A method and device for power balance regulation of distributed renewable energy microgrids

Abstract

This invention relates to the field of power grid control technology, specifically to a method and apparatus for power balance control in a distributed renewable energy microgrid, solving the technical problem that existing technologies struggle to meet the dynamic supply and demand balance of electricity within the power grid. The method includes: collecting electrical data, environmental data, and actual power supply data from each node in the distributed renewable energy microgrid; predicting the power supply data for each node based on the electrical data, environmental data, and actual power supply data; evaluating the prediction accuracy index of each node based on the predicted power supply data and the actual power supply data; the prediction accuracy index includes prediction difficulty and prediction accuracy; and selecting target nodes from the nodes of the distributed renewable energy microgrid to install capacitors based on the prediction accuracy index; the capacitors are used to balance the power supply and demand differences in the distributed renewable energy microgrid.

Description

Technical Field

[0001] This invention relates to the field of power grid control technology, specifically to a method and apparatus for power balance control of a distributed new energy microgrid. Background Technology

[0002] Distributed renewable energy microgrids, as comprehensive energy systems integrating various distributed renewable energy sources such as wind, solar, and biomass energy, achieve flexible interaction with the main power grid through subsystems such as energy storage and dispatching, and have become an important form of renewable energy utilization. Maintaining a dynamic balance between internal power supply and demand is crucial for ensuring the stable operation of the microgrid, as it directly affects the renewable energy absorption efficiency and power supply reliability.

[0003] In existing technologies, various forecasting techniques are typically used to predict the power generation and load demand of microgrids, and then grid power and electricity are dispatched based on the forecast results to achieve balance. However, due to various factors such as the environmental dependence of renewable energy generation and the randomness of user electricity consumption behavior, there is always a deviation between the forecast results and the actual situation, resulting in poor execution of dispatching schemes. This not only disrupts the power balance but also adds extra operational burden to the microgrid, affecting overall operational stability and economy. Summary of the Invention

[0004] To address the technical problem that existing technologies cannot meet the dynamic supply and demand balance of electricity within the power grid, the present invention aims to provide a method and device for power balance regulation in distributed new energy microgrids. The specific technical solution adopted is as follows:

[0005] This application provides a power balance control method for distributed renewable energy microgrids, the method comprising:

[0006] The system collects electrical data, environmental data, and actual power supply data from each node in the distributed renewable energy microgrid. The electrical data is used to characterize at least one electrical parameter among current, voltage, and power at various times in multiple time periods. The environmental data is used to characterize at least one environmental parameter among wind force, wind speed, and solar intensity at various times in multiple time periods. The power supply data includes the power generation and load power consumption of the nodes at various times in multiple time periods.

[0007] Based on electrical data, environmental data, and actual power supply data, the predicted power supply data for each node is obtained.

[0008] Based on the predicted power supply data and actual power supply data of each node, evaluate the prediction accuracy index of each node; the prediction accuracy index includes prediction difficulty and prediction accuracy.

[0009] Based on the prediction accuracy index, target nodes for installing capacitors are selected from each node of the distributed renewable energy microgrid; the capacitors are used to balance the power supply and demand differences of the distributed renewable energy microgrid.

[0010] In one possible implementation, the method includes:

[0011] Electrical sensors and environmental monitoring sensors are installed at each node of the distributed new energy microgrid. Electrical data of each node is collected through electrical sensors, and environmental data of each node is collected through environmental monitoring sensors.

[0012] The actual power supply data of each node is collected by data acquisition equipment; electrical data, environmental data and actual power supply data are stored in multiple time periods divided according to time sequence.

[0013] In one possible implementation, the method includes:

[0014] For each node, the corresponding environmental impact degree is determined based on the node's electrical and environmental data; the environmental impact degree is used to characterize the correlation between changes in environmental parameters and changes in node power supply data.

[0015] For each node, the time period reference degree is determined based on the node's electrical data, environmental data, and environmental impact degree; the time period reference degree is used to characterize the reference value of the electrical data and environmental data of the corresponding time period for the current forecast.

[0016] Using time period reference degree as a weighting factor, the power supply data of each node in the prediction period is predicted by a preset time series analysis algorithm, and the predicted power supply data of each node is obtained.

[0017] In one possible implementation, the method includes:

[0018] For each node, an electrical differential sequence is constructed based on the node's electrical data, and an environmental differential sequence is constructed based on the node's environmental data. The electrical differential sequence includes multiple electrical parameter differential values, which are used to characterize the difference between corresponding electrical parameters at adjacent times. The environmental differential sequence includes multiple environmental parameter differential values, which are used to characterize the difference between corresponding environmental parameters at adjacent times.

[0019] Based on the time sequence matching algorithm, the electrical differential sequence and environmental differential sequence of the node are time-series matched; the electrical parameter difference value and environmental parameter difference value that match in the time-series matched electrical differential sequence and environmental differential sequence correspond to the same time.

[0020] The degree of environmental impact is determined based on the electrical differential sequence and environmental differential sequence after time-series matching.

[0021] In one possible implementation, the method includes:

[0022] For each node, based on the node's environmental data, current environmental data, and the node's environmental impact, determine the similarity between the node and the current weather at different time periods;

[0023] For each node, the preliminary prediction confidence level for each time period is determined based on the node's power supply data.

[0024] For each node, the corresponding time period reference level is determined based on the weather similarity and preliminary prediction confidence of each time period.

