Multi-target planning method and system for source-grid-load-storage cooperation of power system

Abstract

The invention is suitable for the technical field of power distribution network planning, and provides a power system source-network-load-storage coordinated multi-target planning method, which comprises the steps of realizing real-time dynamic coupling simulation of multiple links of source-network-load-storage through a virtual synchronous machine interface, and integrating a refined model library, the refined model library is used for providing a first planning scheme of multi-scene collaboration; re-planning the energy data of the first planning scheme to obtain a second planning scheme for energy complementation and coordinated optimization; generating an extreme scene sample by adopting a Wasserstein generative adversarial network, establishing a multi-dimensional uncertainty joint probability model based on the extreme scene sample, performing dynamic risk assessment on the second planning scheme, and generating a risk assessment report; based on the risk assessment report and in combination with the business mode, a feasible target planning scheme is generated, so that the scientificity, feasibility and efficiency of power system planning are improved, and the intelligent and sustainable development of the power system is promoted.

Description

Technical Field

[0001] This application belongs to the technical field of distribution network planning, and particularly relates to a multi-objective planning method and system for the coordinated operation of power sources, grids, loads, and energy storage in a power system. Background Art

[0002] In the power system, multi-objective planning methods for the coordinated operation of power sources, grids, loads, and energy storage have been widely applied to scenarios such as large-scale grid connection of renewable energy, access of distributed power sources, and microgrids. Through intelligent dispatching and optimal configuration, the stable operation and economy of the power system are ensured.

[0003] However, with the increasing complexity of the power system, such as the access of multiple energy types, the diversification of grid structures, and the diversification of user demands, it has become more difficult to simulate, analyze, and optimize the power system. Traditional power system planning methods are often based on static models and are difficult to accurately reflect the dynamic characteristics and uncertainties of the power system, resulting in poor performance of the planning scheme in practical applications. Summary of the Invention

[0004] Embodiments of this application provide a multi-objective planning method and system for the coordinated operation of power sources, grids, loads, and energy storage in a power system, which can solve one of the above-mentioned existing technical problems.

[0005] In a first aspect, embodiments of this application provide a multi-objective planning method for the coordinated operation of power sources, grids, loads, and energy storage in a power system, including:

[0006] In a digital twin collaborative simulation platform, real-time dynamic coupling simulation of multiple links of power sources, grids, loads, and energy storage is achieved through a virtual synchronous machine interface, and a refined model library is integrated. The refined model library is used to provide a first planning scheme for multi-scenario coordination;

[0007] Based on the simulation scenario, the energy data of the first planning scheme is re-planned to obtain a second planning scheme for energy complementarity and coordinated optimization;

[0008] An extreme scenario sample is generated using a Wasserstein generative adversarial network, a multi-dimensional uncertainty joint probability model is established based on the extreme scenario sample, and dynamic risk assessment is performed on the second planning scheme to generate a risk assessment report;

[0009] Based on the risk assessment report, a feasible target planning scheme is generated in combination with the business model, and the business model includes business requirements, power plant resources, and specification costs.

[0010] Further, the real-time dynamic coupling simulation of multiple links of power sources, grids, loads, and energy storage through a virtual synchronous machine interface in the digital twin collaborative simulation platform includes:

[0011] Define functional modules and data interaction protocols in the digital twin co-simulation platform, obtain real-time operation data of each link of the source network, load, and storage during the simulation process through a virtual synchronous machine, and extract delay parameters and jitter parameters during the data transmission process;

[0012] Calculate the adjustment value of the simulation time step according to the delay parameter and the jitter parameter;

[0013] Input the real-time operation data into a preset dynamic coupling simulation model, calculate the dynamic response characteristics of each operation node, where the dynamic response characteristics are the real-time changes of operation parameters, and the operation parameters include voltage, current, and power;

[0014] Based on the dynamic response characteristics, judge whether the operation status of each operation node is within the safe range;

[0015] If the operation status of the operation node exceeds the safe range, adjust the resource configuration during the simulation process to generate a simulation optimization plan.

[0016] Furthermore, the integrated refined model library, where the refined model library is used to provide a first planning plan for multi-scenario collaboration, including:

[0017] Obtain the equipment parameters and operation parameters of each planning plan in the refined model library, and classify each planning plan using cluster analysis based on the equipment parameters and the operation parameters to obtain multiple classification data sets;

[0018] Establish a matching rule library for the simulation scenario requirements, and based on the matching rule library, screen at least one classification data set from multiple classification data sets as the screening data set, and the corresponding planning plan in the screening data set is the first planning plan.

[0019] Furthermore, based on the simulation scenario, re-plan the energy data of the first planning plan to obtain a second planning plan for energy complementarity and coordinated optimization, including:

[0020] Obtain the energy data of the first planning plan, where the energy data includes thermal power, hydropower, wind power, and photovoltaic power;

[0021] Use the linear programming algorithm to decouple and calculate the simulation scenario to obtain the energy data in the simulation scenario;

[0022] Calculate the correlation coefficient between the energy data in the simulation scenario, and obtain the complementary relationship between different energies by analyzing the correlation coefficient;

[0023] Based on the complementary relationship, calculate the configuration plan after energy complementarity using the genetic algorithm, and based on the configuration plan, re-plan the energy data of the first planning plan to generate the second planning plan.

[0024] Further, the generation of extreme scenario samples using the Wasserstein generative adversarial network includes:

[0025] Obtain historical extreme value data from historical data, perform normalization processing on the data, and establish a data distribution model using the kernel density estimation method;

[0026] According to the data distribution model, design the structure of the Wasserstein generative adversarial network, and set the number of layers, activation functions, and learning rate of the generator and discriminator;

[0027] Train the generative adversarial network through the historical extreme value data to generate extreme scenario samples;

[0028] Extract key feature vectors from the generated extreme scenario samples, and the key feature vectors include extreme points, change trends, and distribution forms;

[0029] According to the key feature vectors, dynamically adjust the learning rate and regularization parameters of the Wasserstein generative adversarial network, and retrain the network to make the Wasserstein generative adversarial network generate more accurate extreme scenario samples.

