Source network load storage capacity dynamic planning method and system based on digital twinning

Abstract

The invention relates to a source network load storage capacity dynamic planning method and system based on digital twinning, and the method comprises the steps: obtaining the operation parameter data of all equipment, and obtaining a final equipment operation parameter set; performing mathematical abstraction on each device by adopting a unified modeling mode to obtain a unified device model; equipment heterogeneous characteristic parameters are extracted from the unified equipment model, and the energy storage capacity is adjusted through a dynamic allocation mechanism; performing real-time capacity allocation according to the energy storage capacity allocation scheme and the real-time load change data by adopting an optimal scheduling algorithm to obtain an equipment operation instruction set; obtaining an equipment cooperative operation state, and obtaining a resource utilization efficiency evaluation result based on the equipment cooperative operation state; and readjusting unified equipment model parameters based on a resource utilization efficiency evaluation result to obtain an optimized capacity configuration scheme. Compared with the prior art, unified modeling and dynamic allocation of different types of equipment are realized, and the method has the advantages of improving the collaborative efficiency and the resource utilization rate of the source network load storage system and the like.

Description

Technical Field

[0001] This invention belongs to the field of smart grid technology and relates to a source-grid-load-storage capacity planning method, and in particular to a source-grid-load-storage capacity dynamic planning method and system based on digital twins. Background Technology

[0002] As the core pillar of modern society's energy supply, the efficient operation and optimized configuration of the power system are crucial for ensuring energy security and promoting green development. With the rapid development of renewable energy and the diversification of electricity loads, the coordinated optimization of the power generation, grid, load, and energy storage (Generation, Grid, Load, Energy Storage) system has become a focus of industry attention. For example, Chinese patent application CN120433182A discloses a coordinated planning method for the power generation, grid, load, and energy storage system, including: obtaining the structural, operational, and economic parameters of the power system; dividing extreme scenarios and establishing a frequency response model; setting frequency security constraints based on the extreme scenarios; constructing a three-layer optimization planning model, including objective functions and constraints, where the third layer includes frequency security constraints; and solving the model to determine the planning scheme. However, existing methods often struggle to achieve unified management and dynamic coordination among equipment when dealing with complex and ever-changing power systems, especially when facing the heterogeneous characteristics of different types of equipment, lacking flexible adaptability and efficient resource allocation mechanisms. This makes it difficult for the system to quickly respond to changes in demand during operation, affecting overall efficiency.

[0003] Against this backdrop, capacity planning for power generation, grid, load, and energy storage systems faces two core technological challenges. First, due to differences in operating characteristics and control logic, different types of equipment (such as wind power, photovoltaics, and energy storage) are difficult to manage collaboratively across devices through a unified modeling approach. For example, in actual operation, the intermittency of wind power and the rapid response requirements of energy storage are difficult to match in real time on the same platform, leading to low resource utilization. Second, due to the lack of a dynamic allocation mechanism, the system cannot flexibly allocate capacity based on real-time load changes and fluctuations in renewable energy output. This further exacerbates the coordination difficulties between devices; for example, during peak load periods, energy storage devices may be unable to effectively support stable grid operation due to a lack of unified scheduling. Summary of the Invention

[0004] The purpose of this invention is to provide a dynamic planning method and system for source-grid-load-storage capacity based on digital twins, which can realize unified modeling and dynamic allocation of different types of equipment, and improve the collaborative efficiency and resource utilization of the source-grid-load-storage system.

[0005] A dynamic planning method for source-grid-load-storage capacity based on digital twins includes the following steps:

[0006] The operating parameter data of each device in the source-grid-load-storage system are obtained to obtain the final set of device operating parameters. The operating parameter data includes wind power intermittent characteristic data, photovoltaic volatility data, energy storage rapid response capability data, and real-time load change data.

[0007] Based on the final set of equipment operating parameters, a unified modeling approach is used to mathematically abstract each piece of equipment to obtain a unified equipment model.

[0008] The heterogeneous characteristic parameters of the equipment are extracted from the unified equipment model and classified to obtain the equipment characteristic classification results. Based on the equipment characteristic classification results, the energy storage capacity is adjusted through a dynamic allocation mechanism to obtain the energy storage capacity allocation scheme.

[0009] An optimized scheduling algorithm is used to perform real-time capacity allocation based on the energy storage capacity allocation plan and real-time load change data, and a set of equipment operation instructions is obtained based on the real-time capacity allocation results.

[0010] Based on the set of equipment operation instructions, a distributed control algorithm is used to coordinate the control of wind power, energy storage and load equipment to obtain the coordinated operation status of the equipment, and the resource utilization efficiency evaluation result is obtained based on the coordinated operation status of the equipment.

[0011] Based on the resource utilization efficiency assessment results, the parameters of the unified equipment model are readjusted to obtain an optimized capacity configuration scheme.

[0012] Furthermore, the process of obtaining the final set of equipment operating parameters includes:

[0013] A preliminary set of operating parameters is constructed based on the operating parameter data of each device in the source-grid-load-storage system;

[0014] The weighted average method is used to integrate the wind power and energy storage related data in the preliminary set of operating parameters to obtain the final set of equipment operating parameters.

[0015] Furthermore, the process of obtaining the unified device model includes:

[0016] The wind power intermittent characteristic data and photovoltaic volatility data in the final equipment operating parameter set are standardized to obtain a standardized parameter set;

[0017] Based on the standardized parameter set, and combined with the attributes of energy storage response speed and load change rate, the characteristic curve of the equipment is fitted by a linear regression algorithm to obtain a characteristic fitting model. The characteristic fitting model uses the attribute values ​​in the standardized parameter set as independent variables and the corresponding equipment output as dependent variables.

