Adaptive Voltage Regulation Method and System for Microgrid with Multi-source Power Generation Characteristics

Abstract

The present application discloses an adaptive voltage regulation method and system for a multi-source power generation characteristic microgrid, which relates to the technical field of smart grids. By collecting the physical characteristic data of each renewable energy source in real time, generating a feature embedding representation, and using a multi-head attention mechanism to predict the voltage fluctuation signal. According to the generated voltage fluctuation prediction signal, the system dynamically adjusts the voltage output of each sub-grid through an adaptive control algorithm. Specifically, it includes calculating the target voltage adjustment amount in real time, adjusting the control parameters of the power conversion device, and updating the parameters of the control algorithm through feedback signals. At the same time, the gradient descent algorithm is used to minimize the voltage regulation error, improving the regulation accuracy and response speed of the system. The present invention has the effects of strong adaptability, being able to handle the fluctuation characteristics of different time scales simultaneously, and can significantly improve the stability and accuracy of microgrid voltage regulation, and is applicable to complex multi-source power generation environments.

Description

Technical Field

[0001] This application relates to the technical field of smart grids, and particularly to an adaptive voltage regulation method and system for a multi-source power generation characteristic microgrid. Background Art

[0002] With the increasing severity of global energy shortages and environmental problems, clean energy has gradually become an important direction for the development of modern power systems. Due to its flexible and efficient characteristics, microgrid technology has been widely used. Especially in multi-source power generation systems that integrate multiple renewable energy power sources, microgrids can effectively integrate renewable energies such as solar energy and wind energy, and at the same time provide coordinated control of distributed power sources to ensure the stable supply of electric energy.

[0003] However, there are many technical problems in the voltage regulation and power distribution of multi-source microgrids. Due to the large differences in the power generation characteristics of different renewable energies and the strong volatility of these energies, traditional voltage regulation methods are difficult to adapt to the dynamic changes of multi-source power generation systems, often resulting in voltage instability or unreasonable power distribution. In addition, fixed control strategies are usually adopted in the existing technologies, which are difficult to respond to complex grid fluctuations in real time. In the case of multiple renewable energy power sources coexisting, this is particularly likely to cause problems such as excessive voltage fluctuations and power distribution imbalance between sub-grids.

[0004] For example, the Chinese patent with the authorization announcement number CN110661248B discloses a multi-source DC microgrid adaptive robust power coordination and distribution method. This method realizes the effective control of the access converters by establishing a continuous and smooth non-linear equivalent resistance regulation law, calculating the stable regulation control quantity of the bus voltage and the power coordination control quantity of the access converters, so as to cope with the problems of disordered power distribution and uneven current distribution in the lines among multiple distributed power sources.

[0005] The above method has the problems proposed in this background art, that is, in a multi-source power generation microgrid system, due to the lack of accurate description of the dynamic characteristics of power sources, especially in renewable energy systems with strong volatility, the existing technologies are difficult to adapt to voltage fluctuations in real time, resulting in insufficient power distribution and voltage regulation accuracy. In addition, traditional methods lack flexible response capabilities to complex dynamic changes, cannot effectively integrate grid fluctuation information at different time scales, and the grid stability and response efficiency need to be improved. Summary of the Invention

[0006] The purpose of the present invention is to provide an adaptive voltage regulation method and system for a multi-source power generation characteristic microgrid, which can collect the physical characteristic data of each renewable energy in real time, generate a feature embedding representation, predict the voltage fluctuation signal through a multi-head attention mechanism, and dynamically adjust the voltage output of each sub-grid based on the predicted signal, so as to achieve stable regulation of the microgrid voltage.

[0007] In a first aspect, the present application provides an adaptive voltage regulation method for a multi-source power generation characteristic microgrid. The microgrid includes a plurality of sub-grids, and each sub-grid is connected to at least one renewable energy power source. The method includes:

[0008] Real-time collecting physical characteristic data of each type of the renewable energy;

[0009] Based on the physical characteristic data, generating a feature embedding representation of each type of renewable energy; wherein, the feature embedding representation is used to determine the dynamic characteristics and volatility of each renewable energy;

[0010] Inputting the feature embedding representation into a multi-head attention mechanism to generate a comprehensive voltage fluctuation prediction signal;

[0011] Based on the comprehensive voltage fluctuation prediction signal, dynamically regulating the voltage output of each sub-grid.

[0012] As an optional implementation manner, the method for generating the feature embedding representation of each type of renewable energy includes:

[0013] Real-time collecting the physical characteristic data to generate an original data set; wherein, the data in the original data set is in the form of time series data;

[0014] Using a convolutional neural network to extract spatial features in the preprocessed original data set;

[0015] Mapping the spatial features into a one-dimensional vector through a fully connected layer to form an intermediate feature representation;

[0016] Performing a non-linear transformation on the intermediate feature representation through an activation function to generate the feature embedding representation of each type of renewable energy.

[0017] As an optional implementation manner, the activation function includes: a parametric exponential linear unit;

[0018] The functional expression of the parametric exponential linear unit is:

[0019]

[0020] represents the input value, that is, each element of the intermediate feature representation after passing through the fully connected layer, is an adjustable parameter used to control the output amplitude of the activation function in the negative value region;

[0021] wherein, the adjustable parameter is determined based on the statistical characteristics of the physical characteristic data.