[0030] Further, the establishment of a multi-dimensional uncertainty joint probability model based on the extreme scenario samples includes:

[0031] Based on the extreme scenario samples, use the Copula function to construct a multi-dimensional uncertainty joint distribution, and the multi-dimensional uncertainty joint distribution is used to describe the correlation between different uncertainty factors;

[0032] Through the non-parametric kernel density estimation method, perform probability modeling on the multi-dimensional uncertain joint distribution to generate a multi-dimensional uncertainty joint probability model;

[0033] Based on the multi-dimensional uncertainty joint probability model, use the scenario reduction method to screen out a set of typical operating scenarios from the generated multiple extreme scenario samples.

[0034] Further, the dynamic risk assessment of the second planning scheme to generate a risk assessment report includes:

[0035] Based on the set of typical operating scenarios, perform extreme scenario simulations on the second planning scheme;

[0036] Extract the failure probability from the simulation results and calculate the failure probability distribution under each extreme scenario;

[0037] Use the linear regression method to quantify the loss degree under each extreme scenario to generate a loss degree matrix;

[0038] Based on the fault probability distribution and the loss degree matrix, use the analytic hierarchy process to calculate the risk index values of the second planning scheme in each extreme scenario;

[0039] Analyze the risk index values to generate a risk assessment report.

[0040] Furthermore, based on the risk assessment report, combined with the business model, generate a feasible target planning scheme. The business model includes business requirements, power plant resources, and specification costs, including:

[0041] Obtain the historical load, use the ARIMA model to calculate the load growth, and obtain the load growth prediction value;

[0042] Obtain the energy data and risk index values of the second planning scheme, and calculate the matching degree between the load growth prediction value and the energy data;

[0043] If the matching degree is greater than the preset matching threshold and the risk index value is less than the preset index threshold, add the corresponding second planning scheme to the target planning scheme set.

[0044] Furthermore, for the second planning scheme added to the target planning scheme set, it also includes:

[0045] Obtain the investment budget and operation and maintenance costs in the power plant resources and specification costs, and establish a cost control model;

[0046] Input the second planning scheme into the cost control model to generate cost evaluation indicators, and determine the target planning scheme based on the cost evaluation indicators.

[0047] In a second aspect, an embodiment of the present application provides a multi-objective planning system for power system source-network-load-storage coordination, including:

[0048] The first processing module is used to realize real-time dynamic coupling simulation of multiple links of source-network-load-storage through a virtual synchronous machine interface in a digital twin collaborative simulation platform, and integrate a refined model library, and the refined model library is used to provide a first planning scheme for multi-scenario coordination;

[0049] The second processing module is used to re-plan the energy data of the first planning scheme based on the simulation scenario to obtain a second planning scheme for energy complementarity and coordinated optimization;

[0050] The third processing module is used to generate extreme scenario samples using a Wasserstein generative adversarial network, establish a multi-dimensional uncertainty joint probability model based on the extreme scenario samples, perform dynamic risk assessment on the second planning scheme, and generate a risk assessment report;

[0051] The fourth processing module is used to generate a feasible target planning scheme based on the risk assessment report and in combination with the business model, where the business model includes business requirements, power plant resources, and specification costs.

[0052] The embodiments of the present application have at least one of the following beneficial effects compared with the prior art:

[0053] 1. Through the virtual synchronous machine interface in the digital twin collaborative simulation platform, real-time dynamic coupling simulation of multiple links such as source (energy supply), network (power grid structure), load (load demand), and storage (energy storage system) is realized, and thus the operation status of the actual energy system can be more accurately simulated and predicted, providing a solid foundation for the formulation of the planning scheme;

[0054] 2. The integrated refined model library provides rich options for the energy system planning under different scenarios. The refined model can take into account various possible variables and factors, thereby generating a first planning scheme for multi-scenario collaboration. By re-planning the energy data, a second planning scheme for energy complementarity and coordinated optimization is obtained, further enhancing the effectiveness and feasibility of the planning scheme;

[0055] 3. The Wasserstein generative adversarial network is used to generate extreme scenario samples, and a multi-dimensional uncertainty joint probability model is established based on these samples. Furthermore, the uncertainty and extreme situations in the energy system can be more comprehensively considered, and a dynamic risk assessment is carried out on the second planning scheme. The generated risk assessment report provides important reference information for decision-makers, helping to reduce potential risks;

[0056] 4. Based on the risk assessment report and the business model, a feasible target planning scheme is generated. The above comprehensive consideration makes the target planning scheme closer to the actual needs, and at the same time more efficient and economical. Description of the Drawings

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0058] Figure 1 is a schematic flowchart of a multi-objective planning method for power system source-network-load-storage collaboration provided by an embodiment of the present invention;

[0059] Figure 2 is a schematic structural diagram of a multi-objective planning system for power system source-network-load-storage collaboration provided by an embodiment of the present invention. Detailed Embodiments

[0060] In the following description, specific details such as specific system architectures, technologies, etc. are presented for purposes of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0061] It should be understood that when used in the specification of the present application and the appended claims, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0062] It should also be understood that the term "and / or" used in the specification of the present application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0063] As used in the specification of the present application and the appended claims, the term "if" can be interpreted as "when" or "once" or "in response to determining" or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if [the described condition or event] is detected" can be interpreted as meaning "once determined" or "in response to determining" or "once [the described condition or event] is detected" or "in response to detecting [the described condition or event]" according to the context.

[0064] In addition, in the description of the specification of the present application and the appended claims, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0065] The reference to "one embodiment" or "some embodiments" or the like described in the specification of the present application means that a specific feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of the present application. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments", etc. that appear in different places in this specification do not necessarily all refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in another way. The terms "comprising", "including", "having", and their variants all mean "including but not limited to", unless otherwise specifically emphasized in another way.

[0066] Please refer to Figure 1 As shown, the present invention is a multi-objective planning method for the coordinated operation of power sources, grids, loads, and energy storage systems in a power system, including the following steps:

[0067] S100. In the digital twin co-simulation platform, real-time dynamic coupling simulation of multiple links of source-grid-load-storage is realized through a virtual synchronous machine interface, and a refined model library is integrated. The refined model library is used to provide a first planning scheme for multi-scenario collaboration.

[0068] In this embodiment, through the virtual synchronous machine interface, real-time dynamic coupling simulation of multiple links of the source (power generation side), grid (power grid), load (load side), and storage (energy storage system) is realized, enhancing the system's real-time perception and simulation ability of the interactive influence of each link, and improving the accuracy and real-time performance of the simulation. In addition, integrating the refined model library provides rich options for the energy system planning under different scenarios. This refined model can take into account various possible variables and factors, thereby generating a first planning scheme for multi-scenario collaboration, providing a solid foundation for subsequent planning optimization.

