t.

[0011] Furthermore, in step S6, a random function is introduced to improve the nested hybrid particle swarm optimization algorithm, where the inertia weight and learning factor are expressed as follows: (30) (31) (32) In the formula, This refers to the number of real-time iterations. This represents the maximum number of iterations. , These are the minimum and maximum values ​​of the inertia weight, respectively. A random value between 0 and 1; learning factor , The range of values ​​for are respectively , .

[0012] The beneficial effects of this invention are: This invention considers data centers as virtual energy storage resources to participate in power grid planning, which can effectively reduce actual energy storage investment and improve the economic efficiency of power grid planning while reducing the peak-valley difference of system net load. At the same time, this invention quantifies the impact of different risk aversion levels on power line construction and actual energy storage capacity configuration.

[0013] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will now be described in further detail with reference to the accompanying drawings, wherein: Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 A flowchart illustrating an embodiment of a specific application; Figure 3 An improved 18-node system architecture diagram; Figure 4 This is a schematic diagram of the power transmission network planning results under different risk aversion coefficients. Detailed Implementation

[0015] The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be understood that the preferred embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0016] like Figure 1 As shown, the method of the present invention specifically includes the following detailed steps: Step S1: Obtain the basic parameters required for power grid optimization planning; In this embodiment, the basic parameters required for power grid optimization planning include the wind farm access location and predicted output curve, the photovoltaic power station access location and predicted output curve, the actual energy storage access location, the conventional power plant access location and rated power, the data center access location, the data center server parameters, the load value of each node and the basic parameter information of the line.

[0017] Step S2: Based on the virtual energy storage characteristics of the data center, establish a virtual charging and discharging model for the data center; the virtual charging and discharging model for the data center is as follows: (1) In the formula, T represents the total time period divided into planning cycles; Let t be the power of the data center at time t; , The charging and discharging status of the virtual energy storage in the data center at time t; , These represent the discharge and charging power of the virtual energy storage in the data center at time t, respectively. The amount of interactive data load that needs to be processed in the data center at time t; This represents the original batch data load in the data center at time t. The data load of batch processing data transferred into the data center at time t; This represents the batch processing data load transferred out of the data center at time t. The operating power of a single server in the data center when idle; This refers to the operating power of a single server in a data center under full load. Data center power efficiency is defined as the ratio of total data center energy consumption to IT facility energy consumption. Let t be the rate at which the data center server processes the data load. Let t be the number of servers running at time t.

[0018] Step S3: Construct an upper-level grid planning model that minimizes transmission line investment costs and maximizes available transmission capacity to determine the construction of transmission lines. The objective function is: (2) (3) In the formula, N B For the set of candidate routes; N u Re represents the scene set; Re represents the set of nodes in the receiving region. For the upper-level network planning model, represents whether or not to construct the candidate line between nodes ij; The investment cost of constructing new lines between nodes ij; The active power of the regular load at node i in scenario u, excluding the data center; The active power of the data center at node i in scenario u; The active power of node i in scenario u under normal load conditions, excluding the data center; This represents the active power of node i in the data center of scenario u under the base state.

[0019] The constraints for constructing the upper-level grid planning model include constraints on the number of new lines, line transmission capacity, power balance, node voltage, and conventional power plant operation.

[0020] Step S4: Introduce conditional value at risk to measure the risk loss caused by the uncertainty of wind and solar power, and establish a mathematical model; conditional value at risk is defined as the risk loss at a certain confidence level. Below, CVaR is the conditional mean when the risk loss exceeds VaR. VaR reflects the maximum potential loss of the grid portfolio at a given confidence level. Considering the uncertainties of wind farm and photovoltaic power plant output, CVaR is used to measure the risk loss of wind and solar curtailment caused by wind and solar forecast errors. The loss function is: (9) In the formula, x represents the decision variables for power grid planning, including whether the lines are built and the actual energy storage capacity; y represents uncertainties, including the output of wind farms and photovoltaic power plants. , These represent the absolute prediction errors for wind power and solar power in scenario u at time t, respectively. Let be the total absolute prediction error for wind power and solar power under scenario u at time t; The negative spinning reserve capacity of the system under scenario u at time t represents the minimum output capacity of the controllable power generation equipment; The unit penalty cost for curtailing wind and solar power.

