Power distribution network resource planning method and system considering flexibility of DG and data center

Abstract

The invention discloses a power distribution network resource planning method and system considering DG and data center flexibility, and the method comprises the steps: determining the output of distributed power generation through a power grid load prediction model, and correcting a power grid load prediction result; constructing an energy storage optimization model, taking an energy storage installation position, capacity and time sequence output as decision variables, and solving to obtain an energy storage capacity configuration and initial power output scheduling scheme under the condition of meeting an energy storage operation constraint condition; constructing a data load transfer optimization model, and solving to obtain an initial optimization result of the transfer quantity of each data center and the target time interval distribution under the condition of meeting the business constraint condition of the data center; optimizing communication network topology according to physical side energy storage configuration, a data load transfer mode and a power grid topology structure; and constructing a collaborative optimization model, introducing a rolling optimization and communication reliability correction mechanism, solving the collaborative optimization model, and obtaining a finally deployed collaborative scheduling scheme. According to the method, the power distribution network resources are optimized cooperatively, and the economy and reliability of the power distribution network are optimal.

Description

Technical Field

[0001] This invention belongs to the field of power distribution network planning technology, and more specifically, relates to a power distribution network resource planning method and system that takes into account the flexibility of distribution networks and data centers. Background Technology

[0002] With the transformation of the global energy structure and the advancement of digitalization, distributed generation and data centers are gradually becoming important components of modern power distribution networks. Distributed generation (DG), such as solar and wind power, has been widely used around the world due to its sustainability and environmentally friendly characteristics. However, while improving energy efficiency, distributed generation also brings new challenges to the stable operation of power distribution networks, particularly in terms of the instability and volatility of energy output. These issues lead to the overreach of power distribution network operations, thereby affecting the stability and reliability of power supply.

[0003] Meanwhile, with the rapid development of the internet and cloud computing, the scale and number of data centers have also increased dramatically. As modern information technology infrastructure, data centers have extremely high electricity demands and experience significant load fluctuations. Peak electricity demand from data centers typically occurs during periods of power shortage, which is highly correlated with the load demand of the power distribution network. Therefore, how to flexibly adjust the load of data centers and coordinate it with the load demand of the power distribution network has become an important direction for current power system optimization.

[0004] To address the above issues, recent research has focused on combining distributed energy storage with data center load regulation. This involves collaboratively planning energy storage configurations, load shifting schemes, and communication network topologies to address the limit-crossing problems caused by distributed generation integration into the distribution network. In particular, the power regulation and voltage regulation characteristics of distributed energy storage systems can effectively balance load fluctuations in the power grid. Meanwhile, data centers, through their potential for spatiotemporal load shifting, can adjust their power demand at different times, thereby avoiding excessive pressure on the power grid during peak load periods.

[0005] Currently, although some distribution network optimization methods involve energy storage systems and load dispatch, most methods are limited to single power load forecasting or energy storage unit configuration, and do not fully consider the combined effects of distributed generation and data center flexibility. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a distribution network resource planning method and system that takes into account the flexibility of distributed generation (DG) and data centers. Addressing the issue that existing algorithms fail to effectively resolve the impact of distributed generation access and data center load shifting on the optimal scheduling of distribution network resources, this invention utilizes a collaborative optimization algorithm for optimal planning of distribution network resources.

[0007] The present invention adopts the following technical solution.

[0008] The first aspect of the present invention provides a distribution network resource planning method that takes into account the flexibility of DG and data centers, comprising the following steps: Step 1: Based on the grid load forecasting model under the condition of distributed generation access, determine the output of distributed generation in different time periods, take into account the impact of uncertain generation and load, and correct the grid load forecasting results; Step 2: Based on the characteristics of distributed energy storage power regulation and voltage support, construct an energy storage optimization model. Using the energy storage installation location, capacity, and time-series output as decision variables, solve for the energy storage capacity configuration and initial power output scheduling scheme under the constraints of energy storage operation. Step 3: Based on the spatiotemporal portability of data center load, construct a data load transfer optimization model. Under the condition of satisfying data center business constraints, solve for the initial optimization results of the transfer volume and target time period allocation of each data center. Step 4: Optimize the communication network topology based on the physical energy storage configuration, data load transfer method, and power grid topology. Step 5: Construct a collaborative optimization model and introduce rolling optimization and communication reliability correction mechanisms. Using the energy storage capacity configuration and initial power output scheduling scheme in Step 2, and the initial optimization results of the data center transfer volume and target time period allocation in Step 3 as initial solutions, solve the collaborative optimization model to obtain the final deployed collaborative scheduling scheme.

[0009] Preferably, in step 1, the power grid load forecasting model is expressed by the following formula:

[0010] In the formula, The predicted power for time period t, including generating power or load power. For the current power, For the first k The change in each predictor factor The weighting coefficient for each predictor. n The number of predictive factors, including weather and load.

