A new energy consumption system based on source network load cooperation

Abstract

The application relates to the technical field of power grid intelligent scheduling, and particularly discloses a new energy consumption system based on source-grid-load cooperation, which is used for solving the problem of how to dynamically adjust the self-regulation operation index and power scheduling strategy of a distribution network through a source-grid-load cooperation mechanism, ensuring stable operation of the power grid, realizing efficient consumption of new energy, and optimizing the utilization of energy storage and load resources, and comprises an energy storage module, a controllable load module, a distribution network self-regulation control module, a main power grid coordination module, a data acquisition and analysis module, a communication scheduling module and a cooperative optimization scheduling module; the new energy consumption system based on source-grid-load cooperation combines real-time data acquisition and analysis, uses a multi-time scale power disturbance model and an improved interval method to perform scenario analysis on wind power generation and photovoltaic power generation output, and realizes self-regulation operation of the distribution network and efficient consumption of new energy through cooperative optimization scheduling of the energy storage module, the controllable load module and the main power grid coordination module.

Description

Technical Field

[0001] This invention relates to the field of intelligent power grid dispatching technology, and in particular to a new energy consumption system based on source-grid-load coordination. Background Technology

[0002] Chinese invention patent CN112018798A discloses a multi-timescale autonomous operation method for distribution networks involving regional energy storage stations in disturbance mitigation. While existing methods for addressing power disturbances caused by renewable energy grid integration often rely on passive management, this patent does not address how to utilize energy storage stations to mitigate power disturbances and thus enable autonomous and disciplined operation of the distribution network. It proposes establishing regional autonomous operation indicators for distribution networks at short, medium, and long time scales, and using these indicators as criteria, proposes a short-timescale method for mitigating power disturbances from energy storage stations, as well as medium- and long-timescale autonomous operation methods for distribution networks based on energy storage, demand response, and network reconfiguration. Furthermore, it proposes a regional energy storage station correction method to improve the state of charge (SOC) of energy storage stations and rationally adjust their SOC. However, the above scheme does not consider the coordinated regulation of the power source, grid, and load. This results in the energy storage station's regulation process during multi-timescale autonomous operation failing to accurately represent the dynamic interaction between the power source, load, and grid. Deviations occur in the calculation of power disturbance amplitude and ramp rate. Particularly during the energy storage station's regulation process, it fails to adapt to the variable power fluctuations in the distribution network in real time. This leads to the inability to accurately match actual demand with power disturbance allocation at short, medium, and long time scales. The short-timescale power smoothing strategy of the energy storage station fails, resulting in its inability to respond promptly to sudden power disturbances and ineffective power disturbance smoothing. This causes a short-term increase in frequency fluctuations in the distribution network. Furthermore, at the medium time scale, the energy storage station cannot accurately... Controlling the power ramp rate prevents the automatic generation control system from responding to changes in power generation in a timely manner, resulting in insufficient power ramp and limiting the adjustment capability of the automatic generation control system. At the same time, the distribution of power disturbances on a long time scale is also unbalanced due to the failure to consider the power fluctuations caused by the coordination of source, grid and load. This causes the charging and discharging strategies of energy storage stations to be out of touch with actual demand, resulting in some areas experiencing overload or insufficient power supply during network reconfiguration. This affects the optimization of power flow in the grid, and the absorption of new energy cannot reach the optimal absorption level, reducing the grid operating efficiency and the utilization rate of new energy. The correction strategies of regional power stations are also affected and cannot charge and discharge under ideal charging conditions. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide a new energy consumption system based on source-grid-load coordination. This new energy consumption system, based on source-grid-load coordination, combines real-time data acquisition and analysis, uses a multi-time-scale power disturbance model and an improved interval method to conduct scenario analysis on wind power and photovoltaic power output, and achieves autonomous operation of the distribution network and efficient consumption of new energy through the coordinated optimization scheduling of energy storage modules, controllable load modules, and main grid coordination modules.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A renewable energy consumption system based on source-grid-load coordination includes an energy storage module, a controllable load module, a distribution network self-regulatory control module, a main grid coordination module, a data acquisition and analysis module, a communication and dispatch module, and a collaborative optimization dispatch module. The collaborative optimization dispatch module, through a source-grid-load coordination mechanism, establishes a multi-timescale power disturbance model based on real-time acquired renewable energy output data, load demand, and energy storage status information. It predicts the uncertainty of renewable energy output in real time and makes dynamic adjustments. An improved interval method is used to perform scenario analysis on the output of wind power and photovoltaic power generation. Combining the dispatch capabilities of energy storage and controllable loads, it optimizes and adjusts the energy storage charging and discharging strategies and load distribution. The load dispatching module collects power, energy storage status, load demand, and renewable energy output data from the distribution network. It uses a predictive model to analyze power disturbance trends and, in conjunction with renewable energy absorption quality indicators, dynamically adjusts the preset constraints of multi-scale autonomous operation indicators. The distribution network autonomous control module monitors the power and energy storage status at the distribution gateway in real time and, in conjunction with the power disturbance model, automatically adjusts the charging and discharging power of the energy storage modules and the dispatching of controllable loads. The main grid coordination module dynamically adjusts the network reconfiguration path and power exchange volume based on the power demand, load status, and power fluctuations of the distribution network. The communication dispatching module transmits information and issues dispatching commands.

[0005] As a further aspect of the present invention, in the collaborative optimization scheduling module, the process of performing scenario analysis on the wind power generation output using the improved interval method includes: Step S1, divide the wind power output scenarios: divide the wind power output into five scenarios, namely the prediction baseline scenario, the two uphill and downhill scenarios from odd to even time points, and the two uphill and downhill scenarios from even to odd time points. Step S2, Longitudinal Analysis: For the wind power output at each moment, based on the real-time collected wind speed factor, calculate and analyze the fluctuation of the wind power output at that moment, determine the magnitude and trend of the change, and determine the upper and lower bounds of the wind power output fluctuation. Step S3, Horizontal Analysis and Scenario Calculation: Perform a horizontal analysis on the fluctuation of wind power output between different time points, calculate the power fluctuation range of each ramping scenario, obtain the upper and lower bounds of the wind power output ramping scenario, and adjust them according to the actual wind power output fluctuation. Step S4, Scenario Supplementation and Adjustment: Based on the actual collected output power data of the wind turbine, if the ramping scenario cannot meet the power fluctuation requirements or the wind power output value does not reach the preset upper and lower limits, the scenario is identified as a missing scenario. The missing power is supplemented according to the existing wind power output prediction model and historical data. The supplemented power value is dynamically adjusted according to the ramping scenario of power fluctuation and the reserve capacity requirements to ensure that the power change meets the AGC regulation capability and reserve capacity requirements. Step S5, Scene Update and Optimization: Based on the supplemented power value and the adjusted ramp scenario, the prediction model for wind power output is updated in real time. Combining real-time data and power fluctuation trends, the subsequent power output is dynamically optimized.

[0006] As a further aspect of the present invention, in step S1 of the collaborative optimization scheduling module, the predicted baseline scenario is a scenario where the wind speed range is from the cut-in wind speed of the wind turbine to the rated wind speed of the wind turbine. The power output of the wind turbine is obtained by analyzing the wind speed at the current moment and the historical wind turbine output power data to obtain the wind turbine output power change curve that is positively correlated with the wind speed. The uphill and downhill scenarios at odd to even times include a first power uphill scenario where the wind speed increases from the cut-in wind speed to the rated wind speed and a second power uphill scenario where the wind speed decreases from the rated wind speed to the cut-in wind speed. The uphill and downhill scenarios at even to odd times include a third power uphill scenario where the wind speed increases from the cut-in wind speed to the rated wind speed and a fourth power uphill scenario where the wind speed decreases from the rated wind speed to the cut-in wind speed.

[0007] As a further aspect of the present invention, in the collaborative optimization scheduling module, the improved interval analysis method for scenario analysis of photovoltaic power generation output includes: Step A1, establish a photovoltaic power generation output model: by collecting real-time environmental data on light intensity and temperature, establish a nonlinear relationship model between photovoltaic power generation output and environmental factors; Step A2, divide the photovoltaic power output scenarios: Divide the collected light intensity into three ranges of low light, medium light and high light according to the preset range, and divide the temperature into low temperature range, medium temperature range and high temperature range according to the preset range. Combine the light intensity range and temperature range to divide the photovoltaic power output scenarios into 9 photovoltaic power output scenarios. Step A3: Predict photovoltaic power output based on a pre-trained LSTM model: By inputting features such as light intensity, temperature, and historical power data, the photovoltaic power output scenario is input into the pre-trained LSTM model for training, capturing the dependence and fluctuation patterns of time-series data, and outputting the predicted photovoltaic power value.