[0018] Based on the aforementioned characteristic fitting model, a unified model library is constructed by integrating parameter unification and model abstraction layer attributes to obtain a unified device model. The unified model library stores the corresponding association between fitting coefficients and abstraction layer attributes.

[0019] Furthermore, the process of obtaining the equipment characteristic classification results includes:

[0020] By combining the heterogeneous characteristic parameters of the equipment with the equipment compatibility index, a clustering algorithm is used to obtain the characteristic classification groups and the coordinates of the cluster centers;

[0021] Based on the cluster center coordinates, the final set of equipment operating parameters and model abstract attributes are integrated to construct equipment grouping mappings and obtain equipment characteristic classification results.

[0022] Furthermore, the heterogeneous characteristic parameters of the equipment include photovoltaic output variation and load demand elasticity.

[0023] Furthermore, the process of obtaining the energy storage capacity allocation scheme includes:

[0024] Based on the equipment characteristic classification results, it is determined whether the wind power intermittent characteristics are significant. If so, a time series analysis algorithm is used to predict the wind power output fluctuation. The adjustment value of the energy storage rapid response capacity is calculated through a dynamic allocation mechanism to obtain a preliminary framework for energy storage capacity allocation. The dynamic allocation mechanism refers to determining the adjustment value by multiplying the predicted fluctuation amplitude by a preset response coefficient.

[0025] Based on the preliminary framework for energy storage capacity allocation, an energy storage capacity allocation scheme is obtained by integrating fluctuation range assessment and energy balance optimization.

[0026] Furthermore, the real-time capacity allocation specifically includes:

[0027] The current status indicators of the energy storage capacity allocation scheme and real-time load change data are obtained from the source-grid-load-storage system. Using these current status indicators as input, an optimization scheduling algorithm is used to obtain a preliminary output ratio framework.

[0028] Based on the aforementioned initial output ratio framework, a dynamic response mechanism is integrated to calculate the output adjustment value of each device, obtain the constraints for energy balance optimization, and obtain an intermediate scheme for real-time capacity allocation.

[0029] Based on the intermediate solution, the output ratio is adjusted to obtain the corrected allocation parameters;

[0030] The real-time capacity allocation results are obtained based on the corrected allocation parameters.

[0031] Furthermore, the process of obtaining the cooperative operating status of the equipment includes:

[0032] Distributed collaborative parameters are obtained from the set of equipment operation instructions, and distributed control algorithms are used to perform preliminary adjustments on the wind power equipment to obtain wind power regulation results.

[0033] Based on the wind power regulation results, relevant instructions are transmitted to the energy storage device controller via a real-time communication protocol to obtain the energy storage optimization status;

[0034] Based on the energy storage optimization status, coordinated control is applied to the load equipment, and the execution of instructions is verified by a communication fusion mechanism to obtain load balance indicators;

[0035] Operational monitoring data is extracted from the load balance index. If the operational monitoring data meets the preset operational threshold, the status feedback information is integrated to obtain the collaborative operation status of the equipment.

[0036] Furthermore, the process of obtaining the resource utilization efficiency assessment results includes:

[0037] Resource utilization efficiency data is obtained from the collaborative operation status of the equipment. Data analysis algorithms are used to statistically analyze the efficiency data. Based on the mean comparison method, it is determined whether the system operating efficiency has reached a preset threshold, and the resource utilization efficiency evaluation result is obtained.

[0038] Furthermore, the adjustment of the unified equipment model parameters specifically includes:

[0039] Fluctuation amplitude data is obtained from the real-time load change data, the mean of the fluctuation amplitude data is calculated, and the intermittent characteristics of wind power and the mean of the fluctuation amplitude data are fused to extract features and obtain the intermittent impact index. The parameters of the unified equipment model are adjusted according to the intermittent impact index using the gradient descent algorithm.

[0040] This invention also provides a source-grid-load-storage capacity dynamic planning system based on digital twins, comprising:

[0041] The data acquisition module is used to acquire the operating parameter data of each device in the source-grid-load-storage system to obtain the final set of device operating parameters. The operating parameter data includes wind power intermittent characteristic data, photovoltaic volatility data, energy storage rapid response capability data, and real-time load change data.

[0042] The unified modeling module is used to perform mathematical abstraction of each device based on the final set of device operating parameters, and obtain a unified device model.

[0043] The feature classification module is used to extract heterogeneous feature parameters of equipment from the unified equipment model and classify them to obtain equipment feature classification results.

[0044] The capacity allocation module is used to adjust the energy storage capacity through a dynamic allocation mechanism based on the equipment characteristic classification results to obtain an energy storage capacity allocation scheme.

[0045] The optimization scheduling module is used to perform real-time capacity allocation of the energy storage capacity allocation scheme and real-time load change data, and to obtain the real-time capacity allocation result by using an optimization scheduling algorithm to perform real-time capacity allocation of the source-grid-load-storage system.

[0046] The instruction issuing module is used to issue the capacity allocation result to each device controller to obtain a set of device operation instructions;

[0047] The collaborative control module is used to obtain the collaborative operation status of wind power, energy storage and load equipment by using a distributed control algorithm based on the set of equipment operation instructions.

[0048] The efficiency evaluation module is used to obtain resource utilization efficiency data based on the collaborative operation status of the equipment, and to evaluate the resource utilization efficiency data to obtain a resource utilization efficiency evaluation result.

[0049] The iterative optimization module is used to readjust the parameters of the unified equipment model based on the resource utilization efficiency evaluation results to obtain an optimized capacity configuration scheme.

[0050] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described above.

[0051] The objective of this invention can be achieved through the following technical solutions:

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] This invention constructs a logical framework that integrates data acquisition, modeling and analysis, dynamic allocation and optimized scheduling, as well as digital twin technologies such as mathematical abstraction and unified model construction. It effectively solves the problem of balancing grid operation efficiency and stability caused by the intermittency of wind power, the rapid response of energy storage and real-time load changes. It realizes unified modeling and dynamic allocation of different types of equipment, and improves the collaborative efficiency and resource utilization of the source-grid-load-storage system.

