Electricity load prediction method and device, computer equipment, storage medium and product

Abstract

The application relates to a power load prediction method and device, computer equipment, a storage medium and a product. The method comprises the following steps: in response to a power load prediction request for a target user, obtaining user information of the target user and target load associated data of the target user in a historical period; determining a user category to which the target user belongs according to the user information; and based on a target load prediction model corresponding to the user category, predicting the power load of the target user in a future period according to the target load associated data and target load influence factors corresponding to the user category. The method can improve the accuracy of the power load prediction result of the user.

Description

Technical Field

[0001] This application relates to the field of power distribution network technology, and in particular to a method, apparatus, computer equipment, storage medium and product for predicting electricity load. Background Technology

[0002] As new energy sources gradually become the mainstay and a major feature of the new power system, the intermittent nature of power generation from new energy sources such as wind and solar power and the real-time balance between power supply and demand create an inherent contradiction. This leads to a significant increase in the system's demand for flexible resources. Therefore, high-precision power user load forecasting is beneficial for fully tapping into the adjustable load resources on the demand side.

[0003] Currently, neural networks are often used to predict the electricity load of power users. However, neural networks are a method that is highly dependent on the amount of data and input factors. Furthermore, the electricity load of users is highly correlated with multiple influencing factors in the user's region. If all these factors are considered, it will inevitably increase the size of the neural network, exacerbate the risk of overfitting, and reduce the accuracy of electricity load prediction. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and product for predicting electricity load, which can improve the accuracy of electricity load prediction results for users.

[0005] Firstly, this application provides a method for predicting electricity load, including:

[0006] In response to a power load forecasting request for a target user, the user information of the target user and the target load association data of the target user in historical time periods are obtained.

[0007] Based on the user information, determine the user category to which the target user belongs;

[0008] Based on the target load prediction model corresponding to the user category, and according to the target load correlation data and the target load influence factor corresponding to the user category, the electricity load of the target user in the future time period is predicted.

[0009] In one embodiment, the method further includes:

[0010] Obtain reference load correlation data of sample users under the user category within the reference time period and electricity load data of the sample users within the target time period corresponding to the reference time period;

[0011] Based on the reference load correlation data and the electricity load data, select the target load influence factor corresponding to the user category from each load influence factor;

[0012] Based on the reference load correlation data, the electricity load data, and the target load influence factor, select the target load prediction model corresponding to the user category from the candidate load prediction models.

[0013] In one embodiment, selecting the target load impact factor corresponding to the user category from each load impact factor based on the reference load association data and the electricity load data includes:

[0014] Based on the reference load correlation data and the electricity load data, determine the mutual information value between each load influence factor and the electricity load of the sample users under the user category;

[0015] The load impact factors whose mutual information value is greater than the mutual information threshold among the load impact factors are taken as the target load impact factors corresponding to the user category.

[0016] In one embodiment, determining the mutual information value between each load influence factor and the electricity load of sample users under the user category based on the reference load association data and the electricity load data includes:

[0017] Extract the factor values ​​corresponding to each load influence factor from the reference load correlation data;

[0018] Based on the factor values ​​of each load influence factor during the reference period, construct the marginal probability density function corresponding to each load influence factor;

[0019] Based on the electricity load data, construct the marginal probability density function of the electricity load of sample users under the user category;

[0020] Based on the electricity load data and the factor values ​​of each load influencing factor during the reference period, a joint probability density function is constructed between each load influencing factor and the electricity load of sample users under the user category.

[0021] Based on the marginal probability density function corresponding to each load influence factor, the marginal probability density function of the electricity load of the sample users under the user category, and the joint probability density function, the mutual information value between each load influence factor and the electricity load of the sample users under the user category is determined.

[0022] In one embodiment, selecting the target load prediction model corresponding to the user category from candidate load prediction models based on the reference load association data, the electricity load data, and the target load influence factor includes:

[0023] Extract the factor values ​​corresponding to the target load influence factors from the reference load correlation data;

[0024] Based on the electricity load data and the factor values ​​corresponding to the target load influence factors, a training sample is constructed;

[0025] Using the training samples, the model parameters of the pre-trained candidate load prediction model are optimized, and the model prediction accuracy of the optimized candidate load prediction model is obtained.

[0026] The candidate load prediction model with the highest prediction accuracy among the optimized candidate load prediction models is taken as the target load prediction model for the user category.

[0027] In one embodiment, the target load prediction model based on the user category, which predicts the electricity load of the target user in a future time period based on the target load association data and the target load influence factor corresponding to the user category, includes:

[0028] Extract the factor values ​​corresponding to the target load influencing factors from the target load associated data;

[0029] The factor values ​​corresponding to the target load influencing factors are input into the target load prediction model to obtain the electricity load of the target user in the future time period.

[0030] Secondly, this application also provides an electricity load forecasting device, comprising:

[0031] The data acquisition module is used to respond to a power load forecasting request for a target user and acquire the user information of the target user and the target load association data of the target user in historical time periods.

[0032] The category determination module is used to determine the user category to which the target user belongs based on the user information.

[0033] The load forecasting module is used to forecast the electricity load of the target user in the future time period based on the target load forecasting model corresponding to the user category, according to the target load association data and the target load influence factor corresponding to the user category.

