Power load prediction method and device, electronic equipment and storage medium

Abstract

The invention provides a power load prediction method and device, electronic equipment and a storage medium. The method comprises the following steps: collecting first multi-dimensional data for power load prediction, wherein the first multi-dimensional data comprises historical load data, meteorological data and holiday and festival data; performing data cleaning and smoothing processing on the multi-dimensional data to obtain second multi-dimensional data; extracting features of the second multi-dimensional data, wherein the features comprise a basic meteorological feature, a meteorological change feature, a meteorological accumulation feature, a time sequence feature and a holiday and seasonal feature; and on the basis of the features, power load prediction is carried out by using a target model, and the target model comprises a dynamic weighting mechanism. Through multi-level preprocessing of data, innovative feature extraction and dynamic weighting of meteorological data, adaptability and precision of the model in a high-frequency prediction scene are enhanced.

Description

Technical Field

[0001] The present application relates to the technical field of power load forecasting, and in particular to a power load forecasting method, device, electronic device and storage medium. Background Art

[0002] In existing technologies, power load forecasting is usually based on historical load data and simple time series modeling, but there is still room for improvement in forecasting accuracy under complex meteorological conditions and adaptability to different regions. Summary of the invention

[0003] In view of this, the purpose of the present application is to propose a power load forecasting method, device, electronic device and storage medium to solve or partially solve the above-mentioned problems.

[0004] Based on the above objectives, in a first aspect, the present application provides a method for predicting power load, comprising:

[0005] Collecting first multi-dimensional data for power load forecasting, wherein the first multi-dimensional data includes historical load data, meteorological data, and holiday data;

[0006] Performing data cleaning and smoothing processing on the multi-dimensional data to obtain second multi-dimensional data;

[0007] Extracting features of the second multi-dimensional data, the features including basic meteorological features, meteorological change features, meteorological accumulation features, time series features, and holiday and seasonal features;

[0008] Based on the characteristics, a target model is used to perform power load forecasting, wherein the target model includes a dynamic weighting mechanism.

[0009] In a second aspect of the present application, a power load prediction device is provided, comprising:

[0010] A collection module is configured to collect first multi-dimensional data for power load forecasting, wherein the first multi-dimensional data includes historical load data, meteorological data, and holiday data;

[0011] A preprocessing module is configured to perform data cleaning and smoothing on the multi-dimensional data to obtain second multi-dimensional data;

[0012] a feature extraction module configured to extract features of the second multi-dimensional data, wherein the features include basic meteorological features, meteorological change features, meteorological accumulation features, time series features, and holiday and seasonal features;

[0013] The prediction module is configured to perform power load prediction based on the features using a target model, wherein the target model includes a dynamic weighting mechanism.

[0014] In a third aspect of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method described in the first aspect is implemented.

[0015] According to a fourth aspect of the present application, a non-transitory computer-readable storage medium is provided, wherein the non-transitory computer-readable storage medium stores computer instructions, wherein the computer instructions are used to enable a computer to execute the method described in the first aspect.

[0016] From the above, it can be seen that the present application provides a method, device, electronic device and storage medium for power load forecasting. The method includes: collecting first multi-dimensional data for power load forecasting, the first multi-dimensional data includes historical load data, meteorological data and holiday data; performing data cleaning and smoothing on the multi-dimensional data to obtain second multi-dimensional data; extracting features of the second multi-dimensional data, the features include basic meteorological features, meteorological change features, meteorological accumulation features, time series features and holiday and seasonal features; based on the features, using a target model to forecast power load, the target model includes a dynamic weighting mechanism. Through multi-level preprocessing of data, innovative feature extraction and dynamic weighting of meteorological data, the adaptability and accuracy of the model in high-frequency prediction scenarios are enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present application or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A schematic diagram of an exemplary data acquisition and preprocessing module according to an embodiment of the present application is shown.

[0019] Figure 2 A schematic diagram of an exemplary feature extraction module according to an embodiment of the present application is shown.

[0020] Figure 3 A schematic diagram of an exemplary target model according to an embodiment of the present application is shown.