[0054] This invention first acquires operating parameters from equipment sensors and the power grid monitoring platform, and uses a linear regression algorithm to construct a unified equipment model, accurately abstracting the operating characteristics of wind power, photovoltaics, energy storage, and loads. Addressing the heterogeneous characteristics such as wind power output fluctuations and energy storage response times, it employs a clustering algorithm for classification and predicts wind power fluctuations based on time series analysis, dynamically adjusting energy storage capacity. A particle swarm optimization algorithm is used to calculate real-time capacity allocation ratios to ensure stable grid operation. If operating efficiency falls below a preset threshold, this invention updates the capacity planning framework and reconfigures resources through an iterative optimization algorithm. This invention achieves efficient collaborative operation of equipment through coordinated control and real-time communication protocols, significantly improving resource utilization efficiency and grid stability, solving the dynamic balance problem in multi-energy collaborative optimization, and providing efficient and stable operational guarantees for smart grids. Attached Figure Description

[0055] Figure 1 This is a schematic flowchart of the method of the present invention;

[0056] Figure 2 This is a schematic diagram of the system of the present invention. Detailed Implementation

[0057] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0058] Example 1

[0059] This embodiment provides a dynamic planning method for source-grid-load-storage capacity based on digital twins, including: acquiring the operating parameter data of each device in the source-grid-load-storage system to obtain a final set of device operating parameters; based on the final set of device operating parameters, using a unified modeling approach to mathematically abstract each device to obtain a unified device model; extracting heterogeneous characteristic parameters of the devices from the unified device model and classifying them to obtain device characteristic classification results; adjusting the energy storage capacity through a dynamic allocation mechanism based on the device characteristic classification results to obtain an energy storage capacity allocation scheme; using an optimized scheduling algorithm to perform real-time capacity allocation based on the energy storage capacity allocation scheme and real-time load change data, and obtaining a set of device operating instructions based on the real-time capacity allocation results; using a distributed control algorithm to coordinate the control of wind power, energy storage, and load devices based on the set of device operating instructions to obtain the coordinated operating status of the devices, and obtaining resource utilization efficiency evaluation results based on the coordinated operating status of the devices; and readjusting the parameters of the unified device model based on the resource utilization efficiency evaluation results to obtain an optimized capacity configuration scheme.

[0060] like Figure 1 As shown, the planning method in this embodiment specifically includes the following steps:

[0061] Step S1: Obtain the operating parameter data of each device in the source-grid-load-storage system, including wind power intermittent characteristic data, photovoltaic volatility data, energy storage rapid response capability data, and real-time load change data. Use a data acquisition system to extract the data from the device sensors and the grid monitoring platform to obtain the final set of device operating parameters.

[0062] Furthermore, wind power intermittent characteristic data, photovoltaic volatility data, and energy storage rapid response capability data are acquired from equipment sensors, and real-time load change data are extracted through the power grid monitoring platform to obtain a preliminary set of operating parameters. For this preliminary set of operating parameters, a weighted average method is used to integrate the wind power and energy storage-related data. This weighted average method involves summing the wind power intermittent characteristic data and the energy storage rapid response capability data according to preset weights to determine the final set of equipment operating parameters.

[0063] Furthermore, as a specific implementation of this embodiment, in the actual operation of a wind farm, acquiring data on the intermittent characteristics of wind power from equipment sensors typically involves wind speed sensors and power output sensors installed on the wind turbines. These sensors monitor wind speed fluctuations and changes in power generation in real time, thereby capturing the intermittent characteristics of wind power, such as a sudden increase in wind speed leading to a sharp rise in power or a decrease in output caused by a weakening of wind. This data acquisition process relies on high-precision sampling by the sensors, for example, acquiring wind speed data once per second and transmitting it to the central control system via a wireless transmission module. Simultaneously, data on the rapid response capability of energy storage comes from sensors in the battery energy storage system. These sensors record the battery's charge and discharge rates, response time, and capacity status, such as measuring the delay time from receiving a command to output power, typically in the millisecond range, to assess the energy storage system's ability to quickly compensate for grid fluctuations. Extracting real-time load change data through a grid monitoring platform refers to using a SCADA system or an advanced EMS platform to collect load curve data from substations and user terminals, such as monitoring sudden increases in electricity demand during peak periods in industrial areas or decreases in load during off-peak periods in residential areas. This data is then integrated to form a preliminary set of operating parameters.

[0064] Furthermore, as a specific implementation of this embodiment, assuming a wind-storage integrated microgrid scenario, sensors collect data on wind power intermittent characteristics (power fluctuation rate of 20%), energy storage response capability (response time of 0.5 seconds), and load change (peak increase of 15%). These form a preliminary parameter set. Next, applying a weighted average method, with weights of 0.7 for wind power and 0.3 for energy storage, the final comprehensive fluctuation index of the equipment operating parameter set can be summed to approximately 15.5%. This value guides the energy storage discharge strategy to avoid overload. Through this integration, the system can achieve a more stable power balance, automatically increasing the energy storage response weight during high wind speed periods to further optimize energy utilization efficiency.

[0065] Step S2: Based on the final set of equipment operating parameters, a unified modeling approach is used to mathematically abstract wind power, photovoltaic, energy storage, and load equipment. Based on the linear regression algorithm, an equipment operating characteristic model is established to obtain a unified equipment model.

