A source network load storage collaborative planning method and system

Abstract

This invention relates to the field of power system technology, specifically a source-grid-load-storage collaborative planning method and system. The method includes: generating multiple initial planning scenarios based on historical renewable energy output and load data of a target region using probability distribution fitting and Monte Carlo simulation; reducing these initial planning scenarios using a clustering algorithm to obtain a set of typical scenarios, and assigning weights to each typical scenario; constructing a multi-scenario stochastic planning model based on the set of typical scenarios, and solving it using an optimization algorithm to obtain a preliminary planning scheme; importing the preliminary planning scheme into a pre-constructed source-grid-load-storage digital twin for simulation, obtaining performance indicators of the planning scheme under various simulated operating conditions; and evaluating the planning scheme based on the performance indicators until a final planning scheme meeting the requirements is obtained. This invention achieves dynamic simulation verification and closed-loop optimization of the planning scheme, significantly improving the scientific rigor and reliability of new power system planning.

Description

Technical Field

[0001] This invention belongs to the field of power system technology, and in particular relates to a source-grid-load-storage coordinated planning method and system. Background Technology

[0002] Against the backdrop of energy transition, the scope of power system planning has expanded from traditional fossil fuel-based power sources and grids to complex "source-grid-load-storage" systems encompassing large-scale wind power, photovoltaics, diversified loads, and energy storage. The challenges facing planning have intensified unprecedentedly: First, both renewable energy output and load demand exhibit significant randomness and volatility, making traditional planning methods based on single or a few deterministic scenarios extremely risky, potentially leading to uneconomical or unsafe planning schemes in most real-world situations. Second, the deep coupling among the "source-grid-load-storage" links necessitates comprehensive consideration of their mutual influences and synergistic effects during planning, resulting in high global optimization difficulty. Finally, the effectiveness of planning schemes needs to be verified over a decades-long lifespan, making traditional evaluation methods based on static indicators or simplified simulations insufficient to comprehensively and dynamically reflect their long-term performance.

[0003] In existing technologies, stochastic programming addresses uncertainty by generating a large number of scenarios, but the reasonable generation and efficient reduction of scenarios, as well as ensuring the representativeness of the reduced scenario set, remain challenges. Furthermore, existing planning processes typically stop at optimizing the mathematical solution of the model, lacking a verification step that enables high-fidelity, full-cycle, multi-condition dynamic simulation of the planning scheme. While digital twin technology has been applied in the operational phase, it has not yet been systematically and deeply integrated with the early planning stages, failing to form a closed loop of "planning-simulation verification-feedback optimization." Patent application CN117277433A discloses a power system planning method, apparatus, equipment, and medium considering carbon emission reduction. It involves collecting relevant configuration parameters of the power system; constructing a main problem optimization model for the power system with the goal of minimizing investment and operating costs; and constructing sub-problem optimization models for the power system with the goal of minimizing the system's operational safety domain offset. A loose-tight decoupling method is used to decouple the sub-problem optimization models into a safety feasibility detection problem model and a low-carbon feasibility detection problem model. The main problem optimization model is solved to obtain a planning scheme and an operating scheme that simultaneously satisfy both the safety feasibility detection problem model and the low-carbon feasibility detection problem model, which are then output as the planning result. This patent application also implements the power system planning scheme through optimization models, but it cannot achieve multi-condition dynamic simulation, exhibiting the same drawbacks as existing technologies.

[0004] Therefore, how to provide a new integrated planning method that can combine uncertainty modeling, collaborative optimization and high-fidelity dynamic simulation verification is a problem that urgently needs to be solved by researchers in this field. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this invention is to provide a source-grid-load-storage collaborative planning method to solve the problem that existing power systems cannot perform high-fidelity, full-cycle, multi-condition dynamic simulation verification of planning schemes, resulting in poor robustness of planning results and low reliability of the power system. In addition, this invention also provides a source-grid-load-storage collaborative planning system.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0007] Firstly, S10, scenario generation: Based on the historical renewable energy output and load data of the target area, the probability distribution fitting and Monte Carlo simulation methods are used to generate multiple initial planning scenarios that include the renewable energy installed capacity increment, output curve and load demand curve within the future planning period, providing a solid input foundation for stochastic planning.