[0011] Preferably, the prediction results are corrected for errors, and the power grid load prediction model is adjusted. The error correction process is expressed by the following formula:

[0012] In the formula, This is the corrected prediction error. For prediction error, This represents the actual error. This is a correction factor.

[0013] Preferably, in step 2, the objective function of the energy storage optimization model is... F 1. This includes minimizing network losses, energy storage operating costs, and energy storage investment costs, expressed by the following formula:

[0014] In the formula, For time period t Network loss, It is used for energy storage and output, where discharging is positive and charging is negative. For the rated energy of energy storage, , , These are the energy storage network loss coefficient, energy storage operating cost coefficient, and energy storage investment cost coefficient, respectively. S A collection of energy storage units. T This represents the number of time periods.

[0015] Preferably, in step 2, the constraints of the energy storage optimization model include overall system energy storage power balance constraints, node power balance constraints, and energy storage power and energy constraints. The overall energy storage power balance constraint of the system is expressed by the following formula:

[0016] In the formula, This refers to the total energy storage regulation power at the system level. To contribute to energy storage, S A collection of energy storage units; The node power balance constraint is expressed by the following formula:

[0017] In the formula, S ( n ) is the access node n Energy storage collection, , The inflow and outflow power at the nodes are determined by the power flow equations. For distributed generation power prediction, Based on the predicted power of the basic load; Energy storage capacity and energy constraints are expressed by the following formula:

[0018]

[0019] In the formula, For energy storage i Maximum power, For energy storage i The state of charge, and For energy storage i The minimum and maximum capacity.

[0020] Preferably, in step 3, the optimization objective of the load transfer optimization model is set as minimizing the electricity cost and migration cost under time-of-use pricing, and the objective function is... F 2 is expressed by the following formula:

[0021] In the formula, For time-of-use electricity pricing, For data centers j The transferred load, This represents the cost coefficient for data center load migration. D For data center collection, T This represents the number of time periods.

[0022] Preferably, in step 3, the constraints of the load transfer optimization model include load transfer summary constraints, maximum flexibility range constraints, and task duration constraints. The load transfer aggregation constraint is expressed by the following formula:

[0023] In the formula, for t Overall load transfer during the period for t The overall transferable limit of the time period system For load transfer amount, For data centers j The transferred load, This represents the baseline IT load for the data center. The maximum flexibility range constraint is expressed by the following formula:

[0024] In the formula: This is the load transfer factor; The task deadline constraint is expressed by the following formula:

[0025] In the formula, For data centers j The task completion time window.

[0026] Preferably, the proportion of transferable load to the baseline IT load of the data center is used as the load transfer factor. It can be expressed by the following formula:

[0027] In the formula, For data centers j Transferable load, The upper limit of the load transfer factor is given by the service level agreement or business policy. This serves as the baseline IT load for the data center.

[0028] Preferably, step 4 includes: Step 4.1: Real-time forecasting of power demand and load for each distribution node, and calculation of the optimal communication path for each node; Step 4.2: Optimize the communication network topology in the distribution network based on graph theory models; Step 4.3: Adjust the connection method of the communication node according to the power demand and load changes of each power distribution node.

[0029] Preferably, in step 5, the main objective of the collaborative optimization model is constructed using energy storage output, energy storage capacity, and data center load transfer as joint decision variables, expressed by the following formula:

[0030] In the formula, For time period t Network loss, For energy storage i of efforts, For energy storage i Rated energy, For data centers j The transferred load, For data centers j Load transfer amount, , , These are the energy storage network loss coefficient, energy storage operating cost coefficient, and energy storage investment cost coefficient, respectively. For time-of-use electricity pricing, This represents the cost coefficient for data center load migration. T The number of time periods. S A collection of energy storage units. D For data center collections.

[0031] Preferably, in step 5, energy storage and data center are explicitly coupled to the same node to obtain the node power balance constraint of the collaborative optimization model, expressed by the following formula:

[0032] In the formula, For energy storage i of efforts, For data centers j The transferred load, , For node inflow and outflow power, For distributed generation power prediction, Based on the predicted power of the basic load, S ( n ) is the access node n Energy storage collection, D ( n ) is the access node n A collection of data centers.

[0033] Preferably, in step 5, the rolling optimization mechanism includes: Define prediction error , Net deviation It can be expressed by the following formula:

[0034]

[0035] In the formula: Based on the measured power of the base load, Based on the predicted power of the basic load, This represents the measured power output of distributed generation. Predicted power output for distributed generation; The system imbalance degree, which reflects load balance, is defined by the following formula:

[0036] In the rolling time domain, the joint optimization results of energy storage and data center load are used as input, and Re-engage in online optimization as an additional objective.

[0037] Preferably, in step 5, the communication reliability correction mechanism includes: Define communication adjustment factor It can be expressed by the following formula:

[0038] In the formula, clip(*,0,1) represents a function that truncates the value of * to [0,1]. As a communication reliability indicator, This represents the minimum communication reliability standard. The optimal solutions for energy storage output and data center load are corrected using the following formulas to obtain the final collaborative scheduling scheme for deployment:

[0039]

[0040] In the formula, To contribute to the final deployment of energy storage, For the final deployed data center load, Energy storage output for security strategies during communication degradation The minimum load for the business is determined based on the importance of the business itself and expert experience. and These are the optimal solutions for energy storage and data center load after rolling optimization, respectively.