[0069] In some embodiments, the real-time dynamic coupling simulation of multiple links of source-grid-load-storage through the virtual synchronous machine interface in the digital twin co-simulation platform includes:

[0070] Define functional modules and data interaction protocols in the digital twin co-simulation platform, and obtain the real-time operation data of each link of the source-grid-load-storage during the simulation process through the virtual synchronous machine, and extract the delay parameters and jitter parameters during the data transmission process.

[0071] Calculate the adjustment value of the simulation time step according to the delay parameter and the jitter parameter.

[0072] Input the real-time operation data into a preset dynamic coupling simulation model, and calculate the dynamic response characteristics of each operation node. The dynamic response characteristics are the real-time changes of operation parameters, and the operation parameters include voltage, current, and power.

[0073] Based on the dynamic response characteristics, judge whether the operation state of each operation node is within the safe range.

[0074] If the operation state of the operation node exceeds the safe range, adjust the resource configuration during the simulation process to generate a simulation optimization scheme. The resource configuration adjustment includes adjusting the power generation output, adjusting the load demand, and optimizing the charge and discharge strategies of the energy storage device.

[0075] In this embodiment, a digital twin collaborative simulation platform is constructed using a preset digital twin platform framework. The boundaries and interfaces of functional modules are defined in the digital twin collaborative simulation platform. For the interfaces of functional modules, a data interaction protocol is designed, data formats and transmission rules are formulated, and an operation logic is established according to the data interaction protocol, and a data processing flow and execution order are set. A data acquisition module is embedded in the operation logic to obtain the status information of the physical entity in real time, map the acquired status information to the digital model, generate a data synchronization instruction, update the status of the digital model, and ensure that the digital model is synchronized with the physical entity. According to the data synchronization instruction, the data management module is called to store the updated digital model data. During the data storage process, a data security mechanism is triggered to encrypt and back up the stored data, thereby establishing the operation logic and data storage structure of the digital twin collaborative simulation platform.

[0076] Specifically, a virtual synchronous machine is a device for obtaining real-time operation data of each link of the source network, load, and energy storage during the simulation process. In some possible embodiments, the virtual synchronous machine is a simulation version of a software module or a hardware device, which can simulate the power generation behavior of the synchronous power generation end and provide real-time data similar to the physical system, where the power generation end can be thermal power generation, hydropower generation, wind power generation, solar power generation, etc. based on the simulation requirements.

[0077] In this embodiment, to calculate the adjustment value of the simulation time step, a basic simulation time step needs to be determined in advance. The basic simulation time step is set based on the physical characteristics and calculation requirements of each simulation model in the digital twin collaborative simulation platform to ensure that the simulation process can accurately reflect the dynamic changes of the power system without being affected by delay and jitter. For the measurement delay parameter and jitter parameter, they can be estimated through actual tests or based on historical data. It can be understood that the delay parameter refers to the time required for data to travel from the sending end to the receiving end, and the jitter parameter refers to the range of variation of the delay parameter. In addition, to determine whether the delay and jitter have a significant impact on the simulation accuracy, preset delay thresholds and jitter thresholds need to be set. The above delay thresholds and jitter thresholds are set based on the accuracy requirements and real-time requirements of the simulation model. When the delay parameter or jitter parameter exceeds the preset threshold, the simulation time step is adjusted by an adjustment factor.

[0078] Specifically, the specific calculation formula for the simulation time step is: T_adjusted = T_base * F_adjust, where T_adjusted represents the simulation time step, T_base represents the basic simulation time step, and F_adjust represents the adjustment factor. For the adjustment factor, its specific calculation formula is: F_adjust = (1 + k_D * (D - T_delay_thresh)) * (1 + k_J * (J - T_jitter_thresh)), where k_D represents the weight coefficient of the delay, which is used to control the influence degree of the delay on the adjustment factor, D represents the actual delay, and T_delay_thresh represents the preset delay threshold; k_J represents the weight coefficient of the jitter; J represents the actual jitter; T_jitter_thresh represents the preset jitter threshold.

[0079] Specifically, the simulation model is a mathematical model or computational model used in the digital twin co-simulation platform to simulate the behavior of the actual power system. In this embodiment, the simulation model may include detailed models of various links such as the source (power generation end), network (power grid), load (load end), and storage (energy storage system), which are used to capture and predict the dynamic behavior of these links during actual operation. In this embodiment, by dynamically coupling the above-mentioned simulation models, a dynamic coupling simulation model of the source-network-load-storage is established. This dynamic coupling simulation model can comprehensively consider the interactions among the source, network, load, and storage. When real-time operation data is input, the dynamic response characteristics of each operation node can be obtained. Among them, the dynamic response characteristics are specifically the real-time changes of parameters such as voltage, current, and power, and the operation nodes are the four links of the source (power generation end), network (power grid), load (load end), and storage (energy storage system) in the source-network-load-storage.

[0080] Specifically, for the source-side simulation model, it is a power generation resource model including various power generation resources such as thermal power generation, hydropower generation, wind power generation, and solar power generation. Each power generation resource model should include the dynamic change characteristics of its power generation power, voltage, current, etc.; for the grid simulation model, it is a grid model including grid components such as transmission lines, transformers, and switches. The grid model can simulate the characteristics such as voltage, current, and power loss during the power transmission and distribution process; for the load-side simulation model, it is a load model including various electrical equipment and users. The load model should be able to reflect the change law of the load demand and its impact on the power system; for the energy storage system simulation model, it is an energy storage system model including various energy storage devices. The energy storage system model can simulate its regulation role in the power system and also includes characteristics such as the charge-discharge process, energy storage, and release.

[0081] Based on the above simulation models, a dynamic coupling relationship is established, such as the relationships between the source and the grid, the grid and the load, the grid and the energy storage, and the load and the energy storage. Specifically, in the relationship between the source and the grid, the output power of the power generation resources is transmitted to the load side and the energy storage system through the power grid, and parameters such as the voltage and current of the power grid are affected by the output power of the power generation resources and the load demand. In the relationship between the grid and the load, the power grid adjusts the power distribution and transmission in a timely manner according to the change of the load demand, and the distribution and change of the load will also affect the planning and construction of the power grid. In the relationship between the grid and the energy storage, the energy storage system participates in functions such as peak shaving, frequency modulation and voltage regulation of the power grid through connection with the power grid, and the operating state of the power grid will also affect the charge and discharge strategy of the energy storage system. In the relationship between the load and the energy storage, the change of the load demand directly affects the charge and discharge strategy of the energy storage system, and the energy storage system can meet the power supply during the peak load demand and charge during the low load demand period.