[0022] As an alternative implementation, the method for generating a comprehensive voltage fluctuation prediction signal includes:

[0023] Copy the feature embedding representation into multiple parallel inputs and input them into multiple attention heads respectively;

[0024] Each attention head corresponds to a preset time scale, weights the feature embedding representation, and extracts time correlation features corresponding to the time scale;

[0025] Use the self-attention mechanism to calculate the correlation between elements in the feature embedding representation and generate an attention weight matrix;

[0026] Fuse the outputs of each attention head to generate a comprehensive voltage fluctuation prediction signal.

[0027] As an alternative implementation, the fusing of the outputs of each attention head includes:

[0028] Calculate corresponding fusion weights based on the feature differences of the outputs of each attention head;

[0029] The calculation of the fusion weights includes:

[0030] Use a weight generation network to analyze the outputs of each attention head and generate dynamic fusion weights;

[0031] Among them, the weight generation network is trained based on the real-time volatility index of renewable energy to adapt to different volatility characteristics;

[0032] Multiply the outputs of each attention head by the corresponding fusion weights to obtain weighted outputs;

[0033] Sum up the weighted outputs to generate a comprehensive voltage fluctuation prediction signal.

[0034] As an alternative implementation, based on the comprehensive voltage fluctuation prediction signal, dynamically adjusting the voltage output of each sub-grid includes:

[0035] Use an adaptive control algorithm to calculate the target voltage adjustment amount of each sub-grid in real time based on the comprehensive voltage fluctuation prediction signal;

[0036] Adjust the control parameters of the power conversion device of each sub-grid according to the target voltage adjustment amount;

[0037] Monitor the actual voltage output of each sub-grid, compare it with the target voltage adjustment amount, and generate a feedback signal;

[0038] Update the parameters of the adaptive control algorithm based on the feedback signal.

[0039] As an alternative implementation, updating the parameters of the adaptive control algorithm based on the feedback signal includes:

[0040] Using the gradient descent algorithm, calculating the gradient of the loss function according to the feedback signal;

[0041] Adjusting the parameters of the adaptive control algorithm according to the gradient to minimize the voltage regulation error.

[0042] In a second aspect, the present application also provides an adaptive voltage regulation system for a multi-source power generation characteristic microgrid. The system includes: a collection unit, a first processing unit, a second processing unit, and an adjustment unit; wherein:

[0043] The collection unit is configured to collect the physical characteristic data of each type of the renewable energy in real time;

[0044] The first processing unit is configured to generate a characteristic embedding representation of each renewable energy based on the physical characteristic data; wherein, the characteristic embedding representation is used to determine the dynamic characteristics and volatility of each renewable energy;

[0045] The second processing unit is configured to input the characteristic embedding representation into a multi-head attention mechanism to generate a comprehensive voltage fluctuation prediction signal;

[0046] The adjustment unit is configured to dynamically adjust the voltage output of each sub-grid based on the comprehensive voltage fluctuation prediction signal.

[0047] In a third aspect, the present application also provides a computer device, including a processor and a memory. The memory stores machine-readable instructions executable by the processor. The processor is configured to execute the machine-readable instructions stored in the memory. When the machine-readable instructions are executed by the processor, the machine-readable instructions execute the steps in the first aspect or any possible implementation manner in the first aspect.

[0048] In a fourth aspect, the present application also provides a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is run, it executes the steps in the first aspect or any possible implementation manner in the first aspect.

[0049] Compared with the prior art, the beneficial effects of the present invention are mainly reflected in its strong adaptability and more comprehensive analysis of voltage fluctuation characteristics on multiple time scales. By collecting real-time physical characteristic data of renewable energy and using the multi-head attention mechanism to generate comprehensive voltage fluctuation prediction signals, it can dynamically adjust the voltage output according to the fluctuation characteristics of different power sources. In addition, the dynamic fusion of multi-head attention output is realized through the weight generation network, making the voltage fluctuation prediction under different time scales more accurate. The gradient descent algorithm is used for parameter optimization, enabling the adaptive control algorithm to be updated in real time, minimizing the voltage regulation error, and significantly improving the stability and regulation accuracy of the power grid. Compared with the prior art, it can better cope with the fluctuation characteristics of complex multi-source power generation systems, ensure the stability and fast response ability of the system, and can improve the voltage regulation accuracy, enhance the robustness and flexibility of the system. Brief Description of the Drawings

[0050] Figure 1 It is a flowchart of the adaptive voltage regulation method for a multi-source power generation characteristic microgrid provided by an embodiment of the present application;

[0051] Figure 2 It is a flowchart of the method for generating the characteristic embedding representation of each renewable energy provided by an embodiment of the present application;

[0052] Figure 3 It is a schematic diagram of the adaptive voltage regulation system for a multi-source power generation characteristic microgrid provided by an embodiment of the present application;

[0053] Figure 4 It is a schematic diagram of a microgrid structure provided by an embodiment of the present application.