[0025] In one possible implementation, the method includes:

[0026] For each node, the prediction difficulty of the node is assessed based on the data difference between the predicted power supply data and the actual power supply data at various times in multiple time periods; the prediction difficulty is positively correlated with the data difference between the predicted power supply data and the actual power supply data at various times in multiple time periods.

[0027] For each node, the time period reference degree of multiple time periods is used as the corresponding weight. The data difference between the predicted power supply data and the actual power supply data of the node at various times in multiple time periods is weighted and calculated to obtain the average prediction error of the node. The prediction accuracy of the node is determined based on the average prediction error.

[0028] In one possible implementation, the method includes:

[0029] Nodes in the distributed renewable energy microgrid whose prediction difficulty is greater than or equal to a preset difficulty threshold, and / or whose prediction accuracy is less than or equal to a preset accuracy threshold, are designated as target nodes.

[0030] In one possible implementation, the method further includes:

[0031] The nominal capacity of the corresponding capacitor is determined based on the data differences between the predicted power supply data and the actual power supply data at various times in multiple time periods, as well as the electrical data.

[0032] In one possible implementation, the method further includes:

[0033] Based on the nominal capacity of each capacitor, each capacitor is divided into multiple capacity levels; different capacity levels of capacitors are used to handle different residual charge scales, and the residual charge scale of the capacitor is positively correlated with the nominal capacity of the capacitor.

[0034] A pre-defined clustering algorithm is used to divide the distributed new energy microgrid into multiple control areas, with the target node where each capacitor is located as the cluster center; each control area includes at least one capacitor as the energy residual adjustment center of the control area.

[0035] A hierarchical control strategy is set for capacitors of different capacity levels; wherein, in the case that there is a first capacitor in the control area that cannot cover the corresponding residual charge, the remaining residual charge after the first capacitor is processed by a second capacitor in the control area with a capacity level higher than that of the first capacitor.

[0036] This application provides a power balance regulation device for a distributed renewable energy microgrid, characterized in that it includes: a processor and a communication interface; the communication interface is coupled to the processor, and the processor is used to run computer programs or instructions to implement the methods described in any of the above embodiments. This power balance regulation device for the distributed renewable energy microgrid can be an electronic device or a chip within an electronic device.

[0037] The present invention has the following beneficial effects:

[0038] Based on the above technical solution, this application systematically collects multi-dimensional data such as electrical data, environmental data, and actual power supply data of each node in the microgrid, providing comprehensive basic support for power supply forecasting. By comparing the forecast data with the actual data to evaluate the forecast accuracy, it can accurately locate nodes prone to power imbalance. Target nodes are selected from the grid based on the forecast accuracy, and capacitors are installed at the target nodes. The rapid charging and discharging characteristics of capacitors are used to specifically balance the power supply and demand differences caused by forecast deviations. This can effectively improve the shortcomings of existing forecasting and scheduling methods in terms of insufficient ability to cope with fluctuations, enhance the stability of the microgrid power balance, and reduce the additional operating burden caused by frequent adjustments to generator output, thus meeting the dynamic power supply and demand balance requirements of the grid. Attached Figure Description

[0039] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart illustrating a power balance control method for a distributed new energy microgrid, as provided in an embodiment of the present invention.

[0041] Figure 2This is a flowchart illustrating another method for power balance regulation of a distributed new energy microgrid, provided in an embodiment of the present invention.

[0042] Figure 3 This is a flowchart illustrating another method for power balance regulation of a distributed new energy microgrid, provided in an embodiment of the present invention.

[0043] Figure 4 This is a flowchart illustrating another method for power balance regulation of a distributed new energy microgrid, provided in an embodiment of the present invention.

[0044] Figure 5 This is a schematic diagram of the structure of a power balance control device for a distributed new energy microgrid provided in one embodiment of the present invention. Detailed Implementation

[0045] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a power balance control method and apparatus for a distributed new energy microgrid proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0047] Given the technical limitations of existing technologies in meeting the dynamic supply and demand balance requirements of electricity within the power grid, this application systematically collects multi-dimensional data, including electrical data, environmental data, and actual power supply data from each node of the microgrid. This provides comprehensive foundational support for power supply forecasting. By comparing the forecast data with actual data to assess forecast accuracy, it can accurately locate nodes prone to power imbalance. Target nodes are then selected from the power grid based on forecast accuracy, and capacitors are installed at these target nodes. The rapid charging and discharging characteristics of capacitors are used to specifically balance the power supply and demand differences caused by forecast deviations. This approach effectively addresses the shortcomings of existing forecasting and scheduling methods in handling fluctuations, enhances the stability of microgrid power balance, and reduces the additional operational burden caused by frequent adjustments to generator output, thus meeting the dynamic supply and demand balance requirements of the power grid.

[0048] The following description, in conjunction with the accompanying drawings, details the specific scheme of the power balance control method and device for a distributed new energy microgrid provided by the present invention.

[0049] Please see Figure 1 The diagram illustrates a flowchart of a power balance control method for a distributed new energy microgrid according to an embodiment of the present invention. The method includes the following steps:

[0050] Step 101: Collect electrical data, environmental data, and actual power supply data of each node in the distributed new energy microgrid.

[0051] In a distributed new energy microgrid, each node includes a power generation node and a power consumption node (also known as a user-side node), and the corresponding power supply data is divided into power generation data and power consumption data.

[0052] Electrical data is used to characterize at least one of the electrical parameters of a node, including current, voltage, and power, at various times across multiple time periods. These electrical parameters are core indicators that reflect the power operation status of a node, and their accuracy directly affects the effectiveness of subsequent forecasting and regulation.