[0008] As a further aspect of this invention, in the collaborative optimization scheduling module, in step A3, the LSTM model dynamically adjusts the size of the input time window based on the changing trends of real-time light intensity, temperature, and historical power data to adapt to data features at different time scales. The input layer receives light intensity, temperature, and historical power data from multiple time series, extracts features through convolutional layers to obtain local features at different time scales, and fuses short-term and long-term features using time windows. The fused features are then input into the LSTM layer for time-series learning. The intervals of light intensity and temperature are encoded as numerical features and input into the LSTM network as additional input features. In the loss function, the gradient rate of change between adjacent time steps of the predicted power sequence at each time point is calculated, and the absolute values ​​of the gradient rates of change at each time point are summed as a penalty term, with an adjustment factor added. Adjusting the relative weights between the volatility stationarity loss term and the predicted loss term, the adjustment factor Preset according to the photovoltaic power output scenario category.

[0009] As a further aspect of the present invention, the process by which the data acquisition and analysis module dynamically adjusts the preset constraints of multi-scale self-regulating operation indicators includes: Step B1, Power Disturbance Trend Analysis: Using historically collected data on power distribution network, energy storage status, load demand, and new energy output, a time series prediction model is used to analyze the future power disturbance trend. Short-term, medium-term, and long-term fluctuations of power disturbance are modeled separately, and the characteristics of power disturbance fluctuations at different time scales are identified. Using the sliding window method, the mean, variance, maximum, and minimum values ​​of power disturbance are calculated based on historical power data and data at the current time step. Based on the prediction model, the trend of power disturbance changes within a preset time period is predicted using real-time data. Step B2, determine the quality indicators of new energy consumption: simulate different wind power and photovoltaic consumption capacity ratios using time-series simulation method, and calculate the wind power and photovoltaic consumption capacity ratios respectively. New energy consumption quality indicators when equal to 0, 1, 1 / 2, 3 / 4, and 1 / 4 , , , , The quality indicators for new energy consumption are one of the following: new energy utilization rate, new energy self-consumption rate, annual fossil energy substitution ratio, and voltage adaptability. Then, based on the improved binary search method, the new energy consumption capacity within the preset threshold range and the optimal consumption capacity of each node are obtained. Step B3: Dynamically adjust the preset constraints of the self-regulatory operation indicators: Based on the predicted short-timescale power disturbance trend in Step B1, dynamically adjust the upper and lower limits of the short-timescale power disturbance amplitude; based on the new energy consumption quality indicators and the mid-timescale trend analysis of power disturbance obtained in Step B2, dynamically adjust the mid-timescale power disturbance amplitude and ramp rate; based on the long-timescale power disturbance trend in Step B1, combined with the new energy consumption quality indicators, dynamically adjust the long-timescale power fluctuation amplitude.

[0010] As a further aspect of the present invention, in step B3 of the data acquisition and analysis module dynamically adjusting the preset constraints of multi-scale autonomous operation indicators, the adjustment of short-time-scale autonomous operation indicators specifically includes: when the predicted power disturbance amplitude exceeds ±10%, automatically reducing the charging and discharging power of the energy storage module; when the energy storage module is in a charging state, reducing the charging power by 20%; when the energy storage module is in a discharging state, reducing the discharging power by 15%; when the power disturbance amplitude is less than ±5% and the new energy consumption quality index is above 95%, relaxing the short-time-scale power disturbance amplitude limit, allowing power fluctuations within the range of ±8%, and restoring the charging and discharging power of the energy storage module to the original set value and adaptively adjusting within the range of ±10% of the original set value.

[0011] As a further aspect of the present invention, in step B3 of the data acquisition and analysis module dynamically adjusting the preset constraints of multi-scale autonomous operation indicators, the adjustment of the autonomous operation indicators in the medium time scale specifically includes: when the fluctuation amplitude of new energy power exceeds ±15% and the new energy absorption quality index is lower than 80%, automatically optimizing energy storage scheduling and reducing the charging and discharging power of the energy storage module; when the energy storage module is in the charging state, reducing the charging power by 25%; when the energy storage module is in the discharging state, reducing the discharging power by 20%; when the new energy absorption quality index is improved to 85% or above, relaxing the power disturbance amplitude and ramp rate restrictions in the medium time scale, allowing power fluctuations to be within the range of ±12% to ±15%, and restoring the charging and discharging power of the energy storage module to the original set value, but not exceeding the set ±15% range for adjustment.

[0012] As a further aspect of the present invention, in step B3 of dynamically adjusting the preset constraints of the multi-scale self-regulating operation indicators in the data acquisition and analysis module, the adjustment of the long-term scale self-regulating operation indicators specifically includes: when the power fluctuation amplitude exceeds ±20% and the new energy consumption quality index is lower than 75%, automatically increasing the storage capacity of the energy storage module and reducing the charging and discharging power of the energy storage module; when the energy storage module is in the charging state, reducing the charging power by 30%; when the energy storage module is in the discharging state, reducing the discharging power by 25%; when the new energy consumption quality index remains above 80%, appropriately relaxing the limitation on the power fluctuation amplitude, allowing the power fluctuation range to reach ±18%, and restoring the charging and discharging power of the energy storage module to the original set value, but not exceeding the set ±18% range for adjustment.

[0013] As a further aspect of the present invention, the main grid coordination module dynamically adjusts the power exchange between upper and lower level grids by acquiring real-time power demand, load status, and power fluctuation data of the distribution network, combined with power, energy storage status, load demand, and new energy output data acquired by the data acquisition and analysis module. Specifically, the main grid coordination module analyzes future power disturbance trends based on a time-series prediction model according to real-time power fluctuations and load changes, and dynamically adjusts the power exchange. When the power disturbance amplitude exceeds ±10%, the main grid coordination module adjusts the main grid power output to 110%–115% of the distribution network demand, keeping power fluctuations within ±8%. When the value is below ±5% and the renewable energy absorption quality index is above 95%, the main grid coordination module will restore the power exchange to the original set value and allow power fluctuations within ±8%. When the power fluctuation is between ±5% and ±10%, the main grid coordination module will adjust the power exchange to 105% to 110% of the distribution network demand to ensure that the power fluctuation is between ±5% and ±8%. When the renewable energy absorption quality index is below 85%, the main grid coordination module will increase the power exchange to 115% of the distribution network demand to optimize the grid load distribution. The main grid coordination module, through a real-time feedback mechanism, cooperates with the energy storage module and the load regulation module to adjust the energy storage charging and discharging power.

[0014] The technical effects of this invention are as follows: This invention, based on a source-grid-load coordination mechanism, combines real-time data acquisition, time-series prediction models, and multi-time-scale power disturbance analysis to achieve dynamic optimization of grid power dispatch and improve the capacity for renewable energy absorption. Through a collaborative optimization dispatch module, combined with real-time collected wind power generation, photovoltaic power generation data, energy storage status, and load demand, it can predict power disturbance trends in real time and adjust the charging and discharging strategies of energy storage modules and controllable load dispatch when power fluctuations are large, ensuring the stability and efficient operation of the grid. Simultaneously, through an improved interval analysis method, scenario analysis of wind power generation and photovoltaic power generation output is performed to optimize power dispatch under different power scenarios, reducing the impact of power fluctuations on the grid. When the quality of renewable energy absorption is poor, the main grid coordination module dynamically adjusts the power exchange and network reconfiguration path to optimize the power flow between upstream and downstream grids and maximize renewable energy absorption. The data acquisition and analysis module dynamically adjusts the preset constraints of multi-scale autonomous operation indicators based on the power disturbance change trend and renewable energy absorption quality indicators to ensure that the power exchange and power fluctuations between the distribution network and the main grid are within a reasonable range. This invention helps to improve the grid's regulation capability and adaptability in the face of complex power fluctuations and uncertainties in renewable energy output, optimize the utilization efficiency of grid resources, reduce the impact of power fluctuations on grid stability, ensure the maximum absorption of renewable energy, and improve the flexibility and stability of grid operation. Attached Figure Description

[0015] Figure 1 This is a diagram of a new energy consumption system based on source-grid-load coordination according to the present invention; Figure 2 This is a schematic diagram of the power distribution network structure tested in this invention. Detailed Implementation