[0021] Figure 4 A schematic diagram of an exemplary dynamic meteorological weighting module according to an embodiment of the present application is shown.

[0022] Figure 5 A flow chart of an exemplary power load forecasting method according to an embodiment of the present application is shown.

[0023] Figure 6 A schematic diagram of an exemplary power load prediction device according to an embodiment of the present application is shown.

[0024] Figure 7 A schematic diagram of an exemplary electronic device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0025] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0026] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should be the usual meanings understood by people with ordinary skills in the field to which the present application belongs. The "first", "second" and similar words used in the embodiments of the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing in front of the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0027] In the prior art, power load forecasting is usually based on historical load data and simple time series modeling. Common methods include autoregressive integrated moving average model (ARIMA), support vector machine (SVM) and traditional decision tree model. In recent years, gradient boosting tree models such as XGBoost (eXtremeGradient Boosting, a machine learning algorithm) have performed well in feature extraction and nonlinear modeling, but there is still room for improvement in prediction accuracy under complex meteorological conditions and adaptability to different regions. In addition, the 96-point load forecast involves high-frequency forecasts at the 15-minute level within a day, requiring the model to have a stronger ability to capture short-term load fluctuations, meteorological influences and periodic laws.

[0028] The existing technology has the following deficiencies in the business application of 96-point load forecasting:

[0029] 1. Difficult to dynamically adapt to meteorological changes: In extreme weather conditions (such as heavy rain and high temperatures), traditional models are not able to respond adequately to changes in load demand, resulting in increased prediction errors.

[0030] 2. Weak regional adaptability: The existing models have insufficient generalization capabilities and it is difficult to maintain stable prediction accuracy in different climate regions.

[0031] 3. Incomplete feature extraction: Potential features such as meteorological factors, short-term fluctuations and periodic patterns of load are not fully utilized, which reduces the accuracy of the 96-point load forecast.

[0032] In order to at least solve the above problems, the present application provides a method, device, electronic device and storage medium for power load forecasting. The method includes: collecting first multi-dimensional data for power load forecasting, the first multi-dimensional data includes historical load data, meteorological data and holiday data; performing data cleaning and smoothing on the multi-dimensional data to obtain second multi-dimensional data; extracting features of the second multi-dimensional data, the features include basic meteorological features, meteorological change features, meteorological accumulation features, time series features and holiday and seasonal features; based on the features, using a target model to forecast power load, the target model includes a dynamic weighting mechanism. Through multi-level preprocessing of data, innovative feature extraction and dynamic weighting of meteorological data, the adaptability and accuracy of the model in high-frequency prediction scenarios are enhanced.

[0033] The technical solution of this application mainly includes the following modules:

[0034] First, the data collection and preprocessing module is responsible for collecting and processing historical load, meteorological data and holiday data, and ensures the consistency and integrity of the data through multi-level data cleaning, outlier repair, time alignment and standardization. Next, the feature extraction module extracts multi-dimensional features such as meteorology, time series and holidays based on the preprocessed data, dynamically adjusts the feature construction strategy to adapt to the needs of different meteorological conditions, and introduces rolling mean and cumulative features to capture the complex changes of load. Subsequently, in the model construction and training module, the XGBoost model implements load forecasting at the core algorithm layer, establishes a nonlinear mapping relationship between features and load demand through a hierarchical combination of decision tree structures, and introduces a dynamic weighting mechanism in the dynamic meteorological weighting module to adaptively adjust the meteorological feature weights according to the current meteorological conditions, so that the model has higher prediction accuracy in extreme weather. Finally, the model optimization and parameter adjustment module optimizes the hyperparameters of the XGBoost model through grid search and cross-validation to ensure the stability and generalization ability of the model in different regions and under various meteorological conditions.

[0035] Figure 1 A schematic diagram of an exemplary data acquisition and preprocessing module according to an embodiment of the present application is shown.

[0036] like Figure 1As shown, in the data acquisition and preprocessing module, in some embodiments, multi-dimensional data required for 96-point load forecasting can be collected, including historical load data, meteorological data, and holiday data. The historical load data covers the load values ​​of 96 time nodes (every 15 minutes interval) of the current day, and is combined with the load data at the same time point in the past 7 days to provide a comprehensive understanding of short-term load fluctuation trends. Meteorological data includes temperature, humidity, wind speed and precipitation, which are obtained on a daily basis through the meteorological interface to ensure the timeliness of the data. Special holiday situations are used to identify abnormal fluctuations in load demand.

