Distribution network power supply method and device based on machine learning

Abstract

The invention provides a distribution network power supply method and device based on machine learning, and relates to the field of distribution network power supply. The method comprises the following steps: acquiring historical electricity consumption information of a target user in a first preset time period in response to a distribution network power supply operation for the target user; obtaining a user electricity consumption feature set in the historical electricity consumption information through a first preprocessing operation; obtaining historical power supply data, and obtaining a historical power supply feature set in the historical power supply data through a second preprocessing operation; performing feature fusion on the user power utilization feature set and the historical power supply feature set through third preprocessing operation, and constructing a comprehensive power utilization feature set; constructing an intelligent distribution network power supply model according to the comprehensive power utilization feature set; and obtaining the predicted power consumption demand of the target user in the second preset time period through the intelligent distribution network power supply model. According to the method and the device, the problem that the rapid change of the power utilization demand of the user cannot be flexibly handled due to the lack of deep analysis on the real-time power utilization dynamic state and consideration on the personalized demand of the user is solved.

Description

Technical Field

[0001] The present application relates to the field of power distribution networks, and in particular to a method and device for power distribution networks based on machine learning. Background Art

[0002] With the development of social economy and the improvement of residents' living standards, the demand for electricity is growing, and the requirements for distribution network power supply systems are getting higher and higher. The distribution network power supply system not only needs to meet the increasingly complex load requirements, but also needs to maintain stable operation in response to changes in external factors such as the environment and climate. At the same time, the rise of emerging power modes such as the access of new energy and distributed generation has also increased the operational complexity of the distribution network power supply system. Therefore, in order to improve the reliability and efficiency of power supply, traditional load control methods are facing huge challenges.

[0003] Traditional distribution network power supply methods mainly rely on fixed load control strategies and empirical judgments. The above methods usually make simple load forecasts based on historical electricity consumption data, and then provide distribution network power supply to users based on the predicted results. However, using the above methods alone, due to the lack of in-depth analysis of real-time electricity consumption dynamics and consideration of user personalized needs, there is a problem of being unable to flexibly respond to rapid changes in user electricity demand.

[0004] Therefore, there is an urgent need for a distribution network power supply method and device based on machine learning. Summary of the invention

[0005] The present application provides a distribution network power supply method and device based on machine learning, which solves the problem of being unable to flexibly respond to rapid changes in user power demand due to a lack of in-depth analysis of real-time power consumption dynamics and consideration of user personalized needs.

[0006] In a first aspect of the present application, a distribution network power supply method based on machine learning is provided, the method comprising: in response to a distribution network power supply operation for a target user, obtaining historical power consumption information of the target user in a first preset time period; obtaining a user power consumption feature set in the historical power consumption information through a first preprocessing operation, the user power consumption feature set including user power consumption time features, user type features, user power consumption peak period features, and user power consumption change trend features; obtaining historical power supply data, and obtaining a historical power supply feature set in the historical power supply data through a second preprocessing operation, the historical power supply feature set including power supply quality index features and power supply quantity features; fusing the user power consumption feature set and the historical power supply feature set through a third preprocessing operation, and constructing a comprehensive power consumption feature set; constructing an intelligent distribution network power supply model based on the comprehensive power consumption feature set; obtaining the predicted power demand of the target user in the second preset time period through the intelligent distribution network power supply model, and performing a distribution network power supply operation based on the predicted power demand.

[0007] Optionally, a user electricity consumption feature set in historical electricity consumption information is obtained through a first preprocessing operation, specifically including: obtaining a user type corresponding to a target user, and constructing a user type feature matrix, where the user types include residential type users, commercial type users, and industrial type users; obtaining the electricity consumption time distribution corresponding to different user types, and constructing an electricity consumption time feature vector corresponding to the target user through time series analysis; obtaining the electricity consumption of different user types in a corresponding preset time period, where the preset time period is a peak electricity consumption time period, and constructing a peak electricity consumption feature vector corresponding to the target user based on the electricity consumption in the preset time period; and performing feature fusion on the user electricity consumption feature set, the electricity consumption time feature vector, and the peak electricity consumption feature vector through the first preprocessing operation to obtain the user electricity consumption feature set.