[0016] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described content is only a part of the embodiments of this invention, and not all of them. Based on the content of this invention, all other technical solutions obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0017] like Figure 1As shown, this invention proposes a renewable energy consumption system based on source-grid-load coordination, comprising an energy storage module, a controllable load module, a distribution network self-regulatory control module, a main grid coordination module, a data acquisition and analysis module, a communication and dispatch module, and a collaborative optimization dispatch module. The collaborative optimization dispatch module, through a source-grid-load coordination mechanism, establishes a multi-timescale power disturbance model based on real-time collected renewable energy output data, load demand, and energy storage status information. It predicts the uncertainty of renewable energy output in real time and makes dynamic adjustments. An improved interval method is used to perform scenario analysis on the output of wind power and photovoltaic power generation. Combined with the dispatch capabilities of energy storage and controllable loads, it optimizes and adjusts the charging and discharging of energy storage. The power strategy and load dispatching module collects power, energy storage status, load demand, and renewable energy output data from the distribution network. It uses a predictive model to analyze power disturbance trends and, in conjunction with renewable energy absorption quality indicators, dynamically adjusts the preset constraints of multi-scale autonomous operation indicators. The distribution network autonomous control module monitors the power and energy storage status at the distribution network gateway in real time and, in conjunction with the power disturbance model, automatically adjusts the charging and discharging power of the energy storage module and the dispatching of controllable loads. The main grid coordination module dynamically adjusts the network reconfiguration path and power exchange volume based on the power demand, load status, and power fluctuations of the distribution network. The communication dispatching module transmits information and issues dispatching commands.

[0018] like Figure 2 The diagram shown is a schematic of the distribution network structure for testing according to the present invention. To clearly illustrate the technical solution of the present invention, the technical effects brought about by the adoption of the technical solution of the present invention in the following test distribution network are explained.

[0019] The test distribution network contains 23 nodes, with node 1 equipped with an energy storage module having a maximum power of 5. The maximum capacity is 21 Adjustable loads are installed at nodes 3, 9, 11, and 14, with an adjustable capacity limit of 4.2. The time-adjustable load is located at nodes 6, 9, 17, and 19. Agreements are pre-signed between operators and users, allowing for the transfer of adjustable load during the controlled time period from 12:00 to 14:00 daily, with a transfer limit of 2.5. , This is a transferable area, connected to the test distribution network via a normally closed switch. As a transferable zone, it is connected to the test distribution network via a normally open switch. - To reconfigure the switches, the average annual peak load for the region where this distribution network is located is set at 29.6. The system has various load types, including flexible and adjustable loads, and abundant solar energy resources. It is equipped with photovoltaic (PV) power generation equipment, with PV installation nodes set at 4, 7, 8, 12, 16, and 18. Solar illumination and load data are based on local annual average statistics. In the distribution network self-regulatory control module, based on the upstream grid demand and the distribution network's own operating conditions, the power output is set to [-15, 10]. The allowable power variation value is [-6, 6]. The standby capacity adjustment margin is 2%, the initial energy storage module in the controllable load module is set to 30%, and the network reconfiguration switch is enabled. - It can switch up to 5 times a day, with a renewable energy utilization rate of 95%, and the maximum connectable capacity of photovoltaic installation nodes is 10.5. With voltage constraints of 95% to 105% and line current carrying capacity constraints of 10% to 20%, the following is the implementation process of the technical solution of this invention: (1) Data acquisition and real-time monitoring: The power, energy storage status, load demand and new energy output data of the distribution network are monitored in real time through the data acquisition and analysis module. The system provides real-time feedback on the status of each node through the acquired data and inputs the real-time data into the collaborative optimization scheduling module for the next step of power disturbance trend analysis and adjustment decision. (2) Power disturbance trend analysis: Using a time series prediction model, combined with real-time collected wind power generation, photovoltaic output data and load demand, the power disturbance change trend is predicted in the short, medium and long time scales. The data acquisition and analysis module analyzes the mean, variance, maximum and minimum of the power disturbance and provides early warning of the power fluctuation trend, providing a basis for adjustment. (3) Wind power generation and photovoltaic output scenario analysis and energy storage scheduling: The collaborative optimization scheduling module performs scenario analysis on the output of wind power generation and photovoltaic power generation based on the improved interval method. It uses real-time collected data such as light intensity, temperature, and wind speed to divide the new energy output into scenarios, adjust the power fluctuation to the optimal range, and automatically optimize the energy storage charging and discharging and load scheduling in combination with the charging and discharging strategy of the energy storage module to ensure grid stability and maximize the absorption of new energy. (4) Power fluctuation adjustment and multi-scale autonomous operation: Based on the power disturbance trend analysis and the quality indicators of new energy consumption, the data acquisition and analysis module dynamically adjusts the preset constraints of the multi-scale autonomous operation indicators. When the predicted power fluctuation exceeds the preset range, the charging and discharging power of the energy storage module, the scheduling strategy of the controllable load module, and the power exchange amount are automatically adjusted. The adjustment range is dynamically corrected according to the power disturbance amplitude and real-time feedback of the system. (5) Main grid coordination and power exchange adjustment: The main grid coordination module dynamically adjusts the power exchange between upstream and downstream grids based on real-time data and power disturbance models, according to the power demand, load conditions, and power fluctuations of the distribution network. When power fluctuations are large, the main grid coordination module adjusts the network reconfiguration path ( - (Switching) and power exchange quantity, optimize the operation of the power grid, and ensure the effective allocation of power resources. Adjust the power flow path according to changes in load demand to achieve stable operation of the power grid; (6) Network reconfiguration and load dispatch optimization: The main power grid coordination module controls the network reconfiguration switch ( - Necessary network adjustments will be made to achieve reasonable power load distribution under conditions of large power fluctuations. The network reconfiguration switch will reconfigure the grid topology according to power flow requirements to ensure that the grid load distribution and energy storage dispatch can cooperate with each other and improve the grid load carrying capacity.

[0020] (7) System feedback and optimization adjustment: The collaborative optimization scheduling module updates the system's scheduling strategy in real time based on the adjustment results of the main power grid coordination module and the actual operating data of the distribution network self-regulatory control module. Through the feedback mechanism, the scheduling decision is continuously optimized to ensure the power flow balance of each node and further improve the renewable energy absorption rate and the operating efficiency of the distribution network.

[0021] It should be noted that in the collaborative optimization scheduling module, the process of performing scenario analysis on wind power output using the improved interval method includes: Step S1, divide the wind power output scenarios: divide the wind power output into five scenarios, namely the prediction baseline scenario, the two uphill and downhill scenarios from odd to even time points, and the two uphill and downhill scenarios from even to odd time points. Step S2, Longitudinal Analysis: For the wind power output at each moment, based on the real-time collected wind speed factor, calculate and analyze the fluctuation of the wind power output at that moment, determine the magnitude and trend of the change, and determine the upper and lower bounds of the wind power output fluctuation. Step S3, Horizontal Analysis and Scenario Calculation: Perform a horizontal analysis on the fluctuation of wind power output between different time points, calculate the power fluctuation range of each ramping scenario, obtain the upper and lower bounds of the wind power output ramping scenario, and adjust them according to the actual wind power output fluctuation. Step S4, Scenario Supplementation and Adjustment: Based on the actual collected output power data of the wind turbine, if the ramping scenario cannot meet the power fluctuation requirements or the wind power output value does not reach the preset upper and lower limits, the scenario is identified as a missing scenario. The missing power is supplemented according to the existing wind power output prediction model and historical data. The supplemented power value is dynamically adjusted according to the ramping scenario of power fluctuation and the reserve capacity requirements to ensure that the power change meets the AGC regulation capability and reserve capacity requirements. Step S5, Scene Update and Optimization: Based on the supplemented power value and the adjusted ramp scenario, the prediction model for wind power output is updated in real time. Combining real-time data and power fluctuation trends, the subsequent power output is dynamically optimized.