[0054] Reference Signs: 10, acquisition unit; 20, first processing unit; 30, second processing unit; 40, regulation unit. Detailed Embodiments

[0055] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0056] Research has found that traditional microgrid voltage regulation methods usually rely on fixed rules or empirical formulas to regulate the system voltage, lacking accurate capture of the dynamic characteristics of various renewable energies. For example, solar energy and wind energy have obvious random fluctuations, and their power generation characteristics are greatly affected by environmental conditions (such as irradiance, wind speed). This uncertainty makes it difficult for traditional regulation methods to flexibly respond when facing multi-source power generation systems, and the regulation accuracy and speed are limited.

[0057] Therefore, refer to Figure 1 and Figure 4As shown Figure 1 is a flowchart of an adaptive voltage regulation method for a multi-source power generation characteristic microgrid provided by an embodiment of the present application, Figure 4 is a schematic diagram of a microgrid structure provided by an embodiment of the present application.

[0058] The microgrid includes multiple sub-grids, and each sub-grid is connected to at least one renewable energy power source; the method includes steps S101 to S104, where:

[0059] S101: Real-time collect the physical characteristic data of each type of the renewable energy.

[0060] S102: Based on the physical characteristic data, generate a characteristic embedding representation for each type of renewable energy; wherein, the characteristic embedding representation is used to determine the dynamic characteristics and volatility of each renewable energy.

[0061] S103: Input the characteristic embedding representation into a multi-head attention mechanism to generate a comprehensive voltage fluctuation prediction signal.

[0062] S104: Dynamically adjust the voltage output of each sub-grid based on the comprehensive voltage fluctuation prediction signal.

[0063] The present invention aims to solve the following technical problems: How to achieve adaptive voltage regulation in a multi-source power generation environment: Through the embedding representation of physical characteristic data, the present invention can accurately capture the fluctuation characteristics of each sub-grid in a multi-source power generation system, generate a voltage fluctuation prediction signal, and dynamically adjust the voltage output accordingly, solving the problem that traditional methods cannot adapt to the multi-source power generation environment.

[0064] How to improve the accuracy and stability of voltage regulation: The present invention analyzes and weights the characteristic embedding representations of each sub-grid through a multi-head attention mechanism to generate a high-precision voltage fluctuation prediction signal, thereby achieving more precise voltage regulation and enhancing the stability of the microgrid.

[0065] Exemplarily, as Figure 4 shown, the microgrid consists of multiple sub-grids (sub-grid 1, sub-grid 2, sub-grid 3), where sub-grid 1 is connected to source 1, sub-grid 2 is connected to source 2, and sub-grid 3 is connected to source 3, that is, each sub-grid is connected to at least one renewable energy power source. The renewable energy power sources may include: solar power generation, wind power generation, biomass power generation, etc. Each sub-grid can operate independently, and all sub-grids are connected to the main grid or other sub-grids through power electronic devices (such as inverters).

[0066] Regarding the above S101:

[0067] To achieve adaptive voltage regulation, it is first necessary to collect the physical characteristics of each renewable energy source in real time. To ensure that the dynamic characteristics and volatility of each renewable energy source are monitored in real time, the physical characteristic data of each renewable energy source can be collected through sensors. Specific physical characteristic data includes, but is not limited to: data such as irradiance, ambient temperature, component temperature, and power generation of a solar power generation system; data such as wind speed, wind direction, blade rotation speed, and power generation of a wind power generation system; data such as fuel consumption rate, power generation efficiency, and power generation of a biomass power generation system.

[0068] After these physical characteristic data are collected by sensors, they are transmitted to the central control system in real time for further analysis. The collection frequency can be set based on the volatility of the power source. For example, the characteristic data of solar energy and wind energy can be set to be updated once per second.

[0069] Regarding the above S102:

[0070] Through the analysis of the physical characteristic data, a characteristic embedding representation of each renewable energy source is generated. This characteristic embedding representation is used to quantify the dynamic characteristics and volatility of each renewable energy source.

[0071] In a specific implementation, the physical characteristic data collected in S101 can be filtered and denoised through preprocessing to eliminate possible environmental interference and acquisition errors. In addition, to eliminate the differences between different physical quantities, all data needs to be normalized so that the physical characteristics of different types of power sources can be compared under the same standard.

[0072] Based on the normalized data, specific algorithms are applied to extract the dynamic characteristics of renewable energy sources. For example, principal component analysis (PCA) or other machine learning algorithms can be used to extract the most significant features in the data, and these features can represent the change laws and volatility of each power source.

[0073] The extracted key feature data is converted into an embedding representation, that is, the physical characteristic data is compressed into a high-dimensional vector representation to summarize all the important characteristics of the renewable energy source at the current moment. This vector will be used as the input data in the subsequent steps.

[0074] Please refer to Figure 2 , Figure 2 The flowchart of the method for generating the characteristic embedding representation of each renewable energy source. As an alternative implementation, the method for generating the characteristic embedding representation of each renewable energy source includes steps S201 to S204, where:

[0075] S201: Collect the physical characteristic data in real time to generate an original data set; where the data in the original data set is in the form of time series data;

[0076] S202: Extract the spatial features from the preprocessed original dataset using a convolutional neural network;

[0077] S203: Map the spatial features into a one-dimensional vector through a fully connected layer to form an intermediate feature representation;

[0078] S204: Perform a non-linear transformation on the intermediate feature representation through an activation function to generate a feature embedding representation for each renewable energy source.