[0066] Further, wind power intermittency and photovoltaic volatility data are obtained from the final set of equipment operating parameters. These data are then normalized using standardization processing, which involves calculating the mean and standard deviation of each data point, subtracting the mean, and dividing by the standard deviation to obtain a standardized parameter set. Based on this standardized parameter set, and combining the energy storage response speed and load change rate attributes, a linear regression algorithm is used to fit the equipment characteristic curve. The linear regression algorithm takes the attribute values ​​from the standardized parameter set as independent variables and the corresponding equipment output as dependent variables, with the output being the fitting coefficients, thus determining the characteristic fitting model. Based on this characteristic fitting model, the parameter unification and model abstraction layer attributes are integrated to construct a unified model library. Constructing a unified model library means storing the fitting coefficients and abstraction layer attributes in correspondence, resulting in a unified equipment model.

[0067] Furthermore, as a specific implementation of this embodiment, for the standardized parameter set, combined with the energy storage response speed and load change rate attributes, a linear regression algorithm is used to fit the device characteristic curve. In a smart grid demonstration project, the energy storage response speed attribute can be the charging and discharging response time of the battery system, such as a delay of 0.5 seconds from the issuance of an instruction to the actual output change, while the load change rate represents the fluctuation rate of user electricity demand, such as a load increase of 10 kilowatts per minute. The input of the linear regression algorithm is the attribute values ​​in the standardized parameter set as independent variables, such as standardized wind speed fluctuations and photovoltaic radiation changes as input variables, and the corresponding device output, such as power generation, as the dependent variable. By solving using the least squares method, the algorithm calculates the fitting coefficients, for example, coefficient 'a' represents the weight of wind power intermittency on power with a weight of 0.7, and coefficient 'b' represents the weight of photovoltaic volatility with a weight of 0.6, thereby outputting a characteristic curve model described by a linear equation.

[0068] Furthermore, as a specific implementation of this embodiment, when constructing a unified model library based on the aforementioned characteristic fitting model and integrating parameter unification and model abstraction layer attributes, it can be further extended in the aforementioned wind-solar hybrid power station. Specifically, parameter unification maps all standardized data to a unified numerical range, such as between 0 and 1, while model abstraction layer attributes include general interface definitions, such as abstracting "response delay" and "fluctuation amplitude" as high-level attributes. The specific method for constructing the unified model library is to store the fitting coefficients corresponding to these abstraction layer attributes. For example, associating the wind power fitting coefficient of 0.7 with the abstract attribute "intermittency index" and storing it in the database forms a reusable model set. In this way, the unified equipment model can support simulation of different scenarios.

[0069] Step S3: Extract heterogeneous characteristic parameters of equipment from the unified equipment model, including the fluctuation range of wind power output and energy storage response time. Use clustering algorithm to classify the heterogeneous characteristics of equipment and obtain the equipment characteristic classification results.

[0070] Furthermore, a heterogeneous parameter vector is constructed from the photovoltaic output variation and load demand elasticity obtained from the unified equipment model. This heterogeneous parameter vector, combined with the equipment compatibility index, is used to divide the characteristics into classification groups using a K-means clustering algorithm. The K-means clustering algorithm takes the normalized heterogeneous parameter vector combined with the equipment compatibility index as data points and outputs the characteristic classification group and cluster center coordinates. By randomly initializing the cluster center coordinates, each data point is assigned to the group containing the nearest cluster center coordinate. The cluster center coordinates are then updated to the mean of the data points within the group. This assignment and update process is repeated until the cluster center coordinates no longer change, thus determining the cluster center coordinates. Based on these cluster center coordinates, the set of operating parameters and model abstract attributes are integrated to construct an equipment grouping mapping. Constructing the equipment grouping mapping involves associating the cluster center coordinates with the set of operating parameters and integrating them into the model abstract attributes to obtain the equipment characteristic classification results.

[0071] Furthermore, as a specific implementation of this embodiment, in the process of dividing characteristic categories using the K-means clustering algorithm by combining heterogeneous parameter vectors with device compatibility indices, the device compatibility index can be understood as a quantitative indicator for evaluating the collaborative capabilities between different devices. For example, the compatibility index between photovoltaic and energy storage devices is 0.8, indicating that they have a high degree of matching in energy transmission. The K-means clustering algorithm is used to divide characteristic categories, with the input being these vectors plus the compatibility index as data points. For example, a data point might include a photovoltaic variation normalized value of 1, a load resilience normalized value of 0.5, and a compatibility index of 0.8. The algorithm randomly initializes the cluster center coordinates, such as initial centers being (0,0,0.5) and (1,1,0.9). Then, it assigns each data point to the group containing the nearest center, calculates the Euclidean distance, groups the data, updates the centers to the group mean, and repeats this process until the centers are stable. For example, after 5 iterations, the center coordinates are fixed at (0.2,0.3,0.7) and (0.9,0.8,0.85), thereby determining the characteristic classification groups such as "high volatility group" and "stable cooperative group".

[0072] Furthermore, as a specific implementation of this embodiment, when integrating the set of operating parameters and the model abstract attributes based on the cluster center coordinates, the set of operating parameters includes actual parameters such as wind speed data and battery capacity, while the model abstract attributes refer to the higher-level description of the equipment, such as the response time abstract layer. The cluster center coordinates are associated with these parameters, for example, associating the center (0.2, 0.3, 0.7) with the photovoltaic output variation parameter set, and incorporating abstract attributes such as "fast response model," thereby constructing an equipment grouping mapping. This yields equipment characteristic classification results, such as mapping photovoltaic equipment to a "flexible load group," facilitating subsequent system optimization.

[0073] Step S4: Determine whether the wind power intermittent characteristics are significant based on the equipment characteristic classification results. If the equipment characteristic classification results show that the wind power intermittent characteristics are significant, then adjust the energy storage fast response capacity through a dynamic allocation mechanism, and predict the wind power output fluctuation based on the time series analysis algorithm to obtain the energy storage capacity allocation scheme.