[0022] The following implementation method is described in the test of the distribution network: Step S1, Defining Wind Power Output Scenarios: In the test distribution network, wind power generation equipment was installed at nodes 4, 7, 8, 12, 16, and 18, and the real-time wind speed data and historical power generation data are as follows: (1) Predicted baseline scenario: Assume that the current wind speed is within the range of the wind speed at which the wind turbine generator cuts in and the rated wind speed. The predicted baseline scenario is the ideal power output curve of the wind turbine generator at this moment, and the corresponding power output curve is the ideal power change curve that is positively correlated with the wind speed. (2) Two uphill and downhill scenarios from odd to even time: Suppose that from time 1 to time 2, the wind speed increases from the cut-in wind speed to the rated wind speed, then the power output change in this segment is the first uphill scenario; from time 3 to time 4, the wind speed decreases from the rated wind speed to the cut-in wind speed, then it is the second uphill scenario. (3) Two uphill and downhill scenarios from even to odd times: For example, from time 5 to time 6, the wind speed increases, which leads to an increase in power fluctuation, forming the third uphill scenario; from time 7 to time 8, the wind speed decreases, and the power fluctuation decreases, forming the fourth uphill scenario. These scenarios will be divided based on wind speed data and historical power curves, forming a preliminary scenario division framework; Step S2, Longitudinal Analysis: For the wind power output at each time point, the real-time collected wind speed data is input into the analysis model. The wind speed at time 1 is 6 m / s, while the wind speed at time 2 is 10 m / s. The trend of wind power output can be analyzed in the following way: Calculate the fluctuation range: By comparing the changes in wind power output at different wind speeds, the power fluctuation range is analyzed. When the wind speed increases from 6 m / s to 10 m / s, the power fluctuation range is 40%. Determine the trend of change: If the fluctuation range and trend of power exceed the preset range, the system will mark the wind power output power fluctuation during this period as abnormal and make further adjustments. Step S3, Lateral Analysis and Scenario Calculation: By laterally analyzing the power output fluctuations of wind power generation at different time points, the power fluctuation range for different ramping scenarios is obtained. The wind speed increase from time 1 to time 2 corresponds to the first uphill scenario, with power fluctuations ranging from 20% to 50%. The wind speed decrease from time 3 to time 4 corresponds to the second climbing scenario, with power fluctuations ranging from 50% to 30%. During this process, through horizontal analysis and comparison, it is confirmed whether the power fluctuation range meets the design requirements of the ramping scenario. If the actual power fluctuation exceeds these ranges, the system will automatically adjust the scenario parameters. Step S4, Scenario Supplementation and Adjustment: Assume that the actual collected power output data shows insufficient wind power generation at time 4, failing to meet the preset fluctuation requirements, and the power output value is below the predetermined upper and lower boundaries. In this case, the system will: Identifying missing scenarios: By comparing historical data and using predictive models, the system can determine that the power fluctuation at the current moment is missing, thus identifying it as a missing scenario. Supplementing missing power: Based on historical data and wind power output prediction models, the missing power is supplemented. In this example, the supplemented power is 3MW to meet the requirements of reserve capacity and power fluctuations. Step S5: Scene Update and Optimization: The supplemented power value and the adjusted ramp scenario will be input into the wind power output prediction model. The system will dynamically optimize the subsequent power output based on the real-time collected wind speed data, historical power data and power fluctuation trends. The updated model predicts that the wind power output will increase by 10% between time 5 and time 6. Therefore, the system will automatically adjust the power output of the wind turbine to adapt to this fluctuation, thereby ensuring the stable operation of the power grid.

[0023] It should be specifically noted that in step S1 of the collaborative optimization scheduling module, the predicted baseline scenario is a scenario where the wind speed range is from the cut-in wind speed of the wind turbine to the rated wind speed of the wind turbine. The power output of the wind turbine is obtained by analyzing the wind speed at the current moment and the historical wind turbine output power data to obtain the wind turbine output power change curve that is positively correlated with the wind speed. The uphill and downhill scenarios at odd to even times include the first power uphill scenario where the wind speed increases from the cut-in wind speed to the rated wind speed and the second power uphill scenario where the wind speed decreases from the rated wind speed to the cut-in wind speed. The uphill and downhill scenarios at even to odd times include the third power uphill scenario where the wind speed increases from the cut-in wind speed to the rated wind speed and the fourth power uphill scenario where the wind speed decreases from the rated wind speed to the cut-in wind speed.

[0024] Based on the scenario division in step S1, wind power output is divided into five scenarios: Predicted baseline scenario: The current wind speed is 10 m / s, which is within the rated wind speed range of the wind turbine generator. Therefore, the corresponding wind power output is the power value of the predicted baseline scenario, and the power output of the wind turbine generator is 4.8 MW (the predicted value obtained based on the positive correlation curve between wind speed and power output). Scenarios of going up and down slopes from odd to even numbers: First power ramp-up scenario: From odd to even time, the wind speed increases from the cut-in wind speed (3m / s) to the rated wind speed (12m / s). At this time, the power output of the wind turbine increases from 0MW to the maximum power of 5MW. The power change shows certain ramp-up characteristics. In this ramp-up scenario, the power output needs to be adjusted according to the proportion of wind speed change. Second power ramping scenario: From an even-numbered moment to the next odd-numbered moment, the wind speed decreases from the rated wind speed (12m / s) to the cut-in wind speed (3m / s). At this time, the power output of the wind turbine generator drops from the maximum value (5MW) to 0MW, showing a downward ramping characteristic. The scenario of climbing uphill from even to odd time points: The third power ramp-up scenario: from even-numbered hours to odd-numbered hours, the wind speed increases from the cut-in wind speed (3m / s) to the rated wind speed (12m / s). At this time, the power of the wind turbine increases from 0MW to the maximum power value. The scenario is similar to the first power ramp-up scenario, but due to the different timing, it involves different wind speed data and power output relationships. Fourth power ramp-up scenario: From an odd-numbered hour to the next even-numbered hour, the wind speed decreases from the rated wind speed (12m / s) to the cut-in wind speed (3m / s). The wind turbine power decreases from the maximum value (5MW) to 0MW, similar to the second power ramp-up scenario. The adjustment method is dynamically adjusted based on the actual wind speed and power output data.

[0025] It should be noted that in the collaborative optimization scheduling module, the improved interval analysis method for scenario analysis of photovoltaic power generation output includes: Step A1, establish a photovoltaic power generation output model: by collecting real-time environmental data on light intensity and temperature, establish a nonlinear relationship model between photovoltaic power generation output and environmental factors; Step A2, divide the photovoltaic power output scenarios: Divide the collected light intensity into three ranges of low light, medium light and high light according to the preset range, and divide the temperature into low temperature range, medium temperature range and high temperature range according to the preset range. Combine the light intensity range and temperature range to divide the photovoltaic power output scenarios into 9 photovoltaic power output scenarios. Step A3: Predict photovoltaic power output based on a pre-trained LSTM model: By inputting features such as light intensity, temperature, and historical power data, the photovoltaic power output scenario is input into the pre-trained LSTM model for training, capturing the dependence and fluctuation patterns of time-series data, and outputting the predicted photovoltaic power value.

[0026] In the aforementioned test distribution network, based on the scenario analysis of photovoltaic power generation output, the following specific implementation examples illustrate how to analyze photovoltaic power generation output using the improved interval analysis method: Step A1, establish a photovoltaic power generation output model: (1) Acquisition of light intensity and temperature data: The photovoltaic power generation equipment in the distribution network is located at nodes 4, 7, 8, 12, 16 and 18. The real-time light intensity and temperature data are: light intensity (unit: W / m²) and temperature (unit: °C). At node 4, the light intensity is 600 W / m² and the temperature is 28 °C. (2) Establishment of nonlinear relationship model: Using the collected data, a nonlinear relationship model between photovoltaic power generation output and light intensity and temperature is established through regression analysis or machine learning methods. Multinomial regression, support vector machine regression (SVR) and other methods are used to establish the mathematical relationship between photovoltaic power generation output and environmental factors. Step A2: Determine photovoltaic power output scenarios: (1) Divide the light intensity range: Low light range: light intensity between 0-500W / m²; Medium illumination range: Illuminance between 500 and 800 W / m²; High illumination range: Illuminance between 800 and 1200 W / m² (2) Divide the temperature range: Low temperature range: Temperature between 0-15°C; Medium temperature range: temperatures between 15-30°C; High temperature range: temperatures between 30-45°C; (3) Combined light intensity and temperature range: By combining light intensity and temperature range, nine photovoltaic power output scenarios are obtained: low light intensity + low temperature; low light intensity + medium temperature; low light intensity + high temperature; medium light intensity + low temperature; medium light intensity + medium temperature; medium light intensity + high temperature; high light intensity + low temperature; high light intensity + medium temperature; high light intensity + high temperature. These combinations form nine photovoltaic power output scenarios, covering photovoltaic power generation output under different light and temperature conditions; Step A3: Predict photovoltaic power output based on a pre-trained LSTM model: LSTM model training: Based on historical data on light intensity, temperature, and photovoltaic power generation, an LSTM (Long Short-Term Memory) network model is pre-trained. This model is able to learn the temporal relationship between light intensity, temperature, and power. At any moment With a light intensity of 650W / m², a temperature of 22°C, and historical power data of 3.8MW, the LSTM model will predict the photovoltaic power output value at the current moment and predict the power at future moments based on the previous time series change patterns. Scene Input and Output: Real-time collected light intensity, temperature data, and historical power data are used as input. After training an LSTM model, the predicted photovoltaic power output value for the next time step is obtained. The predicted photovoltaic power output is 4.2 MW; Prediction updates: Over time, the LSTM model continuously receives new light intensity and temperature data, automatically updating the prediction results to adjust the photovoltaic power generation output scheduling strategy in real time.