[0034] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0035] In response to a power load forecasting request for a target user, the user information of the target user and the target load association data of the target user in historical time periods are obtained.

[0036] Based on the user information, determine the user category to which the target user belongs;

[0037] Based on the target load prediction model corresponding to the user category, and according to the target load correlation data and the target load influence factor corresponding to the user category, the electricity load of the target user in the future time period is predicted.

[0038] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0039] In response to a power load forecasting request for a target user, the user information of the target user and the target load association data of the target user in historical time periods are obtained.

[0040] Based on the user information, determine the user category to which the target user belongs;

[0041] Based on the target load prediction model corresponding to the user category, and according to the target load correlation data and the target load influence factor corresponding to the user category, the electricity load of the target user in the future time period is predicted.

[0042] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0043] In response to a power load forecasting request for a target user, the user information of the target user and the target load association data of the target user in historical time periods are obtained.

[0044] Based on the user information, determine the user category to which the target user belongs;

[0045] Based on the target load prediction model corresponding to the user category, and according to the target load correlation data and the target load influence factor corresponding to the user category, the electricity load of the target user in the future time period is predicted.

[0046] The aforementioned electricity load forecasting method, apparatus, computer equipment, storage medium, and product, upon receiving an electricity load forecasting request for a target user, acquire the target user's user information and target load correlation data for the target user in historical time periods. Based on the user information, they determine the user category to which the target user belongs, ensuring the accuracy of the determined user category. Furthermore, based on the target load forecasting model corresponding to the user category, and according to the target load correlation data and the target load influence factor corresponding to the user category, they forecast the target user's electricity load in future time periods. Each user category corresponds to its own target load influence factor, making the target load influence factor more targeted. At the same time, each user category corresponds to a target load forecasting model, making the predicted target user's electricity load in future time periods more targeted and improving the accuracy of the forecast results. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart illustrating an electricity load forecasting method in one embodiment;

[0049] Figure 2 This is a flowchart illustrating the process of determining a target load prediction model in one embodiment;

[0050] Figure 3 This is a flowchart illustrating the process of selecting the target load impact factor in one embodiment;

[0051] Figure 4 This is a flowchart illustrating the process of determining mutual information values ​​in one embodiment;

[0052] Figure 5 This is a flowchart illustrating the process of determining the target load prediction model in another embodiment;

[0053] Figure 6 This is a flowchart illustrating the electricity load forecasting method in another embodiment;

[0054] Figure 7 This is a flowchart illustrating the electricity load forecasting method in yet another embodiment;

[0055] Figure 8 This is a structural block diagram of an electricity load prediction device in one embodiment;

[0056] Figure 9This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0058] The electricity load forecasting method provided in this application can be applied to application environments that forecast users' electricity load. The electricity load forecasting method provided in this application can be executed by a computer device, which can be a server or a terminal with powerful computing capabilities.

[0059] In one embodiment, such as Figure 1 As shown, a method for predicting electricity load is provided. Taking the application of this method to a server as an example, the method includes the following steps:

[0060] S101, in response to the electricity load forecasting request for the target user, obtains the user information of the target user and the target load association data of the target user in the historical period.

[0061] The target user can be any user; the electricity load forecasting request is used to request an electricity load forecast; user information is information that characterizes the target user's features. In this embodiment, user information includes, but is not limited to, the target user's age, occupation, electricity usage habits, etc.; load-related data is data associated with the electricity load, i.e., data that affects the electricity load, such as climate data and electricity usage duration data of the geographical area where the target user is located; target load-related data is the target user's load-related data in historical time periods; historical time periods are the historical time periods of the current time period.

[0062] Optionally, upon receiving a power load forecasting request for a target user, if the power load forecasting request carries the target user's user information and the target load association data for the target user in historical periods, the user information and the target load association data for the target user in historical periods can be directly obtained from the power load forecasting request; if the power load forecasting request does not carry the target user's user information and the target load association data for the target user in historical periods, the user information and the target load association data for the target user in historical periods can be obtained from the database based on the target user's identification information in the power load forecasting request.

[0063] S102, Based on the user information, determine the user category to which the target user belongs.

[0064] The user category is obtained by classifying users based on their attributes.

[0065] Optionally, target users can be categorized based on attributes such as age, electricity usage habits, and occupation in the user information, and the target users can be divided into the categories with the highest similarity as the user category to which the target user belongs.

[0066] S103, based on the target load prediction model corresponding to the user category, predicts the electricity load of the target user in the future period according to the target load correlation data and the target load influence factor corresponding to the user category.

[0067] The target load forecasting model is used to predict the electricity load of target users. Load influence factors are factors that affect electricity load; the target load influence factor is the load influence factor corresponding to the user category. In this embodiment, the load influence factor may be temperature, holidays, UV intensity, humidity, sunshine duration, etc.

[0068] Optionally, data related to the target load influencing factors in the target load association data can be input into the target load prediction model corresponding to the user category. The target load prediction model can then calculate the electricity load of the target user in the future period based on the model parameters.