[0074] Furthermore, significant indicators of wind power intermittent characteristics and historical output data are obtained from the equipment characteristic classification. The historical output data is processed using the ARIMA time series analysis algorithm, where ARIMA is an autoregressive integral moving average model. The input is a sequence of historical output data, and the output is a prediction of future output fluctuations. The model is fitted and predicted values ​​are generated by calculating the difference order d, the autoregressive order p, and the moving average order q, resulting in the output fluctuation prediction result. Based on the output fluctuation prediction result, an adjustment value for the rapid response capacity of energy storage is calculated using a dynamic allocation mechanism. This dynamic allocation mechanism determines the adjustment value by multiplying the predicted fluctuation amplitude by a preset response coefficient, establishing a preliminary framework for the capacity allocation scheme. Based on this preliminary framework, fluctuation range assessment and energy balance optimization are integrated. Fluctuation range assessment calculates the maximum and minimum differences between the predicted results, while energy balance optimization adjusts the capacity to match the fluctuation differences, resulting in an energy storage capacity allocation scheme.

[0075] Furthermore, as a specific implementation of this embodiment, the process of processing historical power output data using the ARIMA time series analysis algorithm includes: the input historical power output data sequence may be the daily average power output over the past 30 days, such as 200MW on the first day, 180MW on the second day, and so on. To fit the model, the autoregression order p is first determined by observing the autocorrelation function and partial autocorrelation function. For example, if the partial autocorrelation function is significant at lag 1 and then decays rapidly, then p is set to 1. Next, the difference order d is calculated, and the sequence is made stationary by performing a first-order difference. If the original sequence fluctuates greatly, and becomes a stationary sequence after the first-order difference, then d = 1. Finally, the moving average order q is determined, and the value of q is set according to the decay of the autocorrelation function after lag q, such as q = 1. After the model is fitted, it is used to generate a prediction of future power output fluctuations, for example, predicting that the power output range for the next 24 hours will be between 150MW and 250MW.

[0076] Furthermore, as a specific implementation of this embodiment, based on the preliminary framework, fluctuation range assessment and energy balance optimization are further integrated. Fluctuation range assessment involves calculating the difference between the maximum and minimum values ​​of the predicted results. For example, if the predicted maximum output is 250MW and the minimum is 150MW, the range is 100MW. Energy balance optimization matches this difference by adjusting energy storage capacity. For instance, if the current energy storage capacity of 80MW is insufficient to cover a 100MW fluctuation, the optimization algorithm will suggest adding 20MW of capacity to ensure the system maintains energy supply and demand balance during fluctuations, ultimately resulting in a complete energy storage capacity allocation scheme.

[0077] Step S5: Based on the energy storage capacity allocation plan and real-time load change data, an optimized scheduling algorithm is used to perform real-time capacity allocation for the source-grid-load-storage system. In this embodiment, the output ratio of each device is calculated based on the particle swarm optimization algorithm to obtain the real-time capacity allocation result.

[0078] Furthermore, the current state indicators of the energy storage system are obtained from the energy storage capacity allocation and real-time load change data. These indicators are then processed using a particle swarm optimization algorithm (PSO), which is a swarm intelligence optimization method that iteratively updates particle positions and velocities to find the optimal solution. The input is the current state indicator, and the output is a preliminary output ratio framework. Based on this preliminary output ratio framework, a dynamic response mechanism is integrated to calculate the output adjustment values ​​for each device. This dynamic response mechanism obtains the adjustment values ​​by multiplying real-time load changes by a preset response coefficient, thus acquiring constraints for energy balance optimization. Energy balance optimization involves solving a linear constraint equation that minimizes the supply-demand difference to obtain the constraints and determine an intermediate scheme for real-time capacity allocation. Based on this intermediate scheme, potential deviations are predicted by combining fluctuation range assessment. Fluctuation range assessment calculates the difference between the maximum and minimum output values ​​in the intermediate scheme as the deviation. If the deviation exceeds a preset threshold, the output ratio is adjusted to obtain the corrected allocation parameters. The final configuration is generated by merging the ratio calculation results using the corrected allocation parameters. The ratio calculation results refer to the ratio obtained by dividing the corrected allocation parameters by the total capacity, thus obtaining the real-time capacity allocation results.

[0079] Furthermore, as a specific implementation of this embodiment, when calculating the output adjustment values ​​of each device using the integrated dynamic response mechanism within the initial output ratio framework, the dynamic response mechanism can be understood as a real-time adjustment strategy. Based on real-time load changes, such as a sudden increase of 100MW in demand, the adjustment value is obtained by multiplying by a preset response coefficient of 0.8, for example, an increase of 45MW in wind power, 35MW in photovoltaic power, and 20MW in energy storage discharge. Next, the constraints for energy balance optimization are obtained. This optimization is achieved by solving linear constraint equations, such as equations ensuring total output equals the load plus a minimum deviation, to obtain constraints such as energy storage discharge not exceeding the capacity limit, thereby determining an intermediate scheme for real-time capacity allocation. This intermediate scheme includes 550MW of wind power output, 335MW of photovoltaic power, and 180MWh of energy storage response, ensuring short-term system stability.

[0080] Based on the aforementioned intermediate scheme, and in the process of assessing and predicting potential deviations in the fluctuation range assessment, the fluctuation range assessment specifically refers to calculating the difference of 200MW between the maximum output value (e.g., wind power peak of 600MW) and the minimum value (e.g., 400MW) as the deviation. If this deviation exceeds a preset threshold of 150MW, the output ratio is adjusted, for example, by reducing the wind power ratio to 0.4 and increasing energy storage to 0.25, resulting in corrected allocation parameters such as a total adjusted wind power of 480MW, photovoltaic of 320MW, and energy storage of 200MWh.