[0021] The expression for CVaR is: (10) In the formula, Number of scenes; Confidence level; Let u be the probability of scene u. The loss function; The value at which the risk loss exceeds the maximum loss. express ; The value of VaR can be calculated using equation (12): (11) (12) In the formula, Let be the probability density function of wind farms and photovoltaic power plants; The loss function is no greater than the boundary value. The probability distribution function.

[0022] Step S5: Construct a lower-level actual energy storage planning model by minimizing the actual energy storage investment cost, the compensation cost for the data center, and the risk loss cost. This optimizes the actual energy storage configuration and improves the investment decision for upper-level transmission lines. The objective function is: (13) In the formula, This represents the actual investment cost of energy storage. To compensate for the costs of data centers; L is the risk aversion coefficient, representing the degree of investor's aversion to risk. The larger the L value, the more risk factors the investor considers; For the CVaR of the power grid portfolio; The actual energy storage investment cost is: (14) In the formula, This refers to the investment cost per unit capacity of an actual energy storage system. The planned capacity of the actual energy storage system deployed at node i; The discount rate; This refers to the actual service life of the energy storage system. The compensation cost given to the data center is: (15) In the formula, To compensate the data center for its unit cost; Let t be the data center power of node i in scenario u at time t.

[0023] Step S6: An improved nested hybrid particle swarm optimization algorithm is used to iteratively solve the bi-level programming model to obtain the optimal solution for the power grid network planning. The nested hybrid particle swarm optimization algorithm is improved by introducing a random function; the inertia weight and learning factor are expressed as follows: (30) (31) (32) In the formula, This refers to the number of real-time iterations. This represents the maximum number of iterations. , These are the minimum and maximum values ​​of the inertia weight, respectively. A random value between 0 and 1; learning factor , The range of values ​​for are respectively , .

[0024] Based on the overall methodological design described above, Figure 2 The flowchart below is a specific embodiment of the present invention. The method of the present invention will be further described by a specific implementation process, which includes the following detailed steps.

[0025] Step 1: Use Monte Carlo to generate the initial scene and use the k-means method to reduce the scene to obtain a typical scene of landscape and power output; Step 2: Input the data center load level and power grid system parameters; Step 3: Initialize the actual energy storage capacity; Step 4: Encode the decision variables of the upper-level optimization model into binary form as x = x1, x2, ..., x D ; Step 5: Construct an upper-level grid planning model with the objectives of minimizing transmission line investment cost and maximizing ATC expected value, and make decisions on the construction of transmission lines. The objective function is shown in the following formula. (2) (3) In the formula, N B For the set of candidate routes; N u Re represents the scene set; Re represents the set of nodes in the receiving region. For the upper-level network planning model, represents whether or not to construct the candidate line between nodes ij; The investment cost of constructing new lines between nodes ij; The active power of the regular load at node i in scenario u, excluding the data center; The active power of the data center at node i in scenario u; The active power of node i in scenario u under normal load conditions, excluding the data center; This represents the active power of node i in the data center of scenario u under the base state.

[0026] The constraints of the upper-level space frame planning model include: ①Constraints on the number of newly built lines: (4) In the formula, This represents the maximum number of lines that can be constructed between nodes ij; ② Line transmission capacity constraints: (5) In the formula, This represents the power transmitted through the line between node i and node j. This represents the maximum transmission capacity of the line between node i and node j. ③ Power balance constraints: (6) In the formula, , These represent the active and reactive power outputs of the conventional power plant at node i in scenario u, respectively. , These represent the regular active and reactive loads of node i in scenario u, excluding the data center. These represent the active and reactive data center loads of node i in scenario u, respectively. , These represent the active and reactive power outputs of the wind farm at node i in scenario u, respectively. , These represent the active and reactive power outputs of the photovoltaic power station at node i in scenario u, respectively. , These represent the active and reactive power outputs of the actual energy storage at node i in scenario u, respectively. , These are the voltage amplitudes at nodes i and j in scenario u, respectively. , These represent the conductance and susceptance between nodes ij, respectively; The phase angle difference between nodes ij and u in scene u; ④ Node voltage constraints: (7) In the formula, Let be the minimum voltage at node i; The maximum voltage at node i; ⑤ Operational constraints of conventional power plants: (8) In the formula, , These represent the minimum and maximum output power of the conventional power plant located at node i, respectively.