[0041] A second aspect of the present invention provides a distribution network resource planning system that takes into account the flexibility of distribution generation (DG) and data centers, and the method for running the distribution network resource planning system that takes into account the flexibility of distribution generation (DG) and data centers includes: The load forecasting module is used to determine the output of distributed generation in different time periods based on the grid load forecasting model under the condition of distributed generation access, taking into account the impact of uncertain generation and load, and correcting the grid load forecasting results. The energy storage optimization module is used to construct an energy storage optimization model based on the power regulation and voltage support characteristics of distributed energy storage. It uses the energy storage installation location, capacity and time-series output as decision variables, and solves the energy storage capacity configuration and initial power output scheduling scheme under the condition of satisfying the energy storage operation constraints. The load transfer module is used to build a data load transfer optimization model based on the spatiotemporal portability characteristics of data center load. Under the condition of satisfying data center business constraints, it solves the initial optimization results of the transfer volume and target time period allocation for each data center. The communication optimization module is used to optimize the communication network topology based on the physical energy storage configuration, data load transfer method, and power grid topology. The collaborative optimization module is used to construct a collaborative optimization model and introduces rolling optimization and communication reliability correction mechanisms. It uses the initial optimization results of energy storage capacity configuration and initial power output scheduling scheme, and the transfer volume and target time period allocation of each data center as the initial solution to solve the collaborative optimization model and obtain the final deployed collaborative scheduling scheme.

[0042] Compared with the prior art, the beneficial effects of the present invention include at least the following: 1. The distribution network resource planning method and system of the present invention, which takes into account the flexibility of distributed generation and data centers, achieves optimal economy and reliability of distribution network by coordinating and optimizing energy storage configuration, data load transfer and communication network topology optimization of distribution network resource allocation, while effectively solving the operational challenges caused by distributed generation and data center load fluctuations.

[0043] 2. The distribution network resource planning method and system of the present invention, which takes into account the flexibility of distributed energy and data centers, can effectively reduce the risk of distribution network operation exceeding limits under the conditions of distributed energy access and data center load fluctuations, while improving the utilization rate of renewable energy and ensuring the stability and security of the distribution network in complex operating environments. Attached Figure Description

[0044] Figure 1 This is a flowchart of a power distribution network resource planning method that takes into account the flexibility of DG and data centers, provided according to an embodiment of the present invention. Figure 2 This is a power distribution network topology diagram provided in accordance with an embodiment of the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0046] like Figure 1 As shown, Embodiment 1 of the present invention provides a distribution network resource planning method that takes into account the flexibility of DG and data centers, including the following steps: Step 1: Based on the grid load forecasting model under distributed generation access, determine the output of distributed generation and its impact on the distribution network load in different time periods. Taking into account the impact of uncertain generation and load, the grid load forecasting results are corrected.

[0047] In a preferred but non-limiting embodiment of the present invention, step 1 specifically includes: Step 1.1: Based on the multi-period forecast of power grid operation status and load demand, construct a power grid load forecasting model to predict the output changes of distributed generation and the load changes of data centers in different time periods.

[0048] More preferably, the power grid load forecasting model is expressed by the following formula:

[0049] In the formula, The predicted power for time period t, including generating power or load power. For the current power, For the first k The change in each predictor factor The weighting coefficient for each predictor. n The number of predictive factors, including weather, load, etc.

[0050] Step 1.2 involves error correction of the prediction results. An adaptive algorithm is used to adjust the power grid load prediction model from Step 1.1, reducing load prediction errors and optimizing the distribution network dispatching strategy. The error correction process is expressed by the following formula:

[0051] In the formula, This is the corrected prediction error. For prediction error, This represents the actual error. This is a correction factor.

[0052] Step 1.3: Solve the power grid load forecasting model and perform error correction to obtain the distributed generation forecast power. and base load forecast power .

[0053] Step 2: Based on the characteristics of distributed energy storage power regulation and voltage support, construct an energy storage optimization model. Using the energy storage installation location, capacity, and timing output as decision variables, solve for the energy storage capacity configuration and initial power output scheduling scheme under the constraints of energy storage operation.

[0054] In a preferred but non-limiting embodiment of the present invention, step 2 specifically includes: Step 2.1, Define the objective function of the energy storage optimization model. F 1. This includes minimizing network losses, energy storage operating costs, and energy storage investment costs. The objective function is expressed by the following formula:

[0055] In the formula, For time period t Network loss, It is used for energy storage and output, where discharging is positive and charging is negative. For the rated energy of energy storage, , , These are the energy storage network loss coefficient, energy storage operating cost coefficient, and energy storage investment cost coefficient, respectively. S A collection of energy storage units. T This represents the number of time periods.