[0081] Step S6: If the real-time capacity allocation result meets the preset threshold for stable operation of the power grid, the capacity allocation result is sent to each equipment controller through the power grid dispatching system to obtain a set of equipment operation instructions.

[0082] Furthermore, grid stability indicators are obtained from the real-time capacity allocation results. If the grid stability indicators meet a preset threshold, the equipment address information is integrated through the grid dispatch system. Integration refers to matching the equipment address information with the allocation results one by one to generate a unified parameter list, thus obtaining the deployment preparation parameters. These deployment preparation parameters are transmitted to each equipment controller using a communication protocol, where the communication protocol refers to standard data transmission rules. A response confirmation mechanism is then integrated to generate a set of equipment operation instructions, where the response confirmation mechanism refers to the process of verifying successful transmission based on equipment feedback signals.

[0083] Furthermore, as a specific implementation of this embodiment, the power grid dispatching system is an integrated control platform responsible for coordinating the operation of various components including power generation, grid, load, and storage. The integration process involves matching the IP addresses or physical identifiers of devices such as photovoltaic inverters and energy storage battery packs with allocation results. For example, the address of a photovoltaic device, 192.168.1.10, is associated with a power output ratio of 30%, generating a unified parameter list. This list includes device ID, allocated capacity, and priority, thereby obtaining the preparation parameters for issuing commands. This integration not only simplifies data management but also improves the accuracy of command issuance, achieving the goal of rapid response to load changes in business operations.

[0084] Transmitting the obtained preparation parameters to the controllers of each device using a communication protocol is a crucial step in ensuring the effective implementation of the commands. Communication protocols such as Modbus or IEC 61850 define data packet formats, verification mechanisms, and transmission rules. For example, under the Modbus protocol, parameters are sent in a frame structure, including start bits, address, function code, and data area. The system packages the preparation parameters and transmits them to the controller, such as the BMS controller of an energy storage system, via Ethernet or wireless link.

[0085] Step S7: Based on the set of equipment operation instructions, a distributed control algorithm is used to coordinate the control of wind power, energy storage and load equipment. Instructions are transmitted through a real-time communication protocol to obtain the coordinated operation status of the equipment.

[0086] Further, distributed coordination parameters are obtained from the set of equipment operation instructions, and a distributed control algorithm is used to initially adjust the wind power equipment. This distributed control algorithm is a consensus algorithm based on a multi-agent system. The algorithm takes distributed coordination parameters as input and outputs the adjusted equipment state, thus obtaining the wind power adjustment result. Based on the wind power adjustment result, energy storage balance attributes are integrated. Integration means matching the adjustment result with energy storage capacity data one by one, and transmitting relevant instructions to the energy storage equipment controller via a real-time communication protocol. This real-time communication protocol refers to standard rules supporting low-latency data exchange, resulting in an optimized energy storage state. Based on the optimized energy storage state and combined with load adjustment attributes, coordinated control is applied to the load equipment. A communication fusion mechanism is used to verify instruction execution. This communication fusion mechanism integrates multiple source signals to confirm the complete execution process, resulting in a load balance index. Operation monitoring data is extracted from the load balance index. If the operation monitoring data meets a preset threshold, status feedback information is integrated. Integration means associating and matching the feedback information with the monitoring data to obtain the equipment's coordinated operation state.

[0087] Furthermore, as a specific implementation of this embodiment, the input of the algorithm is distributed collaborative parameters such as the current power generation and demand deviation. The process involves iteratively calculating the consensus value among agents. First, each agent calculates its local deviation and then broadcasts it to neighboring agents through the communication network. This process is repeated until the outputs of all agents tend to be consistent. Finally, the adjusted equipment state is output, such as adjusting the blade angle to match the power target, thereby obtaining the wind power regulation result.

[0088] When integrating wind power regulation results with energy storage balance attributes, the integration process involves matching regulation results, such as excess power values, with energy storage capacity data one by one. For example, if wind power regulation generates 100MW of excess power, the available capacity of the battery pack is matched for storage planning. Commands are transmitted via a real-time communication protocol that supports standard rules for low-latency data exchange. This protocol uses time-sensitive networking to ensure data packets arrive within milliseconds. For instance, commands may include charging rate and duration. After being transmitted to the energy storage device controller, the controller adjusts the battery charging and discharging state based on these commands to achieve an optimized energy storage state, such as a stable battery SOC level of 80%.

[0089] When applying coordinated control to load devices by combining load adjustment attributes, the priority of loads, such as industrial electricity consumption, can be adjusted based on the energy storage optimization status, such as remaining capacity. For example, non-critical loads can be prioritized to balance the power grid. A communication fusion mechanism is used to verify command execution. This mechanism integrates multi-source signals, such as device response signals and sensor data, to confirm complete execution. The process includes sending commands and collecting feedback. If all signals match, the verification is successful, and load balance indicators such as voltage fluctuation rate below 5% are obtained.

[0090] Operational monitoring data, such as real-time power balance values, are extracted from load balance indicators. If a preset threshold is met, such as a deviation of less than 2%, the status feedback information is integrated, and the equipment temperature is correlated and matched with the monitoring data to obtain the equipment's collaborative operation status, ensuring the efficient operation of the entire power grid.

[0091] Step S8: Obtain resource utilization efficiency data from the equipment collaborative operation status, use data analysis algorithms to statistically analyze the efficiency data, and determine whether the system operating efficiency has reached the preset threshold based on the mean comparison method to obtain the efficiency evaluation result.

[0092] Furthermore, energy storage capacity data is obtained from the wind power output, and a data analysis algorithm is used to statistically analyze the energy storage capacity data. This data analysis algorithm refers to a calculation process based on descriptive statistics. A mean comparison method is used to determine whether the load demand has reached a preset threshold. The mean comparison method involves calculating the average value of the energy storage capacity data and comparing it with the preset threshold to obtain efficiency data. Based on this efficiency data, resource balance attributes are integrated, and relevant indicators are transmitted to a preset operation monitoring module via a real-time communication protocol. This real-time communication protocol refers to a standard protocol that supports data exchange, resulting in a resource utilization efficiency assessment.