[0079] Regarding the above S201:

[0080] In a specific implementation, a sensor device with high-frequency sampling is adopted, and the specific sampling frequency can be set according to the actual application scenario. For example, for a solar panel, it can be sampled 1 - 10 times per second; for a wind turbine, higher sampling frequencies are required for the changes in wind speed and direction.

[0081] Among them, all the collected data is stored in a distributed database in the form of time series for subsequent processing and analysis. The storage structure of time series data allows for efficient historical backtracking and trend analysis of the data, and can ensure real-time performance.

[0082] It can be understood that time series data has a clear temporal relationship, which is very important in renewable energy systems. For example, the power change of solar power generation is not only related to the current irradiance but also to the irradiance change in a previous period of time. Therefore, time series data can reflect the dynamic characteristics of the energy system and provide richer background information than static data. This data form enables the model to capture short-term fluctuations and long-term trends, which is beneficial for subsequent feature extraction and voltage regulation.

[0083] Exemplarily, a low-latency communication protocol (such as Modbus, IEC 61850, etc.) can be used to transmit the collected physical characteristic data to the processing part of the main power grid in real time.

[0084] Regarding the above S202:

[0085] Use a convolutional neural network (Convolutional Neural Networks, CNN) to extract spatial features from time series data. Among them, spatial features refer to the hidden local patterns in time series data, such as the influence pattern of wind speed on power generation, the influence of irradiance on current output, etc.

[0086] Exemplarily, the specific structure of CNN includes:

[0087] Input layer: Receives the preprocessed time series data. The format of the input data is a two-dimensional array, where the rows represent different physical characteristics (such as wind speed, temperature, etc.), and the columns represent data at different time points.

[0088] Convolutional layer: The first convolutional layer can be composed of several one-dimensional convolutional kernels. The size of each convolutional kernel can be set according to the data characteristics, such as a convolutional window with a size of 5 or 7. The convolutional kernel slides through the time series to extract local patterns at each moment and within a certain time before and after it.

[0089] Activation function: The ReLU (Rectified Linear Unit) activation function is used after each convolutional layer. The main purpose is to introduce non-linearity and enhance the network's ability to recognize complex features.

[0090] Pooling layer: A max pooling layer can follow each convolutional layer for dimensionality reduction. The size of the pooling window can be set to 2 or 3, depending on the frequency and characteristics of the input data.

[0091] In this way, through the CNN, important patterns in the time series data are extracted, such as the impact of wind speed changes at specific moments on power generation output. The convolution operation enables local change features to be captured and used for subsequent analysis, providing a basis for feature embedding representation.

[0092] Regarding the above S203:

[0093] After the convolutional layer extracts the spatial features of the time series data, these high-dimensional spatial features are mapped to a one-dimensional vector through a fully connected layer.

[0094] In a specific implementation, the output of the CNN is a multi-dimensional feature map. This feature map undergoes a flattening operation to form a one-dimensional input vector, which is input into the fully connected layer. The fully connected layer consists of several neurons, and each neuron receives all the inputs from the previous layer.

[0095] The fully connected layer is embedded at the end of the entire CNN as a structure after the convolutional layer. After the convolutional layer extracts local features, the fully connected layer combines all local features through a weighted summation operation to generate a one-dimensional vector that can reflect the dynamic changes of the entire time series, achieving efficient dimensionality reduction and preparing for subsequent generation of feature embedding representation.

[0096] Regarding the above S204:

[0097] After the linear transformation of the fully connected layer, a non-linear transformation is performed through the activation function to further enhance the model's expressive power.

[0098] For example, a ReLU (rectified linear unit) activation function or a more complex PELU (parameterized exponential linear unit) activation function may be selected. For example, PELU can effectively process negative input features, making the model more flexible in processing negative inputs.

[0099] Through nonlinear activation functions, the model's ability to capture complex relationships can be enhanced. Especially in multi-source power generation systems, the dynamic changes of various energy sources are not simple linear relationships, so nonlinear transformation is crucial.

[0100] The output after activation function processing is the feature embedding representation of each renewable energy source. This embedding representation can fully reflect the dynamic characteristics of each energy system and provide accurate input basis for subsequent voltage regulation.

[0101] As an optional implementation, the activation function includes: a parameterized exponential linear unit (PELU); the function expression of the parameterized exponential linear unit is:

[0102]

[0103] represents the input value, that is, each element of the intermediate feature representation after the fully connected layer, It is an adjustable parameter used to control the output amplitude of the activation function in the negative region;

[0104] Among them, the adjustable parameters Determined based on statistical characteristics of the physical property data.

[0105] In the specific implementation, for positive input, PELU maintains linear output; for negative input, PELU generates smooth nonlinear output through exponential function. The main purpose of this design is to enhance the model's ability to handle negative input, especially in complex time series data. When the input contains more negative values, PELU can better retain negative information and prevent the gradient disappearance problem.

[0106] also, Controls the output amplitude of the activation function in the negative region. This parameter is not fixed, but is dynamically determined based on the statistical characteristics of the physical characteristics data of each renewable energy source. For example, when dealing with the irradiance and power output of a solar power generation system, It can be dynamically adjusted according to the variance or mean of the data to adapt to different volatility characteristics. Similarly, in wind power generation systems, It can be optimized based on historical data of wind speed and power fluctuations to more accurately describe the characteristics of its negative part.