[0027] In the actual operation of the distribution network, combined with the technical means of the above steps, when the real-time collected irradiance is 800W / m² and the temperature is 25°C, the system will input these data into the LSTM model. The model will predict the corresponding photovoltaic power output. If the prediction result is 4.0MW and the current load demand is 3.5MW, the system will dynamically adjust the output of photovoltaic power generation according to the prediction result to adapt to the current load demand and power disturbance requirements.

[0028] It should be specifically noted that in the collaborative optimization scheduling module, in step A3, the LSTM model dynamically adjusts the size of the input time window based on the changing trends of real-time light intensity, temperature, and historical power data to adapt to data features at different time scales. The input layer receives light intensity, temperature, and historical power data from multiple time series, extracts features through convolutional layers to obtain local features at different time scales, and fuses short-term and long-term features using time windows. The fused features are then input into the LSTM layer for time-series learning. The intervals of light intensity and temperature are encoded as numerical features and added as additional input features to the LSTM network. In the loss function, the gradient rate of change between adjacent time steps of the predicted power sequence at each time point is calculated, and the absolute values ​​of the gradient rates of change at each time point are summed as a penalty term, with an adjustment factor added. Adjusting the relative weights between the volatility stationarity loss term and the predicted loss term, the adjustment factor Preset according to the photovoltaic power output scenario category.

[0029] The test distribution network includes multiple photovoltaic installation nodes (nodes 4, 7, 8, 12, 16, and 18), which are equipped with photovoltaic power generation equipment. The irradiance and temperature are obtained through real-time sensor data. The real-time power demand and photovoltaic power output data of the distribution network are collected by the data acquisition and analysis module and provided to the collaborative optimization scheduling module for scheduling decisions.

[0030] Step A3: Predict photovoltaic power output based on a pre-trained LSTM model: (1) Input and feature extraction of the LSTM model: (11) Input data preparation: Illumination intensity: Real-time data collected, the illuminance of node 4 is 800W / m²; Temperature: Real-time temperature data; the temperature at node 4 is 25°C. Historical power data: Historical photovoltaic power output data, the photovoltaic power in the past hour was 4.1MW, 4.0MW, 3.9MW, and 4.2MW respectively; (12) Dynamically adjust the size of the input time window: Short-term characteristics: Assuming the current time window is the data within the past 30 minutes, the LSTM model receives the light intensity, temperature and historical power data within the last 30 minutes to capture short-term trends. Long-term features: If a longer time scale is considered (data from the past hour), LSTM will adjust the size of the input layer according to the data within this time window to adapt to the data features of a longer time scale; (3) Feature extraction from convolutional layers: Local feature extraction: Convolutional layers are used to extract local features from the input data. The trend of light intensity change over the past 30 minutes is captured through convolution operations. At the same time, temperature and power output data are also processed by convolutional layers to extract useful local features (such as rate of change and volatility). (4) Feature fusion and LSTM layer processing: (41) Integration of short-term and long-term characteristics: Feature fusion: By using a sliding time window, features at different time scales (short-term and long-term features) are fused. Short-term features of 30 minutes and long-term features of 1 hour are fused by weighting to obtain the final fused feature vector.

[0031] (42) Input to LSTM layer: The fused feature vector is input to the LSTM layer. During the time-series learning process, the LSTM model learns the future photovoltaic power output based on the time-series features of historical data (such as power change trends, wind speed changes, etc.). (43) Encoding the range information of illumination and temperature: Interval Encoding: To use the interval information of light intensity and temperature as input features, the system divides light intensity and temperature into different intervals (low light, medium light, high light; low temperature, medium temperature, high temperature) and encodes each interval with a numerical value (low light is 1, medium light is 2, and high light is 3). This encoded interval information is then fed into the LSTM network as additional input features. (44) Loss function and optimization: Loss function design: Gradient change rate calculation: In order to ensure that the predicted power change is smooth, the loss function will include the calculation of the gradient change rate between adjacent time steps. If the predicted power output changes significantly between two adjacent time steps, the system will impose a penalty.

[0032] Summation of absolute values: For each time point, LSTM calculates the gradient rate of change of the predicted power sequence at adjacent time steps and sums the absolute values ​​of these rates of change. This summation is used as a term in the loss function as a penalty term to adjust the stability of fluctuations. Introduction of regulatory factors: Regulatory factors To balance the weights of the fluctuation stationarity loss term and the prediction error loss term, the system will introduce an adjustment factor. Adjustment factor according to photovoltaic power output scenario (e.g., low light, high light, etc.) It can adaptively adjust, prioritizing prediction error in low-light scenarios and power fluctuation stability in high-light scenarios. (5) Update and optimize prediction results: Optimization and Updates: Prediction Results: Based on the above feature fusion and LSTM learning, the system can predict the photovoltaic power output at future times. The predicted photovoltaic power is 4.1MW; Result optimization: Based on the prediction results, the collaborative optimization scheduling module will adjust the energy storage charging and discharging strategy and load scheduling. The predicted photovoltaic power is 4.1MW, while the current load demand is 3.5MW. Therefore, the system increases the charging capacity of the energy storage module to balance the power demand. It should be noted that the process by which the data acquisition and analysis module dynamically adjusts the preset constraints of multi-scale self-regulatory operation indicators includes: Step B1, Power Disturbance Trend Analysis: Using historically collected data on power distribution network, energy storage status, load demand, and new energy output, a time series prediction model is used to analyze the future power disturbance trend. Short-term, medium-term, and long-term fluctuations of power disturbance are modeled separately, and the characteristics of power disturbance fluctuations at different time scales are identified. Using the sliding window method, the mean, variance, maximum, and minimum values ​​of power disturbance are calculated based on historical power data and data at the current time step. Based on the prediction model, the trend of power disturbance changes within a preset time period is predicted using real-time data. Step B2, determine the quality indicators of new energy consumption: simulate different wind power and photovoltaic consumption capacity ratios using time-series simulation method, and calculate the wind power and photovoltaic consumption capacity ratios respectively. New energy consumption quality indicators when equal to 0, 1, 1 / 2, 3 / 4, and 1 / 4 , , , , The quality indicators for new energy consumption are one of the following: new energy utilization rate, new energy self-consumption rate, annual fossil energy substitution ratio, and voltage adaptability. Then, based on the improved binary search method, the new energy consumption capacity within the preset threshold range and the optimal consumption capacity of each node are obtained. Step B3: Dynamically adjust the preset constraints of the self-regulatory operation indicators: Based on the predicted short-timescale power disturbance trend in Step B1, dynamically adjust the upper and lower limits of the short-timescale power disturbance amplitude; based on the new energy consumption quality indicators and the mid-timescale trend analysis of power disturbance obtained in Step B2, dynamically adjust the mid-timescale power disturbance amplitude and ramp rate; based on the long-timescale power disturbance trend in Step B1, combined with the new energy consumption quality indicators, dynamically adjust the long-timescale power fluctuation amplitude.

[0033] It should be specifically noted that in step B3 of the data acquisition and analysis module's dynamic adjustment of the preset constraints of multi-scale self-regulatory operation indicators, the adjustment of the self-regulatory operation indicators at the medium time scale specifically includes: when the fluctuation amplitude of new energy power exceeds ±15% and the new energy consumption quality index is lower than 80%, automatically optimize energy storage scheduling and reduce the charging and discharging power of the energy storage module; when the energy storage module is in charging state, reduce the charging power from the original set value to 75%; when the energy storage module is in discharging state, reduce the discharging power from the original set value to 80%; when the new energy consumption quality index improves to 85% or above, relax the power disturbance amplitude and ramp rate restrictions at the medium time scale, allowing power fluctuations to be within the range of ±12% to ±15%, and restore the charging and discharging power of the energy storage module to the original set value, but not exceeding the set ±15% range for adjustment.