[0069] In the aforementioned electricity load forecasting method, upon receiving an electricity load forecasting request for a target user, the method acquires the target user's user information and target load correlation data for the target user in historical time periods. Based on the user information, it determines the user category to which the target user belongs, ensuring the accuracy of the determined user category. Furthermore, based on the target load forecasting model corresponding to the user category, and according to the target load correlation data and the target load influence factor corresponding to the user category, it forecasts the target user's electricity load in future time periods. Each user category corresponds to its own target load influence factor, making the target load influence factor more targeted. At the same time, each user category corresponds to a target load forecasting model, making the predicted electricity load of the target user in future time periods more targeted and improving the accuracy of the forecast results.

[0070] Optionally, in one embodiment, such as Figure 2 As shown, a process for determining a target load forecasting model is provided, which specifically includes the following steps:

[0071] S201, Obtain the reference load correlation data of sample users under the user category within the reference time period and the electricity load data of sample users within the target time period corresponding to the reference time period.

[0072] The sample users are all users under the user category. The reference load correlation data is the load correlation data of the sample users within the reference time period; the reference time period can be any historical time period; the electricity load data is the electricity load of the sample users within the target time period.

[0073] Optionally, the electricity load data of sample users under the user category in the target time period corresponding to the reference time period can be downloaded from the power-side website; at the same time, the reference load association data of sample users under the user category in the reference time period can be obtained from the database.

[0074] S202, based on the reference load correlation data and electricity load data, select the target load influence factor corresponding to the user category from each load influence factor.

[0075] Optionally, statistical analysis can be performed on the reference load correlation data and electricity load data to determine the dependence between each load influencing factor and the electricity load, and load influencing factors with high dependence can be used as target load influencing factors corresponding to user categories.

[0076] S203. Based on the reference load correlation data, electricity load data, and target load influence factors, select the target load prediction model corresponding to the user category from the candidate load prediction models.

[0077] Among them, the candidate load prediction model is an existing model that can predict electricity load, and there are multiple candidate load prediction models.

[0078] Optionally, factor values ​​corresponding to the target load influencing factors can be extracted from the reference load correlation data, and the prediction accuracy of the candidate load prediction model can be obtained based on the factor values ​​corresponding to the target load influencing factors and the electricity load data. The candidate load prediction model with the highest prediction accuracy can be used as the target load prediction model corresponding to the user category.

[0079] In this embodiment, by using reference load correlation data, electricity load data, and target load influence factors, the accuracy of selecting the target load prediction model corresponding to the user category from the candidate load prediction models is ensured, and the selected target load prediction model is more targeted.

[0080] Optionally, to ensure the accuracy of the determined target load influence factor, in one embodiment, such as Figure 3 As shown, a method for selecting target load impact factors corresponding to user categories is provided, which specifically includes the following steps:

[0081] S301, Based on the reference load correlation data and electricity load data, determine the mutual information value between each load influence factor and the electricity load of sample users under the user category.

[0082] The mutual information value describes the correlation between the load influence factor and the electrical load. If the mutual information value is 0, it means that the load influence factor and the electrical load are independent of each other. The larger the mutual information value, the higher the correlation between the load influence factor and the electrical load.

[0083] Optionally, the factor values ​​corresponding to each load influencing factor can be extracted from the reference load correlation data first; further, statistical analysis can be performed on the factor values ​​corresponding to each load influencing factor and the user load data to calculate the mutual information value between each load influencing factor and the electricity load of sample users under the user category.

[0084] S302, among all load impact factors, load impact factors with mutual information values ​​greater than the mutual information threshold are used as the target load impact factors corresponding to the user category.

[0085] The mutual information threshold is a pre-set threshold for mutual information.

[0086] Optionally, the mutual information value of each load impact factor is compared with a mutual information threshold. In this embodiment, the mutual information threshold can be set to 0.25. Furthermore, load impact factors that are greater than the mutual information threshold are used as the target load impact factors corresponding to the user category.

[0087] In this embodiment, by introducing mutual information values, since mutual information values ​​can characterize the correlation between load impact factors and electricity load, the target load impact factor determined based on mutual information values ​​is more in line with the characteristics and needs of sample users under the user category.

[0088] Optionally, to ensure the accuracy of the determined mutual information value, in one embodiment, such as Figure 4 As shown, a method for determining the mutual information value between each load influence factor and the electricity load of sample users under the user category is provided, specifically including the following steps:

[0089] S401, extract the factor values ​​corresponding to each load influence factor from the reference load correlation data.

[0090] Optionally, factor values ​​corresponding to each load influence factor can be extracted from the reference load correlation data.

[0091] S402, based on the factor values ​​of each load influencing factor within the reference period, construct the marginal probability density function corresponding to each load influencing factor.

[0092] Among them, the marginal probability density function corresponding to the load influence factor is a function used to describe the marginal distribution of the load influence factor.

[0093] Optionally, for each load impact factor, based on the factor value of the load impact factor within the reference period, the probability that the factor value of the load impact factor falls within a preset range is calculated, and the marginal probability density function corresponding to the load impact factor is constructed based on the obtained probability.

[0094] S403, Based on the electricity load data, construct the marginal probability density function of the electricity load of sample users under the user category.

[0095] The marginal probability density function of the electrical load is a function used to describe the marginal distribution of the electrical load.