[0027] Step 6: Determine whether the obtained solution meets the upper-level convergence condition. If yes, output the current upper-level optimal solution; otherwise, execute step 5. Step 7: Use the output of the upper-level optimal solution as the input data for the lower-level model; Step 8: Optimize the actual energy storage capacity with the goal of minimizing the actual energy storage investment cost, the compensation cost for the data center, and the risk loss cost (i.e., construct a lower-level actual energy storage planning model, optimize the actual energy storage configuration, and improve the upper-level transmission line investment decision), as shown in the following formula; (13) In the formula, This represents the actual investment cost of energy storage. To compensate for the costs of data centers; L is the risk aversion coefficient, representing the degree of investor's aversion to risk. The larger the L value, the more risk factors the investor considers; CVaR for the power grid portfolio.

[0028] The actual investment cost of energy storage is: (14) In the formula, This refers to the investment cost per unit capacity of an actual energy storage system. The planned capacity of the actual energy storage system deployed at node i; The discount rate; This refers to the actual service life of the energy storage system.

[0029] The compensation cost for the data center is: (15) In the formula, To compensate the data center for its unit cost; Let t be the data center power of node i in scenario u at time t.

[0030] The constraints of the lower-level actual energy storage planning model include system power balance constraints, wind and solar power output constraints, actual energy storage system constraints, wind and solar-load matching constraints, data load constraints, data center operation constraints, and data center user response time constraints, specifically: ①System power balance constraints; (16) ② Wind and solar power output constraints: (17) In the formula, This represents the maximum output of the photovoltaic power station at node i. , These are the upper and lower limits of the power output of the wind farm at node i, respectively. ③Constraints of actual energy storage systems: (18) (19) (20) In the formula, The rated power of the actual energy storage system deployed at node i; , These are the minimum and maximum states of charge of the actual energy storage system deployed at node i, respectively. ④ Wind-solar-load matching constraints (twenty one) (twenty two) (twenty three) (twenty four) In the formula, Let be the sum of the virtual energy storage and actual energy storage output of the data center at node i at time t in scenario u; Let i be the original load of node i at time t in scenario u. Let be the equivalent load of node i at time t in scenario u, where the virtual energy storage and actual energy storage of the data center are involved. The wind-solar load matching degree of node i in scenario u; The desired level of wind-solar load matching; ⑤ Data load constraints: (25) In the formula, The maximum latency limit for batch data load processing in data centers; ⑥ Data center operation constraints: (26) In the formula, This represents the maximum number of servers in a data center. ⑦ Data center user response time constraints: (27) (28) (29) In the formula, The service rate at time t represents the data center's ability to handle interactive data loads at time t. The average service rate of interactive data load at time t; , These represent the average queuing delay and average processing delay of the interactive data load at time t, respectively. The maximum tolerable time for users.

[0031] Step 9: Determine whether the obtained solution meets the convergence condition of the lower level. If yes, output the current optimal solution of the lower level. If no, jump to step 8.

[0032] Step 10: Use the output of the lower-level optimal solution as part of the upper-level objective function and jump to step 5.

[0033] To further understand this invention, a modified nested hybrid particle swarm optimization algorithm is used to solve and analyze a case study in the improved 18-node system. The structure diagram of the improved 18-node system is attached. Figure 3 As shown in the diagram. Nodes 11, 16, and 18 in this system are connected to a 500MW wind farm, a 360MW photovoltaic power station, actual energy storage, and a data center. The data center load includes batch processing loads and interactive data loads. It is assumed that the processing time for batch processing loads is within 4 hours, the data center load allocation delay is ignored, and the maximum delay time limit for all batch processing loads is the same.