[0008] S20. Scene Reduction: Clustering algorithm is used to reduce the multiple initial planning scenarios to obtain a set of typical scenarios, and weights are assigned to each typical scenario. This can greatly reduce the computational complexity of subsequent optimization problems while retaining key statistical features and temporal morphological features.

[0009] S30. Model Construction and Solution: Based on the set of typical scenarios, a multi-scenario stochastic programming model is constructed with the goal of minimizing the weighted total life cycle cost and considering various constraints. An optimization algorithm is used to solve the model and obtain a preliminary planning scheme. The model comprehensively considers the economic efficiency of the entire process from investment and construction to operation and maintenance and even decommissioning, and integrates multiple constraints such as technology, policy, and reliability to ensure the comprehensiveness and compliance of the planning scheme.

[0010] S40. Digital Twin Evaluation and Iterative Optimization: The preliminary planning scheme is imported into a pre-constructed source-grid-load-storage digital twin for simulation operation to obtain the performance indicators of the planning scheme under various simulated operating conditions. By conducting long-term, multi-condition simulation "trial runs" of the "static" scheme obtained by mathematical optimization in the "dynamic laboratory" of the digital twin, problems such as dynamic safety and timing matching that are difficult to discover by optimization models alone can be exposed. Based on simulation feedback, parameter adjustments and scheme iterations are carried out to achieve closed-loop optimization and self-evolution of the planning process.

[0011] S50. Evaluate the planning scheme based on the performance indicators. If the evaluation does not meet the preset requirements, adjust the planning parameters and return to S30 to solve and evaluate again until a final planning scheme that meets the requirements is obtained.

[0012] Furthermore, in S10, the probability distribution fitting specifically involves:

[0013] The probability distributions of wind power output, photovoltaic power output, and load demand are fitted using the kernel density estimation method. The number of initial planning scenarios generated by the Monte Carlo simulation is no less than 500, and the scenarios include extreme operating conditions.

[0014] Furthermore, in S20, the clustering algorithm is an improved k-means clustering algorithm. Each scenario feature vector includes the total installed capacity of new energy, the annual peak load, and the typical daily power output curve. The curve similarity is measured by a dynamic time warping algorithm, and the number of typical scenarios after reduction is 10-30.

[0015] Furthermore, in S30, the objective function of the multi-scenario stochastic programming model is:

[0016]

[0017] Where S represents the total number of typical scenarios. For the weights of scene s, LCC s The total lifecycle cost under scenario s includes investment costs, operation and maintenance costs, and decommissioning costs.

[0018] Furthermore, the constraints of the objective function include resource constraints, technological constraints, policy constraints, and reliability constraints. The resource constraint is that the installed capacity of new energy sources does not exceed the regional development limit. The technological constraints include power flow constraints of transmission lines, state of charge constraints of energy storage, and minimum technical output constraints of conventional units. The policy constraints include the proportion of new energy power generation at the end of the planning period and carbon emission intensity constraints. The reliability constraints include the expected value of system power shortage rate or insufficient power supply.

[0019] Furthermore, in S30, the optimization algorithm is an improved particle swarm optimization algorithm, which includes a particle encoding method that incorporates decision variables such as wind power / photovoltaic installed capacity, line length, and energy storage capacity; introduces a penalty term for particles that violate constraints in the fitness function; and adopts an adaptive inertia weight strategy.

[0020] Furthermore, in S40, the source-grid-load-storage digital twin includes a high-fidelity virtual model of the physical system's geometric model, electrical model, control model, and behavioral rules, which can map and simulate the state of the physical system in real time.

[0021] Furthermore, the simulated operating conditions include one or more of the following: large-scale generation of new energy sources, peak load, equipment N-1 failure, and extreme weather.