[0037] To ensure the validity and consistency of the data, in some embodiments, multi-level data cleaning and smoothing processing can be introduced in the preprocessing process, specifically including:

[0038] Anomaly detection and repair: The seasonal weighted moving average (SWMA) method is used to detect anomalies in historical load data, meteorological data, and holiday data, and the interpolation method is used to repair the anomaly data.

[0039] Time alignment and missing value filling: For data with missing time nodes, multiple interpolation methods are used to fill them; for data with different time resolutions, linear interpolation alignment is performed to ensure data integrity.

[0040] Data smoothing and standardization: All data are smoothed using a sliding window to remove high-frequency noise, and are standardized so that features of different dimensions have consistent influence in the model.

[0041] Figure 2 A schematic diagram of an exemplary feature extraction module according to an embodiment of the present application is shown.

[0042] like Figure 2 As shown in the figure, in the 96-point load forecasting business scenario, the feature extraction of the second dimension data (e.g., meteorological data and load data) is crucial to the prediction effect of the model. In view of this, the embodiment of the present application has made innovations in feature extraction so that the XGBoost model can make full use of the periodicity of the load and the complex impact of the weather on the load demand. In the feature extraction module, the main features are as follows:

[0043] Basic meteorological characteristics: directly extract the original values ​​of temperature, humidity, wind speed, and precipitation to reflect the impact of current meteorological conditions on load.

[0044] Meteorological change characteristics: Calculate the change rate of meteorological variables (such as temperature and humidity change rate) to capture meteorological fluctuations, which is particularly important in 96-point high-frequency forecasts. The specific calculation formula is as follows:

[0045]

[0046] Among them, T t and H t They represent the temperature and humidity at a certain point in time, Δt represents the time interval, t represents the time, ΔT and ΔH represent the temperature change rate and humidity change rate respectively.

[0047] Meteorological accumulation characteristics: In order to capture the impact of meteorological accumulation in the short term, precipitation accumulation and average wind speed are introduced. The precipitation accumulation is calculated by the sum of precipitation in the past 7 days, and the formula is as follows:

[0048]

[0049] Among them, R i represents the precipitation at time point i to capture the continuous impact of precipitation on load, R cumulative Indicates the amount of precipitation accumulated.

[0050] Time series characteristics: Based on historical load data, the load at the same time point of the previous day, the load average at the same time point of the past week, and the 48-point rolling average are constructed to capture the cyclical laws and short-term trends of the load. The specific formula is as follows:

[0051] L day-1 =L t-96

[0052]

[0053] Among them, L day-1 represents the load at the same time point the previous day, Indicates the average load at the same time point in the past week, L rolling represents the 48-point rolling mean, k represents the value of the day, j represents the time point, L j Indicates the actual load value at a point in time.

[0054] Holiday and seasonal characteristics: By introducing holiday identifiers and seasonal factors, the model can capture the load characteristics of holidays and seasonal changes. Holiday characteristics are identified by binary variables, and seasonal characteristics are encoded by sine (sin) and cosine (cos) functions. The formula is as follows:

[0055]

[0056] Among them, day o f y ear represents the number of days in a year for the current date, 2π represents a complete cycle, mathrmSeason sin Represents the seasonal sine value of a day in a year, Season cos Represents the seasonal cosine value of a day of the year. The days of the year are normalized to the interval [0,1], that is, the first day is mapped to 0 and the 365th day is mapped to 1. Then the date is mapped to the period of a complete circle (unit circle) by multiplying by 2π.

[0057] In the model building and training module, in some embodiments, the XGBoost model can be used as the core algorithm for load forecasting. XGBoost is a tree model based on gradient boosting, which has efficient feature extraction capabilities and strong nonlinear modeling capabilities. During the model training process, XGBoost fits nonlinear relationships through layer-by-layer modeling, combining the input multidimensional meteorological features, time series features, and holiday features to achieve accurate prediction of load demand.