[0082] In this embodiment, by inputting real-time operation data and prediction data into the dynamic coupling simulation model and running the simulation program, the simulation program should be able to simulate the real-time operation state of the power system. By analyzing the simulation results, including the real-time change characteristics of parameters such as voltage, current and power, the dynamic response characteristics of each operation node are obtained, and then the interaction and coordination effect between the source, grid, load and energy storage, as well as the regulation effect of the energy storage system in the power system are evaluated. Among them, the real-time operation data includes the output power of the power generation resources, the load demand, the grid voltage and current, etc., and the prediction data includes load prediction, renewable energy power generation prediction, etc. It can be understood that the dynamic coupling simulation model is also pre-set with simulation parameters for each node of the source, grid, load and energy storage, specifically including parameters of grid components, such as resistance, inductance, capacitance, etc.; parameters of power generation resources, such as rated power, efficiency, etc.; parameters of the load, such as power factor, load curve, etc. and parameters of the energy storage system, such as energy storage capacity, charge and discharge rate, etc.

[0083] In this embodiment, in the dynamic coupling simulation model, a dynamic response characteristic algorithm is adopted to calculate the dynamic response characteristics of each operating node. Specifically, methods such as the finite difference method, the finite element method, and the state space method can be used for the dynamic response characteristic algorithm. The selection of the corresponding method depends on the complexity of the power system and the requirements of simulation accuracy. In addition, after calculating the dynamic response characteristics of each operating node, the operating state of each operating node is verified, which includes comparing with the measurement data of the actual power system to evaluate the accuracy and reliability of the simulation results. If there is a large error between the simulation results and the actual data, the simulation program needs to adjust and optimize the simulation parameters or the dynamic response characteristic algorithm in the dynamic coupling simulation model to improve the simulation accuracy and precision. Specifically, the dynamic response characteristics of the source-grid-load-storage are obtained through the dynamic coupling simulation model, and a preset threshold is used to judge whether the operating state of each node is within the safe range. The threshold is set based on the safety standards and operating requirements of the power system. If the operating state of the node exceeds the safe range, the resource configuration of the source-grid-load-storage is adjusted according to the simulation results and actual requirements, and the adjusted resource configuration is refitted to the dynamic coupling simulation model to verify whether the adjusted operating state meets the safety requirements.

[0084] For the equipment parameters and operating parameters of each first planning scheme provided in the refined model library, simulation calculations are carried out through the above-mentioned scheme to meet the above-mentioned safety requirements and improve the safety of the finally generated target planning scheme.

[0085] In some embodiments, the integrated refined model library, the refined model library is used to provide a first planning scheme for multi-scenario collaboration, including:

[0086] Obtain the equipment parameters and operating parameters of each planning scheme in the refined model library, and classify each planning scheme by using cluster analysis based on the equipment parameters and the operating parameters to obtain multiple classification data sets;

[0087] Establish a matching rule library for the simulation scenario requirements, and based on the matching rule library, screen at least one classification data set from the multiple classification data sets as the screening data set, and the corresponding planning scheme in the screening data set is the first planning scheme.

[0088] In this embodiment, the planning schemes in the refined model library cover various possibilities for power system optimization. For example, on the power generation side, the planning scheme considers the complementarity between different energy sources and their characteristics changing with time and seasons. On the power grid side, it considers the layout of transmission lines, the location and capacity selection of substations, and also considers the configuration of energy storage systems. By classifying each planning scheme in the refined model library through cluster analysis, the set of planning schemes similar to the simulation scenario requirements can be quickly located, thereby reducing the number of schemes to be evaluated during the simulation process and improving the simulation efficiency.

[0089] In this embodiment, in the clustering analysis, a density-based clustering method can be used to classify the planning schemes. Taking the distribution network planning as an example, power supply reliability, economy, and power quality can be used as key features for clustering. Through clustering analysis, each planning scheme with similar features is grouped into a category, such as the high-reliability and low-cost category, the medium-reliability and medium-cost category, etc. The establishment of the matching rule library needs to consider the actual requirements of the specific simulation scenario. Taking the new energy access planning as an example, if the simulation scenario focuses on the security of the power system, the matching rules should focus on indicators such as voltage stability margin and power fluctuation response ability; if it focuses on economy, the matching rules should focus on indicators such as equipment investment cost and operation and maintenance cost. Specifically, the matching rules can be set at multiple levels, and the importance of each indicator is reflected through weight configuration.

[0090] In a simple application, several planning schemes in the refined model library are formed into three classification data sets through clustering analysis. The first category is characterized by high reliability, with a power supply reliability reaching 99.99%, but the investment cost is relatively high; the second category balances reliability and economy, with a power supply reliability of 79.9% and a moderate investment cost; the third category focuses on economy, with the lowest investment cost but relatively low reliability. If the simulation scenario requirement is the planning of the power supply scheme for an industrial park, considering that there are a large number of manufacturing loads with high requirements for power supply reliability in the park, through screening by the matching rule library, the planning schemes in the first category data set can be preferentially selected. Although the initial investment of such schemes is relatively high, they can effectively reduce the power outage loss and have better comprehensive benefits in the long run. In specific applications, to ensure the rationality of the screening results, the matching rules can also introduce a dynamic weight adjustment mechanism. For example, in different seasons and different time periods, the weights of indicators such as reliability and economy can be dynamically adjusted according to factors such as load characteristics and new energy output characteristics, so that the selected planning schemes can better adapt to the actual operation requirements and improve the adaptability and practicability of the planning schemes.

[0091] S200. Based on the simulation scenario, re-plan the energy data of the first planning scheme to obtain a second planning scheme with energy complementarity and coordinated optimization;

[0092] In this embodiment, by re-planning the first planning scheme, a second planning scheme with energy complementarity and coordinated optimization is obtained, thereby promoting the coordinated optimization between different energy systems, improving the energy utilization efficiency, and reducing the system operation cost.