[0034] It should be specifically noted that in step B3 of the data acquisition and analysis module's dynamic adjustment of the preset constraints of multi-scale self-regulatory operation indicators, the adjustment of long-term scale self-regulatory operation indicators specifically includes: when the power fluctuation amplitude exceeds ±20% and the new energy consumption quality index is below 75%, automatically increasing the storage capacity of the energy storage module and reducing the charging and discharging power of the energy storage module; when the energy storage module is in charging state, reducing the charging power from the original set value to 70%; when the energy storage module is in discharging state, reducing the discharging power from the original set value to 75%; when the new energy consumption quality index remains above 80%, appropriately relaxing the power fluctuation amplitude limit, allowing the power fluctuation range to reach ±18%, and restoring the charging and discharging power of the energy storage module to the original set value, but not exceeding the set ±18% range for adjustment.

[0035] To illustrate the implementation of the aforementioned technologies in a test distribution network, this network comprises multiple nodes and is equipped with energy storage modules, adjustable load modules, photovoltaic power generation equipment, and wind power generation equipment. The data acquisition and analysis module collects real-time data on the distribution network's power output, energy storage status, load demand, and output from new energy sources (wind and photovoltaic). The following details how to dynamically adjust autonomous operation indicators across short, medium, and long timescales based on this test distribution network.

[0036] Step B1, Power disturbance change trend analysis: at time... The power demand of the distribution network is 28MW, the energy storage status is 30% (i.e., the current energy storage module capacity is 6.3MWh), the load demand is 27MW, the photovoltaic output is 3.2MW, and the wind power output is 2.8MW. Using a time series forecasting model, historically collected power data and real-time data are used to predict the trend of power disturbance changes. The system predicts that the power disturbance amplitude will be ±12% within the next 15 minutes, with short-term fluctuations within ±12%, medium-term fluctuations based on 1-hour forecast data predicting ±15%, and long-term fluctuations based on 6-hour forecast data predicting ±20%. Step B2, determine the quality indicators of new energy consumption: Calculate the quality indicators of new energy consumption, assuming a new energy utilization rate of 92% (the proportion of actual wind and solar new energy power generation to theoretical power generation), a new energy self-consumption rate of 95% (the proportion of local consumption of new energy to total new energy power generation), a new energy replacement rate of 50% for traditional fossil energy, and a voltage adaptability of 90% (the impact of new energy penetration on the voltage quality of the distribution network). The optimal wind power consumption capacity obtained through the binary search method is 15MW, the optimal photovoltaic consumption capacity is 10MW, and the total new energy consumption capacity is 25MW. Step B3: Dynamically adjust the preset constraints of the self-regulatory operation indicators: Short-term self-regulatory operation indicator adjustment Adjustment when power disturbance amplitude exceeds ±10%: At any moment The predicted power disturbance amplitude is ±12%, which exceeds the ±10% limit, so the charging and discharging power of the energy storage module is automatically adjusted. Energy storage module charging status: The current charging power of the energy storage module is 5MW. The system will reduce the charging power to 80% of the original setting, i.e., 4MW. Energy storage module discharge status: Assuming the current discharge power of the energy storage module is 5MW, the system will reduce the discharge power to 85% of the original setting, i.e., 4.25MW; Adjustments for power disturbance amplitude less than ±5% and renewable energy consumption quality index above 95%: at time... The power disturbance amplitude is reduced to ±4%, and the quality index of new energy consumption is improved to over 95%. The system relaxes the power disturbance amplitude limit on a short time scale, allowing power fluctuations to be within ±8%, and restores the charging and discharging power of the energy storage module to the original set value, but not exceeding ±10% of the set threshold adjustment range.

[0037] Adjustment of self-regulatory operation indicators on a medium time scale: Adjustments when the fluctuation amplitude of renewable energy power exceeds ±15% and the renewable energy consumption quality index is below 80%: at time... If the fluctuation amplitude of new energy power exceeds ±15% and the new energy consumption quality index is 75%, the system will automatically optimize energy storage scheduling to reduce the charging and discharging power of energy storage modules. Energy storage module charging status: The charging power is reduced from the original setting of 5MW to 3.75MW (a reduction of 25%). Energy storage module discharge status: Reduce the discharge power from the original setting of 5 MW to 4MW (a reduction of 20%). Adjustments when the quality index for renewable energy consumption is raised to 85%: At any moment The quality index of new energy consumption has been improved to over 85%. The system relaxes the power disturbance amplitude and ramp rate restrictions on the medium time scale, allowing power fluctuations to be within the range of ±12% to ±15%, and restores the charging and discharging power of the energy storage module to the original set value, but not exceeding the set range of ±15% for adjustment. Adjustment of long-term self-regulatory operation indicators: Adjustments when power fluctuation amplitude exceeds ±20% and renewable energy consumption quality index is below 75%: at time... The power fluctuation amplitude exceeds ±20%, and the quality index of new energy consumption is 70%. The system automatically increases the energy storage module's reserve capacity and reduces the energy storage module's charging and discharging power. Energy storage module charging status: The charging power is reduced from the original setting of 5MW to 3.5MW (a reduction of 30%). Energy storage module discharge status: Reduce the discharge power from the original setting of 5MW to 3.75MW (a reduction of 25%). Adjustments when the quality indicators for renewable energy consumption remain above 80%: At any moment The quality index of new energy consumption has been improved to over 80%. The system has appropriately relaxed the restrictions on power fluctuation range, allowing power fluctuation range to reach ±18%. The charging and discharging power of the energy storage module has been restored to the original set value, but not exceeding the set range of ±18%.

[0038] It should be noted that the main grid coordination module dynamically adjusts the power exchange between upstream and downstream grids by acquiring real-time power demand, load status, and power fluctuations in the distribution network, combined with power, energy storage status, load demand, and renewable energy output data obtained by the data acquisition and analysis module. Specifically, the main grid coordination module analyzes future power disturbance trends based on a time-series prediction model, considering real-time power fluctuations and load changes, and dynamically adjusts the power exchange accordingly. When the power disturbance amplitude exceeds ±10%, the main grid coordination module adjusts the main grid power output to 110%–115% of the distribution network demand, keeping power fluctuations within ±8%. When the power fluctuation amplitude is lower than... When the renewable energy absorption quality index is above 95% and the power exchange rate is ±5%, the main grid coordination module will restore the power exchange to the original set value and allow power fluctuations within ±8%. When the power fluctuation is between ±5% and ±10%, the main grid coordination module will adjust the power exchange to 105% to 110% of the distribution network demand to ensure that the power fluctuation is between ±5% and ±8%. When the renewable energy absorption quality index is below 85%, the main grid coordination module will increase the power exchange to 115% of the distribution network demand to optimize the grid load distribution. The main grid coordination module, through a real-time feedback mechanism, cooperates with the energy storage module and the load regulation module to adjust the energy storage charging and discharging power.

[0039] In summary, this invention, based on a source-grid-load coordination mechanism, combines real-time data acquisition, time-series prediction models, and multi-timescale power disturbance analysis to achieve dynamic optimization of grid power dispatch and enhance the capacity for renewable energy absorption. Through a collaborative optimization dispatch module, combined with real-time collected wind power, photovoltaic power generation data, energy storage status, and load demand, it can predict power disturbance trends in real time and adjust the charging and discharging strategies of energy storage modules and controllable load dispatch when power fluctuations are significant, ensuring grid stability and efficient operation. Simultaneously, by using an improved interval analysis method to perform scenario analysis on wind power and photovoltaic power output, it achieves power dispatch optimization under different power scenarios, reducing the impact of power fluctuations on the grid and enhancing the absorption capacity of renewable energy. When the source absorption quality is poor, the main grid coordination module dynamically adjusts the power exchange and network reconfiguration path to optimize the power flow between upstream and downstream grids and maximize the absorption of new energy. The data acquisition and analysis module dynamically adjusts the preset constraints of multi-scale autonomous operation indicators based on the power disturbance change trend and new energy absorption quality indicators to ensure that the power exchange and power fluctuation between the distribution network and the main grid are within a reasonable range. This invention helps to improve the grid's regulation capability and adaptability in the face of complex power fluctuations and uncertainties in new energy output, optimize the utilization efficiency of grid resources, reduce the impact of power fluctuations on grid stability, ensure the maximum absorption of new energy, and improve the flexibility and stability of grid operation.