[0093] Step S9: Judge the resource utilization efficiency assessment results. If the efficiency assessment results are lower than the preset threshold, then readjust the unified equipment model parameters based on real-time load changes and wind power intermittent characteristic data, update the capacity planning framework based on iterative optimization algorithm, and obtain the optimized capacity configuration scheme.

[0094] Fluctuation amplitude data is obtained from the real-time load change data. The mean of the fluctuation amplitude data is calculated using statistical methods, where the statistical method involves summing all fluctuation amplitude values ​​and dividing by the number of data points. Feature extraction is performed on the mean of the fluctuation amplitude data, which is then integrated with the wind power intermittency characteristics. Feature extraction involves multiplying the mean by the standard deviation of the intermittency characteristic data to quantify the degree of influence, resulting in an intermittency impact index. The equipment model parameters are adjusted based on the intermittency impact index. Adjustment involves updating parameter values ​​using a gradient descent algorithm to minimize error. The gradient descent algorithm takes the intermittency impact index and initial parameter values ​​as input and outputs iterated parameter values. The error is the squared difference between the intermittent impact index and the predicted parameter values. The convergence of the updated equipment model parameters is determined; if convergence is achieved, the adjusted parameter set is determined. The capacity planning framework is updated based on the adjusted parameter set. Updating involves iteratively calculating the adjusted parameter set against the framework's constraints. The capacity planning framework is a pre-defined linear programming model with constraints including upper and lower capacity limits. Iterative calculation involves repeatedly solving the linear programming model until the constraints are satisfied, resulting in a revised version of the framework. Extract the capacity allocation ratio from the revised version of the framework, where extraction refers to selecting the ratio value from the solution results of the revised version, and verifying it by integrating the efficiency evaluation results, where verification refers to comparing whether the deviation between the ratio value and the efficiency evaluation results is lower than a preset threshold, and obtaining the optimized capacity configuration scheme.

[0095] If the above methods are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0096] Example 2

[0097] This embodiment provides a source-grid-load-storage capacity dynamic planning system based on digital twins, including a data acquisition module 1, a unified modeling module 2, a characteristic classification module 3, a capacity allocation module 4, an optimization scheduling module 5, an instruction issuance module 6, a collaborative control module 7, an efficiency evaluation module 8, and an iterative optimization module 9, wherein:

[0098] Data acquisition module 1 is used to acquire the operating parameter data of each device in the source-grid-load-storage system to obtain the final set of device operating parameters;

[0099] Unified Modeling Module 2 is used to perform mathematical abstraction of each device based on the final set of device operating parameters, and obtain a unified device model.

[0100] Feature classification module 3 is used to extract heterogeneous feature parameters of equipment from the unified equipment model and classify them to obtain equipment feature classification results;

[0101] The capacity allocation module 4 is used to adjust the energy storage capacity through a dynamic allocation mechanism based on the equipment characteristic classification results, so as to obtain an energy storage capacity allocation scheme.

[0102] The optimized scheduling module 5 is used to perform real-time capacity allocation of the source-grid-load-storage system based on the energy storage capacity allocation plan and real-time load change data, and obtain the real-time capacity allocation result.

[0103] Command issuance module 6 is used to issue capacity allocation results to each device controller to obtain a set of device operation commands;

[0104] The collaborative control module 7 is used to obtain the collaborative operation status of wind power, energy storage and load equipment by using a distributed control algorithm based on the set of equipment operation instructions.

[0105] The efficiency evaluation module 8 is used to obtain resource utilization efficiency data based on the collaborative operation status of equipment, and to evaluate the resource utilization efficiency data to obtain the resource utilization efficiency evaluation result.

[0106] Iterative optimization module 9 is used to readjust the parameters of the unified equipment model based on the resource utilization efficiency assessment results to obtain an optimized capacity configuration scheme.

[0107] The rest is the same as in Example 1.

[0108] This embodiment also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method.

[0109] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0110] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A source network load storage capacity dynamic programming method based on digital twinning, characterized in that, The method comprises the following steps: obtaining operation parameter data of each device in a source grid load storage system to obtain a final device operation parameter set, wherein the operation parameter data comprises wind power intermittence data, photovoltaic fluctuation data, energy storage fast response capability data and real-time load change data; based on the final device operation parameter set, performing mathematical abstraction on each device by using a unified modeling method to obtain a unified device model; extracting device heterogeneous characteristic parameters from the unified device model and classifying to obtain a device characteristic classification result, adjusting energy storage capacity based on the device characteristic classification result by using a dynamic adjustment mechanism to obtain an energy storage capacity adjustment scheme; performing real-time capacity allocation according to the energy storage capacity adjustment scheme and the real-time load change data by using an optimal scheduling algorithm, and obtaining a device operation instruction set according to the real-time capacity allocation result; based on the device operation instruction set, performing collaborative control on wind power, energy storage and load devices by using a distributed control algorithm to obtain a device collaborative operation state, and obtaining a resource utilization efficiency evaluation result based on the device collaborative operation state; re-adjusting the unified device model parameters based on the resource utilization efficiency evaluation result to obtain an optimized capacity configuration scheme.

2. The digital-twin-based source-network-load-storage capacity dynamic planning method according to claim 1, characterized in that, The process of obtaining the final device operation parameter set comprises: constructing a preliminary operation parameter set based on the operation parameter data of each device in the source grid load storage system; integrating wind power and energy storage related data in the preliminary operation parameter set by using a weighted average method to obtain the final device operation parameter set.