[0093] In some of these embodiments, the above step S200 includes:

[0094] Obtain the energy data of the first planning scheme, where the energy data includes thermal power, hydropower, wind power, and photovoltaic power;

[0095] The decoupling calculation of the simulation scenario is carried out by using the linear programming algorithm to obtain the energy data in the simulation scenario;

[0096] Calculate the correlation coefficients between the energy data in the simulation scenario, and obtain the complementary relationship between different energies by analyzing the correlation coefficients;

[0097] Based on the complementary relationship, calculate the configuration plan after energy complementarity by using the genetic algorithm. Based on the configuration plan, re-plan the energy data of the first planning plan to generate the second planning plan.

[0098] In this embodiment, in a specific scenario, combined dispatching is carried out between thermal power generation and hydropower generation, while the power generation characteristics of wind power generation and solar power generation are affected by weather conditions and need to be further combined with thermal power generation or hydropower generation.

[0099] In this embodiment, for specific scenario requirements, the energy data in the scenario requirements are obtained by using the linear programming algorithm. During the decoupling calculation, the independent characteristics between different energies can be revealed, such as the base load characteristics of thermal power, the peak shaving ability of hydropower, and the volatility characteristics of wind power and photovoltaic power. The calculation of the correlation coefficient helps to discover the complementary relationship between energies. In a specific simulation scenario, the area where it is located has sufficient sunlight during the day and strong wind at night. Therefore, in this scenario, specifically, there can be two energy data of solar energy and wind energy. By calculating the correlation between solar power generation and wind power generation, the correlation coefficient between wind power generation and solar power generation is -0.6, indicating that the two have good complementarity. Then, based on this complementary relationship, a genetic algorithm is used to formulate a configuration plan for the first planning plan. Specifically, the genetic algorithm searches for the optimal solution by simulating the natural evolution process.

[0100] Specifically, when using the linear programming algorithm to carry out the decoupling calculation of the simulation scenario, according to the simulation scenario and simulation requirements, a linear programming model is established. The linear programming model needs to establish an objective function and constraint conditions. The objective function can be to minimize costs, maximize energy utilization efficiency, or meet specific energy demands, etc. The constraint conditions can be energy supply and demand balance, energy conversion efficiency limit, energy storage capacity limit, etc. Use a linear programming solver, such as the simplex method, interior point method, etc. to solve the linear programming model and output the decoupled energy data, that is, the power generation and supply and demand situations of each energy at different time points.

[0101] In this embodiment, by reasonably configuring the proportions of different energies according to the energy data in the simulation scenario requirements, not only the stability of power supply is guaranteed, but also the consumption capacity of renewable energy is improved, realizing the unity of economic benefits and environmental benefits.

[0102] S300. Use the Wasserstein generative adversarial network to generate extreme scenario samples, establish a multi-dimensional uncertainty joint probability model based on the extreme scenario samples, conduct dynamic risk assessment on the second planning scheme, and generate a risk assessment report.

[0103] In this embodiment, the Wasserstein generative adversarial network is used to generate extreme scenario samples, and a multi-dimensional uncertainty joint probability model is established based on these samples. Thus, it is possible to more comprehensively consider the uncertainties and extreme situations in the energy system, conduct dynamic risk assessment on the second planning scheme, and the generated risk assessment report provides important reference information for decision-makers, helping to reduce potential risks.

[0104] In some of these embodiments, using the Wasserstein generative adversarial network to generate extreme scenario samples includes:

[0105] Obtain historical extreme value data from historical data, perform normalization processing on the data, and establish a data distribution model using the kernel density estimation method.

[0106] According to the data distribution model, design the structure of the Wasserstein generative adversarial network, and set the number of layers, activation functions, and learning rate of the generator and discriminator.

[0107] Train the generative adversarial network with the historical extreme value data to generate extreme scenario samples.

[0108] Extract key feature vectors from the generated extreme scenario samples, where the key feature vectors include extreme points, change trends, and distribution patterns.

[0109] According to the key feature vectors, dynamically adjust the learning rate and regularization parameters of the Wasserstein generative adversarial network, and retrain the network to make the Wasserstein generative adversarial network generate more accurate extreme scenario samples.

[0110] In this embodiment, extreme value data is screened out from the historical data of the power system in the area where the simulation scenario is located. These extreme value data represent the extreme situations that the power system may face, such as peak and trough of electricity consumption load, fluctuations in renewable energy power generation, etc. The screened extreme value data is normalized to make data with different characteristics at the same magnitude, which is convenient for subsequent model processing. The kernel density estimation method is used to establish a data distribution model for the extreme value data to describe the statistical characteristics and distribution laws of the data. During the model establishment process, appropriate kernel functions and bandwidth parameters are selected to improve the fitting accuracy and generalization ability of the data distribution model. At the same time, the data distribution model is verified and calibrated to ensure the accuracy and reliability of the data distribution model. When designing the generative adversarial network, the generator adopts a four-layer neural network structure, and each layer uses the rectified linear unit as the activation function to enhance the nonlinear expression ability of the network. The discriminator adopts a three-layer structure and uses the hyperbolic tangent function as the activation function, which is beneficial to stabilizing the training process. The learning rate is initially set to 0.001 and gradually decreases as the training progresses. Then, the historical extreme value data is used to train the generative adversarial network to generate extreme scenario samples that conform to the data distribution.

[0111] For the generated extreme scenario samples, it is also necessary to extract their key feature vectors, and then adjust the model parameters of the Wasserstein generative adversarial network for adaptive training to improve the reliability and accuracy of the extreme scenario samples.

[0112] Specifically, taking the specific simulation scenario where there is sufficient daylight during the day and strong wind at night as an example, the historical data of the corresponding area is obtained. A batch of extreme scenario samples are generated through the trained Wasserstein generative adversarial network. The above extreme scenario samples include situations such as a sudden drop in power generation caused by rainy weather and power generation interruption caused by equipment failures. The feature vectors extracted from these extreme scenario samples include key information such as the moment when the lowest power generation occurs, the power decline rate, the recovery time, etc., and key feature vectors such as a sudden drop in power caused by a sudden change in wind speed and long-term power restriction caused by continuous strong wind. By dynamically adjusting the network parameters, the ability to depict such extreme situations is improved.

[0113] In some embodiments, establishing the multi-dimensional uncertainty joint probability model based on the extreme scenario samples includes:

[0114] Based on the extreme scenario samples, a Copula function is used to construct a multi-dimensional uncertainty joint distribution, and the multi-dimensional uncertainty joint distribution is used to describe the correlation between different uncertainty factors;

[0115] Through the non-parametric kernel density estimation method, probability modeling is performed on the multi-dimensional uncertainty joint distribution to generate a multi-dimensional uncertainty joint probability model;

[0116] Based on the multi-dimensional uncertainty joint probability model, the scenario reduction method is used to screen out a set of typical operation scenarios from multiple generated extreme scenario samples.