[0053] Environmental data is used to characterize at least one of the environmental parameters, namely wind force, wind speed, and solar intensity, at various times in multiple time periods. This type of environmental data affects the efficiency of new energy power generation. For example, wind speed determines the output power of wind turbine generators, and solar intensity affects the power generation of photovoltaic modules.

[0054] Power supply data includes the power generation and load consumption of nodes at various times across multiple time periods, used to assess the power supply and demand status of nodes.

[0055] In one possible implementation, this application can set up electrical sensors and environmental monitoring sensors at each node of the distributed new energy microgrid, and collect electrical data of each node through electrical sensors, collect environmental data of each node through environmental monitoring sensors, and collect actual power supply data of each node through data acquisition equipment.

[0056] Electrical data, environmental data, and actual power supply data are stored in multiple time periods divided according to time sequence.

[0057] For example, electrical sensors such as current sensors, voltage sensors, and power sensors, as well as environmental monitoring sensors such as anemometers and light sensors, can be deployed at each node of the microgrid. The division of time periods can be flexibly adjusted according to the time characteristics of the data and the control requirements. In addition to the conventional hourly division, shorter time periods can be used for nodes with large fluctuations, such as 15 minutes per period, in order to more accurately capture the data change patterns.

[0058] For example, a day can be divided into 24 time periods, each lasting one hour. After the sensor collects data at a preset frequency (such as once per second or once per minute), the data is stored in a cloud database or a local server according to the corresponding time period. At the same time, a timestamp and node identifier are added to each data entry to facilitate subsequent data analysis and prediction calculations by time period and node, ensuring the traceability and manageability of the data.

[0059] Step 102: Based on electrical data, environmental data, and actual power supply data, predict the power supply data for each node.

[0060] In some embodiments, the future power supply data can be predicted by combining the changing patterns of electrical parameters in historical time-series data and the influence of environmental parameters on power generation, using appropriate prediction algorithms. The predicted power supply data refers to the estimated results of node power generation and load power consumption within a preset future time period, which is used to provide data support for the formulation of subsequent power regulation strategies and avoid passively responding to power imbalance problems.

[0061] For example, power supply data typically changes over time and repeats continuously throughout the day. This application can select algorithms such as Long Short-Term Memory (LSTM) and Least Squares Support Vector Machine (LS-SVM), which can effectively mine the temporal features and correlation patterns in the data and improve prediction accuracy.

[0062] The preset time period for prediction can be flexibly set according to the microgrid's dispatch cycle. For example, short-term predictions can be set to 1 hour, and medium-term predictions can be set to 24 hours to adapt to the control needs under different scenarios. In one example, for a photovoltaic node, the historical trend of light intensity and current environmental data can be combined to obtain the power generation change curve for the next 2 hours through a prediction algorithm. At the same time, the load consumption pattern of the node in the same period in the past can be combined to complete the prediction of load power consumption.

[0063] Step 103: Evaluate the prediction accuracy index of each node based on the predicted power supply data and the actual power supply data of each node.

[0064] The prediction accuracy indicators include prediction difficulty and prediction accuracy.

[0065] It should be noted that the power generation of new energy sources such as solar and wind power is environmentally related, and environmental factors are difficult to predict. Furthermore, while there are certain trends in electricity consumption among individual users, these trends are also difficult to predict accurately. The combined effect of these two factors leads to various errors in the prediction of the balance between electricity consumption and power generation in distributed new energy microgrids.

[0066] The predicted power supply data for each node in the above steps for a certain period (such as the prediction period) usually has a certain error compared with the actual data. This is because the power supply in the renewable energy microgrid is greatly affected by the environment, making it difficult for the relevant prediction methods to make accurate predictions. Therefore, this application can further analyze the prediction accuracy indicators of each node in order to formulate reasonable relevant processing measures.

[0067] In this context, prediction difficulty refers to the degree to which node power supply data is difficult to predict accurately due to factors such as environmental fluctuations and load changes. Essentially, it reflects the stability and regularity of the power supply data; the greater the data fluctuation and the weaker the regularity, the higher the prediction difficulty. Prediction accuracy refers to the degree of agreement between the predicted power supply data and the actual power supply data, used to measure the prediction effectiveness.

[0068] This application can analyze the differences between the predicted power supply data and the actual power supply data of a node at the same time, and then quantify the prediction difficulty and prediction accuracy of that node.

[0069] Step 104: Based on the prediction accuracy index, select target nodes for capacitor installation from the nodes of the distributed new energy microgrid.

[0070] Capacitors are used to balance the power supply and demand differences in distributed renewable energy microgrids. Target nodes are those nodes that are prone to power imbalance due to insufficient forecast accuracy. Targeted capacitor configuration for these nodes can optimize the overall power balance at the lowest cost.

[0071] As an energy storage element, the capacitor has the characteristics of rapid charging and discharging. It can absorb excess energy when the power generation at a node is greater than the power consumption of the load, and release the stored energy when the power generation is less than the power consumption of the load, thereby smoothing the difference between power supply and demand in real time. It is especially suitable for dealing with instantaneous power fluctuations caused by prediction deviations.

[0072] In one possible implementation, this application may select nodes in the distributed new energy microgrid whose prediction difficulty is greater than or equal to a preset difficulty threshold, and / or whose prediction accuracy is less than or equal to a preset accuracy threshold, as target nodes.