[0056] Step 2.2: Set the constraints of the energy storage optimization model, namely the energy storage operation constraints, including the overall system energy storage power balance constraints, node power balance constraints, and energy storage power and energy constraints.

[0057] More preferably, the overall energy storage power balance constraint of the system is expressed by the following formula:

[0058] In the formula, This refers to the total energy storage regulation power at the system level.

[0059] More preferably, the node power balance constraint is expressed by the following formula:

[0060] In the formula, S ( n ) is the access node n Energy storage collection, , The inflow and outflow power at the nodes are determined by the power flow equations. For distributed generation power prediction, The predicted power for the base load is obtained from step 1.

[0061] More preferably, the energy storage power and energy (SOC) constraint is expressed by the following formula:

[0062]

[0063] In the formula, For energy storage i Maximum power, For energy storage i The state of charge, and For energy storage i The minimum and maximum capacity.

[0064] Step 2.3: Solve the energy storage optimization model to obtain the energy storage capacity configuration and the initial optimized output of the energy storage. This refers to the initial power output scheduling scheme.

[0065] Step 3: Based on the spatiotemporal portability of data center load, construct a data load transfer optimization model. Under the condition of satisfying data center business constraints, solve for the transfer volume and target time period allocation of each data center, thereby achieving peak shaving and valley filling and limit overrun mitigation.

[0066] In a preferred but non-limiting embodiment of the present invention, step 3 specifically includes: Step 3.1: Define the load transfer coefficient and set the optimization objective of the load transfer optimization model.

[0067] More preferably, the load transfer coefficient Defined as a data center j During the period t The proportion of transferable load to the baseline IT load of the data center is expressed by the following formula:

[0068] In the formula, For data centers j Transferable load, The upper limit of the load transfer factor is given by the service level agreement or business policy. This serves as the baseline IT load for the data center.

[0069] More preferably, the load transfer optimization objective is set as minimizing the electricity cost and migration cost under time-of-use pricing, with the objective function... F 2 is expressed by the following formula:

[0070] In the formula, For time-of-use electricity pricing, For data centers j The transferred load, This represents the cost coefficient for data center load migration. D For data center collections.

[0071] Step 3.2: Construct the constraints of the load transfer optimization model based on the load transfer coefficient, namely the data center business constraints, including the load transfer aggregation constraint, the maximum flexibility range constraint, and the task deadline constraint.

[0072] More preferably, the load transfer aggregation constraint is expressed by the following formula:

[0073] In the formula, for t Overall load transfer during the period for t The overall transferable limit of the time period system This represents the load transfer amount.

[0074] More preferably, the maximum flexibility range constraint is expressed by the following formula:

[0075] More preferably, the task deadline constraint is expressed by the following formula:

[0076] In the formula, For data centers j The task completion time window.

[0077] Step 3.3: Solve the data load transfer optimization model to obtain the initial optimization results of the data center load. This refers to the initial optimization results of the transfer volume and target time period allocation for each data center.

[0078] Step 4: Optimize the communication network topology based on the physical energy storage configuration, data load transfer method, and power grid topology to ensure real-time information exchange and remote control capabilities between the data center and the distribution network.

[0079] In a preferred but non-limiting embodiment of the present invention, step 4 specifically includes: Step 4.1: Real-time forecasting of power demand and load for each distribution node, and calculation of the optimal communication path for each node.

[0080] Step 4.2: Based on the graph theory model, optimize the communication network topology in the distribution network to ensure the efficiency of information exchange and avoid power dispatching problems caused by communication delays.

[0081] Step 4.3: Adjust the connection method of the communication node according to the power demand and load changes of each distribution node to improve the information transmission rate and ensure real-time scheduling and data monitoring.

[0082] Step 5: By coordinating the optimization of energy storage configuration, data load transfer, and communication network topology, a coordinating optimization model is constructed. Under the premise of satisfying the energy storage operation constraints in Step 2 and the data center business constraints in Step 3, the overall economy and reliability are optimized. Rolling optimization and communication reliability correction mechanisms are introduced into the coordinating optimization model so that real-time load changes, distributed generation output, and communication status are directly included in the scheduling calculation. The initial optimization results of the energy storage capacity configuration and initial power output scheduling scheme in Step 2 and the transfer volume and target time period allocation of each data center in Step 3 are used as the initial solutions to solve the coordinating optimization model and obtain the final deployed coordinating scheduling scheme.

[0083] In a preferred but non-limiting embodiment of the present invention, step 5 specifically includes: Step 5.1: Using energy storage output, energy storage capacity, and data center load transfer as joint decision variables, construct the main objective of the collaborative optimization model, expressed by the following formula:

[0084] In the formula, For time period t Network loss, For energy storage i of efforts, For energy storage i Rated energy, For data centers j The transferred load, For data centers j Load transfer amount, , , These are the energy storage network loss coefficient, energy storage operating cost coefficient, and energy storage investment cost coefficient, respectively. For time-of-use electricity pricing, This represents the cost coefficient for data center load migration.