[0117] Specifically, the Copula function is a tool used to describe the dependence structure between multiple random variables. It can "connect" multiple one-dimensional marginal distributions to form a multi-dimensional joint distribution. In this embodiment, the generated extreme scenario samples are used as inputs, and the Copula function is used to construct the joint distribution between different uncertainty factors in these extreme scenario samples, thereby capturing the correlation between these uncertainty factors, where the uncertainty factors are the uncertainties of wind speed, rainfall, and temperature.

[0118] Specifically, non-parametric kernel density estimation is a non-parametric method for estimating the probability density function of a random variable. It does not require prior assumption of the data distribution form, but estimates the density function based on the data samples themselves. Thus, by using kernel density estimation to estimate the density function of this multi-dimensional distribution, a multi-dimensional uncertainty joint probability model is obtained.

[0119] In practical applications, the number of generated extreme scenario samples may be huge. Directly processing these extreme scenario samples is very time-consuming and computationally intensive. Therefore, through the scenario reduction method, a set of representative typical operation scenarios is screened out from a large number of generated scenarios. These typical operation scenario sets can well represent the main characteristics and probability distributions of multiple extreme scenario samples, thus greatly simplifying the subsequent analysis and decision-making processes while retaining key uncertainty information.

[0120] In some of the embodiments, the dynamic risk assessment of the second planning scheme to generate a risk assessment report includes:

[0121] Based on the set of typical operation scenarios, perform extreme scenario simulation on the second planning scheme;

[0122] Extract the failure probability from the simulation results and calculate the failure probability distribution under each extreme scenario;

[0123] Use the linear regression method to quantify the loss degree under each extreme scenario and generate a loss degree matrix;

[0124] Based on the failure probability distribution and the loss degree matrix, use the analytic hierarchy process to calculate the risk index value of the second planning scheme under each extreme scenario;

[0125] Analyze the risk index values to generate a risk assessment report.

[0126] In this embodiment, the second planning scheme is input into the multi-dimensional uncertainty joint probability model, and the Monte Carlo simulation method is used to calculate the risk index values under different extreme scenarios. The risk indices include the failure probability and the economic loss evaluation value. The extreme scenarios are the extreme scenario samples in the above-mentioned typical operation scenario set, and the failure probability is specifically the probability of power supply interruption.

[0127] In this embodiment, the loss data under each historical extreme scenario is collected, including direct economic losses, indirect economic losses, etc. The characteristic variables of the extreme scenario and the dependent variable of the loss degree are defined. The characteristic variables include disaster intensity, duration, influence range, etc., and the dependent variable of the loss degree is the loss value to be quantified. According to the characteristics of the data and the requirements of the problem, an appropriate linear regression model is selected, such as simple linear regression, multiple linear regression, etc. The characteristic variable values under the extreme scenario are substituted into the trained linear regression model, and then the corresponding loss degree is quantified. According to the loss degree, a loss degree matrix is constructed. Each row of the matrix represents an extreme scenario, and each column represents a type of loss, such as direct economic loss, indirect economic loss, etc. The elements in the matrix are the loss degree values corresponding to the extreme scenario and the loss type.

[0128] In this embodiment, a hierarchical structure model is constructed, and the problem is divided into an objective layer, a criterion layer, and a scheme layer. Specifically, the objective layer is to calculate the risk index values of the second planning scheme under each extreme scenario. The criterion layer is the various criteria for evaluating the risk index values, such as failure probability, loss degree, etc. The scheme layer is the specific performance of the second planning scheme under each extreme scenario. The factors at the same level are compared pairwise to determine their relative importance. Specifically, the 1-9 scale method is used to represent the relative importance degree between factors. According to the pairwise comparison results, a judgment matrix for each layer is constructed. According to the judgment matrix, the eigenvector method is used to calculate the weight vector corresponding to the layer. Specifically, each column of the judgment matrix is normalized to obtain a normalized matrix. It can be understood that each element of the normalized matrix is equal to the corresponding element of the judgment matrix divided by the sum of the elements in that column. Then, the sum of each row of the normalized matrix is calculated to obtain the weight vector. According to the weight vectors of each layer, the total sorting weight of the second planning scheme under each extreme scenario is calculated. The weighted summation method is used to combine the data in the failure probability distribution and the loss degree matrix with the total sorting weight to obtain the risk index value.

[0129] S400. Based on the risk assessment report and combined with the business model, a feasible target planning scheme is generated. The business model includes business requirements, power plant resources, and specification costs.

[0130] In this embodiment, a feasible target planning scheme generated based on a risk assessment report and a business model combines the risk assessment results with actual business requirements, ensuring the feasibility and practicality of the planning scheme. At the same time, by comprehensively considering business requirements, resource limitations, and cost factors, the generated target planning scheme can better meet the actual needs of the system, improving the overall performance and economic benefits of the system.

[0131] In some of these embodiments, the above step S400 includes:

[0132] Obtain the historical load volume, calculate the load growth amount using the ARIMA model, and obtain the load growth prediction value;

[0133] Obtain the energy data and risk index values of the second planning scheme, and calculate the matching degree between the load growth prediction value and the energy data;

[0134] If the matching degree is greater than a preset matching threshold and the risk index value is less than a preset index threshold, add the corresponding second planning scheme to the target planning scheme set.

[0135] In this embodiment, to obtain the load growth prediction value of the area where the simulation scenario is located, it can be carried out through the historical load volume of the area. Specifically, using the time series analysis method, obtain the historical load data, organize it into a time series, conduct a stationarity test on the time series. If the data is not stationary, perform differencing processing, determine the differencing order, calculate the autocorrelation coefficient and partial autocorrelation coefficient based on the differenced sequence, determine the autoregressive order and moving average order, estimate the ARIMA model parameters using the least squares method, construct a prediction model, fit the historical data through the prediction model, calculate the residual term, and evaluate the prediction model accuracy. Based on the constructed prediction model, predict the load growth amount in the future period and calculate the confidence interval. Adjust the prediction model parameters according to the prediction results and error values to optimize the prediction accuracy.