[0008] Optionally, a historical power supply feature set in the historical power supply data is obtained through a second preprocessing operation, specifically including: obtaining historical voltage fluctuation values, historical power supply frequency change values ​​and historical harmonic frequencies in a first preset time period; constructing a historical power supply quality feature vector based on the voltage fluctuation values, historical power supply frequency change values ​​and historical harmonic frequencies; obtaining the historical total power supply and historical peak power supply in the first preset time period; constructing a historical power supply feature vector based on the historical total power supply and historical peak power supply; and performing feature fusion on the historical power supply quality feature vector and the historical power supply feature vector through a second preprocessing operation to obtain a historical power supply feature set.

[0009] Optionally, the user's electricity consumption feature set and the historical power supply feature set are feature fused through a third preprocessing operation, and a comprehensive electricity consumption feature set is constructed, specifically including: obtaining the spatiotemporal mapping relationship between the user's electricity consumption feature set and the historical power supply feature set; based on the spatiotemporal mapping relationship, the user's electricity consumption feature set and the historical power supply feature set are feature fused through a third preprocessing operation, and a comprehensive electricity consumption feature set is constructed.

[0010] Optionally, obtaining the predicted power demand of the target user in the second preset time period through the smart distribution network power supply model specifically includes: calculating the predicted power demand of the target user through the following formula:

[0011]

[0012] Where P(t) is the predicted electricity demand at time t, time t is any time in the second preset time period, f is the preset activation function, U i (t') is the i-th feature in the electricity consumption feature set, w i is the feature weight corresponding to the i-th feature, S j (t') is the jth feature in the historical power supply feature set, v jis the feature weight corresponding to the jth feature, C(t') is the time-space mapping relationship, γ is the weight of the time-space mapping relationship, and t' is the moment corresponding to moment t in the first preset time period.

[0013] Optionally, obtaining the spatiotemporal mapping relationship between the user power consumption feature set and the historical power supply feature set specifically includes: obtaining the spatiotemporal mapping relationship according to the following formula:

[0014]

[0015] Wherein, exp is a preset exponential function, which is used to capture the time decay effect, τ is the time point associated with the user's power consumption characteristics, σ is the standard deviation of the control Gaussian function, and is used to represent the speed of time decay. Indicates the degree of periodic impact of the j-th historical power supply characteristics.

[0016] Optionally, after obtaining the predicted electricity demand of the target user in the second preset time period through the intelligent distribution network power supply model and performing the distribution network power supply operation according to the predicted electricity demand, the method also includes: obtaining the actual electricity demand of the target user in the second preset time period; calculating whether the electricity difference between the actual electricity demand and the predicted electricity demand is greater than the preset difference; if the electricity difference is greater than the preset difference, obtaining multiple prediction error factors; and improving the intelligent distribution network power supply model according to the multiple prediction error factors.

[0017] In a second aspect of the present application, a distribution network power supply device based on machine learning is provided, the device comprising an acquisition module, a preprocessing module and a prediction module, wherein:

[0018] The acquisition module is used to obtain historical power consumption information of the target user in a first preset time period in response to the power distribution network power supply operation for the target user.

[0019] A preprocessing module is used to obtain a user power consumption feature set in historical power consumption information through a first preprocessing operation, the user power consumption feature set including user power consumption time features, user type features, user power consumption peak period features and user power consumption change trend features; obtain historical power supply data, and obtain a historical power supply feature set in the historical power supply data through a second preprocessing operation, the historical power supply feature set including power supply quality index features and power supply quantity features; and fuse the user power consumption feature set and the historical power supply feature set through a third preprocessing operation, and construct a comprehensive power consumption feature set.

[0020] The prediction module is used to build an intelligent distribution network power supply model based on the comprehensive power consumption feature set; obtain the predicted power demand of the target user in the second preset time period through the intelligent distribution network power supply model, and perform distribution network power supply operations according to the predicted power demand.

[0021] In the third aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes any one of the methods described above.

[0022] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to perform any of the above methods.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0024] 1. Obtain historical electricity consumption information of target users in a first preset time period, and obtain the user electricity consumption feature set in the historical electricity consumption information through a first preprocessing operation, obtain historical power supply data, and obtain the historical power supply feature set in the historical power supply data through a second preprocessing operation, based on feature extraction and fusion of historical electricity consumption data and power supply data, thereby forming a comprehensive electricity consumption feature set, and through the comprehensive electricity consumption feature set, predict the user's electricity demand in a specific time period, i.e., a second preset time period, according to the intelligent distribution network power supply model, so as to be able to adjust the distribution network power supply in real time and efficiently, and thus flexibly respond to the rapid changes in user electricity demand.