[0073] For example, the values ​​for prediction difficulty and prediction accuracy both range from 0 to 1. The closer the prediction difficulty value is to 1, the greater the impact of environmental / load fluctuations on the power supply data of that node, and the more difficult it is to predict. The closer the prediction accuracy value is to 0, the lower the degree of agreement between the prediction result and the actual data. The above-mentioned prediction difficulty threshold and prediction accuracy threshold can be determined by combining specific application scenario data such as microgrid node type (e.g., photovoltaic, wind power, user side), historical prediction data fluctuation characteristics, grid stability requirements, and capacitor configuration costs. For example, the prediction difficulty threshold can be set to 0.5, and the prediction accuracy threshold can be set to 0.8.

[0074] Regarding the prediction difficulty, if the prediction difficulty of a certain node is greater than or equal to the preset difficulty threshold, it means that the electricity consumption (power generation) of the user (or power source) corresponding to that node is difficult to predict (e.g., the user's electricity consumption habits are irregular). This node is an unstable node, and a capacitor needs to be installed at this node to balance the prediction difference.

[0075] Regarding prediction accuracy, if the prediction accuracy of a certain node is less than or equal to the preset accuracy threshold, it indicates that the predicted value of electricity consumption (power generation) at that node generally differs significantly from the actual value. This node is considered to be a node with excessively large differences, and a capacitor needs to be installed at this node to balance the prediction difference.

[0076] Based on the above technical solution, this application systematically collects multi-dimensional data such as electrical data, environmental data, and actual power supply data of each node in the microgrid, providing comprehensive basic support for power supply forecasting. By comparing the forecast data with the actual data to evaluate the forecast accuracy, it can accurately locate nodes prone to power imbalance. Target nodes are selected from the grid based on the forecast accuracy, and capacitors are installed at the target nodes. The rapid charging and discharging characteristics of capacitors are used to specifically balance the power supply and demand differences caused by forecast deviations. This can effectively improve the shortcomings of existing forecasting and scheduling methods in terms of insufficient ability to cope with fluctuations, enhance the stability of the microgrid power balance, and reduce the additional operating burden caused by frequent adjustments to generator output, thus meeting the dynamic power supply and demand balance requirements of the grid.

[0077] Furthermore, in this embodiment, the prediction data can be analyzed and predicted based on environmental factors and the time variation pattern of the power grid, thereby improving the prediction accuracy.

[0078] As one possible embodiment of this application, combined with Figure 1 ,like Figure 2 As shown, step 102 above can be achieved through the following steps:

[0079] Step 201: For each node, determine the corresponding environmental impact degree based on the node's electrical and environmental data.

[0080] Among them, the environmental impact degree is used to characterize the degree of correlation between changes in environmental parameters and changes in node power supply data.

[0081] In one possible implementation, this application can construct an electrical differential sequence based on the electrical data of each node and an environmental differential sequence based on the environmental data of each node. Based on a time-series sequence matching algorithm, the electrical differential sequence and the environmental differential sequence of the node are time-series matched, and the environmental impact degree is determined based on the time-series matched electrical differential sequence and environmental differential sequence.

[0082] The electrical differential sequence comprises multiple electrical parameter difference values, which characterize the differences between adjacent time points for the corresponding electrical parameters. The electrical differential sequence reflects the rate and trend of change of electrical parameters. The environmental differential sequence comprises multiple environmental parameter difference values, which characterize the differences between adjacent time points for the corresponding environmental parameters. The environmental differential sequence reflects the fluctuations of environmental factors. There can be one or more electrical and environmental differential sequences; for example, they can be divided into time periods based on the electrical and environmental data, with one differential sequence corresponding to each time period.

[0083] The electrical parameter difference values ​​and environmental parameter difference values ​​that match in the time-matched electrical differential sequence and environmental differential sequence correspond to the same time.

[0084] For example, the time series matching algorithm can be a dynamic time warping (DWT) algorithm. The environmental impact degree satisfies the following formula:

[0085]

[0086] in, Environmental parameters The corresponding environmental impact level, The number of sequences in the electrical differential sequence or the number of sequences in the environmental differential sequence (the number of sequences in the electrical differential sequence and the number of sequences in the environmental differential sequence are consistent, and there is a one-to-one correspondence between the electrical differential sequence and the environmental differential sequence). Electrical parameters In the The electrical differential sequence corresponding to each time period contains time points. The corresponding electrical parameter differential values, Environmental parameters In the The environmental difference sequence corresponding to each time period is time 1. The corresponding environmental parameter difference values, Electrical parameters In the The electrical differential sequence corresponding to each time period contains time points. Corresponding electrical parameter differential values ​​and environmental parameters In the The environmental difference sequence corresponding to each time period is time 1. The ratio of the corresponding environmental parameter differences is normalized to eliminate differences based on different electrical parameters. Received The problem of inconsistent dimensions between them Based on various electrical parameters Received The average value, For the first Time in each period The time difference with the previous moment, For the first The average of the time differences between each time point and the previous time point in each time period. This represents the function that takes the absolute value. This represents an exponential function with the natural constant e as the base, used for inverse proportional normalization. Used for proportional normalization.

[0087] For example, the normalization function or dimensional processing function (i.e., the norm function) in the embodiments of this application can be a minimum-maximum normalization function.

[0088] It should be noted that, to ensure the calculation results are meaningful, in the embodiments of this application, when performing fractional operations, when encountering a denominator (e.g., When the denominator is 0, a parameter adjustment factor can be added to the denominator to prevent the denominator from being 0. This parameter adjustment factor is a very small positive number. For example, the value of this parameter adjustment factor can be 0.01. Its specific value can be set by the implementer according to the actual situation, and this application embodiment does not make a specific limitation.