[0040] The aforementioned test distribution network includes multiple nodes, with node 1 equipped with an energy storage module having a maximum power of 5MW and a maximum capacity of 21MWh. Data on the distribution network's power demand, energy storage status, load demand, and renewable energy output (such as wind and solar power) are collected in real-time through a data acquisition and analysis module, providing a basis for dynamic adjustments to the main grid coordination module. The distribution network's average annual peak load is 29.6MW. Solar power generation is installed at nodes 4, 7, 8, 12, 16, and 18, with the energy storage module initially charged at 30% capacity. The maximum grid-connected capacity for solar power is 10.5MW.

[0041] The main power grid coordination module dynamically adjusts the power exchange between upstream and downstream power grids by acquiring real-time power demand, load conditions, and power fluctuations in the distribution network, combined with the output data from the data acquisition and analysis module. The specific implementation process is as follows: (1) Obtain real-time data: at time The power demand of the distribution network is 28MW, the energy storage module capacity is 6.3MWh (30% capacity), the photovoltaic power output is 3.2MW, and the wind power output is 2.8MW; the power fluctuation amplitude of the distribution network is predicted to be ±12%, while the quality index of new energy consumption is 92%. (2) Time-series forecasting analysis: Based on historically collected data on power distribution, energy storage status, load demand, and renewable energy output, the main grid coordination module uses a time-series forecasting model to predict future power disturbance trends. At time... It is predicted that the power disturbance amplitude will exceed ±10% (i.e. ±12%) within the next 15 minutes. (3) Dynamically adjust the power exchange quantity: When the power disturbance amplitude exceeds ±10%, the main grid coordination module will adjust the main grid power output to 110% to 115% of the distribution network demand (i.e., 32.8MW to 34.2MW) based on the predicted power disturbance, so as to keep the power fluctuation within ±8%. The current demand of the distribution network is 28MW, and the main grid power output is adjusted to 32.8 MW (28MW×1.15). When power fluctuation is below ±5% and the quality index of new energy consumption is above 95%, at time... If the predicted power fluctuation amplitude is ±4% and the quality index of new energy consumption is 96%, the main grid coordination module will restore the power exchange to the original set value, assuming the original set value is 28MW, and allow the power fluctuation to be within ±8%. Adjustment when power fluctuations are between ±5% and ±10%: at time With a power fluctuation amplitude of ±8%, the main grid coordination module will adjust the power exchange to 105% to 110% of the distribution network demand (i.e., 29.4MW to 30.8MW) to ensure that the power fluctuation is between ±5% and ±8%. Adjustments when the quality index of new energy consumption is below 85%: at any given time When the quality index of renewable energy consumption is lower than 85% (out of 80%), the main grid coordination module increases the power exchange capacity to 115% of the distribution network demand (i.e., 32.2MW) to optimize grid load distribution, ensure stable grid operation and higher renewable energy consumption efficiency. (4) Coordination between real-time feedback mechanism and energy storage module: The main grid coordination module monitors the power exchange in real time and adjusts the charging and discharging power of the energy storage module through a feedback mechanism. When the energy storage module is in a charging state, the system will adjust the charging power of the energy storage module according to the predicted power fluctuations in order to balance the power demand of the grid and the energy storage capacity. The charging power of the energy storage module is reduced from the original setting of 5MW to 4MW (a reduction of 20%) to adapt to the changes in power fluctuations.

[0042] Through the above process, the main grid coordination module realizes dynamic adjustment of the power exchange between the upper and lower level grids, ensuring the stable operation of the distribution network under different power fluctuation conditions. By combining real-time data prediction and new energy consumption quality indicators, it flexibly adjusts the power exchange and energy storage scheduling, thereby optimizing the operating efficiency of the distribution network and the capacity for new energy consumption.

[0043] In summary, this invention achieves dynamic optimization of grid power dispatch and enhances the absorption capacity of new energy sources by combining a source-grid-load coordination mechanism with real-time data acquisition, time-series prediction models, and multi-time-scale power disturbance analysis. Through a collaborative optimization dispatch module, combined with real-time collected wind power, photovoltaic power generation data, energy storage status, and load demand, it can predict power disturbance trends in real time and adjust the charging and discharging strategies of energy storage modules and controllable load dispatch when power fluctuations are significant, ensuring grid stability and efficient operation. Simultaneously, by using an improved interval analysis method to perform scenario analysis on wind power and photovoltaic power output, it achieves power dispatch optimization under different power scenarios, reducing the impact of power fluctuations on the grid and enhancing the absorption capacity of new energy sources. When the absorption quality is poor, the main grid coordination module dynamically adjusts the power exchange and network reconfiguration path to optimize the power flow between upstream and downstream grids and maximize the absorption of new energy. The data acquisition and analysis module dynamically adjusts the preset constraints of multi-scale autonomous operation indicators based on the power disturbance change trend and new energy absorption quality indicators to ensure that the power exchange and power fluctuation between the distribution network and the main grid are within a reasonable range. This invention helps to improve the grid's regulation capability and adaptability in the face of complex power fluctuations and uncertainties in new energy output, optimize the utilization efficiency of grid resources, reduce the impact of power fluctuations on grid stability, ensure the maximum absorption of new energy, and improve the flexibility and stability of grid operation.

[0044] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0045] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A new energy consumption system based on source-grid-load coordination, characterized in that, It includes an energy storage module, a controllable load module, a distribution network self-regulatory control module, a main grid coordination module, a data acquisition and analysis module, a communication and dispatch module, and a collaborative optimization dispatch module. The collaborative optimization dispatch module, through a source-grid-load coordination mechanism, establishes a multi-timescale power disturbance model based on real-time collected renewable energy output data, load demand, and energy storage status information. It predicts the uncertainty of renewable energy output in real time and makes dynamic adjustments. It uses an improved interval method to conduct scenario analysis on the output of wind power and photovoltaic power generation, and combines the dispatch capabilities of energy storage and controllable loads to optimize and adjust energy storage charging and discharging strategies and load dispatch. The data acquisition module... The analysis module collects power, energy storage status, load demand, and renewable energy output data from the distribution network. It uses a predictive model to analyze power disturbance trends and, in conjunction with renewable energy absorption quality indicators, dynamically adjusts the preset constraints of multi-scale autonomous operation indicators. The distribution network autonomous control module monitors the power and energy storage status at the distribution gateway in real time and, in conjunction with the power disturbance model, automatically adjusts the charging and discharging power of the energy storage module and the scheduling of controllable loads. The main grid coordination module dynamically adjusts the network reconfiguration path and power exchange volume based on the power demand, load status, and power fluctuations of the distribution network. The communication and scheduling module transmits information and issues scheduling commands.

2. The new energy consumption system based on source-grid-load coordination according to claim 1, characterized in that, In the collaborative optimization scheduling module, the improved interval method for scenario analysis of wind power generation output includes: Step S1, divide the wind power output scenarios: divide the wind power output into five scenarios, namely the prediction baseline scenario, the two uphill and downhill scenarios from odd to even time points, and the two uphill and downhill scenarios from even to odd time points. Step S2, Longitudinal Analysis: For the wind power output at each moment, based on the real-time collected wind speed factor, calculate and analyze the fluctuation of the wind power output at that moment, determine the magnitude and trend of the change, and determine the upper and lower bounds of the wind power output fluctuation. Step S3, Horizontal Analysis and Scenario Calculation: Perform a horizontal analysis on the fluctuation of wind power output between different time points, calculate the power fluctuation range of each ramping scenario, obtain the upper and lower bounds of the wind power output ramping scenario, and adjust them according to the actual wind power output fluctuation. Step S4, Scenario Supplementation and Adjustment: Based on the actual collected output power data of the wind turbine, if the ramping scenario cannot meet the power fluctuation requirements or the wind power output value does not reach the preset upper and lower limits, the scenario is identified as missing. The missing power is supplemented according to the existing wind power output prediction model and historical data. The supplemented power value is dynamically adjusted according to the ramping scenario of power fluctuation and the reserve capacity requirements to ensure that the power change meets the AGC regulation capability and reserve capacity requirements. Step S5, Scene Update and Optimization: Based on the supplemented power value and the adjusted ramp scenario, the prediction model for wind power output is updated in real time. Combining real-time data and power fluctuation trends, the subsequent power output is dynamically optimized.