[0085] Step 5.2, explicitly couple energy storage and data center to the same node to obtain the node power balance constraint of the collaborative optimization model, expressed by the following formula:

[0086] In the formula, , For node inflow and outflow power, For distributed generation power prediction, The predicted power for the base load is obtained from step 1. S ( n ) is the access node n Energy storage collection, D ( n ) is the access node n A collection of data centers.

[0087] Understandably, this formula allows the energy storage output and the data center transfer volume to work together on the node's net load, thereby achieving coordinated scheduling in the same optimization.

[0088] Step 5.3: Solve the objective function from Step 5.1 and the constraints from Step 5.2 together, and then apply the initial optimized output results of energy storage and data center load from Steps 2 and 3. and Substituting these values ​​as initial inputs, we can obtain the joint optimization result, which is the output of energy storage and data center load. and .

[0089] Step 5.4, taking into account real-time load and distributed generation output, performs rolling optimization on the joint optimization results in the rolling time domain.

[0090] In a further preferred embodiment of the present invention, step 5.4 specifically includes: Step 5.4.1, define the prediction error , Net deviation It can be expressed by the following formula:

[0091]

[0092] In the formula: Based on the measured power of the base load, Based on the predicted power of the basic load, This represents the measured power output of distributed generation. Predicted power output for distributed generation.

[0093] Step 5.4.2 defines the system imbalance degree to reflect load balance, expressed by the following formula:

[0094] Step 5.4.3: In the rolling time domain, combine the joint optimization results of energy storage and data center load. and As input, As an additional objective, it re-enters the online optimization process, thereby making load balancing explicitly quantifiable and computable. Rolling optimization then yields the optimal solutions for energy storage and data center loads. and .

[0095] Step 5.5: Correct the rolling optimization results based on the communication reliability index to obtain the final collaborative scheduling scheme.

[0096] In a further preferred embodiment of the present invention, step 5.5 specifically includes: Step 5.5.1, define the communication adjustment factor. It can be expressed by the following formula:

[0097] In the formula, clip(*,0,1) represents a function that truncates the value of * to [0,1]. This is a communication reliability index, a normalized value in the range [0,1]. Higher communication availability generally indicates better reliability. The closer to 1, the greater the reliance on online collaborative optimization instructions; when communication degrades, A value close to 0 indicates a greater adoption of local security strategies. This is the minimum communication reliability indicator, which can be set to 0.7 based on expert experience.

[0098] Step 5.5.2, using the communication adjustment factor Correcting the optimal solution for energy storage and data center load and This leads to the final collaborative scheduling scheme for deployment.

[0099] More preferably, the optimal solution for energy storage output and data center load is corrected according to the following formula:

[0100]

[0101] In the formula, For energy storage output under a security strategy during communication degradation, the security strategy can remain unchanged from the energy storage output scheme of the previous moment; The minimum load for the business is determined based on the importance of the business itself and the experience of experts.

[0102] It is understandable that this is due to the same communication regulation factor. Unified control refines the optimal solution for energy storage output and data center load, resulting in the final energy storage output and data center load scheme for deployment. and It can achieve overall coordinated scheduling of energy storage, data center and communication.

[0103] In a preferred but non-limiting embodiment of the present invention, the distribution network resource planning method taking into account the flexibility of DG and data centers further includes verifying the final co-optimization result, including the following steps: Step A.1: Perform simulation analysis on the optimized energy storage configuration scheme to evaluate the power loss and load balancing effect under different configurations.

[0104] The formula for power loss analysis is:

[0105] in, For total power loss, Let i be the power output of the i-th energy storage unit. Here, N represents the loss factor of the corresponding energy storage unit, and N is the number of energy storage units.

[0106] Step A.2: By constructing a simulation system, the data load transfer scheme is tested to verify the stability and reliability of the system under different load conditions.

[0107] The stability verification formula for data load balancing is:

[0108] in, As a measure of system stability, For the current load, For the transferred load, T This represents the number of time periods.

[0109] Step A.3: Apply this method in multiple real-world power distribution network scenarios to verify its feasibility in actual operation, and further optimize the algorithm based on the experimental results.

[0110] The formula for verification in a real-world scenario is:

[0111] in, For the feasibility of the method, S For the number of scenes, For the first s Performance metrics for each scenario.

[0112] Step A.4: If the verification passes, output the final optimal planning scheme for distribution network resources; otherwise, return to step 2.