[0136] In this embodiment, the energy data of the second planning scheme is obtained. Specifically, the energy data is the energy types included in the second planning scheme, as well as the rated power and adjustment range planned to be generated by the corresponding energy types, etc., thereby constituting the power characteristic set of the second planning scheme. Specifically, data extraction technology is used to obtain the energy data from the second planning scheme, classify the energy types in the energy data, and extract the rated power and adjustment range of each energy according to the classification results. Through preset data construction rules, the extracted rated power and adjustment range are integrated to form a power characteristic description of each energy, and a power characteristic set is generated based on the power characteristic descriptions of all energies. If there are multiple energy types in the power characteristic set, the total rated power and adjustment range interval of each energy are calculated respectively to obtain the overall power characteristics of the second planning scheme. It can be understood that the overall power characteristics can determine the power supply capacity of the second planning scheme.

[0137] In this embodiment, the matching degree between the power characteristic set and the load growth prediction value is calculated. Specifically, for the total rated power of each energy in the power characteristic set, the total rated power is compared with the load growth prediction value. If the total rated power is greater than the load growth prediction value, it means that the energy data of the second planning scheme meets the power generation requirements in the simulation scenario. At this time, the matching degree between the total rated power and the load growth prediction value is calculated to generate a matching degree. Specifically, the matching degree calculation formula is where P match represents the matching degree percentage, P rated,i represents the rated power of the i-th energy, n represents the total number of energies, and L pred represents the load growth prediction value.

[0138] In this embodiment, if the matching degree is higher than the preset matching threshold and the risk index value is lower than the preset index threshold, it means that the second planning scheme meets the overall power distribution requirements and corresponding safety requirements of the power system. Therefore, the corresponding second planning scheme can be added to the target planning scheme set for the decision-maker to select.

[0139] In some of these embodiments, for the second planning scheme added to the target planning scheme set, it further includes:

[0140] Obtain the power plant resources and the investment budget and operation and maintenance costs in the specification costs, and establish a cost control model;

[0141] Input the second planning scheme into the cost control model to generate a cost evaluation index, and determine the target planning scheme based on the cost evaluation index.

[0142] In this embodiment, the investment budget and operation and maintenance costs in the power plant resources and specification costs are obtained, and the key values in the investment budget and operation and maintenance costs are extracted. Specifically, the investment budget in the specification costs includes power generation equipment procurement costs, infrastructure project costs, technical renovation costs, etc., and the operation and maintenance costs cover daily inspection costs, labor costs, fuel consumption, etc. In the construction of the cost control model, it is necessary to comprehensively consider fixed asset investment, operating costs, and maintenance expenditures. Specifically, the cost control model includes a cost control function, and the specific calculation formula is as follows: Among them, C t represents the total cost prediction value, β 0 represents the basic cost constant term, β i represents the cost coefficient of the i-th factor, x i represents the value of the i-th influencing factor, ε represents the random error term, and n represents the total number of influencing factors.

[0143] In this embodiment, when the second planning scheme is input into the cost control model, the cost control model analyzes the second planning scheme, converts the costs involved in the second planning scheme into a numerical format so that it can be input into the cost control function in the cost control model, and generates a cost evaluation index. It can be understood that in the construction process, the cost control model collects cost-related data in the power system from channels such as historical data, market research, and internal reports. The data should cover various cost elements, such as raw material costs, labor costs, equipment depreciation, etc. According to the needs of cost control, the costs are divided into fixed costs, variable costs, semi-variable costs, etc., and clear calculation methods and formulas are defined for each cost type to facilitate the conversion of the costs involved in the second planning scheme into a numerical format.

[0144] In this embodiment, the cost evaluation indicators in the second planning scheme are sorted according to the scores, and the decision-maker can select and determine the target planning scheme according to the scores.

[0145] Please refer to Figure 2 as shown. The present invention also provides a multi-objective planning system for the coordinated operation of the power system's source, grid, load, and storage. The system includes:

[0146] A first processing module 201, configured to implement real-time dynamic coupling simulation of multiple links of the source, grid, load, and storage through a virtual synchronous machine interface in a digital twin collaborative simulation platform, and integrate a refined model library, where the refined model library is used to provide a first planning scheme for multi-scenario coordination;

[0147] A second processing module 202, configured to re-plan the energy data of the first planning scheme based on the simulation scenario to obtain a second planning scheme for energy complementarity and coordinated optimization;

[0148] The third processing module 203 is configured to generate extreme scenario samples by using a Wasserstein generative adversarial network, establish a multi-dimensional uncertainty joint probability model based on the extreme scenario samples, perform dynamic risk assessment on the second planning scheme, and generate a risk assessment report;

[0149] The fourth processing module 204 is configured to generate a feasible target planning scheme based on the risk assessment report and in combination with the business model, where the business model includes business requirements, power plant resources, and specification costs.

[0150] It can be understood that the content in the embodiment of the multi-objective planning method for the coordinated operation of the power generation, transmission, load, and energy storage in the power system as Figure 1 shown is applicable to the embodiment of the multi-objective planning system for the coordinated operation of the power generation, transmission, load, and energy storage in the power system. The functions specifically implemented in the embodiment of the multi-objective planning system for the coordinated operation of the power generation, transmission, load, and energy storage in the power system are the same as those in the embodiment of the multi-objective planning method for the coordinated operation of the power generation, transmission, load, and energy storage in the power system as Figure 1 shown, and the beneficial effects achieved are also the same as those in the embodiment of the multi-objective planning method for the coordinated operation of the power generation, transmission, load, and energy storage in the power system as Figure 1 shown.

[0151] It should be noted that for the information interaction, execution process, etc. between the above systems, since they are based on the same concept as the method embodiment of the present invention, the specific functions and the technical effects brought thereby can be specifically referred to the method embodiment part, and will not be elaborated herein.

[0152] Those skilled in the art can clearly understand that for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In practical applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the system is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of the present application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiment, and will not be elaborated herein.

[0153] The foregoing embodiments are only used to illustrate the technical solutions of the present application, rather than to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included within the protection scope of the present application.