[0089] Step 202: For each node, determine the time period reference degree for each time period based on the node's electrical data, environmental data, and environmental impact degree.

[0090] Among them, the time period reference degree is used to characterize the reference value of electrical and environmental data for the corresponding time period for the current forecast.

[0091] In one possible implementation, this application can determine the similarity between a node and the current weather at various time periods based on the node's environmental data, current environmental data, and the node's environmental influence.

[0092] Weather similarity refers to the degree of agreement between historical and current environmental conditions. The higher the environmental impact of a node, the more significant the impact of weather similarity on prediction.

[0093] For example, weather similarity satisfies the following formula:

[0094]

[0095] in, For time period Similarity to the current weather The number of environmental parameters in the environmental data. For time period The average parameter values ​​of the environmental parameters at that time. For the current time period The average parameter values ​​of the environmental parameters at that time. For time period The corresponding environmental difference sequence time The corresponding environmental parameter difference values, For the current time period The corresponding environmental difference sequence time The corresponding environmental parameter difference values, Environmental parameters The corresponding environmental impact level, This represents the function that takes the absolute value. This represents an exponential function with the natural constant e as the base, used for inverse proportional normalization. This is a dimensional processing function used to eliminate the dimensional inconsistency between the average parameter value and the difference value of the environmental parameter. It can be achieved by utilizing... Divide by and Take the absolute value after finding the minimum value in the range, and then use... Divide by and This is achieved by taking the absolute value after finding the minimum value in the set. For time period and time period The difference in environmental parameters at each time point is first processed by dimensioning and then averaged by inverse proportional normalization.

[0096] Then, for each node, the preliminary prediction confidence level for each time period is determined based on the node's power supply data.

[0097] The preliminary prediction confidence level is used to characterize the reliability of the data for this period for the current prediction.

[0098] For example, the initial prediction confidence level satisfies the following formula:

[0099]

[0100] in, For time period The initial prediction confidence level, For time period The moment in The corresponding predicted electrical parameter values, The time in the current time period The corresponding actual electrical parameter values. express The z-score is a function used to describe the accuracy of this prediction relative to other predictions at that time point. This represents the function that takes the absolute value. This represents the hyperbolic tangent function, used for proportional normalization.

[0101] Thus, this application can determine the corresponding time period reference level for each node based on the weather similarity and preliminary prediction confidence of each time period.

[0102] The reference value for this time period is used to characterize the reference value of electrical and environmental data for the corresponding time period for the current forecast.

[0103] For example, the time-period reference degree satisfies the following formula:

[0104]

[0105] in, For time period The time period reference value, For time period Similarity to the current weather For time period The initial prediction confidence level.

[0106] Step 203: Using time period reference degree as a weighting factor, predict the power supply data of each node in the prediction period through a preset time series analysis algorithm to obtain the predicted power supply data of each node.

[0107] For example, the preset time series analysis algorithm can be a seasonal autoregressive integrated moving average (SARIMA) model.

[0108] In some embodiments, data from different historical periods can be weighted according to the corresponding period reference degree. The higher the reference degree of the period data, the greater its contribution to the prediction result, thereby making the prediction result more consistent with the current actual environment and node operation characteristics.

[0109] For example, for photovoltaic nodes with obvious diurnal cycle characteristics, data from the past 7 days with high similarity and reference value to the current weather can be selected as the main training samples. The parameters are trained through the SARIMA model to finally obtain the power generation prediction curve for the next 12 hours.

[0110] Based on the above technical solution, this application can determine the corresponding environmental impact degree based on the electrical and environmental data of the nodes, thereby quantifying the degree of influence of environmental factors on power supply data and providing a scientific basis for the subsequent allocation of prediction weights. Subsequently, based on the electrical data, environmental data, and environmental impact degree of the nodes, the time period reference degree is determined and used as a weight factor in the prediction algorithm. This ensures that the prediction process fully considers the differences in reference value of data from different time periods, effectively improving the accuracy and reliability of power supply data prediction, reducing the problem of control failure caused by excessive prediction deviations, and providing stronger technical support for subsequent precise power balance control.

[0111] As one possible embodiment of this application, combined with Figure 2 ,like Figure 3 As shown, step 103 above can be achieved through the following steps:

[0112] Step 301: For each node, assess the prediction difficulty of the node based on the data differences between the predicted power supply data and the actual power supply data at various times in multiple time periods.

[0113] Among them, the prediction difficulty is positively correlated with the difference between the predicted power supply data and the actual power supply data at various times in multiple time periods.

[0114] For example, the difference between the predicted power supply data and the actual power supply data at various times across multiple time periods can be represented by the difference between the predicted data and the actual data, as shown in the following formula:

[0115]

[0116] in, For nodes At any moment The difference between the predicted power supply data and the actual power supply data. For nodes At any moment The parameter values ​​of the predicted power supply data. For nodes At any moment The difference between the actual power supply data and the predicted value reflects the magnitude of the deviation. This application can describe the dispersion of these prediction differences using statistical measures such as standard deviation. A larger standard deviation indicates more drastic fluctuations in the prediction difference, meaning that the power supply data of that node is more susceptible to external factors and more difficult to control accurately through prediction algorithms. This application can perform a proportional normalization process on this dispersion to obtain the prediction difficulty of that node.