3. A new energy consumption system based on source-grid-load coordination according to claim 1, characterized in that, In step S1 of the collaborative optimization scheduling module, the predicted baseline scenario is a scenario in which the wind speed range is from the cut-in wind speed of the wind turbine to the rated wind speed of the wind turbine. The power output of the wind turbine is obtained by analyzing the wind speed at the current moment and the historical wind turbine output power data to obtain the wind turbine output power change curve that is positively correlated with the wind speed. The uphill and downhill scenarios at odd to even times include the first power uphill scenario when the wind speed increases from the cut-in wind speed to the rated wind speed and the second power uphill scenario when the wind speed decreases from the rated wind speed to the cut-in wind speed. The uphill and downhill scenarios at even to odd times include the third power uphill scenario when the wind speed increases from the cut-in wind speed to the rated wind speed and the fourth power uphill scenario when the wind speed decreases from the rated wind speed to the cut-in wind speed.

4. A new energy consumption system based on source-grid-load coordination according to claim 2, characterized in that, In the collaborative optimization scheduling module, the improved interval analysis method for scenario analysis of photovoltaic power generation output includes: Step A1, establish a photovoltaic power generation output model: by collecting real-time environmental data on light intensity and temperature, establish a nonlinear relationship model between photovoltaic power generation output and environmental factors; Step A2, divide the photovoltaic power output scenarios: Divide the collected light intensity into three ranges of low light, medium light and high light according to the preset range, and divide the temperature into low temperature range, medium temperature range and high temperature range according to the preset range. Combine the light intensity range and temperature range to divide the photovoltaic power output scenarios into 9 photovoltaic power output scenarios. Step A3: Predict photovoltaic power output based on a pre-trained LSTM model: By inputting features such as light intensity, temperature, and historical power data, the photovoltaic power output scenario is input into the pre-trained LSTM model for training, capturing the dependence and fluctuation patterns of time-series data, and outputting the predicted photovoltaic power value.

5. A new energy consumption system based on source-grid-load coordination according to claim 4, characterized in that, In the collaborative optimization scheduling module, in step A3, the LSTM model dynamically adjusts the size of the input time window based on the changing trends of real-time light intensity, temperature, and historical power data to adapt to data features at different time scales. The input layer receives light intensity, temperature, and historical power data from multiple time series, extracts features through convolutional layers to obtain local features at different time scales, and fuses short-term and long-term features using time windows. The fused features are then input into the LSTM layer for time-series learning. The intervals of light intensity and temperature are encoded as numerical features and added as additional input features to the LSTM network. In the loss function, the gradient rate of change between adjacent time steps of the predicted power sequence at each time point is calculated, and the absolute values ​​of the gradient rates of change at each time point are summed as a penalty term, with an adjustment factor added. Adjusting the relative weights between the volatility stationarity loss term and the predicted loss term, the adjustment factor Preset according to the photovoltaic power output scenario category.

6. A new energy consumption system based on source-grid-load coordination according to claim 1, characterized in that, The process by which the data acquisition and analysis module dynamically adjusts the preset constraints of multi-scale self-regulatory operation indicators includes: Step B1, Power Disturbance Trend Analysis: Using historically collected data on power distribution network, energy storage status, load demand, and new energy output, a time series prediction model is used to analyze the future power disturbance trend. Short-term, medium-term, and long-term fluctuations of power disturbance are modeled separately, and the characteristics of power disturbance fluctuations at different time scales are identified. Using the sliding window method, the mean, variance, maximum, and minimum values ​​of power disturbance are calculated based on historical power data and data at the current time step. Based on the prediction model, the trend of power disturbance changes within a preset time period is predicted using real-time data. Step B2, determine the quality indicators of new energy consumption: simulate different wind power and photovoltaic consumption capacity ratios using time-series simulation method, and calculate the wind power and photovoltaic consumption capacity ratios respectively. New energy consumption quality indicators when equal to 0, 1, 1 / 2, 3 / 4, and 1 / 4 , , , , The quality indicators for new energy consumption are one of the following: new energy utilization rate, new energy self-consumption rate, annual fossil energy substitution ratio, and voltage adaptability. Then, based on the improved binary search method, the new energy consumption capacity within the preset threshold range and the optimal consumption capacity of each node are obtained. Step B3: Dynamically adjust the preset constraints of the self-regulatory operation indicators: Based on the predicted short-timescale power disturbance trend in Step B1, dynamically adjust the upper and lower limits of the short-timescale power disturbance amplitude; based on the new energy consumption quality indicators and the mid-timescale trend analysis of power disturbance obtained in Step B2, dynamically adjust the mid-timescale power disturbance amplitude and ramp rate; based on the long-timescale power disturbance trend in Step B1, combined with the new energy consumption quality indicators, dynamically adjust the long-timescale power fluctuation amplitude.

7. A new energy consumption system based on source-grid-load coordination according to claim 6, characterized in that, In step B3 of the data acquisition and analysis module's dynamic adjustment of preset constraints on multi-scale self-regulatory operation indicators, the adjustment of short-time-scale self-regulatory operation indicators specifically includes: when the predicted power disturbance amplitude exceeds ±10%, automatically reducing the charging and discharging power of the energy storage module; when the energy storage module is in charging state, reducing the charging power by 20%; when the energy storage module is in discharging state, reducing the discharging power by 15%; when the power disturbance amplitude is less than ±5% and the new energy consumption quality index is above 95%, relaxing the short-time-scale power disturbance amplitude limit, allowing power fluctuations within ±8%, and restoring the charging and discharging power of the energy storage module to the original set value and adaptively adjusting within ±10% of the original set value.

8. A new energy consumption system based on source-grid-load coordination according to claim 6, characterized in that, In step B3 of the data acquisition and analysis module's dynamic adjustment of preset constraints on multi-scale self-regulatory operation indicators, the adjustment of self-regulatory operation indicators at the medium time scale specifically includes: when the fluctuation amplitude of new energy power exceeds ±15% and the new energy absorption quality index is below 80%, automatically optimize energy storage scheduling and reduce the charging and discharging power of energy storage modules; when the energy storage module is in charging state, reduce the charging power by 25%; when the energy storage module is in discharging state, reduce the discharging power by 20%; when the new energy absorption quality index improves to 85% or above, relax the power disturbance amplitude and ramp rate restrictions at the medium time scale, allowing power fluctuations to be within the range of ±12% to ±15%, and restore the charging and discharging power of the energy storage module to the original set value and adaptively adjust within the range of ±15% of the original set value.

9. A new energy consumption system based on source-grid-load coordination according to claim 6, characterized in that, In step B3 of the data acquisition and analysis module's dynamic adjustment of preset constraints for multi-scale self-regulatory operation indicators, the adjustment of long-term self-regulatory operation indicators specifically includes: when the power fluctuation amplitude exceeds ±20% and the new energy consumption quality index is below 75%, automatically increasing the energy storage module's reserve capacity and reducing the energy storage module's charging and discharging power; when the energy storage module is in charging mode, reducing the charging power by 30%; when the energy storage module is in discharging mode, reducing the discharging power by 25%; when the new energy consumption quality index remains above 80%, appropriately relaxing the power fluctuation amplitude limit, allowing the power fluctuation range to reach ±18%, and restoring the energy storage module's charging and discharging power to the original set value and adaptively adjusting within the original set value ±18%.

10. A new energy consumption system based on source-grid-load coordination according to claim 1, characterized in that, The main grid coordination module dynamically adjusts the power exchange between upstream and downstream grids by acquiring real-time power demand, load status, and power fluctuations in the distribution network. This is done in conjunction with data from the data acquisition and analysis module, which includes distribution network power, energy storage status, load demand, and renewable energy output. Specifically, the main grid coordination module analyzes future power disturbance trends based on a time-series forecasting model, considering real-time power fluctuations and load changes, and dynamically adjusts the power exchange accordingly. When the power disturbance amplitude exceeds ±10%, the main grid coordination module adjusts the main grid power output to 110%–115% of the distribution network demand, keeping power fluctuations within ±8%. When the power fluctuation amplitude is below ±5%, the module adjusts the power exchange accordingly. Furthermore, when the renewable energy absorption quality index is above 95%, the main grid coordination module restores the power exchange to the original set value and allows power fluctuations within ±8%; when the power fluctuation is between ±5% and ±10%, the main grid coordination module adjusts the power exchange to 105% to 110% of the distribution network demand to ensure that the power fluctuation is between ±5% and ±8%; when the renewable energy absorption quality index is below 85%, the main grid coordination module increases the power exchange to 115% of the distribution network demand to optimize the grid load distribution; the main grid coordination module, through a real-time feedback mechanism, cooperates with the energy storage module and the load regulation module to adjust the energy storage charging and discharging power.

Images