[0107] Exemplarily, taking a wind power generation system as an example, the physical characteristic data includes wind speed and power generation . In order to enable the parameters of the activation function to adapt to the different fluctuations of wind speed and power, the statistical characteristics of these data can be calculated in real time to dynamically adjust the value.

[0108] For example, high-precision sensors installed on the wind turbine collect wind speed and power generation in real time, and the sampling frequency is set to once per second.

[0109] Use a sliding window of length N (for example, data in the last 60 seconds) to perform statistical analysis on the collected data.

[0110] Exemplarily, the calculation formula of can be defined as:

[0111]

[0112] where is the adjustment coefficient, and the value range can be set to , which is used to control the value of, so that it fluctuates within a reasonable range. and are the standard deviations of power generation and wind speed respectively.

[0113] When the variance of power generation is large while the variance of wind speed is small, it indicates that under the condition of little change in wind speed, the power generation has large fluctuations, and there may be abnormalities or faults. At this time, the value increases, enhancing the response of the activation function in the negative value region, making the model more sensitive to capturing abnormal features.

[0114] When the variance of wind speed is large, indicating that the wind speed itself fluctuates greatly. In order to prevent the model from overfitting to too much noise, the value decreases, reducing the output amplitude of the activation function in the negative value region, and enhancing the robustness of the model. Use the dynamically calculated value to perform non-linear transformation on at each moment in PELU.

[0115] Compared with the traditional ReLU activation function, the PELU activation function can handle negative value inputs more flexibly and adjust the parameter , Dynamically optimize the output of the activation function according to the specific physical characteristics of renewable energy, thereby enhancing the non-linear expression ability of the model. When dealing with different types of renewable energy power sources, the PELU activation function can adapt to different physical data distributions and achieve more accurate feature extraction effects. This flexibility is of great significance for the adaptive voltage regulation of microgrids in a multi-source power generation environment, ensuring that the model shows better robustness and accuracy when dealing with complex fluctuations.

[0116] In specific implementation, to determine "large variance" or "small variance", a sliding window for calculating variance can be set first (such as the most recent thirty seconds or sixty seconds), and the real-time variances of wind speed and power generation are calculated separately within each sliding window. To determine the specific threshold to judge whether a certain variance belongs to "large" or "small", one or more representative reference values can be determined based on historical operation data. For example, by statistically analyzing the variances of past wind speed and power generation, their average levels and standard deviations over a long period can be obtained, and the judgment criteria can also be selected by observing different percentiles of their distributions.

[0117] In addition, if several "variance threshold" data sets have been formed during system initialization or when historical data is relatively sufficient, after calculating the wind speed variance or power generation variance at a certain moment in real time, compare it with the pre-selected reference value: if the real-time variance is higher than the corresponding threshold, it can be determined as "large variance"; if it is lower than the threshold, it can be determined as "small variance". When the real-time variance of power generation is much higher than its threshold while the wind speed variance is lower than its threshold, it can be understood that there are large fluctuations in power generation under the condition of insignificant wind speed fluctuations, indicating that there may be system anomalies or faults, and further detection or processing is required; on the contrary, when the detected wind speed variance is much higher than the corresponding threshold, it means that the current wind speed fluctuates greatly, and it should be retained in the control algorithm to avoid overfitting or frequent adjustment in a high-fluctuation environment, thus ensuring the stability and robustness of system operation.

[0118] As an optional implementation method, the method for generating a comprehensive voltage fluctuation prediction signal includes: copying the feature embedding representation into multiple parallel inputs and respectively inputting them into multiple attention heads; each attention head corresponds to a preset time scale, performs weighted processing on the feature embedding representation, and extracts time correlation features corresponding to the time scale; uses the self-attention mechanism to calculate the correlation between elements in the feature embedding representation to generate an attention weight matrix; fuses the outputs of each attention head to generate a comprehensive voltage fluctuation prediction signal.

[0119] Regarding the above S103:

[0120] In a specific implementation, the feature embedding representations of each renewable energy source are replicated as multiple parallel inputs. These embedding representations will be separately input into multiple attention heads.

[0121] The number of attention heads can be set according to actual needs. For example, 4, 8, or 12, and the specific number can be optimized through model training.

[0122] Each attention head corresponds to a different time scale. The time scale can be set based on the volatility characteristics of renewable energy sources. For example, a short time scale can capture voltage fluctuations at the second level, while a long time scale can capture trends at the minute and hour levels.

[0123] For each specific time scale, the attention head weights the feature embedding representations through calculation to extract the correlation features at that time scale.

[0124] The core of the self-attention mechanism lies in calculating the correlation between elements in the feature embedding representation. In a specific implementation, for the input feature embedding representation y, first, three groups of vectors, namely query, key, and value, are generated through linear transformation:

[0125]

[0126] Among them, , , are the weight matrices of the query, key, and value respectively, which are determined through model training. Then, the attention scores S are calculated through the dot product of the query vector and the key vector, and the scores are normalized using the Softmax function:

[0127]

[0128] Among them, is the dimension of the query vector and the key vector.