[0113] Embodiment 2 of the present invention provides a distribution network resource planning system that takes into account the flexibility of distribution generation (DG) and data centers, and runs the distribution network resource planning method for considering the flexibility of distribution generation (DG) and data centers described in Embodiment 1, including: The load forecasting module is used to determine the output of distributed generation in different time periods based on the grid load forecasting model under the condition of distributed generation access, taking into account the impact of uncertain generation and load, and correcting the grid load forecasting results. The energy storage optimization module is used to construct an energy storage optimization model based on the power regulation and voltage support characteristics of distributed energy storage. It uses the energy storage installation location, capacity and time-series output as decision variables, and solves the energy storage capacity configuration and initial power output scheduling scheme under the condition of satisfying the energy storage operation constraints. The load transfer module is used to build a data load transfer optimization model based on the spatiotemporal portability characteristics of data center load. Under the condition of satisfying data center business constraints, it solves the initial optimization results of the transfer volume and target time period allocation for each data center. The communication optimization module is used to optimize the communication network topology based on the physical energy storage configuration, data load transfer method, and power grid topology. The collaborative optimization module is used to construct a collaborative optimization model and introduces rolling optimization and communication reliability correction mechanisms. It uses the initial optimization results of energy storage capacity configuration and initial power output scheduling scheme, and the transfer volume and target time period allocation of each data center as the initial solution to solve the collaborative optimization model and obtain the final deployed collaborative scheduling scheme.

[0114] The present invention will now be further described with reference to specific embodiments and accompanying drawings.

[0115] This embodiment uses an IEEE 33-node distribution network system as the experimental object, which includes multiple distributed generation units and several large-scale data centers. Based on this, simulation analysis of distribution network optimization scheduling was conducted to verify the feasibility and effectiveness of the proposed method in a real-world scenario.

[0116] In this specific embodiment, the distribution network overview is as follows, and the topology is as follows: Figure 2 As shown: This power distribution network is located in an area of ​​a city, covering a total of 33 distribution nodes, including multiple solar and wind power generation units. The load demand in this area exhibits significant spatiotemporal fluctuations, especially during peak summer and winter periods, when load fluctuations are substantial. Meanwhile, the three major data centers in the area have high power demands, and their load fluctuations vary depending on their business operations.

[0117] The optimal planning method for distribution network resources, which considers distributed generation and data center flexibility as described in Example 1, includes the following steps: Distributed generation forecasting and load analysis are performed according to step 1 of Example 1. Specifically, historical load data and meteorological data are used to predict distributed generation using a machine learning-based forecasting model. The Support Vector Machine (SVM) algorithm is used to predict the power generation of solar and wind power. After training on data from the past 5 years, the model can accurately predict hourly power generation fluctuations.

[0118] Following step 2 of Example 1, energy storage system scheduling and grid optimization are performed. Specifically, a distributed energy storage system is used to balance power fluctuations in the distribution network. Based on the power regulation characteristics of distributed energy storage, a power scheduling model for the energy storage system is designed. The scheduling of the energy storage system must not only meet the load demand of the distribution network but also consider the maximum output power and charging / discharging efficiency of the energy storage units. The optimization objective of energy storage power scheduling is to minimize power losses and ensure grid reliability.

[0119] According to step 3 of Example 1, data center load transfer and scheduling are carried out. Specifically, considering that the data center load fluctuates greatly, a load transfer algorithm is used to transfer the power demand of the data center in time and space. By analyzing the historical load data of the data center, the time and space pattern of its load fluctuation is determined, and then a suitable load transfer strategy is designed based on these patterns.

[0120] Following step 4 of Example 1, communication network topology optimization is performed. Specifically, real-time data transmission of the communication network is crucial during load scheduling in power distribution networks and data centers. Therefore, this example employs graph theory to optimize the communication network topology in the power distribution network. First, the optimal communication path for each node is calculated by real-time prediction of power demand and load changes. Then, based on these calculation results, the shortest path algorithm in graph theory is used to optimize the communication network topology, ensuring efficient data transmission and real-time scheduling.

[0121] Following step 5 of Example 1, collaborative optimization and overall scheduling are performed. Specifically, a collaborative optimization algorithm is used to jointly schedule all resources. This algorithm combines the optimization results of energy storage configuration, data center load shifting, and communication network topology to maximize the economy and reliability of the distribution network. By optimizing the objective function, the system ensures that it meets load demand while minimizing power loss, thereby improving the overall operating efficiency of the system.

[0122] The following are the optimal planning results for a 33-node distribution network system, including the optimized configuration of energy storage, distributed generation, and communication nodes, as shown in Tables 1-3 below: Table 1: Optimal Configuration of Energy Storage System

[0123] Table 2: Optimal Configuration of Distributed Generation

[0124] Table 3: Optimal Configuration of Communication Nodes

[0125] Using the MATLAB simulation platform, the above optimization process was employed to comprehensively optimize the scheduling of the distribution network. Simulation results show that through collaborative optimization, the economic efficiency of the distribution network was significantly improved, while power loss was reduced by 15%. Furthermore, load shifting from the data center reduced the system's load fluctuation amplitude, ensuring the reliability of the power grid. Specifically, the system's load fluctuation was reduced by 20% after optimization, and the utilization efficiency of the energy storage system was also effectively improved. Simulation results are shown in Tables 4-6. Table 4: Comparison of Distribution Network Load Fluctuations Before and After Optimization

[0126] Table 5: Power Dispatch Process of Energy Storage System

[0127] Table 6: Data Center Load Transfer Status

[0128] This embodiment verifies the effectiveness of the optimization method of this invention in solving the load fluctuation problems of distributed generation and data centers by applying it to a real distribution network. Simulation results show that this invention can significantly improve the economy and reliability of the distribution network, and has broad application prospects and practical value. This embodiment demonstrates the practical application of this invention in distribution network optimization through simulation analysis, and further verifies its advantages in distributed generation and data center load regulation.