[0025] 2. Obtain the spatiotemporal mapping relationship between the user's electricity consumption feature set and the historical power supply feature set, and based on the spatiotemporal mapping relationship, perform feature fusion on the user's electricity consumption feature set and the historical power supply feature set through a third preprocessing operation, and construct the comprehensive electricity consumption feature set, thereby helping the intelligent distribution network power supply model to more accurately understand the mutual influence between user electricity consumption behavior and power supply conditions under different time and space conditions, and improve prediction accuracy.

[0026] 3. Obtain the actual electricity demand of the target user in the second preset time period, and calculate whether the difference between the actual electricity demand and the predicted electricity demand is greater than the preset difference. When the electricity difference is greater than the preset difference, obtain multiple prediction error factors, so as to improve the intelligent distribution network power supply model according to the multiple prediction error factors, thereby enhancing the adaptability of the intelligent distribution network power supply model, making the prediction of the future intelligent distribution network power supply model more accurate, and achieving more efficient and reliable distribution network power supply. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flow chart of a distribution network power supply method based on machine learning provided in an embodiment of the present application;

[0028] Figure 2 It is a module schematic diagram of a distribution network power supply device based on machine learning provided in an embodiment of the present application;

[0029] Figure 3 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application.

[0030] Explanation of the reference numerals: 21, acquisition module; 22, preprocessing module; 23, prediction module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. DETAILED DESCRIPTION

[0031] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0032] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more listed items.

[0033] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0034] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0035] With the development of social economy and the improvement of residents' living standards, the demand for electricity is growing, and the requirements for distribution network power supply systems are getting higher and higher. The distribution network power supply system not only needs to meet the increasingly complex load requirements, but also needs to maintain stable operation in response to changes in external factors such as the environment and climate. At the same time, the rise of emerging power modes such as the access of new energy and distributed generation has also increased the operational complexity of the distribution network power supply system. Therefore, in order to improve the reliability and efficiency of power supply, traditional load control methods are facing huge challenges. .

[0036] Traditional distribution network power supply methods mainly rely on fixed load control strategies and empirical judgments. The above methods usually make simple load forecasts based on historical electricity consumption data, and then provide distribution network power supply to users based on the predicted results. However, using the above methods alone, due to the lack of in-depth analysis of real-time electricity consumption dynamics and consideration of user personalized needs, there is a problem of being unable to flexibly respond to rapid changes in user electricity demand.

[0037] Therefore, there is an urgent need for a distribution network power supply method and device based on machine learning.

[0038] Please refer to Figure 1 , which shows a flow chart of a distribution network power supply method based on machine learning provided in an embodiment of the present application, and the flow chart mainly includes the following steps: S101 to S106.

[0039] Step S101, in response to a power distribution network power supply operation for a target user, obtaining historical power consumption information of the target user in a first preset time period.

[0040] Specifically, the distribution network power supply, as the last link of the power system, directly connects the user and is responsible for the final power distribution. In the distribution network power supply operation, the target user can be any of different types of users such as residential users, commercial users and industrial users. The historical power consumption information of the target user can be obtained through power consumption monitoring equipment, and the power consumption monitoring equipment includes but is not limited to smart meters, load monitoring terminals, data collectors, etc. The embodiments of this application do not limit the method of obtaining historical power consumption information.

[0041] Step S102, obtaining a user power usage feature set in the historical power usage information through a first preprocessing operation, wherein the user power usage feature set includes user power usage time features, user type features, user power usage peak period features, and user power usage change trend features.

[0042] Specifically, the user type corresponding to the target user is obtained, and a user type feature matrix is ​​constructed. The user types include residential type users, commercial type users, and industrial type users. When constructing the user type feature matrix, more segmentation features can be considered, such as the user's scale (small, medium, large), the type of electrical equipment (such as household appliances, office equipment, industrial machinery), etc.; the electricity consumption time distribution corresponding to different user types is obtained, and the electricity consumption time feature vector corresponding to the target user is constructed through time series analysis. The electricity consumption time feature vector refers to the user's electricity consumption habits in different time periods, which can be further subdivided into weekday electricity consumption time characteristics and weekend electricity consumption time characteristics; the electricity consumption of different user types in the corresponding preset time period is obtained. The preset time period is the peak electricity consumption time period, and the peak electricity consumption feature vector corresponding to the target user is constructed based on the electricity consumption in the preset time period. The peak period of electricity consumption can be identified based on the results of historical data analysis, which may be morning, lunch time or evening. The peak period feature vector is formed by integrating the electricity consumption data in different time periods to reflect the user's electricity consumption intensity and pattern during the peak period. The user's electricity consumption feature set, the electricity consumption time feature vector and the electricity peak period feature vector are subjected to feature fusion through the first preprocessing operation to obtain the user's electricity consumption feature set. The first preprocessing operation can use principal component analysis (PCA) and feature selection algorithms (such as LASSO, random forest, etc.) to select the most representative features to improve the accuracy and robustness of the model. Feature fusion can be performed through weighted averaging, feature splicing or through deep learning models (such as neural networks). The fused feature set will more comprehensively reflect the comprehensive characteristics of the user's electricity consumption, providing a more reliable foundation for subsequent analysis and model building.