[0034] Table 1 compares the planning results of the traditional power grid planning method that only considers the joint planning of transmission lines and actual energy storage with the power grid planning method of this study that considers virtual energy storage in data centers. The planning method proposed in this invention reduces the actual energy storage capacity configuration by 14% and the transmission line investment cost by 4.7% compared to the traditional method. This demonstrates that the strategy in this paper effectively reduces the investment cost of the power transmission system. This is because the proposed planning method considers both actual energy storage and virtual energy storage in data centers. Compared to considering only actual energy storage, it reduces the power generation corresponding to the smoothing of net load fluctuations by actual energy storage, thereby reducing actual energy storage investment. Therefore, configuring actual energy storage systems can reduce the construction of transmission lines in power grid planning, increase available transmission capacity, and, considering virtual energy storage in data centers, utilize its batch processing data load transfer characteristics on a time scale to smooth net load fluctuations, which can reduce the output of actual energy storage for smoothing net load fluctuations, thereby reducing actual energy storage investment and improving the economics of power grid planning.

[0035] Table 1 Comparison of transmission network planning results under the two methods Comparison Projects Traditional methods This article's method Actual energy storage capacity / MW 492 / 586 / 507 412 / 530 / 428 Transmission line investment cost / 10,000 yuan 316950 301950 Risk cost of wind and solar power curtailment / 10,000 yuan 2165.3 2121.2 Actual energy storage investment cost (ten thousand yuan) 37234.7 32183.9 Compensation cost for data centers / 10,000 RMB / 41.15 Appendix Figure 4 The results of power transmission network planning under different risk aversion coefficients are presented. From Figure 4 It can be seen that when investors are aggressive (L=0.1 to L=0.5), the maximum tolerable wind and solar curtailment risk decreases from 45% to 17%. The sensitivity of reducing wind and solar curtailment risk by increasing actual energy storage capacity is 0.86, and the sensitivity of reducing wind and solar curtailment risk by adding transmission lines is 0.14. The actual energy storage capacity increases significantly, while the number of newly built transmission line circuits changes little. When investors are conservative (L=1.5), the maximum tolerable wind and solar curtailment risk is 12%. The sensitivity of reducing wind and solar curtailment risk by increasing actual energy storage capacity and adding transmission lines is 0.57 and 0.43, respectively. In this case, compared with the aggressive investor, the actual energy storage capacity and the number of newly built transmission line circuits increase significantly. This is because the different risk aversion coefficients lead investors to consider the degree of wind and solar curtailment risk losses caused by uncertain wind and solar planning to different extents, thus affecting the transmission network planning scheme and the actual energy storage optimization configuration. Therefore, the construction of transmission lines, the configuration of actual energy storage capacity, and the risks and costs associated with wind and solar power curtailment primarily depend on the relative relationship between the cost coefficient and risk aversion coefficient of the transmission system. Consequently, investors should select an appropriate risk aversion coefficient based on their expected risk level to weigh the economics and security of transmission network planning. This allows them to reduce the risks and costs associated with wind and solar power curtailment by increasing actual energy storage capacity and adding transmission lines. Furthermore, increasing actual energy storage capacity is more sensitive to reducing the risks of wind and solar power curtailment compared to adding transmission lines.

[0036] It should be noted that any process or method description in the flowchart or otherwise described herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which the embodiments of the invention pertain.