[0129] Compared with the prior art, the beneficial effects of the present invention include at least the following: 1. The distribution network resource planning method and system of the present invention, which takes into account the flexibility of distributed generation and data centers, achieves optimal economy and reliability of distribution network by coordinating and optimizing energy storage configuration, data load transfer and communication network topology optimization of distribution network resource allocation, while effectively solving the operational challenges caused by distributed generation and data center load fluctuations.

[0130] 2. The distribution network resource planning method and system of the present invention, which takes into account the flexibility of distributed energy and data centers, can effectively reduce the risk of distribution network operation exceeding limits under the conditions of distributed energy access and data center load fluctuations, while improving the utilization rate of renewable energy and ensuring the stability and security of the distribution network in complex operating environments.

[0131] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0132] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A power distribution network resource planning method considering DG and data center flexibility, characterized in that, The method comprises the following steps: Step 1, based on the power grid load prediction model under the condition of distributed power generation access, determine the output of distributed power generation in different time periods, considering the influence of uncertain power generation and load, correct the power grid load prediction result; Step 2, based on the power regulation and voltage support characteristics of distributed energy storage, build an energy storage optimization model, take the installation location, capacity and time sequence output of energy storage as decision variables, solve the energy storage capacity configuration and initial power output scheduling scheme under the condition of meeting the energy storage operation constraints; Step 3, based on the space-time migratory characteristics of data center load, build a data load transfer optimization model, and solve the initial optimization result of each data center transfer amount and target period allocation under the condition of meeting the data center business constraints; Step 4, according to the physical side energy storage configuration, data load transfer mode and power grid topology, optimize the communication network topology; Step 5, build a collaborative optimization model, and introduce a rolling optimization and communication reliability correction mechanism, take the energy storage capacity configuration and initial power output scheduling scheme of step 2 and the initial optimization result of each data center transfer amount and target period allocation of step 3 as the initial solution, solve the collaborative optimization model, and get the final deployed collaborative scheduling scheme.

2. The power distribution network resource planning method considering the flexibility of DG and data center according to claim 1, wherein: In step 1, the power grid load prediction model is represented by the following formula: In the formula, The predicted power for time period t, including generating power or load power. For the current power, For the first k The change in each predictor factor The weighting coefficient for each predictor. n The number of predictive factors, including weather and load.

3. The power distribution network resource planning method considering the flexibility of DG and data center according to claim 2, wherein: The error correction process is represented by the following formula: In the formula, is the corrected prediction error, is the prediction error, is the actual error, is the correction coefficient.

4. The power distribution network resource planning method considering the flexibility of DG and data center according to claim 1, wherein: In step 2, the objective function of the energy storage optimization model F 1includes minimizing the network loss, the energy storage operating cost and the energy storage investment cost, which is expressed as follows: wherein, is the net loss of the time period t is the net loss of the time period is the energy storage output, where discharging is positive and charging is negative, is the energy storage rated energy, , , are the energy storage net loss coefficient, energy storage operation cost coefficient, and energy storage investment cost coefficient, respectively, S is the set of energy storage units, T is the number of time periods.

5. The power distribution network resource planning method considering the flexibility of DG and data center according to claim 1, wherein: In step 2, the constraint conditions of the energy storage optimization model include system overall energy storage power balance constraint, node power balance constraint and energy storage power and energy constraint; The system overall energy storage power balance constraint is represented by the following formula: In the formula, is the total regulating power of the energy storage at the system level, is the energy storage output, S is the energy storage unit set; The node power balance constraint is represented by the following formula: wherein S (0) is the energy storage set of the access node n , n , , are the node power inflow and outflow determined by the distribution power flow equation, is the predicted power of distributed generation, is the predicted power of basic load; The energy storage power and energy constraint is represented by the following formula: wherein the energy storage i maximum power, the energy storage i state of charge, and the energy storage i minimum and maximum capacity.

6. The power distribution network resource planning method considering the flexibility of DG and data center according to claim 1, wherein: In step 3, the optimization objective of the load transfer optimization model is set to minimize the electricity cost under the time-of-use electricity price, the migration cost, the objective function F 2 is expressed by the following equation: In the formula, is a time-of-use electricity price, is a data center j is a load after transfer, is a data center load transfer cost coefficient, D is a set of data centers, T is the number of time periods.