[0043] Step S103, acquiring historical power supply data, and acquiring a historical power supply feature set in the historical power supply data through a second preprocessing operation, wherein the historical power supply feature set includes power supply quality index features and power supply quantity features.

[0044] Specifically, the historical voltage fluctuation value, the historical power supply frequency change value and the historical harmonic frequency in the first preset time period are obtained; according to the voltage fluctuation value, the historical power supply frequency change value and the historical harmonic frequency, a historical power supply quality feature vector is constructed, wherein, for the voltage fluctuation value, the voltage fluctuation value includes the standard deviation and fluctuation amplitude of the voltage, the standard deviation of the voltage can be calculated by the historical voltage data, and the fluctuation amplitude can be evaluated by the difference between the maximum voltage and the minimum voltage. For the historical power supply frequency change value, the historical power supply frequency change value can be calculated by analyzing the historical power supply frequency data, calculating the average value, standard deviation and fluctuation of the frequency. For the historical harmonic frequency, the harmonic is the frequency component higher than the fundamental frequency in the current and voltage waveforms, which is usually caused by nonlinear loads, so it can be calculated by analyzing the amplitude and number of harmonics; the historical total power supply and the historical peak power supply in the first preset time period are obtained, wherein the historical total power supply is the total power provided to the target user in the first preset time period, and the first preset time period can be a fixed period in the past, such as one day, one week or one month, and the specific selection should be based on business needs and data availability. The historical peak power supply is the maximum power supply within the first preset time period; a historical power supply feature vector is constructed based on the historical total power supply and the historical peak power supply; and the historical power supply quality feature vector and the historical power supply feature vector are feature fused through a second preprocessing operation to obtain a historical power supply feature set. Similarly, the second preprocessing operation can use weighted averaging, principal component analysis (PCA) and other techniques to fuse the historical power supply quality feature vector and the power supply feature vector to extract the most representative features.

[0045] Step S104: The user's power consumption feature set and the historical power supply feature set are subjected to feature fusion through a third preprocessing operation, and a comprehensive power consumption feature set is constructed.

[0046] Specifically, the spatiotemporal mapping relationship between the user's power consumption feature set and the historical power supply feature set is obtained; according to the spatiotemporal mapping relationship, the user's power consumption feature set and the historical power supply feature set are feature-fused through the third preprocessing operation, and a comprehensive power consumption feature set is constructed. The spatiotemporal mapping relationship refers to the correspondence between the user's power consumption features and the historical power supply features, considering the time and space dimensions. For the time dimension: before feature fusion, it is necessary to ensure that the data time periods of the user's power consumption features and the historical power supply features are consistent. This can be achieved by selecting the same time interval (such as hours, days, weeks, etc.); for the space dimension: if there are multiple target users or multiple power supply areas, it is necessary to consider how to match the power consumption features of different target users with the corresponding power supply features to ensure the effectiveness of feature fusion. Among them, the spatiotemporal mapping relationship can be obtained according to the following formula:

[0047]

[0048] Wherein, exp is a preset exponential function, which is used to capture the time decay effect, τ is the time point associated with the user's power consumption characteristics, σ is the standard deviation of the control Gaussian function, and is used to represent the speed of time decay. represents the periodic impact degree of the jth historical power supply characteristics, ω j is the frequency of the jth historical power supply feature, The phase of the jth historical power supply characteristic.

[0049] Step S105: constructing a smart distribution network power supply model based on the comprehensive power consumption feature set.