[0096] Optionally, based on the electricity load data, the probability that the electricity load falls within a preset range is calculated, and the marginal probability density function of the electricity load of sample users under the user category is constructed based on the obtained probability.

[0097] S404. Based on the electricity load data and the factor values ​​of each load influencing factor during the reference period, construct the joint probability density function between each load influencing factor and the electricity load of sample users under the user category.

[0098] The joint probability density function is used to describe the probability distribution of the joint values ​​of electricity load and load influence factor.

[0099] Optionally, for each load impact factor, based on the electricity load data and the factor values ​​of each load impact factor within the reference period, the probability that the electricity load data falls within a preset value range and that the factor value corresponding to the load impact factor falls within the preset value range is simultaneously determined, and a suitable function is selected as the joint probability density function between the load impact factor and the electricity load of the sample users under the user category.

[0100] S405. Based on the marginal probability density function corresponding to each load influence factor, the marginal probability density function of the electricity load of sample users under the user category, and the joint probability density function, determine the mutual information value between each load influence factor and the electricity load of sample users under the user category.

[0101] Optionally, based on the calculation rules of mutual information, the mutual information value between each load influence factor and the electricity load of the sample users under the user category can be determined according to the marginal probability density function corresponding to each load influence factor, the marginal probability density function of the electricity load of the sample users under the user category, and the joint probability density function.

[0102] Specifically, for each load influence factor, the mutual information value can be determined by the following formula (1):

[0103] (1)

[0104] in, For user load; Load impact factor; Let be the joint probability density function; The marginal probability density function of the electrical load; The marginal probability density function of the load influence factor; This represents the mutual information value between electrical load and load influence factor.

[0105] In this embodiment, by introducing marginal probability density functions and joint probability density functions, the accuracy of the mutual information values ​​between the determined load influence factors and the electricity load of sample users under the user category is ensured.

[0106] Optionally, to ensure the accuracy and relevance of the determined target load forecasting model, in one embodiment, such as Figure 5 As shown, a method for determining a target load forecasting model is provided, which specifically includes the following steps:

[0107] S501, extract the factor values ​​corresponding to the target load influence factors from the reference load correlation data.

[0108] Optionally, the factor values ​​corresponding to the target load influence factors can be extracted from the reference load correlation data.

[0109] S502, construct training samples based on electricity load data and the factor values ​​corresponding to the target load influencing factors.

[0110] The training samples are those used to train the pre-trained candidate load prediction model.

[0111] Optionally, the training samples include multiple samples, and each sample includes a set of input data and output data; specifically, the factor value corresponding to the target load influencing factor can be used as input data, and the corresponding electricity load data can be used as output data.

[0112] S503 uses training samples to optimize the model parameters of the pre-trained candidate load prediction model and obtains the model prediction accuracy of the optimized candidate load prediction model.

[0113] Here, model prediction accuracy refers to the precision of the results predicted by the model.

[0114] Optionally, training samples can be used to optimize the model parameters of the pre-trained candidate load prediction model. For example, a global optimization algorithm based on a sequential model can be used to optimize the model parameters in the pre-trained candidate load prediction model to obtain the optimal parameters of the pre-trained candidate load prediction model, and then the pre-trained candidate load prediction model can be optimized using the optimal parameters.

[0115] Furthermore, for the optimized pre-trained candidate load prediction models, the model prediction accuracy of each candidate load prediction model will be output. At this point, the model prediction accuracy of the optimized candidate load prediction model can be obtained.

[0116] S504 selects the candidate load prediction model with the highest prediction accuracy from the optimized candidate load prediction models as the target load prediction model for the user category.

[0117] Understandably, the higher the model's prediction accuracy, the more accurate the prediction results. Therefore, to ensure the accuracy of the prediction results, it is necessary to select candidate load prediction models with high prediction accuracy.

[0118] Optionally, the candidate load prediction model with the highest prediction accuracy among the optimized candidate load prediction models can be used as the target load prediction model corresponding to the user category to ensure the accuracy of the prediction results.

[0119] In this embodiment, training samples are constructed, and the model parameters of the pre-trained candidate load prediction model are optimized based on the constructed training samples. Then, the candidate load prediction model with the highest prediction accuracy is selected as the target load prediction model, thus ensuring the prediction accuracy of the target load prediction model.

[0120] Optionally, in one embodiment, such as Figure 6 As shown, a method for predicting the electricity load of a target user in a future time period is provided, which specifically includes the following steps:

[0121] S601, extract the factor values ​​corresponding to the target load influencing factors from the target load correlation data.

[0122] Optionally, it is necessary to select the corresponding value from the target load association data according to the name of the target load influencing factor, and use it as the factor value corresponding to the target load influencing factor.

[0123] S602, input the factor values ​​corresponding to the target load influencing factors into the target load prediction model to obtain the electricity load of the target user in the future period.

[0124] Optionally, the factor values ​​corresponding to the target load influencing factors can be input into the target load prediction model, so that the target load prediction model can calculate the factor values ​​corresponding to the target load influencing factors based on the model parameters, and obtain the electricity load of the target user in the future period.