[0129] Finally, the attention output is obtained through weighted summation:

[0130]

[0131] The output results of multiple attention heads are fused through weighting to generate a comprehensive voltage fluctuation prediction signal. The purpose of fusion is to comprehensively process information at different time scales to obtain a unified voltage fluctuation prediction signal.

[0132] As an alternative implementation, fusing the outputs of each attention head includes:

[0133] The outputs of each attention head will have certain differences due to different time scales and features. These differences are adjusted through fusion weights to ensure that the contribution of each attention head to the final output is adaptive.

[0134] In specific implementation, a weight generation network can be used to calculate the fusion weights. The weight generation network is a small neural network, and its input is the output features of each attention head.

[0135] The network dynamically generates fusion weights by analyzing the feature differences output by different attention heads. The network structure can include several fully connected layers, with a ReLU activation function connected after each layer, and the output is the fusion weight corresponding to each attention head. 。

[0136] The training process of the weight generation network is based on the real-time volatility index of renewable energy. For example, in a solar power generation system, the volatility of irradiance and power are key factors, and the training data can include the power fluctuations under different weather conditions.

[0137] The network adjusts its parameters by minimizing the voltage prediction error to ensure that the weights can be adaptively adjusted under different volatility characteristics.

[0138] The calculated fusion weights will be used to weight the outputs of each attention head. The specific weighting formula is:

[0139]

[0140] where n is the number of attention heads, represents the output of the i-th attention head.

[0141] Finally, by summing up the weighted outputs, a comprehensive voltage fluctuation prediction signal is generated. This signal will be used to dynamically regulate the voltage outputs of each sub-grid in the microgrid.

[0142] In this way, through multi-time scale feature extraction, correlation analysis of the self-attention mechanism, and dynamic fusion of the weight generation network, the accuracy of voltage fluctuation prediction is effectively improved. The strategy of fusing different time scale information ensures that the system can balance between short-term fluctuations and long-term trends, thereby improving the response speed and stability of voltage regulation. Especially the weight generation network dynamically trained based on real-time volatility further enhances the adaptability of the system under multi-source power generation conditions, making the voltage regulation more flexible and accurate.

[0143] As an alternative implementation, dynamically adjusting the voltage output of each sub-grid based on the comprehensive voltage fluctuation prediction signal includes: using an adaptive control algorithm to calculate the target voltage adjustment amount of each sub-grid in real time based on the comprehensive voltage fluctuation prediction signal; adjusting the control parameters of the power conversion devices of each sub-grid according to the target voltage adjustment amount; monitoring the actual voltage output of each sub-grid, comparing it with the target voltage adjustment amount, and generating a feedback signal; updating the parameters of the adaptive control algorithm based on the feedback signal.

[0144] As an alternative implementation, updating the parameters of the adaptive control algorithm based on the feedback signal includes: using the gradient descent algorithm to calculate the gradient of the loss function according to the feedback signal; adjusting the parameters of the adaptive control algorithm according to the gradient to minimize the voltage regulation error.

[0145] In a specific implementation, receive the comprehensive voltage fluctuation prediction signal generated by the multi-head attention mechanism, which can reflect the possible voltage fluctuation trend of the micro-grid in a future period of time.

[0146] Based on this signal, use an adaptive control algorithm (e.g., adaptive PID control algorithm, adaptive fuzzy control algorithm) to calculate the target voltage adjustment amount of each sub-grid in real time. The target voltage adjustment amount is used to indicate the voltage output that each sub-grid needs to increase or decrease.

[0147] According to the calculated target voltage adjustment amount, adjust the control parameters of the power conversion devices (such as inverters, transformers) of each sub-grid in real time.

[0148] Among them, the control parameters may include the output voltage amplitude or phase angle of the inverter, the tap position of the transformer, etc. The adjustment of these control parameters directly affects the voltage output of each sub-grid to reach the desired voltage level.

[0149] In a specific implementation, the actual voltage output of each sub-grid can be monitored in real time through voltage sensors installed in each sub-grid.

[0150] Compare the actual voltage output with the target voltage adjustment amount, calculate the voltage deviation, and this voltage deviation is the feedback signal.

[0151] Update the parameters of the adaptive control algorithm dynamically according to the above feedback signal.

[0152] Among them, the update of the adaptive control algorithm can be realized by the gradient descent algorithm. For example, the loss function is defined as the square of the voltage regulation error:

[0153]

[0154] Among them, represents the parameter set of the adaptive control algorithm, is the voltage deviation of each sub-grid.

[0155] By calculating the gradient of the loss function with respect to the control parameter the direction of parameter adjustment can be determined. The gradient calculation formula is: Based on the calculated gradient, the parameters of the adaptive control algorithm are updated using the gradient descent algorithm to minimize the loss function

[0156]

[0157] where, is the learning rate, which is used to control the step size of each parameter update.

[0158]

[0159] This step gradually reduces the voltage regulation error by repeatedly adjusting the parameters of the adaptive control algorithm through iteration, ensuring that the difference between the actual voltage and the predicted value of each sub-grid is minimized. is the learning rate, which is used to control the step size of each parameter update.