[0050] Specifically, the main objectives of the model are clarified, including but not limited to: predicting the power demand of users, optimizing the power supply efficiency of the distribution network, improving the power supply quality, etc. According to the modeling objectives, multiple model types can be selected. Divide the comprehensive power consumption feature set into a training set and a test set. For example, 70%-80% of the data can be used for training and 20%-30% of the data can be used for testing. Use the training set data to train the selected model, and adjust the model parameters through the optimization algorithm to minimize the prediction error. Use the cross-validation method to evaluate the performance of the model to avoid overfitting and improve the generalization ability of the model. According to the selected model type, select appropriate evaluation indicators, such as: regression model: root mean square error (RMSE), mean absolute error (MAE), etc. Classification model: accuracy, recall rate, F1-score, etc. Use the test set to verify the trained model, and then adjust the model's hyperparameters based on the evaluation results to optimize the model performance. Deploy the trained smart distribution network power supply model to the actual application environment to ensure that the smart distribution network power supply model can access real-time data, perform dynamic prediction and adjustment, and adapt to changing power supply needs.

[0051] Step S106, obtaining the predicted power demand of the target user in the second preset time period through the intelligent distribution network power supply model, and performing the distribution network power supply operation according to the predicted power demand.

[0052] Specifically, the relevant features in the comprehensive power consumption feature set are input into the smart distribution network power supply model. These features should be closely related to the historical power consumption behavior and power supply characteristics of the target user. The target user's predicted power demand can be calculated by the following formula:

[0053]

[0054] Where P(t) is the predicted electricity demand at time t, time t is any time in the second preset time period, f is the preset activation function, U i (t') is the i-th feature in the electricity consumption feature set, w i is the feature weight corresponding to the i-th feature, S j(t') is the jth feature in the historical power supply feature set, v j is the feature weight corresponding to the jth feature, C(t') is the time-space mapping relationship, γ is the weight of the time-space mapping relationship, and t' is the moment corresponding to moment t in the first preset time period.

[0055] In a possible implementation, step S106 also includes: obtaining the actual electricity demand of the target user in a second preset time period; calculating whether the electricity demand difference between the actual electricity demand and the predicted electricity demand is greater than the preset difference; if the electricity demand difference is greater than the preset difference, obtaining multiple prediction error factors; and improving the smart distribution network power supply model based on the multiple prediction error factors.

[0056] Specifically, in the second preset time period, the actual power consumption data of the target user is monitored and recorded in real time, which is a time period corresponding to the first preset time period. For example, if the first preset time period is a specific time period in the past week, such as 8:00 to 10:00 every Monday in the past, the second preset time period can be set to the same time period in the current week, such as 8:00 to 10:00 on the current Monday; if the first preset time period is the first full week in a month, the second preset time period can be the same time period in the next month, etc. The power difference is calculated by comparing the predicted power demand with the actual power demand, D(t)=|P(t)-A(t)|, where D(t) is the power difference, P(t) is the predicted power demand, A(t) is the actual power demand, and the preset difference threshold is obtained to determine whether D(t) is greater than the preset difference. If the power difference D(t) is greater than the preset difference, it indicates that the prediction is inaccurate and further analysis is required. At this time, all input features and their weights at this moment should be recorded for subsequent analysis. Analyze potential factors that lead to forecast differences, including but not limited to: Feature correlation: Check which changes in input features are most relevant to changes in actual electricity demand. External factors: Consider the impact of external factors such as weather changes, holidays, and emergencies on electricity consumption. Adjust the smart distribution network power supply model based on the analysis results, including but not limited to: Feature selection: Add or remove features with greater influence to improve the model's predictive ability; Parameter optimization: Readjust model parameters to improve the model's accuracy through retraining; Model retraining: After collecting enough new data, retrain the model to ensure that it adapts to new electricity consumption patterns and environmental changes. Establish a feedback mechanism to continuously monitor the difference between forecast results and actual electricity demand, so as to continuously adjust and optimize the smart distribution network power supply model to maintain its accuracy in long-term operation. It should be noted that the preset difference can be set according to multiple factors, including but not limited to: user power consumption characteristics: different users' power consumption habits and volatility may affect the selection of the preset difference, for example, industrial users' power consumption fluctuations may be large, while residential users are relatively stable; historical data analysis: through statistical analysis of historical power consumption data, determine the common fluctuation range of user power demand, so as to set a reasonable difference threshold; real-time monitoring capability: according to the accuracy and response time of the actual monitoring equipment, adjust the preset difference to better reflect the changes in real-time data; business needs: set according to business goals and service level requirements, for example, stricter difference limits may be required during critical periods to ensure the reliability of power supply, etc. This embodiment does not limit the setting of the preset difference.