[0117] For example, the prediction difficulty satisfies the following formula:

[0118]

[0119] in, For nodes The corresponding prediction difficulty For nodes At any moment The difference between the predicted power supply data and the actual power supply data. This indicates that the standard deviation is calculated for the differences at each time point. It is used for proportional normalization. The normalized prediction difficulty typically ranges from 0 to 1, facilitating horizontal comparisons between different nodes. Prediction difficulty is positively correlated with the dispersion of prediction differences; that is, the greater the dispersion, the higher the prediction difficulty.

[0120] For example, the normalization function or dimensional processing function (i.e., the norm function) in the embodiments of this application can be a minimum-maximum normalization function.

[0121] Step 302: For each node, the time period reference degree of multiple time periods is used as the corresponding weight. The data difference between the predicted power supply data and the actual power supply data of the node at each time of multiple time periods is weighted and calculated to obtain the average prediction error of the node. The prediction accuracy of the node is determined based on the average prediction error.

[0122] The weighted calculation method described above can highlight the impact of periods with high reference value on the overall prediction effect, making the average prediction error more reflective of the actual prediction quality.

[0123] For example, the average prediction error of a node satisfies the following formula:

[0124]

[0125] in, For nodes The average prediction error, The total number of moments. For nodes At any moment The difference between the predicted power supply data and the actual power supply data. For nodes At any moment The time period reference degree corresponding to the current time period. The higher the average prediction error, the lower the prediction accuracy of the corresponding node. Therefore, this application can perform inverse proportional normalization on the average prediction error to obtain the prediction accuracy of the node. That is, the smaller the average prediction error, the higher the prediction accuracy after normalization.

[0126] It should be noted that, to ensure the calculation results are meaningful, in this embodiment of the application, when performing fractional operations, if the denominator is 0, a parameter adjustment factor can be added to the denominator to prevent the denominator from being 0. This parameter adjustment factor is a very small positive number. For example, the value of this parameter adjustment factor can be 0.01. Its specific value can be set by the implementer according to the actual situation, and this embodiment of the application does not impose a specific limitation.

[0127] For example, the prediction accuracy corresponding to a node satisfies the following formula:

[0128]

[0129] in, For nodes The corresponding prediction accuracy, For nodes The average prediction error, Used for inverse proportional normalization.

[0130] Furthermore, after identifying the target nodes where capacitors need to be installed, this application can further clarify the configuration parameters of the capacitors and achieve efficient processing of microgrid power residuals through a hierarchical control strategy.

[0131] As one possible embodiment of this application, combined with Figure 1 ,like Figure 4 As shown, the method also includes the following steps:

[0132] Step 401: Determine the nominal capacity of the corresponding capacitor based on the data differences between the predicted power supply data and the actual power supply data of each target node at various times in multiple time periods, as well as the electrical data.

[0133] By analyzing the differences between the predicted and actual power supply data for each target node at various times across multiple time periods, the residual power capacity that the capacitors need to cover under extreme conditions can be assessed. Determining the corresponding nominal capacity based on this difference ensures that the capacitors meet the grid balance requirements while avoiding cost waste due to excess capacity. When the power supply at a node exceeds the required power, the excess power can be absorbed by capacitors at nearby nodes; conversely, when the power supply at a node is less than the required power, the missing power can be supplied by nearby capacitors.

[0134] In some embodiments, this application may also configure the relevant operating parameters of the capacitor according to the electrical parameters of the node. For example, the maximum voltage and current in the electrical data can be used to ensure that the electrical parameters of the capacitor match the operating parameters of the node, so as to avoid safety hazards or control failures caused by parameter incompatibility.

[0135] Step 402: Divide each capacitor into multiple capacitance levels according to its nominal capacitance.

[0136] In this system, capacitors of different capacitance levels handle different residual charge sizes, and the residual charge size of a capacitor is positively correlated with its nominal capacitance. That is, the larger the capacitance, the larger the residual charge size it can handle. This tiered approach optimizes resource allocation, avoids inefficiency caused by large-capacity capacitors handling small residual charges, and ensures that large residual charges can be effectively processed.

[0137] For example, this application can divide the nominal capacity of each capacitor into 4 capacity ranges, which correspond to 4 capacity levels respectively. The largest capacity range corresponds to level 1, and the smallest capacity range corresponds to level 4.

[0138] Step 403: Using a preset clustering algorithm, the distributed new energy microgrid is divided into multiple control areas with the target node where each capacitor is located as the cluster center.

[0139] Each control area includes at least one capacitor as the energy residual adjustment center of the control area.

[0140] For example, the preset clustering algorithm can be the K-means clustering algorithm, with the range of each region centered on the cluster center, ensuring that the electrical distance from each node in the region to the regulation center is relatively short, reducing losses in the power transmission process and improving regulation efficiency.

[0141] It should be noted that the above clustering process needs to be divided for each capacity level, and the clustering ranges of different capacity levels may overlap. For example, a certain control area includes one primary clustering range and three secondary clustering ranges. The nodes within each clustering range are balanced by the corresponding capacitors. The primary clustering range covers the three secondary clustering ranges. Therefore, the primary capacitors corresponding to the primary clustering range can also balance the grid for the nodes within the three secondary clustering ranges.

[0142] Step 404: Set up a hierarchical control strategy for capacitors of different capacitance levels.

[0143] In cases where a first capacitor in the control area cannot cover the corresponding residual charge, the remaining residual charge after the first capacitor is processed by a second capacitor in the control area with a higher capacity level than the first capacitor.