[0125] In this embodiment, by extracting the factor values ​​corresponding to the target load influencing factors from the target load associated data, and based on the factor values ​​corresponding to the target load influencing factors, the electricity load of the target user in the future period is predicted based on the target load prediction model, thus ensuring the accuracy and relevance of the prediction results.

[0126] Figure 7 This is a flowchart illustrating the electricity load forecasting method in another embodiment. Based on the above embodiments, this embodiment provides an optional example of the electricity load forecasting method. (Combined with...) Figure 7 The specific implementation process is as follows:

[0127] S701, in response to a request for electricity load forecasting for a target user, obtains user information of the target user and target load association data of the target user in historical time periods.

[0128] S702, based on user information, determines the user category to which the target user belongs.

[0129] S703, obtain the reference load correlation data of sample users under the user category within the reference time period and the electricity load data of sample users within the target time period corresponding to the reference time period.

[0130] S704, based on reference load correlation data and electricity load data, determine the mutual information value between each load influence factor and the electricity load of sample users under the user category.

[0131] Optionally, the factor values ​​corresponding to each load influencing factor are extracted from the reference load correlation data; based on the factor values ​​of each load influencing factor within the reference period, a marginal probability density function corresponding to each load influencing factor is constructed; based on the electricity load data, a marginal probability density function of the electricity load of sample users under the user category is constructed; based on the electricity load data and the factor values ​​of each load influencing factor within the reference period, a joint probability density function between each load influencing factor and the electricity load of sample users under the user category is constructed; based on the marginal probability density function corresponding to each load influencing factor, the marginal probability density function of the electricity load of sample users under the user category, and the joint probability density function, the mutual information value between each load influencing factor and the electricity load of sample users under the user category is determined.

[0132] S705 specifies that load impact factors with mutual information values ​​greater than the mutual information threshold among all load impact factors are used as the target load impact factors corresponding to the user category.

[0133] S706, Extract the factor values ​​corresponding to the target load influence factors from the reference load correlation data.

[0134] S707: Construct training samples based on electricity load data and the factor values ​​corresponding to the target load influencing factors.

[0135] S708 uses training samples to optimize the model parameters of the pre-trained candidate load prediction model and obtains the model prediction accuracy of the optimized candidate load prediction model.

[0136] S709 selects the candidate load prediction model with the highest prediction accuracy from the optimized candidate load prediction models as the target load prediction model for the user category.

[0137] S710, extract the factor values ​​corresponding to the target load influencing factors from the target load correlation data.

[0138] S711 inputs the factor values ​​corresponding to the target load influencing factors into the target load prediction model to obtain the electricity load of the target user in the future period.

[0139] The specific processes of S701-S711 described above can be found in the description of the above method embodiments. Their implementation principles and technical effects are similar, and will not be repeated here.

[0140] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0141] Based on the same inventive concept, this application also provides an electricity load forecasting device for implementing the electricity load forecasting method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more embodiments of the electricity load forecasting device provided below can be found in the limitations of the electricity load forecasting method described above, and will not be repeated here.

[0142] In one exemplary embodiment, such as Figure 8 As shown, an electricity load forecasting device 80 is provided, including: a data acquisition module 10, a category determination module 20, and a load forecasting module 30, wherein:

[0143] The data acquisition module 10 is used to obtain user information of the target user and target load related data of the target user in historical time periods in response to the electricity load forecasting request for the target user.

[0144] The category determination module 20 is used to determine the user category to which the target user belongs based on user information.

[0145] The load forecasting module 30 is used to forecast the electricity load of target users in future periods based on the target load forecasting model corresponding to the user category, according to the target load correlation data and the target load influence factor corresponding to the user category.

[0146] The aforementioned electricity load forecasting device, upon receiving an electricity load forecasting request for a target user, acquires the target user's user information and target load correlation data for the target user within historical time periods. Based on the user information, it determines the user category to which the target user belongs, ensuring the accuracy of the determined user category. Furthermore, based on the target load forecasting model corresponding to the user category, and according to the target load correlation data and the target load influence factor corresponding to the user category, it forecasts the target user's electricity load in future time periods. Each user category corresponds to its own target load influence factor, making the target load influence factor more targeted. At the same time, each user category corresponds to a target load forecasting model, making the predicted electricity load of the target user in future time periods more targeted and improving the accuracy of the forecast results.

[0147] In one embodiment, the power load forecasting device 80 further includes:

[0148] The information acquisition module is used to acquire the reference load correlation data of sample users under the user category within the reference time period and the electricity load data of sample users within the target time period corresponding to the reference time period.

[0149] The factor selection module is used to select the target load influence factor corresponding to the user category from various load influence factors based on the reference load correlation data and electricity load data.

[0150] The model selection module is used to select the target load prediction model corresponding to the user category from the candidate load prediction models based on the reference load correlation data, electricity load data and target load influence factors.

[0151] In one embodiment, the factor selection module includes:

[0152] The value determination unit is used to determine the mutual information value between each load influence factor and the electricity load of sample users under the user category, based on the reference load correlation data and electricity load data.

[0153] The factor selection unit is used to select load influence factors whose mutual information value is greater than the mutual information threshold from among the load influence factors as the target load influence factors corresponding to the user category.