[0057] The present application obtains the historical electricity consumption information of the target user in the first preset time period by adopting the above method, and obtains the user's electricity consumption feature set in the historical electricity consumption information through a first preprocessing operation, obtains the historical power supply data, and obtains the historical power supply feature set in the historical power supply data through a second preprocessing operation, based on the feature extraction and fusion of the historical electricity consumption data and the power supply data, thereby forming a comprehensive electricity consumption feature set, and through the comprehensive electricity consumption feature set, according to the intelligent distribution network power supply model, predicts the user's electricity demand in a specific time period, i.e., the second preset time period, so as to adjust the distribution network power supply in real time and efficiently, and thus flexibly respond to the rapid changes in user electricity demand.

[0058] Please refer to Figure 2 , which shows a module schematic diagram of a distribution network power supply device based on machine learning provided in an embodiment of the present application, the device includes an acquisition module 21, a preprocessing module 22 and a prediction module 23, wherein,

[0059] The acquisition module 21 is used to acquire historical power consumption information of the target user in a first preset time period in response to the power distribution network power supply operation for the target user.

[0060] The preprocessing module 22 is used to obtain a user power consumption feature set in the historical power consumption information through a first preprocessing operation, the user power consumption feature set including user power consumption time features, user type features, user power consumption peak period features and user power consumption change trend features; obtain historical power supply data, and obtain a historical power supply feature set in the historical power supply data through a second preprocessing operation, the historical power supply feature set including power supply quality index features and power supply quantity features; and fuse the user power consumption feature set and the historical power supply feature set through a third preprocessing operation, and construct a comprehensive power consumption feature set.

[0061] The prediction module 23 is used to construct an intelligent distribution network power supply model based on the comprehensive power consumption feature set; obtain the predicted power demand of the target user in the second preset time period through the intelligent distribution network power supply model, and perform distribution network power supply operations according to the predicted power demand.

[0062] In one possible implementation, the preprocessing module 22 is used to obtain a user power consumption feature set in historical power consumption information through a first preprocessing operation, specifically including: obtaining a user type corresponding to a target user, and constructing a user type feature matrix, where the user types include residential type users, commercial type users, and industrial type users; obtaining the power consumption time distribution corresponding to different user types, and constructing a power consumption time feature vector corresponding to the target user through time series analysis; obtaining the power consumption of different user types in a corresponding preset time period, where the preset time period is a peak power consumption time period, and constructing a peak power consumption feature vector corresponding to the target user based on the power consumption in the preset time period; performing feature fusion on the user power consumption feature set, the power consumption time feature vector, and the peak power consumption feature vector through the first preprocessing operation to obtain the user power consumption feature set.

[0063] In one possible implementation, the preprocessing module 22 is used to obtain a historical power supply feature set in the historical power supply data through a second preprocessing operation, specifically including: obtaining historical voltage fluctuation values, historical power supply frequency change values, and historical harmonic frequencies in a first preset time period; constructing a historical power supply quality feature vector based on the voltage fluctuation values, historical power supply frequency change values, and historical harmonic frequencies; obtaining the historical total power supply and the historical peak power supply in the first preset time period; constructing a historical power supply feature vector based on the historical total power supply and the historical peak power supply; and performing feature fusion on the historical power supply quality feature vector and the historical power supply feature vector through a second preprocessing operation to obtain a historical power supply feature set.

[0064] In a possible implementation, the preprocessing module 22 is used to perform feature fusion of the user's power consumption feature set and the historical power supply feature set through a third preprocessing operation, and construct a comprehensive power consumption feature set, specifically including: obtaining the spatiotemporal mapping relationship between the user's power consumption feature set and the historical power supply feature set; based on the spatiotemporal mapping relationship, perform feature fusion of the user's power consumption feature set and the historical power supply feature set through a third preprocessing operation, and construct a comprehensive power consumption feature set.

[0065] In a possible implementation, the prediction module 23 is used to obtain the predicted power demand of the target user in the second preset time period through the smart distribution network power supply model, specifically including: calculating the predicted power demand of the target user through the following formula:

[0066]

[0067] Where P(t) is the predicted electricity demand at time t, time t is any time in the second preset time period, f is the preset activation function, U i (t') is the i-th feature in the electricity consumption feature set, w i is the feature weight corresponding to the i-th feature, S j(t') is the jth feature in the historical power supply feature set, v j is the feature weight corresponding to the jth feature, C(t') is the time-space mapping relationship, γ is the weight of the time-space mapping relationship, and t' is the moment corresponding to moment t in the first preset time period.