[0022] Secondly, the present invention also provides a source-grid-load-storage collaborative planning system employing the above method, comprising:

[0023] The scene generation and processing module is used to generate and reduce multiple initial planning scenes;

[0024] The multi-scenario planning model solving module is used to construct and solve multi-scenario stochastic planning models, and uses optimization algorithms to solve them, outputting preliminary planning schemes.

[0025] The digital twin simulation and evaluation platform is used to build and run digital twins of source-grid-load-storage systems, and to perform simulation operations and calculate performance indicators.

[0026] The scheme evaluation and iterative optimization control module is used to compare the simulation evaluation results with the preset targets, determine whether the scheme meets the targets, and trigger the iterative process of adjusting and optimizing the planning parameters when the targets are not met.

[0027] Furthermore, the digital twin simulation evaluation platform is connected to the multi-scenario planning model solving module through an application programming interface (API) to realize the automatic import of planning schemes, the automatic distribution of simulation tasks, and the automatic feedback of evaluation results, forming an automated closed loop of planning, simulation, evaluation, and optimization.

[0028] Compared with existing technologies, the source-grid-load-storage coordinated planning method and system provided by this invention have at least the following advantages:

[0029] The robustness of the planning is significantly enhanced: Through multi-scenario stochastic planning, various possible (including extreme) situations in the future are clearly considered, and the resulting planning scheme is optimal in a probabilistic sense, avoiding the risk of planning errors caused by a single scenario;

[0030] Comprehensive and in-depth evaluation dimensions: With the help of digital twin technology, the evaluation of planning schemes has been expanded from static and single-point indicators to dynamic, long-term, and multi-condition comprehensive performance simulation, which can more realistically predict the performance of the schemes in complex real-world environments.

[0031] A closed-loop planning mechanism was created: a closed-loop workflow of "mathematical optimization → high-fidelity simulation → evaluation and feedback → re-optimization" was established, which enabled the planning scheme to be continuously corrected and improved in the iteration process, thereby improving the scientific nature and accuracy of the planning.

[0032] Supporting collaborative decision-making: It provides planners with a visual and interactive decision support environment, which can intuitively compare the long-term simulation results of different planning schemes and assist in making multi-objective trade-offs and final scheme selection. Attached Figure Description

[0033] To more clearly illustrate the solution of the present invention, a brief introduction will be given to the drawings used in the description of the embodiments below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0034] Figure 1 A flowchart of a source-grid-load-storage coordinated planning method provided in an embodiment of the present invention;

[0035] Figure 2 This is a framework diagram for source-grid-load-storage coordinated planning provided in an embodiment of the present invention. Detailed Implementation

[0036] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Preferred embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the invention.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0038] This invention provides a source-grid-load-storage coordinated planning method, applicable to the long-term coordinated planning of new power systems with a high proportion of renewable energy and strong uncertainty. The source-grid-load-storage coordinated planning method includes:

[0039] S10, Scenario Generation: Based on historical renewable energy output and load data of the target area, multiple initial planning scenarios are generated using probability distribution fitting and Monte Carlo simulation methods. These scenarios include the increase in renewable energy installed capacity, output curves, and load demand curves within the future planning period. S20, Scenario Reduction: Clustering algorithms are used to reduce the multiple initial planning scenarios to obtain a set of typical scenarios, and weights are assigned to each typical scenario. S30, Model Construction and Solution: Based on the set of typical scenarios, a multi-scenario stochastic planning model is constructed with the goal of minimizing weighted total life cycle cost and considering various constraints. An optimization algorithm is used to solve the model to obtain a preliminary planning scheme. S40, Digital Twin Evaluation and Iterative Optimization: The preliminary planning scheme is imported into a pre-constructed source-grid-load-storage digital twin for simulation to obtain performance indicators of the planning scheme under various simulated operating conditions. S50, The planning scheme is evaluated based on the performance indicators. If the evaluation does not meet the preset requirements, the planning parameters are adjusted and the process returns to S30 for re-solving and evaluation until a final planning scheme that meets the requirements is obtained.

[0040] This invention enables dynamic simulation verification and closed-loop optimization of planning schemes, significantly improving the scientific nature and reliability of new power system planning.