[0154] In one embodiment, the value determination unit is specifically used for:

[0155] From the reference load correlation data, extract the factor values ​​corresponding to each load influencing factor; construct the marginal probability density function corresponding to each load influencing factor based on the factor values ​​of each load influencing factor within the reference period; construct the marginal probability density function of the electricity load of sample users under the user category based on the electricity load data; construct the joint probability density function between each load influencing factor and the electricity load of sample users under the user category based on the electricity load data and the factor values ​​of each load influencing factor within the reference period; determine the mutual information value between each load influencing factor and the electricity load of sample users under the user category based on the marginal probability density function corresponding to each load influencing factor, the marginal probability density function of the electricity load of sample users under the user category, and the joint probability density function.

[0156] In one embodiment, the model selection module is specifically used for:

[0157] Extract the factor values ​​corresponding to the target load influencing factors from the reference load correlation data; construct training samples based on the electricity load data and the factor values ​​corresponding to the target load influencing factors; use the training samples to optimize the model parameters of the pre-trained candidate load prediction model and obtain the model prediction accuracy of the optimized candidate load prediction model; take the candidate load prediction model with the highest model prediction accuracy among the optimized candidate load prediction models as the target load prediction model corresponding to the user category.

[0158] In one embodiment, the load forecasting module 30 is specifically used for:

[0159] Extract the factor values ​​corresponding to the target load influencing factors from the target load associated data; input the factor values ​​corresponding to the target load influencing factors into the target load prediction model to obtain the electricity load of the target user in the future period.

[0160] Each module in the aforementioned electricity load forecasting device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0161] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The database stores user information for target users and target load correlation data for target users over historical periods. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements an electricity load forecasting method.

[0162] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0163] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0164] In response to a request for electricity load forecasting for a target user, obtain the user information of the target user and the target load associated data of the target user in historical time periods;

[0165] Based on user information, determine the user category to which the target user belongs;

[0166] Based on the target load prediction model corresponding to user categories, the electricity load of target users in future time periods is predicted according to the target load correlation data and the target load influencing factors corresponding to user categories.

[0167] In one embodiment, when the processor executes a computer program, it also performs the following steps:

[0168] Obtain reference load correlation data for sample users within a reference time period and electricity load data for sample users within the target time period corresponding to the reference time period under the user category; based on the reference load correlation data and electricity load data, select the target load influence factor corresponding to the user category from each load influence factor; based on the reference load correlation data, electricity load data, and target load influence factor, select the target load prediction model corresponding to the user category from the candidate load prediction models.

[0169] In one embodiment, when the processor executes a computer program to select the target load impact factor corresponding to the user category from various load impact factors based on reference load correlation data and electricity load data, it also performs the following steps:

[0170] Based on the reference load correlation data and electricity load data, the mutual information value between each load influence factor and the electricity load of sample users under the user category is determined; the load influence factors with mutual information values ​​greater than the mutual information threshold are taken as the target load influence factors corresponding to the user category.

[0171] In one embodiment, when the processor executes a computer program to determine the mutual information value between each load influence factor and the electricity load of sample users under a user category based on reference load correlation data and electricity load data, it also performs the following steps:

[0172] From the reference load correlation data, extract the factor values ​​corresponding to each load influencing factor; construct the marginal probability density function corresponding to each load influencing factor based on the factor values ​​of each load influencing factor within the reference period; construct the marginal probability density function of the electricity load of sample users under the user category based on the electricity load data; construct the joint probability density function between each load influencing factor and the electricity load of sample users under the user category based on the electricity load data and the factor values ​​of each load influencing factor within the reference period; determine the mutual information value between each load influencing factor and the electricity load of sample users under the user category based on the marginal probability density function corresponding to each load influencing factor, the marginal probability density function of the electricity load of sample users under the user category, and the joint probability density function.

[0173] In one embodiment, when the processor executes a computer program to select the target load prediction model corresponding to the user category from the candidate load prediction models based on reference load correlation data, electricity load data, and target load influence factors, it also performs the following steps:

[0174] Extract the factor values ​​corresponding to the target load influencing factors from the reference load correlation data; construct training samples based on the electricity load data and the factor values ​​corresponding to the target load influencing factors; use the training samples to optimize the model parameters of the pre-trained candidate load prediction model and obtain the model prediction accuracy of the optimized candidate load prediction model; take the candidate load prediction model with the highest model prediction accuracy among the optimized candidate load prediction models as the target load prediction model corresponding to the user category.

[0175] In one embodiment, when the processor executes a computer program to predict the electricity load of a target user in a future time period based on the target load prediction model corresponding to the user category and according to the target load correlation data and the target load influence factor corresponding to the user category, it also performs the following steps:

[0176] Extract the factor values ​​corresponding to the target load influencing factors from the target load associated data; input the factor values ​​corresponding to the target load influencing factors into the target load prediction model to obtain the electricity load of the target user in the future period.

[0177] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0178] In response to a request for electricity load forecasting for a target user, obtain the user information of the target user and the target load associated data of the target user in historical time periods;

[0179] Based on user information, determine the user category to which the target user belongs;

[0180] Based on the target load prediction model corresponding to user categories, the electricity load of target users in future time periods is predicted according to the target load correlation data and the target load influencing factors corresponding to user categories.