[0068] In a possible implementation manner, the prediction module 23 is used to obtain the spatiotemporal mapping relationship between the user power consumption feature set and the historical power supply feature set, specifically including: obtaining the spatiotemporal mapping relationship according to the following formula:

[0069]

[0070] Wherein, exp is a preset exponential function, which is used to capture the time decay effect, τ is the time point associated with the user's power consumption characteristics, σ is the standard deviation of the control Gaussian function, and is used to represent the speed of time decay. Indicates the degree of periodic impact of the j-th historical power supply characteristics.

[0071] In a possible implementation, the prediction module 23 is used to obtain the predicted electricity demand of the target user in the second preset time period through the intelligent distribution network power supply model, and after that, the distribution network power supply operation is performed according to the predicted electricity demand. The method also includes: obtaining the actual electricity demand of the target user in the second preset time period; calculating whether the electricity difference between the actual electricity demand and the predicted electricity demand is greater than the preset difference; if the electricity difference is greater than the preset difference, obtaining multiple prediction error factors; and improving the intelligent distribution network power supply model according to the multiple prediction error factors.

[0072] It should be noted that: when the device provided in the above embodiment realizes its function, only the division of the above functional modules is used as an example. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0073] The present application also provides an electronic device. Figure 3 , Figure 3 30 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include: at least one processor 301, at least one communication bus 302, a user interface 303, at least one network interface 304, and a memory 305.

[0074] The communication bus 302 is used to realize the connection and communication between these components.

[0075] The user interface 303 may include a display screen (Display) and a camera (Camera). The optional user interface 303 may also include a standard wired interface and a wireless interface.

[0076] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0077] Among them, the processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 301 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 301, and it can be implemented separately through a chip.

[0078] Among them, the memory 305 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may optionally also be at least one storage device located away from the aforementioned processor 301. Refer to Figure 3, the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a distribution network power supply application based on machine learning.

[0079] exist Figure 3 In the electronic device shown, the user interface 303 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 301 can be used to call the distribution network power supply application based on machine learning stored in the memory 305. When executed by one or more processors 301, the electronic device executes one or more of the methods described in the above embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited to the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application.

[0080] The present application also provides a computer-readable storage medium, which stores instructions. When executed by one or more processors, the electronic device executes one or more of the methods described in the above embodiments.

[0081] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0082] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as 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 mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0083] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0084] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0085] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0086] The above are only exemplary embodiments disclosed in this application, and the scope of the disclosure of this application cannot be limited by this. That is, any equivalent changes and modifications made according to the teachings disclosed in this application are still within the scope of the disclosure of this application. After considering the disclosure of the specification, those skilled in the art will easily think of other embodiments disclosed in this application.

[0087] This application is intended to cover any modifications, uses or adaptive changes disclosed in this application, which follow the general principles disclosed in this application and include common knowledge or customary technical means in the technical field not recorded in this application.

Claims

1. A distribution network power supply method based on machine learning, characterized in that: The method comprises: In response to a distribution network power supply operation for a target user, obtaining historical power consumption information of the target user in a first preset time period; Acquire a user power usage feature set in the historical power usage information through a first preprocessing operation, wherein the user power usage feature set includes user power usage time features, user type features, user power usage peak period features, and user power usage change trend features; Acquire historical power supply data, and acquire a historical power supply feature set in the historical power supply data through a second preprocessing operation, wherein the historical power supply feature set includes a power supply quality index feature and a power supply quantity feature; By means of a third preprocessing operation, the user power consumption feature set and the historical power supply feature set are subjected to feature fusion, and a comprehensive power consumption feature set is constructed; Constructing a smart distribution network power supply model according to the comprehensive power consumption feature set; The predicted power demand of the target user in the second preset time period is obtained through the intelligent distribution network power supply model, and the distribution network power supply operation is performed according to the predicted power demand.