[0037] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0038] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0039] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A power transmission network planning method considering wind and solar uncertainty and data center virtual energy storage characteristics, characterized in that: The method includes the following steps: Step S1: Obtain the basic parameters required for power transmission network optimization planning; Step S2: Based on the virtual energy storage characteristics of the data center, establish a virtual charging and discharging model for the data center; in step S2, the virtual charging and discharging model for the data center is as follows: （1） In the formula, T represents the total time period divided into planning cycles; Let t be the power of the data center at time t; , The charging and discharging status of the virtual energy storage in the data center at time t; , These represent the discharge and charging power of the virtual energy storage in the data center at time t, respectively. The amount of interactive data load that needs to be processed in the data center at time t; This represents the original batch data load in the data center at time t. The data load of batch processing data transferred into the data center at time t; This represents the batch processing data load transferred out of the data center at time t. The operating power of a single server in the data center when idle; This refers to the operating power of a single server in a data center under full load. Data center power efficiency is defined as the ratio of total data center energy consumption to IT facility energy consumption. Let t be the rate at which the data center server processes the data load. Let t be the number of servers running. Step S3: Construct an upper-level grid planning model that minimizes transmission line investment costs and maximizes available transmission capacity; Step S4: Introduce Conditional Value at Risk (VaR) to measure the risk loss caused by the uncertainty of wind and solar power, and establish a mathematical model; in Step S4, VaR is defined as the risk value at a certain confidence level. Below, CVaR is the conditional mean when the risk loss exceeds VaR. VaR reflects the maximum potential loss of the grid portfolio at a given confidence level. Considering the uncertainties of wind farm and photovoltaic power plant output, CVaR is used to measure the risk loss of wind and solar curtailment caused by wind and solar forecast errors. The loss function is: （9） In the formula, x represents the decision variables for power grid planning, including whether the lines are built and the actual energy storage capacity; y represents uncertainties, including the output of wind farms and photovoltaic power plants. , These represent the absolute prediction errors for wind power and solar power in scenario u at time t, respectively. Let be the total absolute prediction error for wind power and solar power under scenario u at time t; The negative spinning reserve capacity of the system under scenario u at time t represents the minimum output capacity of the controllable power generation equipment; The unit penalty cost for curtailing wind and solar power; The expression for CVaR is: （10） In the formula, Number of scenes; Confidence level; Let u be the probability of scene u. The loss function; The value at which the risk loss exceeds the maximum loss. express ; The value of VaR is calculated using equation (12): （11） （12） In the formula, Let be the probability density function of wind farms and photovoltaic power plants; The loss function is no greater than the boundary value. The probability distribution function; Step S5: Construct a lower-level actual energy storage planning model with the actual energy storage investment cost, the compensation cost for the data center, and the risk loss cost as the minimum, and optimize the actual energy storage configuration; Step S6: The improved nested hybrid particle swarm optimization algorithm is used to iteratively solve the bi-level programming model to obtain the optimal solution for the power transmission network architecture planning.

2. The power grid planning method considering wind and solar uncertainties and the virtual energy storage characteristics of data centers according to claim 1, characterized in that: In step S1, the basic parameters required for power grid optimization planning include: wind farm access location and predicted output curve, photovoltaic power station access location and predicted output curve, actual energy storage access location, conventional power plant access location and rated power, data center access location, data center server parameters, load values ​​of each node, and basic line parameter information.

3. The power grid planning method considering wind and solar uncertainties and the virtual energy storage characteristics of data centers according to claim 1, characterized in that: In step S3, a superstructure-grid planning model is constructed to minimize transmission line investment costs and maximize available transmission capacity, and the construction of transmission lines is decided upon. The objective function is: （2） （3） In the formula, N B For the set of candidate routes; N u Re represents the scene set; Re represents the set of nodes in the receiving region. For the upper-level network planning model, represents whether or not to construct the candidate line between nodes ij; The investment cost of constructing new lines between nodes ij; The active power of the regular load at node i in scenario u, excluding the data center; The active power of the data center at node i in scenario u; The active power of node i in scenario u under normal load conditions, excluding the data center; This represents the active power of node i in the data center of scenario u under the base state.