[0058] Figure 3 A schematic diagram of an exemplary target model according to an embodiment of the present application is shown.

[0059] like Figure 3 As shown in the figure, the load is modeled based on the input multi-dimensional features (such as temperature, humidity and other meteorological data, historical load, holiday characteristics). The target model constructs multiple decision trees (CATE trees) through the splitting and pruning mechanism, fits the nonlinear relationship between the input features and the load in a hierarchical manner, and uses the gradient boosting optimization strategy to gradually reduce the prediction error. Finally, the outputs of all decision trees are integrated through the model to complete the future 96-point load prediction results.

[0060] Figure 4 A schematic diagram of an exemplary dynamic meteorological weighting module according to an embodiment of the present application is shown.

[0061] like Figure 4 As shown, in the dynamic meteorological weighting module, in order to improve the adaptability of the model under extreme meteorological conditions, in some embodiments, a dynamic weighting mechanism can be introduced to obtain load forecast results based on historical load data and meteorological data using a dynamic weighting mechanism. When meteorological data indicates extreme weather (such as heavy rain or high temperature), the weight of meteorological features is automatically increased to make the model more sensitive to meteorological changes. The calculation formula for dynamic weighting is as follows:

[0062] L pred =α·L hiist +β·L weather

[0063] Among them, L pred represents the final predicted load, L hiist and L weatherThey represent the prediction results of historical load data and meteorological data respectively. α represents the weight coefficient for adjusting the prediction results of historical load data, which is used to balance the contribution of historical load and meteorological conditions to the prediction results. β represents the weight coefficient for adjusting the prediction results of meteorological data. β is automatically increased in extreme weather to strengthen the impact of meteorological factors on the load.

[0064] In the model optimization and parameter adjustment module, in some embodiments, the hyperparameters of the XGBoost model (such as tree depth, minimum sample weight, and learning rate) are optimized through a grid search method, combined with cross-validation to ensure the generalization ability of the model. Finally, the load forecast values ​​of 96 time nodes are generated through the trained model, and the load fluctuation trend is displayed in a curve, providing an intuitive reference for power dispatching.

[0065] Evaluation results after using the method provided in this application:

[0066]

[0067] Figure 5 FIG. 5 is a flow chart showing an exemplary power load forecasting method 500 according to an embodiment of the present application. Figure 5 As shown, method 500 may include the following steps.

[0068] In step 502, first multi-dimensional data for power load forecasting is collected, wherein the first multi-dimensional data includes historical load data, meteorological data, and holiday data.

[0069] In step 504, data cleaning and smoothing are performed on the multi-dimensional data to obtain second multi-dimensional data.

[0070] In some embodiments, the data cleaning and smoothing of the multidimensional data to obtain the second multidimensional data further includes: detecting outliers in the multidimensional data; repairing the outliers using an interpolation method to obtain the second multidimensional data; and / or in response to the presence of data with missing time nodes in the multidimensional data, filling the data with missing time nodes using a multiple interpolation method to obtain the second multidimensional data; in response to the presence of data with different time resolutions in the multidimensional data, aligning the data with different time resolutions using a linear interpolation method to obtain the second multidimensional data; and / or performing sliding window smoothing and standardization on the multidimensional data to obtain the second multidimensional data.

[0071] In step 506, features of the second multi-dimensional data are extracted, wherein the features include basic meteorological features, meteorological change features, meteorological accumulation features, time series features, and holiday and seasonal features.

[0072] In some embodiments, the extracting of features of the second multi-dimensional data further includes: extracting the temperature, humidity, wind speed and precipitation of the second multi-dimensional data to obtain the basic meteorological features; based on the temperature and humidity, calculating the rate of change of meteorological variables to obtain the meteorological change features; based on the precipitation, calculating the accumulated precipitation to obtain the meteorological accumulation features.