[0041] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0042] This invention provides a source-grid-load-storage coordinated planning method, applicable to the long-term coordinated planning of new power systems with high proportions of renewable energy and strong uncertainty. Source-grid-load-storage coordinated planning needs to consider the long-term development needs of each link to achieve coordinated development of source, grid, load, and storage. Embodiments of this invention design, as follows... Figure 2 The collaborative planning framework shown includes a planning objective layer, a constraint layer, and an optimization method layer.

[0043] The planning objective layer aims to minimize the total life cycle cost, maximize the renewable energy absorption rate, and improve system reliability

[18] . The constraint layer includes resource constraints, technical constraints, and policy constraints. The optimization method layer uses a multi-objective optimization algorithm to optimize the planning scheme of source-grid-load-storage.

[0044] like Figure 1 As shown, in this embodiment, the source-grid-load-storage coordinated planning method includes:

[0045] S10. Scenario Generation: Taking a wind and solar power-rich area as an example, with a planning period of 20 years, collect hourly output data of wind and solar power, as well as hourly load data of industry, commerce, and residential sectors for the past 5 years. After cleaning and preprocessing the data, typical annual data is generated.

[0046] Using the kernel density estimation method, the following results were obtained through fitting:

[0047] Wind power output probability distribution: Weibull distribution with shape parameter k=2.1 and scale parameter λ=9.5m / s.

[0048] Photovoltaic output probability distribution: Beta distribution with α=2.8 and β=5.2.

[0049] Daily peak load probability distribution: a normal distribution with mean μ=1250MW and standard deviation σ=85MW.

[0050] Based on the above distribution, 1000 initial scenarios were randomly generated using Monte Carlo simulation. Each scenario includes the annual increase in wind power and photovoltaic installed capacity over the next 20 years (within a range set based on policy objectives and resource potential), as well as a composite output curve and load curve spanning 8760 hours (one year). During the generation process, extreme event scenarios such as "wind power output is less than 10% of installed capacity for 3 consecutive days" and "load surges by 20% for 5 hours in the summer afternoon" were forcibly introduced to ensure the completeness of the scenario set.

[0051] S20. Scene Reduction: The 1000 initial scenes are reduced using an improved k-means clustering algorithm.

[0052] Feature vector construction: For each scene i, construct the feature vector:

[0053]

[0054] in This represents the total installed capacity of new energy vehicles (MW) at the end of the 20th year of the planning period. This represents the maximum annual load (MW) in this scenario. The DTW feature representation of the annual 8760-hour output curve of new energy in this scenario is obtained by dimensionality reduction (e.g., extracting typical daily curves).

[0055] Clustering process: Set the target number of clusters K=20, randomly select 20 scenes as initial cluster centers, use the weighted Euclidean distance between feature vectors (each dimension can be assigned different weights) as the similarity measure, and perform iterative calculations until the cluster centers are stable.

[0056] Weight assignment: For the j-th typical scenario obtained after reduction, its weight is... This equals the proportion of the original 1000 scenes that were classified into this cluster out of the total number of scenes. Ultimately, this ensures... .

[0057] S30. Model Construction and Solution: Construct a multi-scenario mixed integer linear programming (MILP) model.

[0058] The objective function is:

[0059]

[0060] Minimize the weighted total lifecycle cost (LCC) for 20 typical scenarios. The LCC includes:

[0061] Investment costs: The initial investment for wind power (unit cost 6,000 yuan / kW), photovoltaic (unit cost 4,500 yuan / kW), 220kV transmission lines (unit cost 1.5 million yuan / km), and lithium battery energy storage systems (unit cost 1,800 yuan / kWh) is calculated based on a one-time investment at the beginning of the planning period.

[0062] Operation and maintenance costs include fuel costs for conventional units (calculated based on coal prices and coal consumption for power generation), annual maintenance costs for all equipment (calculated at 2% of investment costs), and system network loss costs.

[0063] Decommissioning and disposal costs: At the end of the planning period (end of the 20th year), the costs will be recovered based on the residual value of the equipment (10% of the initial investment) and deducted from the total cost.