4. The power grid planning method considering wind and solar uncertainties and the virtual energy storage characteristics of data centers according to claim 1, characterized in that: It also includes constraints for constructing the upper-level grid planning model. These constraints include constraints on the number of new transmission lines, transmission capacity of transmission lines, power balance, node voltage, and operation of conventional power plants. Specifically, these include: ①Constraints on the number of newly built lines: （4） In the formula, This represents the maximum number of lines that can be constructed between nodes ij; ② Line transmission capacity constraints: （5） In the formula, This represents the power transmitted through the line between node i and node j. This represents the maximum transmission capacity of the line between node i and node j. ③ Power balance constraints: (6) In the formula, , These represent the active and reactive power outputs of the conventional power plant at node i in scenario u, respectively. , These represent the regular active and reactive loads of node i in scenario u, excluding the data center. , These represent the active and reactive data center loads of node i in scenario u, respectively. , These represent the active and reactive power outputs of the wind farm at node i in scenario u, respectively. , These represent the active and reactive power outputs of the photovoltaic power station at node i in scenario u, respectively. , These represent the active and reactive power outputs of the actual energy storage at node i in scenario u, respectively. , These are the voltage amplitudes at nodes i and j in scenario u, respectively. , These represent the conductance and susceptance between nodes ij, respectively; The phase angle difference between nodes ij and u in scene u; ④ Node voltage constraints: （7） In the formula, Let be the minimum voltage at node i; The maximum voltage at node i; ⑤ Operational constraints of conventional power plants: （8） In the formula, , These represent the minimum and maximum output power of the conventional power plant located at node i, respectively.

5. The power grid planning method considering wind and solar uncertainties and the virtual energy storage characteristics of data centers according to claim 1, characterized in that: In step S5, a lower-level actual energy storage planning model is constructed by minimizing the actual energy storage investment cost, the compensation cost for the data center, and the risk loss cost. This optimizes the actual energy storage configuration and improves the investment decision for the upper-level transmission lines. The objective function is: （13） In the formula, This represents the actual investment cost of energy storage. To compensate for the costs of data centers; L is the risk aversion coefficient, representing the degree of investor's aversion to risk. The larger the L value, the more risk factors the investor considers; For the CVaR of the power grid portfolio; The actual energy storage investment cost is: （14） In the formula, This refers to the investment cost per unit capacity of an actual energy storage system. The planned capacity of the actual energy storage system deployed at node i; The discount rate; This refers to the actual service life of the energy storage system. The compensation cost given to the data center is: （15） In the formula, To compensate the data center for its unit cost; Let t be the data center power of node i in scenario u at time t.

6. The power grid planning method considering wind and solar uncertainties and the virtual energy storage characteristics of data centers according to claim 5, characterized in that: Step S5 also includes constructing the constraints of the lower-level actual energy storage planning model. These constraints include system power balance constraints, wind and solar power output constraints, actual energy storage system constraints, wind / solar-load matching constraints, data load constraints, data center operation constraints, and data center user response time constraints. Specifically, these include: ①System power balance constraints; （16） ② Wind and solar power output constraints: （17） In the formula, This represents the maximum output of the photovoltaic power station at node i. , These are the upper and lower limits of the power output of the wind farm at node i, respectively. ③Constraints of actual energy storage systems: （18） （19） （20） In the formula, The rated power of the actual energy storage system deployed at node i; , These are the minimum and maximum states of charge of the actual energy storage system deployed at node i, respectively. ④ Wind-solar-load matching constraints （21） （22） （23） （24） In the formula, Let be the sum of the virtual energy storage and actual energy storage output of the data center at node i at time t in scenario u; Let i be the original load of node i at time t in scenario u. Let be the equivalent load of node i at time t in scenario u, where the virtual energy storage and actual energy storage of the data center are involved. The wind-solar load matching degree of node i in scenario u; The desired level of wind-solar load matching; ⑤ Data load constraints: （25） In the formula, The maximum latency limit for batch data load processing in data centers; ⑥ Data center operation constraints: （26） In the formula, This represents the maximum number of servers in a data center. ⑦ Data center user response time constraints: （27） （28） （29） In the formula, The service rate at time t represents the data center's ability to handle interactive data loads at time t. The average service rate of interactive data load at time t; , These represent the average queuing delay and average processing delay of the interactive data load at time t, respectively. The maximum tolerable time for users.

7. The power grid planning method considering wind and solar uncertainties and the virtual energy storage characteristics of data centers according to claim 1, characterized in that: In step S6, a random function is introduced to improve the nested hybrid particle swarm optimization algorithm. The inertia weight and learning factor are expressed as follows: （30） （31） （32） In the formula, This refers to the number of real-time iterations. This represents the maximum number of iterations. , These are the minimum and maximum values ​​of the inertia weight, respectively. A random value between 0 and 1; Learning factor , The range of values ​​for are respectively , .

Images