Claims

1. A multi-objective planning method for power system source-grid-load-storage coordination, characterized in that: include: In the digital twin collaborative simulation platform, the real-time dynamic coupling simulation of multiple links of source-grid-load-storage is realized through the virtual synchronous machine interface, and a refined model library is integrated. The refined model library is used to provide the first planning scheme for multi-scenario collaboration; Based on the simulation scenario, the energy data of the first planning scheme is replanned to obtain a second planning scheme with energy complementarity and coordinated optimization; Using a Wasserstein generative adversarial network to generate extreme scenario samples, establishing a multidimensional uncertainty joint probability model based on the extreme scenario samples, performing a dynamic risk assessment on the second planning scheme, and generating a risk assessment report; Based on the risk assessment report and combined with the business model, a feasible target planning scheme is generated, wherein the business model includes business requirements, power plant resources and specification costs.

2. The method according to claim 1, characterized in that In the digital twin collaborative simulation platform, the real-time dynamic coupling simulation of multiple links of source-grid-load-storage is realized through the virtual synchronous machine interface, including: Define functional modules and data interaction protocols in the digital twin collaborative simulation platform, obtain real-time operation data of each link of source, grid, load and storage during the simulation process through the virtual synchronizer, and extract delay parameters and jitter parameters during data transmission; Calculating an adjustment value of a simulation time step according to the delay parameter and the jitter parameter; Inputting the real-time operation data into a preset dynamic coupling simulation model to calculate the dynamic response characteristics of each operation node, wherein the dynamic response characteristics are real-time changes in operation parameters, and the operation parameters include voltage, current and power; Based on the dynamic response characteristics, determine whether the operating state of each operating node is within a safe range; If the running status of the running node exceeds the safe range, the resource configuration in the simulation process is adjusted to generate a simulation optimization plan.

3. The method according to claim 1, characterized in that The integrated refined model library, which is used to provide a first planning solution for multi-scenario collaboration, includes: Acquire equipment parameters and operating parameters of each planning scheme in the refined model library, and classify each planning scheme based on the equipment parameters and the operating parameters by cluster analysis to obtain multiple classification data sets; A matching rule library is established according to the simulation scenario requirements. Based on the matching rule library, at least one classified data set is screened from the multiple classified data sets as a screened data set, and the corresponding planning scheme in the screened data set is the first planning scheme.

4. The method according to claim 1, characterized in that The replanning of the energy data of the first planning scheme based on the simulation scenario to obtain a second planning scheme for energy complementarity and coordinated optimization includes: Acquiring energy data of the first planning scheme, the energy data including thermal power, hydropower, wind power and photovoltaic power; A linear programming algorithm is used to decouple the simulation scenario and obtain the energy data in the simulation scenario; Calculate the correlation coefficient between the energy data in the simulation scenario, and obtain the complementary relationship between different energy sources by analyzing the correlation coefficient; Based on the complementary relationship, a configuration scheme after energy complementation is calculated by a genetic algorithm, and based on the configuration scheme, energy data of the first planning scheme is replanned to generate a second planning scheme.

5. The method according to claim 1, characterized in that The Wasserstein generative adversarial network is used to generate extreme scene samples, including: Obtain historical extreme value data from historical data, normalize the data, and use the kernel density estimation method to establish a data distribution model; According to the data distribution model, design the structure of the Wasserstein generative adversarial network, set the number of layers, activation function and learning rate of the generator and discriminator; Generate extreme scenario samples by training a generative adversarial network using the historical extreme value data; Extract key feature vectors from the generated extreme scenario samples, wherein the key feature vectors include extreme value points, change trends, and distribution forms; According to the key feature vector, the learning rate and regularization parameters of the Wasserstein generative adversarial network are dynamically adjusted, and the network is retrained so that the Wasserstein generative adversarial network can generate more accurate extreme scene samples.

6. The method according to claim 5, characterized in that The establishing of a multi-dimensional uncertainty joint probability model based on the extreme scenario samples includes: Based on extreme scenario samples, a Copula function is used to construct a multidimensional uncertainty joint distribution, which is used to describe the correlation between different uncertainty factors; Probabilistic modeling is performed on the multidimensional uncertain joint distribution by a nonparametric kernel density estimation method to generate a multidimensional uncertain joint probability model; Based on the multidimensional uncertainty joint probability model, the scenario reduction method is used to screen out a set of typical operating scenarios from multiple generated extreme scenario samples.

7. The method according to claim 6, characterized in that The step of performing a dynamic risk assessment on the second planning scheme and generating a risk assessment report includes: Based on a set of typical operating scenarios, extreme scenario simulations are performed on the second planning scheme; Extract the failure probability from the simulation results and calculate the failure probability distribution under each extreme scenario; The linear regression method is used to quantify the loss degree in each extreme scenario and generate a loss degree matrix; Based on the failure probability distribution and loss degree matrix, using the hierarchical analysis method, calculate the risk index value of the second planning scheme in each extreme scenario; The risk indicator values ​​are analyzed to generate a risk assessment report.

8. The method according to claim 1, characterized in that Based on the risk assessment report, a feasible target planning scheme is generated in combination with a business model, wherein the business model includes business requirements, power plant resources and specification costs, including: Obtain historical load, calculate load growth using the ARIMA model, and obtain load growth forecasts; Obtaining energy data and risk indicator values ​​of the second planning scheme, and calculating a matching degree between the load growth forecast value and the energy data; If the matching degree is greater than a preset matching threshold, and the risk index value is less than a preset index threshold, the corresponding second planning scheme is added to the target planning scheme set.

9. The method according to claim 8, characterized in that For the second planning scheme added to the target planning scheme set, it also includes: Obtain power plant resources and investment budget and operation and maintenance costs from specification costs, and establish a cost control model; The second planning scheme is input into the cost control model to generate a cost evaluation index, and a target planning scheme is determined based on the cost evaluation index.

10. A multi-objective planning system for power system source-grid-load-storage coordination, characterized in that: include: A first processing module is used to implement real-time dynamic coupling simulation of multiple links of source-grid-load-storage through a virtual synchronous machine interface in a digital twin collaborative simulation platform, and integrate a refined model library, wherein the refined model library is used to provide a first planning scheme for multi-scenario collaboration; A second processing module is used to re-plan the energy data of the first planning scheme based on the simulation scenario to obtain a second planning scheme with energy complementarity and coordinated optimization; A third processing module is used to generate extreme scenario samples using a Wasserstein generative adversarial network, establish a multidimensional uncertainty joint probability model based on the extreme scenario samples, perform dynamic risk assessment on the second planning scheme, and generate a risk assessment report; The fourth processing module is used to generate a feasible target planning solution based on the risk assessment report and in combination with the business model, wherein the business model includes business requirements, power plant resources and specification costs.

Images