[0144] For example, if a fourth-level capacitor in the control area cannot cover the corresponding residual charge, the remaining residual charge after the first capacitor's treatment is handled by a third-level capacitor in the control area, which has a higher capacity level than the first capacitor. Similarly, residual charges that the third-level capacitor cannot handle are handled by a second-level capacitor, and those that the second-level capacitor cannot handle are handled by a first-level capacitor. This hierarchical and coordinated control method can fully leverage the advantages of capacitors with different capacities, achieve step-by-step processing of residual charge, and ensure that all types of residual charges can be handled efficiently.

[0145] In addition, after the power balance of the entire microgrid is achieved, the remaining power in the capacitors can be transferred to distributed batteries and reused during the next regulation. This hybrid energy storage mode can make full use of the fast response capability of capacitors and the large-capacity storage capability of distributed batteries, further improving energy utilization efficiency.

[0146] Based on the above technical solution, this application can determine the nominal capacity of capacitors according to the actual prediction difference of the target node, realizing personalized and precise capacitor configuration. This not only meets the power balance requirements of different nodes but also effectively controls configuration costs and avoids the problems of overcapacity or undercapacity. By dividing the control area into capacity layers and clusters, a multi-level, regionalized power residual processing system is constructed, ensuring that power residuals of different scales can receive matching processing resources, significantly improving the targeting and efficiency of control. The layered collaborative control strategy fully leverages the performance advantages of various capacitors, achieving efficient cooperation among capacitors and solving the problem of insufficient processing capacity of a single capacitor. It can quickly respond to power fluctuations of different scales and effectively mitigate power imbalances caused by prediction deviations.

[0147] This application embodiment also provides a hardware structure diagram of a power balance control device for a distributed new energy microgrid (denoted as power balance control device 50 for a distributed new energy microgrid), see [link to relevant documentation]. Figure 5 The power balance control device 50 of the distributed new energy microgrid includes a processor 501, and optionally, a memory 502 connected to the processor 501.

[0148] In the first possible implementation, see Figure 5 The power balance control device 50 of the distributed new energy microgrid also includes a transceiver 503. The processor 501, memory 502, and transceiver 503 are connected via a bus. The transceiver 503 is used to communicate with other devices or communication networks. Optionally, the transceiver 503 may include a transmitter and a receiver. The device in the transceiver 503 that implements the receiving function can be considered as a receiver, which is used to perform the receiving steps in the embodiments of this application. The device in the transceiver 503 that implements the transmitting function can be considered as a transmitter, which is used to perform the transmitting steps in the embodiments of this application.

[0149] Based on the first possible implementation method Figure 5 The structural diagram shown can be used to illustrate the structure of the power balance control device of the distributed new energy microgrid involved in the above embodiments.

[0150] in, Figure 5 The diagram can also illustrate the system chip in the power balance control device of a distributed renewable energy microgrid. In this case, the actions performed by the power balance control device of the aforementioned distributed renewable energy microgrid can be implemented by this system chip. The specific actions performed can be found above and will not be repeated here.

[0151] In implementation, each step of the method provided in this embodiment can be completed by integrated logic circuits in the processor or by instructions in software form. The steps of the method disclosed in the embodiments of this application can be directly manifested as being executed by a hardware processor, or being executed by a combination of hardware and software modules in the processor.