[0160] Through this mechanism, the target voltage adjustment amount of each sub-grid can be calculated according to the real-time voltage fluctuation situation by using the adaptive control algorithm, and the control parameters can be dynamically updated by monitoring the feedback signal, ensuring the accuracy and response speed of voltage regulation. The gradient descent algorithm is used to continuously optimize the parameters of the control algorithm, reduce the voltage regulation error, and make the system show higher robustness and adaptability in a complex multi-source power generation environment. This mechanism can significantly improve the voltage stability of the micro-grid and ensure the efficient and stable operation of the power system.

[0161] Those skilled in the art can understand that in the above method of the specific implementation manner, the writing order of each step does not mean a strict execution order that constitutes any limitation to the implementation process, and the specific execution order of each step should be determined according to its function and possible internal logic.

[0162] Based on the same inventive concept, an adaptive voltage regulation system for a multi-source power generation characteristic micro-grid is also provided in the embodiments of the present application. Since the principle of solving problems by the system in the embodiments of the present application is similar to that of the above-mentioned adaptive voltage regulation method for a multi-source power generation characteristic micro-grid in the embodiments of the present application, the implementation of the system can refer to the implementation of the method, and the repeated parts will not be elaborated.

[0163] Based on the same inventive concept, an adaptive voltage regulation system for a multi-source power generation characteristic micro-grid is also provided in the embodiments of the present application. Since the principle of solving problems by the system in the embodiments of the present application is similar to that of the above-mentioned adaptive voltage regulation method for a multi-source power generation characteristic micro-grid in the embodiments of the present application, the implementation of the system can refer to the implementation of the method, and the repeated parts will not be elaborated.

[0164] Refer to Figure 3As shown in the figure, it is a schematic diagram of an adaptive voltage regulation system for a multi-source power generation characteristic microgrid provided by an embodiment of the present application. The microgrid includes multiple sub-grids, and each sub-grid is connected to at least one renewable energy power source. The system includes: a collection unit 10, a first processing unit 20, a second processing unit 30, and a regulation unit 40; where:

[0165] The collection unit 10 is configured to collect the physical characteristic data of each type of the renewable energy in real time;

[0166] The first processing unit 20 is configured to generate a characteristic embedding representation of each renewable energy based on the physical characteristic data; wherein, the characteristic embedding representation is used to determine the dynamic characteristics and volatility of each renewable energy;

[0167] The second processing unit 30 is configured to input the characteristic embedding representation into a multi-head attention mechanism to generate a comprehensive voltage fluctuation prediction signal;

[0168] The regulation unit 40 is configured to dynamically regulate the voltage output of each sub-grid based on the comprehensive voltage fluctuation prediction signal.

[0169] For the description of the processing flow of each unit in the system and the interaction flow between each unit, reference can be made to the relevant description in the above method embodiment, which will not be elaborated here.

[0170] An embodiment of the present application also provides a computer device, including:

[0171] A processor and a memory; the memory stores machine-readable instructions executable by the processor, and the processor is configured to execute the machine-readable instructions stored in the memory. When the machine-readable instructions are executed by the processor, the processor executes the following steps:

[0172] Collect the physical characteristic data of each type of the renewable energy in real time;

[0173] Generate a characteristic embedding representation of each renewable energy based on the physical characteristic data; wherein, the characteristic embedding representation is used to determine the dynamic characteristics and volatility of each renewable energy;

[0174] Input the characteristic embedding representation into a multi-head attention mechanism to generate a comprehensive voltage fluctuation prediction signal;

[0175] Dynamically regulate the voltage output of each sub-grid based on the comprehensive voltage fluctuation prediction signal.

[0176] The above-mentioned memory includes an internal memory and an external memory; here, the internal memory is also called the main memory, which is used to temporarily store the operation data in the processor and the data exchanged with the external memory such as the hard disk. The processor exchanges data with the external memory through the internal memory.

[0177] The specific execution process of the above instructions can refer to the steps of the adaptive voltage regulation method for a multi-source power generation characteristic microgrid described in the embodiments of this application, which will not be elaborated here.

[0178] The embodiments of this application also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the steps of the adaptive voltage regulation method for a multi-source power generation characteristic microgrid described in the above method embodiments. Among them, the storage medium can be a volatile or non-volatile computer-readable storage medium.

[0179] The embodiments of this application also provide a computer program product, which carries program codes. The instructions included in the program codes can be used to execute the steps of the adaptive voltage regulation method for a multi-source power generation characteristic microgrid described in the above method embodiments. Specifically, reference can be made to the above method embodiments, which will not be elaborated here.

[0180] Among them, the above computer program product can be specifically implemented by means of hardware, software, or a combination thereof. In an alternative embodiment, the computer program product is specifically embodied as a computer storage medium. In another alternative embodiment, the computer program product is specifically embodied as a software product, such as a Software Development Kit (SDK), etc.

[0181] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems and devices described above can refer to the corresponding processes in the foregoing method embodiments, which will not be elaborated here. In the several embodiments provided by this disclosure, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. Also, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some communication interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0182] When the above-mentioned functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium executable by a processor. Based on such an understanding, the technical solution of the present disclosure, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present disclosure. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs that can store program codes.