2. The method according to claim 1, characterized in that The obtaining of the user power usage feature set in the historical power usage information through the first preprocessing operation specifically includes: Acquire a user type corresponding to the target user and construct a user type feature matrix, wherein the user type includes residential type users, commercial type users and industrial type users; Obtaining the power consumption time distribution corresponding to different user types, and constructing the power consumption time feature vector corresponding to the target user through time series analysis; Acquire the power consumption of different user types in a corresponding preset time period, where the preset time period is a peak power consumption time period, and construct a peak power consumption feature vector corresponding to the target user according to the power consumption in the preset time period; The first preprocessing operation is used to perform the feature fusion on the user power usage feature set, the power usage time feature vector, and the power peak feature vector to obtain the user power usage feature set.

3. The method according to claim 1, characterized in that The obtaining of the historical power supply feature set in the historical power supply data through the second preprocessing operation specifically includes: Obtaining historical voltage fluctuation values, historical power supply frequency change values, and historical harmonic frequencies in the first preset time period; Constructing a historical power supply quality feature vector according to the voltage fluctuation value, the historical power supply frequency change value and the historical harmonic frequency; Obtaining a historical total power supply and a historical peak power supply in the first preset time period; Constructing a historical power supply feature vector according to the historical total power supply and the historical peak power supply; The feature fusion is performed on the historical power supply quality feature vector and the historical power supply quantity feature vector through the second preprocessing operation to obtain the historical power supply feature set.

4. The method according to claim 1, characterized in that: The third preprocessing operation is used to fuse the user power consumption feature set and the historical power supply feature set, and construct a comprehensive power consumption feature set, specifically including: Acquire a spatiotemporal mapping relationship between the user power consumption feature set and the historical power supply feature set; According to the time-space mapping relationship, the user power consumption feature set and the historical power supply feature set are feature-fused through a third preprocessing operation to construct the comprehensive power consumption feature set.

5. The method according to claim 4, characterized in that The obtaining the predicted power demand of the target user in the second preset time period through the smart distribution network power supply model specifically includes: The predicted electricity demand of the target user is calculated by the following formula: Wherein, P(t) is the predicted power demand at time t, the time t is any time in the second preset time period, f is the preset activation function, U i (t') is the i-th feature in the electricity usage feature set, w i is the feature weight corresponding to the i-th feature, S j (t') is the jth feature in the historical power supply feature set, v j is the feature weight corresponding to the j-th feature, C(t') is the space-time mapping relationship, γ is the weight of the space-time mapping relationship, and t' is the moment corresponding to the t moment in the first preset time period.

6. The method according to claim 5, characterized in that The acquiring of the spatiotemporal mapping relationship between the user power consumption feature set and the historical power supply feature set specifically includes: The time-space mapping relationship is obtained according to the following formula: Wherein, exp is a preset exponential function, which is used to capture the time decay effect, τ is a time point associated with the user's power consumption characteristics, σ is the standard deviation of the control Gaussian function, and is used to represent the speed of time decay. It represents the periodic impact degree of the j-th historical power supply feature.

7. The method according to claim 1, characterized in that After obtaining the predicted power demand of the target user in the second preset time period through the intelligent distribution network power supply model and performing the distribution network power supply operation according to the predicted power demand, the method further includes: Obtaining the actual electricity demand of the target user in the second preset time period; Calculating whether the power consumption difference between the actual power consumption demand and the predicted power consumption demand is greater than a preset difference; If the power consumption difference is greater than the preset difference, obtaining a plurality of prediction error factors; The smart distribution network power supply model is improved according to the multiple prediction error factors.

8. A distribution network power supply device based on machine learning, characterized in that: The device comprises an acquisition module, a preprocessing module and a prediction module, wherein: The acquisition module is used to acquire the historical power consumption information of the target user in the first preset time period in response to the power supply operation of the distribution network for the target user; The preprocessing module is used to obtain a user power consumption feature set in the historical power consumption information through a first preprocessing operation, wherein the user power consumption feature set includes user power consumption time features, user type features, user power consumption peak period features, and user power consumption change trend features; obtain historical power supply data, and obtain a historical power supply feature set in the historical power supply data through a second preprocessing operation, wherein the historical power supply feature set includes power supply quality index features and power supply quantity features; and perform feature fusion on the user power consumption feature set and the historical power supply feature set through a third preprocessing operation, and construct a comprehensive power consumption feature set; The prediction module is used to construct an intelligent distribution network power supply model based on the comprehensive power consumption feature set; obtain the predicted power demand of the target user in the second preset time period through the intelligent distribution network power supply model, and perform the distribution network power supply operation according to the predicted power demand.

9. An electronic device, characterized in that: It includes a processor, a communication bus, a user interface, a network interface and a memory, wherein the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 7 is performed.

Images