[0152] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0153] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A power balance regulation method for a distributed new energy micro-grid, characterized in that, The method comprises the following steps: Collecting electrical data, environmental data and actual power supply data of each node in a distributed new energy micro-grid, wherein the electrical data is used to represent at least one electrical parameter of current, voltage and power of each node at each time in multiple time periods, the environmental data is used to represent at least one environmental parameter of wind power, wind speed and light intensity of each node at each time in multiple time periods, and the power supply data comprises power generation and load power consumption of each node at each time in multiple time periods; For each node, determining a corresponding environmental influence degree according to the electrical data and the environmental data of the node, wherein the environmental influence degree is used to represent the correlation between the change of the environmental parameter and the change of the power supply data of the node; For each node, determining a time period reference degree of each time period according to the electrical data, the environmental data and the environmental influence degree of the node; The time period reference degree is used to represent the reference value of the electrical data and the environmental data of the corresponding time period to the current prediction; Using the time period reference degree as a weight factor, predicting the power supply data of each node in a prediction time period by a preset time sequence analysis algorithm to obtain predicted power supply data of each node; According to the predicted power supply data and the actual power supply data of each node, evaluating a prediction accuracy index of each node, wherein the prediction accuracy index comprises a prediction difficulty and a prediction accuracy; Based on the prediction accuracy index, screening a target node to be installed with a capacitor from each node of the distributed new energy micro-grid, wherein the capacitor is used to balance the power supply and demand difference of the distributed new energy micro-grid. 2.The power balance regulation method of the distributed new energy micro-grid according to claim 1, characterized in that, Collecting electrical data, environmental data and actual power supply data of each node in a distributed new energy micro-grid, comprising: Setting electrical sensors and environmental monitoring sensors at each node of the distributed new energy micro-grid, collecting electrical data of each node by the electrical sensors, and collecting environmental data of each node by the environmental monitoring sensors; Collecting actual power supply data of each node by a data collection device; the electrical data, the environmental data and the actual power supply data are stored according to multiple time periods divided in time sequence. 3.The power balance regulation method of the distributed new energy micro-grid according to claim 1, characterized in that, The method comprises the following steps: For each node, respectively constructing an electrical difference sequence according to the electrical data of the node and constructing an environmental difference sequence according to the environmental data of the node; the electrical difference sequence comprises multiple electrical parameter difference values, wherein the electrical parameter difference value is used to represent the difference value of the corresponding electrical parameter between adjacent time points; the environmental difference sequence comprises multiple environmental parameter difference values, wherein the environmental parameter difference value is used to represent the difference value of the corresponding environmental parameter between adjacent time points; Based on a time sequence matching algorithm, performing time sequence matching on the electrical difference sequence and the environmental difference sequence of the node; the matched electrical parameter difference values and environmental parameter difference values in the time sequence matched electrical difference sequence and environmental difference sequence correspond to the same time point; Determining the environmental influence degree according to the time sequence matched electrical difference sequence and environmental difference sequence; The environmental influence degree satisfies the following formula: in, Environmental parameters The corresponding environmental impact level, This refers to the number of sequences in the electrical differential sequence or the number of sequences in the environmental differential sequence. Electrical parameters In the The electrical differential sequence corresponding to each time period contains time points. The corresponding electrical parameter differential values, Environmental parameters In the The environmental difference sequence corresponding to each time period is time 1. The corresponding environmental parameter difference values, For the first Time in each period The time difference with the previous moment, For the first The average of the time differences between each time point and the previous time point in each time period. This represents the function that takes the absolute value. This represents an exponential function with the natural constant e as the base. Used for proportional normalization. 4.The power balance regulation method of the distributed new energy micro-grid according to claim 1, characterized in that, The method comprises the following steps: For each node, a time period reference degree is determined according to electrical data, environmental data and environmental influence degree of the node; For each node, a weather similarity degree in each time period is determined according to environmental data and current environmental data of the node and the environmental influence degree of the node; the weather similarity degree refers to the degree of coincidence of environmental conditions in the historical time period and the current time period; For each node, a preliminary prediction confidence degree in each time period is determined according to power supply data of the node; the preliminary prediction confidence degree is used to represent the credibility of time period data to the current prediction; For each node, a corresponding time period reference degree is determined according to the weather similarity degree and the preliminary prediction confidence degree in each time period of the node; the time period reference degree is used to represent the reference value of electrical data and environmental data in the corresponding time period to the current prediction; wherein, is a time period is a similarity degree to current weather, is a number of environmental parameters in the environmental data, is a time period is an average parameter value of the environmental parameter at the time, is a current time period is an average parameter value of the environmental parameter at the time, is a time period is a time point in the corresponding environmental difference sequence, is a corresponding environmental parameter difference value, is a current time period is a time point in the corresponding environmental difference sequence, is a corresponding environmental parameter difference value, is an environmental parameter is a corresponding environmental influence degree, represents an absolute value function; represents an exponential function with a natural constant e as a base number, is a dimension processing function; The weather similarity degree satisfies the following formula: wherein is a preliminary prediction confidence for the time period , is a time in the time period , is a corresponding predicted electrical parameter value, is a time in the current time period is a corresponding actual electrical parameter value; denotes a calculation function of scores, denotes an absolute value function; denotes a hyperbolic tangent function; The preliminary prediction confidence degree satisfies the following formula: wherein, is a time period a time period reference, is a time period a similarity to current weather, is a time period a preliminary prediction confidence.

5. The power balance regulation method of the distributed new energy micro-grid according to claim 1, characterized in that, The time period reference degree satisfies the following formula: The method further comprises the following steps: For each node, a prediction difficulty degree is evaluated according to the data difference between the predicted power supply data and the actual power supply data of the node at each time point in multiple time periods; the prediction difficulty degree is positively correlated with the data difference between the predicted power supply data and the actual power supply data of the node at each time point in multiple time periods; 6.The power balance regulation method of the distributed new energy micro-grid according to claim 1, characterized in that, For each node, the data difference between the predicted power supply data and the actual power supply data of the node at each time point in multiple time periods is weighted calculated by taking the time period reference degrees in multiple time periods as corresponding weights, to obtain an average prediction error of the node, and the prediction accuracy of the node is determined according to the average prediction error. The method further comprises the following steps: 7.The power balance regulation method of the distributed new energy micro-grid according to claim 1, characterized in that, The nodes in the distributed new energy micro-grid whose prediction difficulty degree is greater than or equal to a preset difficulty threshold and / or whose prediction accuracy degree is less than or equal to a preset accuracy threshold are taken as the target nodes. The method further comprises the following steps: 8.The power balance regulation method of the distributed new energy micro-grid according to claim 7, characterized in that, The nominal capacity of the corresponding capacitor is determined according to the data difference between the predicted power supply data and the actual power supply data of each target node at each time point in multiple time periods and the electrical data. The method further comprises the following steps: According to the nominal capacity of each capacitor, each capacitor is divided into multiple capacity levels; capacitors in different capacity levels correspond to processing different power residual scales; the power residual scale corresponding to the capacitor is positively correlated with the nominal capacity of the capacitor; A preset clustering algorithm is adopted to divide the distributed new energy micro-grid into multiple regulation areas by taking the target nodes in which each capacitor is located as clustering centers; each regulation area includes at least one capacitor as an energy residual adjustment center of the regulation area. A layered regulation strategy is set for capacitors of different capacity levels; wherein, in the case that there is a first capacitor in the regulation area that cannot cover the corresponding power residual, the second capacitor in the regulation area with a higher power level than the first capacitor processes the residual power after the first capacitor.

9. A power balance regulation device for a distributed new energy micro-grid, characterized in that, Comprise: A processor and a communication interface; the communication interface and the processor are coupled, and the processor is used to run computer programs or instructions to realize the power balance regulation method of the distributed new energy micro-grid as claimed in any one of claims 1-8.

Images