[0181] In one embodiment, when the processor executes a computer program, it also performs the following steps:

[0182] Obtain reference load correlation data for sample users within a reference time period and electricity load data for sample users within the target time period corresponding to the reference time period under the user category; based on the reference load correlation data and electricity load data, select the target load influence factor corresponding to the user category from each load influence factor; based on the reference load correlation data, electricity load data, and target load influence factor, select the target load prediction model corresponding to the user category from the candidate load prediction models.

[0183] In one embodiment, when the processor executes a computer program to select the target load impact factor corresponding to the user category from various load impact factors based on reference load correlation data and electricity load data, it also performs the following steps:

[0184] Based on the reference load correlation data and electricity load data, the mutual information value between each load influence factor and the electricity load of sample users under the user category is determined; the load influence factors with mutual information values ​​greater than the mutual information threshold are taken as the target load influence factors corresponding to the user category.

[0185] In one embodiment, when the processor executes a computer program to determine the mutual information value between each load influence factor and the electricity load of sample users under a user category based on reference load correlation data and electricity load data, it also performs the following steps:

[0186] From the reference load correlation data, extract the factor values ​​corresponding to each load influencing factor; construct the marginal probability density function corresponding to each load influencing factor based on the factor values ​​of each load influencing factor within the reference period; construct the marginal probability density function of the electricity load of sample users under the user category based on the electricity load data; construct the joint probability density function between each load influencing factor and the electricity load of sample users under the user category based on the electricity load data and the factor values ​​of each load influencing factor within the reference period; determine the mutual information value between each load influencing factor and the electricity load of sample users under the user category based on the marginal probability density function corresponding to each load influencing factor, the marginal probability density function of the electricity load of sample users under the user category, and the joint probability density function.

[0187] In one embodiment, when the processor executes a computer program to select the target load prediction model corresponding to the user category from the candidate load prediction models based on reference load correlation data, electricity load data, and target load influence factors, it also performs the following steps:

[0188] Extract the factor values ​​corresponding to the target load influencing factors from the reference load correlation data; construct training samples based on the electricity load data and the factor values ​​corresponding to the target load influencing factors; use the training samples to optimize the model parameters of the pre-trained candidate load prediction model and obtain the model prediction accuracy of the optimized candidate load prediction model; take the candidate load prediction model with the highest model prediction accuracy among the optimized candidate load prediction models as the target load prediction model corresponding to the user category.

[0189] In one embodiment, when the processor executes a computer program to predict the electricity load of a target user in a future time period based on the target load prediction model corresponding to the user category and according to the target load correlation data and the target load influence factor corresponding to the user category, it also performs the following steps:

[0190] Extract the factor values ​​corresponding to the target load influencing factors from the target load associated data; input the factor values ​​corresponding to the target load influencing factors into the target load prediction model to obtain the electricity load of the target user in the future period.

[0191] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0192] In response to a request for electricity load forecasting for a target user, obtain the user information of the target user and the target load associated data of the target user in historical time periods;

[0193] Based on user information, determine the user category to which the target user belongs;

[0194] Based on the target load prediction model corresponding to user categories, the electricity load of target users in future time periods is predicted according to the target load correlation data and the target load influencing factors corresponding to user categories.

[0195] In one embodiment, when the processor executes a computer program, it also performs the following steps:

[0196] Obtain reference load correlation data for sample users within a reference time period and electricity load data for sample users within the target time period corresponding to the reference time period under the user category; based on the reference load correlation data and electricity load data, select the target load influence factor corresponding to the user category from each load influence factor; based on the reference load correlation data, electricity load data, and target load influence factor, select the target load prediction model corresponding to the user category from the candidate load prediction models.

[0197] In one embodiment, when the processor executes a computer program to select the target load impact factor corresponding to the user category from various load impact factors based on reference load correlation data and electricity load data, it also performs the following steps:

[0198] Based on the reference load correlation data and electricity load data, the mutual information value between each load influence factor and the electricity load of sample users under the user category is determined; the load influence factors with mutual information values ​​greater than the mutual information threshold are taken as the target load influence factors corresponding to the user category.

[0199] In one embodiment, when the processor executes a computer program to determine the mutual information value between each load influence factor and the electricity load of sample users under a user category based on reference load correlation data and electricity load data, it also performs the following steps:

[0200] From the reference load correlation data, extract the factor values ​​corresponding to each load influencing factor; construct the marginal probability density function corresponding to each load influencing factor based on the factor values ​​of each load influencing factor within the reference period; construct the marginal probability density function of the electricity load of sample users under the user category based on the electricity load data; construct the joint probability density function between each load influencing factor and the electricity load of sample users under the user category based on the electricity load data and the factor values ​​of each load influencing factor within the reference period; determine the mutual information value between each load influencing factor and the electricity load of sample users under the user category based on the marginal probability density function corresponding to each load influencing factor, the marginal probability density function of the electricity load of sample users under the user category, and the joint probability density function.