[0073] In some embodiments, the meteorological change characteristics include a temperature change rate and a humidity change rate, and the temperature change rate is expressed as: The humidity change rate is expressed as: Among them, T t and H t Respectively represent the temperature and humidity at a time point, Δt represents the time interval, t represents the time, ΔT and ΔH represent the temperature change rate and humidity change rate respectively; the meteorological accumulation characteristics are expressed as: Among them, R i represents the precipitation at time point i, R cumulative Indicates the amount of precipitation accumulated.

[0074] In some embodiments, the extracting of the features of the second multi-dimensional data further includes: constructing the load at the same time point of the previous day, the load mean at the same time point of the past week, and the 48-point rolling mean based on the historical load data to obtain the time series features; the load at the same time point of the previous day is expressed as: L day-1 =L t-96 ; The load mean at the same time point in the past week is expressed as: The 48-point rolling mean is expressed as: Among them, L day-1 represents the load at the same time point the previous day, Indicates the average load at the same time point in the past week, L rolling represents the 48-point rolling mean, k represents the value of the day, j represents the time point, L j Indicates the actual load value at a point in time.

[0075] In some embodiments, extracting the features of the second multi-dimensional data further includes: extracting the holiday and seasonal features in the holiday data based on the constructed seasonal model, and the seasonal model is expressed as:

[0076]

[0077] Among them, day o f y ear represents the number of days in a year for the current date, 2π represents a complete cycle, mathrmSeason sinRepresents the seasonal sine value of a day in a year, Season cos Represents the seasonal cosine of the day of the year.

[0078] In step 508, based on the features, a target model is used to perform power load forecasting, wherein the target model includes a dynamic weighting mechanism.

[0079] In some embodiments, the dynamic weighting mechanism is expressed as: pred =α·L hist +β·L weather Among them, L pred represents the final predicted load, L hist and L weather They represent the prediction results of historical load characteristics and meteorological characteristics respectively, α represents the weight coefficient for adjusting the prediction results of historical load data, and β represents the weight coefficient for adjusting the prediction results of meteorological data.

[0080] The present application provides a method, device, electronic device and storage medium for power load forecasting. The method includes: collecting first multi-dimensional data for power load forecasting, the first multi-dimensional data including historical load data, meteorological data and holiday data; performing data cleaning and smoothing on the multi-dimensional data to obtain second multi-dimensional data; extracting features of the second multi-dimensional data, the features including basic meteorological features, meteorological change features, meteorological accumulation features, time series features and holiday and seasonal features; based on the features, using a target model to forecast power load, the target model including a dynamic weighting mechanism. Through multi-level preprocessing of data, innovative feature extraction and dynamic weighting of meteorological data, the adaptability and accuracy of the model in high-frequency forecasting scenarios are enhanced. The method provided in the present application is suitable for high-precision forecasting of load demand at 96 points (every 15 minutes interval) within a day, providing support for stable scheduling and optimized management of power systems.

[0081] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the described method.

[0082] It should be noted that the above describes some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0083] Based on the same technical concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a power load prediction device.

[0084] refer to Figure 6 , the power load forecasting device comprises:

[0085] The acquisition module 601 is configured to acquire first multi-dimensional data for power load forecasting, wherein the first multi-dimensional data includes historical load data, meteorological data, and holiday data.

[0086] The preprocessing module 602 is configured to perform data cleaning and smoothing on the multi-dimensional data to obtain second multi-dimensional data.

[0087] The preprocessing module 602 is also configured to perform data cleaning and smoothing on the multidimensional data to obtain the second multidimensional data, further including: detecting outliers in the multidimensional data; repairing the outliers using an interpolation method to obtain the second multidimensional data; and / or in response to the presence of data with missing time nodes in the multidimensional data, filling the data with missing time nodes using a multiple interpolation method to obtain the second multidimensional data; in response to the presence of data with different time resolutions in the multidimensional data, aligning the data with different time resolutions using a linear interpolation method to obtain the second multidimensional data; and / or performing sliding window smoothing and standardization on the multidimensional data to obtain the second multidimensional data.

[0088] The feature extraction module 603 is configured to extract features of the second multi-dimensional data, wherein the features include basic meteorological features, meteorological change features, meteorological accumulation features, time series features, and holiday and seasonal features.