[0183] Finally, it should be noted that the above-mentioned embodiments are only specific implementation manners of the present disclosure, used to illustrate the technical solutions of the present disclosure, rather than limiting them. The protection scope of the present disclosure is not limited thereto. Although the present disclosure has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present disclosure can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims

1. An adaptive voltage regulation method for a microgrid with multi-source generation characteristics, characterized in that: The microgrid comprises a plurality of subgrids, each subgrid being connected to at least one renewable energy power generation source, and the method comprises: Collecting physical property data of each renewable energy source in real time; Based on the physical property data, generating a feature embedding representation of each renewable energy source; wherein the feature embedding representation is used to determine the dynamic characteristics and volatility of each renewable energy source; The feature embedding representation is input into a multi-head attention mechanism to generate a comprehensive voltage fluctuation prediction signal; Dynamically adjusting the voltage output of each sub-grid based on the comprehensive voltage fluctuation prediction signal; Wherein, the method for generating a comprehensive voltage fluctuation prediction signal comprises: The feature embedding representation is copied into multiple parallel inputs and input into multiple attention heads respectively; Each attention head corresponds to a preset time scale, performs weighted processing on the feature embedding representation, and extracts the time correlation features of the corresponding time scale; The self-attention mechanism is used to calculate the correlation between the elements in the feature embedding representation and generate the attention weight matrix; The outputs of each attention head are fused to generate a comprehensive voltage fluctuation prediction signal; The fusing of the outputs of the attention heads includes: Based on the feature differences output by each attention head, the corresponding fusion weights are calculated; The calculation of the fusion weight includes: Use the weight generation network to analyze the output of each attention head and generate dynamic fusion weights; Wherein, the weight generation network is trained based on the real-time volatility index of renewable energy to adapt to different volatility characteristics; Multiply the output of each attention head by the corresponding fusion weight to get the weighted output; The weighted outputs are summed to generate a comprehensive voltage fluctuation prediction signal; The physical property data include: irradiance, ambient temperature, component temperature, and power generation data of solar power generation system; wind speed, wind direction, blade speed, and power generation data of wind power generation system; fuel consumption rate, power generation efficiency, and power generation data of biomass power generation system; The method for generating a feature embedding representation of each renewable energy source comprises: Collecting the physical property data in real time to generate an original data set; wherein the data in the original data set is in the form of time series data; Extracting spatial features from the preprocessed raw data set using a convolutional neural network; Mapping the spatial features into a one-dimensional vector through a fully connected layer to form an intermediate feature representation; The intermediate feature representation is nonlinearly transformed through an activation function to generate a feature embedding representation of each renewable energy source; The activation function includes: a parameterized exponential linear unit; The function expression of the parameterized exponential linear unit is: ； represents the input value, that is, each element of the intermediate feature representation after the fully connected layer, It is an adjustable parameter used to control the output amplitude of the activation function in the negative region; Among them, the adjustable parameters Determining based on statistical characteristics of the physical property data; Adjustable parameters The dynamic adjustment is used to make the parameters of the activation function Adapt to the different volatility of the physical characteristics data of renewable energy.

2. The adaptive voltage regulation method for a multi-source power generation characteristic microgrid according to claim 1, characterized in that: Based on the comprehensive voltage fluctuation prediction signal, dynamically adjusting the voltage output of each sub-grid includes: Using an adaptive control algorithm, based on the comprehensive voltage fluctuation prediction signal, a target voltage adjustment amount of each sub-grid is calculated in real time; According to the target voltage adjustment amount, adjusting the control parameters of the power conversion device of each sub-grid; Monitor the actual voltage output of each sub-grid, compare it with the target voltage adjustment amount, and generate a feedback signal; Based on the feedback signal, parameters of the adaptive control algorithm are updated.

3. The adaptive voltage regulation method for a multi-source power generation characteristic microgrid according to claim 2, characterized in that: Based on the feedback signal, updating the parameters of the adaptive control algorithm includes: Calculating the gradient of the loss function according to the feedback signal using a gradient descent algorithm; Based on the gradient, parameters of an adaptive control algorithm are adjusted to minimize the voltage regulation error.

4. An adaptive voltage regulation system for a microgrid with multi-source power generation characteristics, used to implement the adaptive voltage regulation method for a microgrid with multi-source power generation characteristics as claimed in any one of claims 1 to 3, characterized in that: The system comprises: a collection unit, a first processing unit, a second processing unit, and an adjustment unit; wherein: The collection unit is used to collect physical characteristic data of each renewable energy source in real time; The first processing unit is used to generate a feature embedding representation of each renewable energy source based on the physical property data; wherein the feature embedding representation is used to determine the dynamic characteristics and volatility of each renewable energy source; The second processing unit is used to embed the feature into a multi-head attention mechanism to generate a comprehensive voltage fluctuation prediction signal; The regulating unit is used to dynamically regulate the voltage output of each sub-grid based on the comprehensive voltage fluctuation prediction signal.

5. A computer device, characterized in that: include: A processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the processor is used to execute the machine-readable instructions stored in the memory. When the machine-readable instructions are executed by the processor, the processor executes the steps of the adaptive voltage regulation method for a microgrid with multi-source power generation characteristics as described in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program. When the computer program is executed by a computer device, the computer device executes the steps of the adaptive voltage regulation method for a microgrid with multi-source power generation characteristics as described in any one of claims 1 to 3.

Images