[0064] Constraints:

[0065] Resource constraints: Total installed wind power capacity ≤ 2500 MW, total installed photovoltaic capacity ≤ 1800 MW.

[0066] Technical constraints: Power flow at critical sections ≤ thermal stability limit (800 MW); daily operating range of energy storage SOC 20%-90%; minimum technical output of coal-fired units is 40% of rated capacity.

[0067] Policy constraints: In the 20th year, the proportion of new energy power generation in total electricity consumption shall be ≥ 50%; the average carbon emission intensity of the system shall be ≤ 0.35 tCO2 / MWh.

[0068] Reliability constraint: Under all typical scenarios, the system's annual power consumption shortfall (EENS) is ≤ 0.01% * total annual power consumption.

[0069] S40. Digital Twin Assessment and Iterative Optimization: To improve the feasibility and effectiveness of the planning scheme, digital twin technology is introduced to construct a digital twin and conduct a comprehensive assessment of the planning scheme. On the digital twin platform, based on the actual power grid geographical wiring diagram, equipment parameters (transformers, line impedance), new energy power station model (including wind / solar resource model, inverter control model), load model (static and dynamic characteristics), and energy storage system model (electrochemical model, BMS management logic) of the region, a high-fidelity virtual model that maps to the physical system in a 1:1 manner is established.

[0070] Import and Simulation: Import the preliminary planning scheme (adding equipment parameters and locations) obtained in step S30 into the digital twin. Set up various simulation tasks in the digital twin:

[0071] Long-term production simulation: Using 8,760 hours of data from 20 typical scenarios throughout the year, the simulation runs for 20 years, recording annual costs, renewable energy consumption rate, and reliability indicators.

[0072] Typical operating condition test: Simulate the extreme situation of sudden drop in photovoltaic output (cloud obstruction) during the summer high load period, and observe the dynamic response of system frequency and voltage.

[0073] N-1 Fault Verification: Simulate faults in critical lines or transformers in the power grid after planning to verify whether the system can still meet safety standards.

[0074] Assessment and Decision-Making: Analyze simulation results. For example, long-term production simulation shows that the renewable energy consumption rate is only 88%, lower than the policy requirement of 90%; N-1 fault verification found that the voltage in a certain area exceeded the limit.

[0075] Iterative optimization: Based on the evaluation results, adjust the constraints or objective function weights of the planning model. For example, increase the grid integration rate requirement in the policy constraints from 90% to 92%, or add an incentive term for grid investment in areas with weak voltage stability to the optimization model. Then, return to step S3, resolve the optimization model, obtain a new planning scheme (e.g., adjust energy storage deployment, add a line), and import it into the digital twin for a new round of evaluation. This process is repeated until the simulation evaluation results meet the preset safety, economic, and environmental goals under all test conditions, thus obtaining the final planning scheme.

[0076] The model is solved using commercial optimization solvers (such as CPLEX and Gurobi) or an improved Particle Swarm Optimization (PSO) algorithm. Improvements to the PSO algorithm include: particle encoding as [new wind power capacity, new solar power capacity, new transmission line length, new energy storage capacity]; fitness values ​​are weighted LCC, with significant penalties imposed on particles violating reliability or carbon emission constraints; and the inertia weight is linearly reduced from 0.9 to 0.4. After solving, a set of Pareto optimal fronts is obtained, allowing decision-makers to select a preliminary planning scheme based on their preferences, for example: 800MW of new wind power, 600MW of new solar power, 120km of new transmission lines, and 300MW / 1200MWh of new energy storage.

[0077] This invention also provides a source-grid-load-storage collaborative planning system employing the method described in the above embodiments, combined with... Figure 1 and Figure 2 In this embodiment, the source-grid-load-storage coordinated planning system includes:

[0078] Scene generation and processing module: It has built-in data interface, probability distribution fitting tool, Monte Carlo simulator and clustering algorithm library, which automatically completes the output of typical scene sets from data.