3. The digital-twin-based source-network-load-storage capacity dynamic planning method according to claim 1, wherein, The process of obtaining the unified device model comprises: performing standardization processing on the wind power intermittence data and the photovoltaic fluctuation data in the final device operation parameter set to obtain a standardized parameter set; based on the standardized parameter set, combining energy storage response speed and load change rate attributes, fitting a device characteristic curve by using a linear regression algorithm to obtain a characteristic fitting model, wherein the characteristic fitting model takes the attribute values in the standardized parameter set as independent variables and takes the corresponding device output as dependent variables; based on the characteristic fitting model, integrating parameter unification and model abstraction layer attributes to construct a unified model library to obtain the unified device model, wherein the unified model library stores the corresponding association between fitting coefficients and abstraction layer attributes.

4. The digital-twin-based source-network-load-storage capacity dynamic planning method according to claim 1, characterized in that, The process of obtaining the device characteristic classification result comprises: combining device heterogeneous characteristic parameters with device compatibility indexes, and obtaining characteristic classification groups and clustering center coordinates by using a clustering algorithm; integrating the final device operation parameter set and model abstraction attributes according to the clustering center coordinates to construct a device grouping mapping to obtain the device characteristic classification result.

5. The digital-twin-based source-network-load-storage capacity dynamic planning method according to claim 1, characterized in that, The process of obtaining the energy storage capacity adjustment scheme comprises: determining whether the wind power intermittence characteristic is significant according to the device characteristic classification result, if so, predicting wind power output fluctuation by using a time series analysis algorithm, calculating an adjustment value of the energy storage fast response capacity by using a dynamic adjustment mechanism to obtain an energy storage capacity adjustment preliminary framework, wherein the dynamic adjustment mechanism refers to determining the adjustment value according to the predicted fluctuation amplitude multiplied by a preset response coefficient; integrating fluctuation range evaluation and energy balance optimization according to the energy storage capacity adjustment preliminary framework to obtain the energy storage capacity adjustment scheme.

6. The digital-twin-based source-network-load-storage capacity dynamic planning method according to claim 1, characterized in that, The real-time capacity allocation specifically comprises: An initial state index of the source-grid-load-storage system is obtained from the energy storage capacity allocation scheme and real-time load change data, and an optimization scheduling algorithm is used to obtain a preliminary output proportion framework based on the initial state index. For the preliminary output proportion framework, a dynamic response mechanism is integrated to calculate the output adjustment value of each device, obtain the constraint condition of energy balance optimization, and obtain an intermediate scheme of real-time capacity allocation. Based on the intermediate scheme, the output proportion is adjusted to obtain a corrected allocation parameter. Based on the corrected allocation parameter, a real-time capacity allocation result is obtained.

7. The digital-twin-based source-network-load-storage capacity dynamic planning method according to claim 1, characterized in that, The process of obtaining the device cooperative operation state includes: Distributed cooperative parameters are obtained from the device operation instruction set, and a distributed control algorithm is used to preliminarily adjust the wind power device to obtain a wind power adjustment result. For the wind power adjustment result, relevant instructions are transmitted to the energy storage device controller through a real-time communication protocol to obtain an energy storage optimization state. According to the energy storage optimization state, cooperative control is applied to the load device, and a communication fusion mechanism is used to verify the instruction execution to obtain a load balance index. From the load balance index, operation monitoring data is extracted, and if the operation monitoring data meets the preset operation threshold, state feedback information is integrated to obtain the device cooperative operation state.

8. The digital-twin-based source-network-load-storage capacity dynamic planning method according to claim 1, characterized in that, The adjustment of the unified device model parameter specifically includes: The fluctuation amplitude data is obtained from the real-time load change data, the mean value of the fluctuation amplitude data is calculated, the intermittent characteristics of wind power and the mean value of the fluctuation amplitude data are fused for feature extraction to obtain an intermittent influence index, and the unified device model parameter is adjusted based on the intermittent influence index using a gradient descent algorithm.

9. A source network load storage capacity dynamic planning system based on digital twinning, characterized in that, It includes: A data acquisition module is configured to obtain operation parameter data of each device in a source-grid-load-storage system to obtain a final device operation parameter set. The operation parameter data includes wind power intermittent characteristic data, photovoltaic fluctuation data, energy storage fast response capability data, and real-time load change data. A unified modeling module is configured to use a unified modeling method to mathematically abstract each device based on the final device operation parameter set to obtain a unified device model. A characteristic classification module is configured to extract device heterogeneous characteristic parameters from the unified device model and classify to obtain a device characteristic classification result. An energy storage capacity allocation module is configured to adjust the energy storage capacity based on the device characteristic classification result through a dynamic allocation mechanism to obtain an energy storage capacity allocation scheme. An optimization scheduling module is configured to use an optimization scheduling algorithm to perform real-time capacity allocation of the source-grid-load-storage system based on the energy storage capacity allocation scheme and real-time load change data to obtain a real-time capacity allocation result. An instruction issuing module is configured to issue the capacity allocation result to each device controller to obtain a device operation instruction set. A cooperative control module is configured to use a distributed control algorithm to cooperatively control wind power, energy storage, and load devices based on the device operation instruction set to obtain a device cooperative operation state. An efficiency evaluation module is configured to obtain resource utilization efficiency data based on the device cooperative operation state and evaluate the resource utilization efficiency data to obtain a resource utilization efficiency evaluation result. An iterative optimization module is configured to adjust the unified device model parameters based on the resource utilization efficiency evaluation result, and obtain an optimized capacity configuration scheme.

10. A computer apparatus comprising a memory, a processor, and a computer program stored on the memory, wherein the computer program, when executed by the processor, causes the processor to perform the method of any one of claims 1 to 9. The processor executes the computer program to implement the steps of the method of any one of claims 1-8.

Images