7. The power distribution network resource planning method considering the flexibility of DG and data center according to claim 1, wherein: In step 3, the constraint conditions of the load transfer optimization model include load transfer summary constraint, maximum flexibility range constraint and task deadline constraint; The load transfer summary constraint is represented by the following formula: wherein is t the overall transfer load of the period, is t the overall transferable upper limit of the period system, is the load transfer amount, is the data center j the load after transfer, is the data center reference IT load; The maximum flexibility range constraint is represented by the following formula: In the formula: is the load transfer coefficient; The task deadline constraint is represented by the following formula: In the formula, is the task completion time window of the data center j .

8. The power distribution network resource planning method considering the flexibility of DG and data center according to claim 7, wherein: Data center j At time period t The proportion of the transferable load to the benchmark IT load of the data center is taken as the load transfer coefficient It is expressed as follows: wherein is the data center j is the transferable load, is the load transfer coefficient upper limit, given by service level agreement or business policy, is the data center base IT load.

9. The power distribution network resource planning method considering the flexibility of DG and data center according to claim 1, wherein: Step 4 comprises: Step 4.1, real-time prediction of power demand and load of each distribution node, and calculation of the optimal communication path of each node; Step 4.2, optimization of communication network topology in the distribution network based on graph theory model; Step 4.3, adjustment of the connection mode of the communication node according to the change of power demand and load of each distribution node.

10. The distribution network resource planning method considering DG and data center flexibility according to claim 1, wherein: In step 5, the energy storage output, energy storage capacity and data center transferred load are taken as joint decision variables to construct the main objective of the collaborative optimization model, which is expressed by the following formula: In the formula, is the net loss of the time period t , is the output of the energy storage i , is the rated energy of the energy storage i , is the load of the data center j after the transfer, is the load transfer amount of the data center j , , , are respectively the energy storage net loss coefficient, the energy storage operation cost coefficient, and the energy storage investment cost coefficient, is the time-of-use electricity price, is the data center load migration cost coefficient, T is the number of time periods, S is the set of energy storage units, D is the set of data centers.

11. The distribution network resource planning method considering DG and data center flexibility according to claim 1, wherein: In step 5, the energy storage and data center are explicitly coupled to the same node to obtain the node power balance constraint of the collaborative optimization model, which is expressed by the following formula: In the formula, For energy storage i of efforts, For data centers j The transferred load, , For node inflow and outflow power, For distributed generation power prediction, Based on the predicted power of the basic load, S ( n ) is the access node n Energy storage collection, D ( n ) is the access node n A collection of data centers.

12. The distribution network resource planning method considering DG and data center flexibility according to claim 1, wherein: In step 5, the rolling optimization mechanism includes: define the prediction error , with the net bias as follows: In the formula: is a base load measured power, is a base load predicted power, is a distributed generation measured power, is a distributed generation predicted power; The system imbalance degree is defined to reflect the load balance, which is expressed by the following formula: In the rolling time horizon, the joint optimization results of energy storage and data center load are taken as input, and Re-engage in online optimization as an additional objective.

13. The distribution network resource planning method considering DG and data center flexibility according to claim 1, wherein: In step 5, the communication reliability correction mechanism includes: Defining a communication regulator is expressed by the following equation: In the formula, clip(*, 0, 1) represents a function of truncating the value of * to [0, 1], is a communication reliability index, is a minimum communication reliability index; The optimal solution of energy storage output and data center load is corrected according to the following formula to obtain the final deployed collaborative scheduling scheme: wherein, is the energy storage output for final deployment, is the data center load for final deployment, is the energy storage output for security policy under communication degradation, is the business bottom line load, which is determined by the importance of the business itself according to expert experience, and are the optimal solutions of the energy storage and data center load after rolling optimization, respectively.

14. A power distribution network resource planning system considering DG and data center flexibility, running the power distribution network resource planning method considering DG and data center flexibility according to any one of claims 1-13. It includes: A load prediction module for determining the output of distributed power generation in different time periods based on a power grid load prediction model under the condition of distributed power generation access, considering the influence of uncertain power generation and load, and correcting the power grid load prediction result; An energy storage optimization module for constructing an energy storage optimization model based on the power regulation and voltage support characteristics of distributed energy storage, taking the installation location, capacity and time sequence output of energy storage as decision variables, and solving the energy storage capacity configuration and initial power output scheduling scheme under the condition of meeting the energy storage operation constraints; A load transfer module for constructing a data load transfer optimization model based on the time and space migratory characteristics of data center load, and solving the initial optimization result of each data center transfer amount and target period allocation under the condition of meeting the data center business constraints; A communication optimization module for optimizing the communication network topology according to the physical side energy storage configuration, data load transfer mode and power grid topology structure; A collaborative optimization module for constructing a collaborative optimization model and introducing a rolling optimization and communication reliability correction mechanism, taking the energy storage capacity configuration and initial power output scheduling scheme, and the initial optimization result of each data center transfer amount and target period allocation as the initial solution, solving the collaborative optimization model, and obtaining the final deployed collaborative scheduling scheme.

Images