[0079] Multi-scenario planning model solving module: Provides model building templates, allows users to customize objective functions and constraints, and integrates multiple optimization solvers (exact solvers and intelligent algorithms).

[0080] Digital twin simulation evaluation platform: Based on professional power system simulation software or self-developed simulation kernel, it has the functions of 3D visualization, multi-task parallel simulation and automatic result analysis.

[0081] Scheme evaluation and iterative optimization control module: As the control center, it defines evaluation rules (such as KPI thresholds), compares simulation results with the rules, and automatically generates adjustment suggestions (such as "strengthen voltage support at XX node") when the rules are not met, and triggers a new round of "planning-simulation" process.

[0082] The digital twin simulation evaluation platform and the planning model solving module are deeply integrated through the application programming interface (API) to realize the automatic import of planning schemes, the automatic distribution of simulation tasks, and the automatic feedback of evaluation results, forming an automated closed loop of "planning-simulation-evaluation-optimization".

[0083] In this embodiment of the invention, a source-grid-load-storage collaborative planning method is adopted to collaboratively plan the source-grid-load-storage system in the region and evaluate the planning scheme.

[0084] The main contents of the planning scheme include: adding 300MW of wind power capacity, 200MW of photovoltaic capacity, constructing 50km of new transmission lines, and adding 200MW / 400MWh of energy storage capacity. The planning scheme was simulated and evaluated using a digital twin, and the results are shown in Table 2.

[0085] Table 2

[0086] Evaluation indicators Planning scheme industry standards Comparison results Total lifecycle cost (100 million yuan) 85 ≤100 Meets requirements New energy consumption rate (%) 96 ≥90 Meets requirements System power supply reliability (%) 99.98 ≥99.9 Meets requirements <![CDATA[Carbon emission intensity (tCO2 / MWh)]]> 0.35 ≤0.5 Meets requirements

[0087] As shown in Table 2, the proposed plan meets industry standards in terms of total life cycle cost, renewable energy absorption rate, system power supply reliability, and carbon emission intensity, indicating that the plan is feasible and effective.

[0088] Compared with existing technologies, the source-grid-load-storage collaborative planning method and system described in the above embodiments cannot perform high-fidelity, full-cycle, multi-condition dynamic simulation verification of planning schemes in existing power systems, resulting in poor robustness of planning results and low reliability of power systems. This invention, through multi-scenario stochastic planning, explicitly considers all possible (including extreme) future situations, and the resulting planning scheme is optimal in a probabilistic sense, avoiding the risk of planning errors due to a single scenario. This invention utilizes digital twin technology to expand the evaluation of planning schemes from static, single-point indicators to dynamic, long-term, multi-condition comprehensive performance simulation, enabling more realistic prediction of scheme performance in complex real-world environments. This invention creates a closed-loop workflow of "mathematical optimization → high-fidelity simulation → evaluation feedback → re-optimization," allowing planning schemes to be continuously corrected and improved through iteration, enhancing the scientific nature and accuracy of planning. This invention provides planners with a visual and interactive decision support environment, allowing for intuitive comparison of long-term simulation results of different planning schemes, assisting in multi-objective trade-offs and final scheme selection.

[0089] Obviously, the embodiments described above are merely preferred embodiments of the present invention, and not all embodiments. The accompanying drawings illustrate preferred embodiments of the present invention, but do not limit the scope of the patent. The present invention can be implemented in many different forms; rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this invention.

Claims

1. A source-grid-load-storage coordinated planning method, characterized in that, Includes the following steps: S10. Scenario Generation: Based on the historical renewable energy output and load data of the target area, the probability distribution fitting and Monte Carlo simulation methods are used to generate multiple initial planning scenarios that include the renewable energy installed capacity increment, output curve and load demand curve within the future planning period. S20. Scene Reduction: The multiple initial planning scenes are reduced using a clustering algorithm to obtain a set of typical scenes, and a weight is assigned to each typical scene. S30. Model Construction and Solution: Based on the set of typical scenarios, a multi-scenario stochastic programming model is constructed with the goal of minimizing the weighted total life cycle cost and considering multiple constraints. An optimization algorithm is used to solve the model and obtain a preliminary planning scheme. S40. Digital Twin Evaluation and Iterative Optimization: The preliminary planning scheme is imported into a pre-constructed source-grid-load-storage digital twin for simulation operation to obtain the performance indicators of the planning scheme under various simulated operating conditions. S50. Evaluate the planning scheme based on the performance indicators. If the evaluation does not meet the preset requirements, adjust the planning parameters and return to S30 to solve and evaluate again until a final planning scheme that meets the requirements is obtained.