[0089] The feature extraction module 603 is also configured to further include: extracting the temperature, humidity, wind speed and precipitation of the second multi-dimensional data to obtain the basic meteorological characteristics; based on the temperature and humidity, calculating the rate of change of meteorological variables to obtain the meteorological change characteristics; based on the precipitation, calculating the accumulated precipitation to obtain the meteorological accumulation characteristics.

[0090] The feature extraction module 603 is further configured such that the meteorological change feature includes a temperature change rate and a humidity change rate, and the temperature change rate is expressed as: The humidity change rate is expressed as: Among them, T t and H t Respectively represent the temperature and humidity at a time point, Δt represents the time interval, t represents the time, ΔT and ΔH represent the temperature change rate and humidity change rate respectively; the meteorological accumulation characteristics are expressed as: Among them, R i represents the precipitation at time point i, R cumulative Indicates the amount of precipitation accumulated.

[0091] The feature extraction module 603 is further configured to extract the features of the second multi-dimensional data further including: constructing the load at the same time point of the previous day, the load mean at the same time point of the past week and the 48-point rolling mean based on the historical load data to obtain the time series features; the load at the same time point of the previous day is expressed as: L day-1 =L t-96 ; The load mean at the same time point in the past week is expressed as: The 48-point rolling mean is expressed as: Among them, L day-1 represents the load at the same time point the previous day, Indicates the average load at the same time point in the past week, L rolling represents the 48-point rolling mean, k represents the value of the day, j represents the time point, L j Indicates the actual load value at a point in time.

[0092] The feature extraction module 603 is further configured to extract the features of the second multi-dimensional data further including: extracting the holiday and seasonal features in the holiday data based on the constructed seasonal model, and the seasonal model is expressed as:

[0093]

[0094] Among them, day o f y ear represents the number of days in a year for the current date, 2π represents a complete cycle, mathrmSeason sin Represents the seasonal sine value of a day in a year, Season cos Represents the seasonal cosine of the day of the year.

[0095] The prediction module 604 is configured to perform power load prediction based on the features using a target model, wherein the target model includes a dynamic weighting mechanism.

[0096] The prediction module 604 is further configured to express the dynamic weighting mechanism as: L pred =α·L hist +β·L weather Among them, L pred represents the final predicted load, L hist and L weather They represent the prediction results of historical load characteristics and meteorological characteristics respectively, α represents the weight coefficient for adjusting the prediction results of historical load data, and β represents the weight coefficient for adjusting the prediction results of meteorological data.

[0097] For the convenience of description, the above device is described in terms of functions divided into various modules. Of course, when implementing the present application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0098] The device of the above embodiment is used to implement the corresponding method 500 in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0099] Based on the same technical concept, corresponding to any of the above-mentioned embodiments, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method 500 described in any of the above embodiments when executing the program.

[0100] Figure 7 A schematic diagram of an exemplary electronic device according to an embodiment of the present application is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other in communication within the device through the bus 1050.

[0101] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0102] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0103] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0104] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).

[0105] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0106] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0107] The electronic device of the above embodiment is used to implement the corresponding method 500 in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0108] Based on the same technical concept, corresponding to any of the above-mentioned embodiments, the present application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute method 500 described in any of the above embodiments.

[0109] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0110] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the method 500 described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0111] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present application (including the claims) is limited to these examples. In line with the concept of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.

[0112] In addition, to simplify the description and discussion, and in order not to make the embodiments of the present application difficult to understand, the known power supply / ground connection with the integrated circuit (IC) chip and other components may or may not be shown in the provided drawings. In addition, the device can be shown in the form of a block diagram to avoid making the embodiments of the present application difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform to be implemented in the embodiments of the present application (that is, these details should be fully within the scope of understanding of those skilled in the art). In the case of elaborating specific details (e.g., circuits) to describe exemplary embodiments of the present application, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details or when these specific details are changed. Therefore, these descriptions should be considered to be illustrative rather than restrictive.

[0113] Although the present application has been described in conjunction with specific embodiments of the present application, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.

[0114] The embodiments of the present application are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the scope of protection of the present application.