[0201] In one embodiment, when the processor executes a computer program to select the target load prediction model corresponding to the user category from the candidate load prediction models based on reference load correlation data, electricity load data, and target load influence factors, it also performs the following steps:

[0202] Extract the factor values ​​corresponding to the target load influencing factors from the reference load correlation data; construct training samples based on the electricity load data and the factor values ​​corresponding to the target load influencing factors; use the training samples to optimize the model parameters of the pre-trained candidate load prediction model and obtain the model prediction accuracy of the optimized candidate load prediction model; take the candidate load prediction model with the highest model prediction accuracy among the optimized candidate load prediction models as the target load prediction model corresponding to the user category.

[0203] In one embodiment, when the processor executes a computer program to predict the electricity load of a target user in a future time period based on the target load prediction model corresponding to the user category and according to the target load correlation data and the target load influence factor corresponding to the user category, it also performs the following steps:

[0204] Extract the factor values ​​corresponding to the target load influencing factors from the target load associated data; input the factor values ​​corresponding to the target load influencing factors into the target load prediction model to obtain the electricity load of the target user in the future period.

[0205] It should be noted that the user information and data involved in this application (including but not limited to data used for analysis, data stored, data displayed, etc.) are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0206] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0207] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0208] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for predicting electricity load, characterized in that, The method includes: In response to a power load forecasting request for a target user, the user information of the target user and the target load association data of the target user in historical time periods are obtained. Based on the user information, determine the user category to which the target user belongs; Based on the target load prediction model corresponding to the user category, the electricity load of the target user in the future time period is predicted according to the target load correlation data and the target load influence factor corresponding to the user category. The target load prediction model is determined in the following way: Obtain reference load correlation data of sample users under the user category within the reference time period and electricity load data of the sample users within the target time period corresponding to the reference time period; Based on the marginal probability density function corresponding to each load influence factor, the marginal probability density function of the electricity load of the sample users under the user category, and the joint probability density function between each load influence factor and the electricity load of the sample users under the user category, the mutual information value between each load influence factor and the electricity load of the sample users under the user category is determined. The load impact factors whose mutual information value is greater than the mutual information threshold among the load impact factors are taken as the target load impact factors corresponding to the user category. Based on the reference load correlation data, the electricity load data, and the target load influence factor, select the target load prediction model corresponding to the user category from the candidate load prediction models.

2. The method according to claim 1, characterized in that, The step of determining the mutual information value between each load influence factor and the electricity load of the sample users under the user category based on the marginal probability density function corresponding to each load influence factor, the marginal probability density function of the electricity load of the sample users under the user category, and the joint probability density function between each load influence factor and the electricity load of the sample users under the user category includes: Extract the factor values ​​corresponding to each load influence factor from the reference load correlation data; Based on the factor values ​​of each load influence factor during the reference period, construct the marginal probability density function corresponding to each load influence factor; Based on the electricity load data, construct the marginal probability density function of the electricity load of sample users under the user category; Based on the electricity load data and the factor values ​​of each load influencing factor during the reference period, a joint probability density function is constructed between each load influencing factor and the electricity load of sample users under the user category.

3. The method according to claim 1, characterized in that, The step of selecting the target load prediction model corresponding to the user category from the candidate load prediction models based on the reference load correlation data, the electricity load data, and the target load influence factor includes: Extract the factor values ​​corresponding to the target load influence factors from the reference load correlation data; Based on the electricity load data and the factor values ​​corresponding to the target load influence factors, a training sample is constructed; Using the training samples, the model parameters of the pre-trained candidate load prediction model are optimized, and the model prediction accuracy of the optimized candidate load prediction model is obtained. The candidate load prediction model with the highest prediction accuracy among the optimized candidate load prediction models is taken as the target load prediction model for the user category.

4. The method according to claim 1, characterized in that, The target load prediction model based on the user category predicts the electricity load of the target user in a future time period based on the target load correlation data and the target load influence factor corresponding to the user category, including: Extract the factor values ​​corresponding to the target load influencing factors from the target load associated data; The factor values ​​corresponding to the target load influencing factors are input into the target load prediction model to obtain the electricity load of the target user in the future time period.

5. An electricity load prediction device, characterized in that, The device includes: The data acquisition module is used to respond to a power load forecasting request for a target user and acquire the user information of the target user and the target load association data of the target user in historical time periods. The category determination module is used to determine the user category to which the target user belongs based on the user information. The load forecasting module is used to forecast the electricity load of the target user in a future time period based on the target load forecasting model corresponding to the user category, according to the target load association data and the target load influence factor corresponding to the user category; The electricity load forecasting device also includes: The information acquisition module is used to acquire reference load correlation data of sample users under the user category within the reference time period and electricity load data of the sample users within the target time period corresponding to the reference time period; The factor selection module is used to determine the mutual information value between each load influence factor and the electricity load of the sample users under the user category based on the marginal probability density function corresponding to each load influence factor, the marginal probability density function of the electricity load of the sample users under the user category, and the joint probability density function between each load influence factor and the electricity load of the sample users under the user category; and to select the load influence factors whose mutual information value is greater than the mutual information threshold as the target load influence factors corresponding to the user category. The model selection module is used to select the target load prediction model corresponding to the user category from the candidate load prediction models based on the reference load association data, the electricity load data, and the target load influence factor.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images