2. The source-grid-load-storage coordinated planning method according to claim 1, characterized in that, In step S10, the probability distribution fitting specifically involves: The probability distributions of wind power output, photovoltaic power output, and load demand are fitted using the kernel density estimation method. The number of initial planning scenarios generated by the Monte Carlo simulation is no less than 500, and the scenarios include extreme operating conditions.

3. A source-grid-load-storage coordinated planning method according to claim 1 or 2, characterized in that, In S20, the clustering algorithm is an improved k-means clustering algorithm. Each scenario feature vector includes the total installed capacity of new energy, the annual peak load, and the typical daily power output curve. The curve similarity is measured by the dynamic time warping algorithm, and the number of typical scenarios after reduction is 10-30.

4. The source-grid-load-storage coordinated planning method according to claim 1, characterized in that, In S30, the objective function of the multi-scenario stochastic programming model is: Where S represents the total number of typical scenarios. For the weights of scene s, LCC s The total lifecycle cost under scenario s includes investment costs, operation and maintenance costs, and decommissioning costs.

5. The source-grid-load-storage coordinated planning method according to claim 4, characterized in that, The constraints of the objective function include resource constraints, technological constraints, policy constraints, and reliability constraints. The resource constraint is that the installed capacity of new energy sources does not exceed the regional development limit. The technological constraints include power flow constraints of transmission lines, state of charge constraints of energy storage, and minimum technical output constraints of conventional units. The policy constraints include the proportion of new energy power generation at the end of the planning period and carbon emission intensity constraints. The reliability constraints include the expected value of system power shortage rate or insufficient power supply.

6. A source-grid-load-storage coordinated planning method according to claim 1, 4, or 5, characterized in that, In S30, the optimization algorithm is an improved particle swarm optimization algorithm, which includes a particle encoding method that incorporates decision variables such as wind power / photovoltaic installed capacity, line length, and energy storage capacity; introduces a penalty term for particles that violate constraints in the fitness function; and adopts an adaptive inertia weight strategy.

7. The source-grid-load-storage coordinated planning method according to claim 1, characterized in that, In S40, the source-grid-load-storage digital twin includes a high-fidelity virtual model of the physical system's geometric model, electrical model, control model, and behavioral rules, which can map and simulate the state of the physical system in real time.

8. The source-grid-load-storage coordinated planning method according to claim 7, characterized in that, The simulated operating conditions include one or more of the following: large-scale generation of new energy sources, peak load, equipment N-1 failure, and extreme weather.

9. A system employing the method as described in any one of claims 1 to 8, characterized in that, include: The scene generation and processing module is used to generate and reduce multiple initial planning scenes; The multi-scenario planning model solving module is used to construct and solve multi-scenario stochastic planning models, and uses optimization algorithms to solve them, outputting preliminary planning schemes. The digital twin simulation and evaluation platform is used to build and run digital twins of source-grid-load-storage systems, and to perform simulation operations and calculate performance indicators. The scheme evaluation and iterative optimization control module is used to compare the simulation evaluation results with the preset targets, determine whether the scheme meets the targets, and trigger the iterative process of adjusting and optimizing the planning parameters when the targets are not met.

10. The system according to claim 9, characterized in that, The digital twin simulation evaluation platform is connected to the multi-scenario planning model solving module through an application programming interface, enabling automatic import of planning schemes, automatic distribution of simulation tasks, and automatic feedback of evaluation results, forming an automated closed loop of planning, simulation, evaluation, and optimization.

Images