Claims

1. A method for predicting power load, comprising: Collecting first multi-dimensional data for power load forecasting, wherein the first multi-dimensional data includes historical load data, meteorological data, and holiday data; Performing data cleaning and smoothing processing on the multi-dimensional data to obtain second multi-dimensional data; Extracting features of the second multi-dimensional data, the features including basic meteorological features, meteorological change features, meteorological accumulation features, time series features, and holiday and seasonal features; Based on the characteristics, a target model is used to perform power load forecasting, wherein the target model includes a dynamic weighting mechanism.

2. The method of claim 1, wherein: The performing data cleaning and smoothing on the multi-dimensional data to obtain second multi-dimensional data further includes: Detecting outliers in the multidimensional data; Repairing the abnormal value by using an interpolation method to obtain the second multi-dimensional data; and / or In response to the presence of data of missing time nodes in the multi-dimensional data, fill the data of the missing time nodes by using a multiple interpolation method to obtain the second multi-dimensional data; In response to the presence of data with different time resolutions in the multi-dimensional data, aligning the data with different time resolutions by using a linear interpolation method to obtain the second multi-dimensional data; and / or The multi-dimensional data is subjected to sliding window smoothing processing and standardization processing to obtain the second multi-dimensional data.

3. The method of claim 1, wherein: The extracting the features of the second multi-dimensional data further comprises: extracting temperature, humidity, wind speed and precipitation from the second multi-dimensional data to obtain the basic meteorological characteristics; Based on the temperature and humidity, calculating the rate of change of the meteorological variable to obtain the meteorological change characteristics; Based on the precipitation, the precipitation accumulation is calculated to obtain the meteorological accumulation characteristics.

4. The method of claim 3, wherein: The meteorological change characteristics include temperature change rate and humidity change rate. The temperature change rate is expressed as: The humidity change rate is expressed as: Among them, T t and H t Respectively represent the temperature and humidity at a time point, Δt represents the time interval, t represents the time, ΔT and ΔH represent the temperature change rate and humidity change rate respectively; The meteorological accumulation characteristics are expressed as: Among them, R i represents the precipitation at time point i, E cumulative Indicates the amount of precipitation accumulated.

5. The method of claim 1, wherein: The extracting the features of the second multi-dimensional data further comprises: Based on the historical load data, the load at the same time point of the previous day, the load average at the same time point of the past week and the 48-point rolling average are constructed to obtain the time series characteristics; The load at the same time point on the previous day is expressed as: L day-1 =L t-96 ; The load mean at the same time point in the past week is expressed as: The 48-point rolling mean is expressed as: Among them, L day-1 represents the load at the same time point the previous day, Indicates the average load at the same time point in the past week, L rolling represents the 48-point rolling mean, k represents the value of the day, j represents the time point, L j Indicates the actual load value at a point in time.

6. The method of claim 1, wherein: The extracting the features of the second multi-dimensional data further comprises: The holiday and seasonal characteristics in the holiday data are extracted based on the constructed seasonal model, and the seasonal model is expressed as: Among them, day o f y ear represents the number of days in a year for the current date, 2π represents a complete cycle, mathrmSeason sin Represents the seasonal sine value of a day in a year, Season cos Represents the seasonal cosine of the day of the year.

7. The method of claim 1, wherein: The dynamic weighting mechanism is expressed as: L pred =α·L hiist +β·L weather ; Among them, L pred represents the final predicted load, L hiist and L weather They represent the prediction results of historical load characteristics and meteorological characteristics respectively, α represents the weight coefficient for adjusting the prediction results of historical load data, and β represents the weight coefficient for adjusting the prediction results of meteorological data.

8. A power load forecasting device, comprising: A collection module is configured to collect first multi-dimensional data for power load forecasting, wherein the first multi-dimensional data includes historical load data, meteorological data, and holiday data; A preprocessing module is configured to perform data cleaning and smoothing on the multi-dimensional data to obtain second multi-dimensional data; a feature extraction module configured to extract features of the second multi-dimensional data, wherein the features include basic meteorological features, meteorological change features, meteorological accumulation features, time series features, and holiday and seasonal features; The prediction module is configured to perform power load prediction based on the features using a target model, wherein the target model includes a dynamic weighting mechanism.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.

Images