Electric quantity prediction method, device and equipment
By acquiring and processing historical electricity consumption, business expansion application volume, and holiday data for the target area, and combining them with temperature data, the X13 algorithm is used to separate the trend term and the seasonal holiday effect term. This solves the problem of insufficient electricity consumption prediction in existing models during holidays and extreme temperatures, and improves the prediction accuracy.
Patent Information
- Application Number
- CN202511507484.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-21
- Publication Date
- 2026-01-23
AI Technical Summary
Existing electricity consumption forecasting models are unable to effectively capture the complex electricity consumption fluctuations caused by holidays, fail to fully reflect the potential for regional electricity consumption growth, and are not adaptable enough to nonlinear load changes under extreme temperature conditions, resulting in insufficient forecast accuracy.
By acquiring historical electricity consumption, business expansion application volume, holiday and temperature data for the target area, holiday windows are determined, and the X13 algorithm is used to process these data to separate historical trend terms, seasonal holiday effect terms and residual random terms. Electricity consumption is then predicted by combining holiday effect values, business expansion application volume and temperature data.
It improves the accuracy of electricity consumption forecasting, better reflects the impact of holidays and temperature on electricity consumption, and enhances the forecasting ability under extreme temperature conditions.
Smart Images
Figure CN121395280A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] Embodiments of the present application relate to the technical field of power system data processing, and in particular to a power consumption prediction method, device and equipment. BACKGROUND
[0002] Power consumption (also known as electricity sales) prediction is a key link for power system planning, resource allocation and market decision. Existing prediction models, especially those based on general seasonal adjustment algorithms (such as X11 or X13), have significant limitations: it is difficult to effectively capture the complex power consumption fluctuation rules caused by holidays; it fails to fully integrate the key leading indicator of industry expansion load to reflect the regional power consumption growth potential; and it lacks adaptability to nonlinear load changes under extreme temperature. These problems result in prediction accuracy that is difficult to meet the needs of fine operation of the power system. SUMMARY
[0003] Embodiments of the present application provide a power consumption prediction method, device and equipment, which can improve the accuracy of power consumption prediction.
[0004] In a first aspect, embodiments of the present application provide a power consumption prediction method, comprising:
[0005] obtaining historical power consumption data, industry expansion load data, holiday data and temperature data of a target region;
[0006] determining a holiday window based on the historical power consumption data and the holiday data; wherein the holiday window includes a pre-holiday sub-window, a holiday sub-window and a post-holiday sub-window;
[0007] determining a holiday effect value based on the holiday window and the historical power consumption data;
[0008] processing the historical power consumption data, exogenous variable data and the holiday data based on an X13 algorithm to obtain a historical trend item, a historical seasonal holiday effect item and a historical residual random item; wherein the exogenous variable data includes the holiday effect value, the industry expansion load data and the temperature data;
[0009] predicting power consumption based on the historical trend item, the historical seasonal holiday effect item and the historical residual random item.
[0010] In a second aspect, embodiments of the present application also provide a power consumption prediction device, comprising:
[0011] a data acquisition module configured to obtain historical power consumption data, industry expansion load data, holiday data and temperature data of a target region;
[0012] a holiday window determination module configured to determine a holiday window based on the historical power consumption data and the holiday data, wherein the holiday window comprises a pre-holiday sub-window, a mid-holiday sub-window and a post-holiday sub-window;
[0013] a holiday effect value determination module configured to determine a holiday effect value based on the holiday window and the historical power consumption data;
[0014] a data processing module configured to process the historical power consumption data, exogenous variable data and the holiday data based on an X13 algorithm to obtain a historical trend item, a historical seasonal holiday effect item and a historical residual random item, wherein the exogenous variable data comprises the holiday effect value, the industry expansion and installation data and the air temperature data;
[0015] a power consumption prediction module configured to predict power consumption based on the historical trend item, the historical seasonal holiday effect item and the historical residual random item.
[0016] In a third aspect, an electronic device is provided, and the electronic device comprises:
[0017] at least one processor; and
[0018] a memory connected with the at least one processor; wherein
[0019] the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the power consumption prediction method according to any one of the embodiments of the present application.
[0020] The embodiment of the present application discloses a power prediction method, device and equipment. Historical power consumption data, industry expansion and installation data, holiday data and temperature data of a target area are acquired; a holiday window is determined based on the historical power consumption data and the holiday data; wherein, the holiday window comprises a pre-holiday sub-window, a holiday sub-window and a post-holiday sub-window; a holiday effect value is determined based on the holiday window and the historical power consumption data; the historical power consumption data, exogenous variable data and holiday data are processed based on an X13 algorithm to obtain a historical trend item, a historical seasonal holiday effect item and a historical residual random item; wherein, the exogenous variable data comprises the holiday effect value, the industry expansion and installation data and the temperature data; power consumption is predicted based on the historical trend item, the historical seasonal holiday effect item and the historical residual random item. The power prediction method provided by the embodiment of the present application, after the holiday effect value is determined, the holiday effect value, the industry expansion and installation data and the temperature data are output as exogenous variables to the X13 algorithm, so that the X13 algorithm splits the historical power consumption data into the historical trend item, the historical seasonal holiday effect item and the historical residual random item based on the exogenous variables to predict the power consumption, that is, the holiday data, the temperature data and the industry expansion and installation data are considered when the power consumption is predicted, so that the accuracy of the power consumption prediction can be improved. BRIEF DESCRIPTION OF DRAWINGS
[0021] Figure 1 is a flowchart of a power prediction method in the embodiment one of the present application;
[0022] Figure 2 is a structural schematic diagram of a power prediction device in the embodiment two of the present application;
[0023] Figure 3 is a structural schematic diagram of an electronic device in the embodiment three of the present application. DETAILED DESCRIPTION
[0024] The present application will be further described below in conjunction with the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application. In addition, it should be noted that, for the convenience of description, only the parts related to the present application are shown in the drawings, but not all the structures.
[0025] Embodiment one
[0026] Figure 1 is a flowchart of a power prediction method provided by the embodiment one of the present application, the embodiment can be applicable to the case of predicting power consumption, the method can be executed by a power prediction device, the device can be realized in the form of software and / or hardware, and can be realized by an electronic device, which can be a mobile terminal, a PC terminal or a server, etc. Specifically, the method comprises the following steps:
[0027] S110, obtain historical power consumption data, industry expansion installation data, holiday data and temperature data of the target area.
[0028] The historical power consumption data can include daily power consumption sequences and monthly power consumption sequences in the past N years; the industry expansion installation data can include monthly industry expansion installation data sequences in the past N years; the holiday data can include holiday data in the past N years with rest rules; and the temperature data can be meteorological data of daily maximum temperature in the past N years.
[0029] In this embodiment, the historical power consumption data, the industry expansion installation data and the temperature data also need to be filled with missing values and corrected for abnormal values.
[0030] For the historical power consumption data, linear interpolation combined with periodic correction terms can be used to process missing values (long holiday data is filled based on historical holiday mean values); for the industry expansion installation data and the temperature data, non-linear interpolation can be used to complete the filling of missing values. The 3σ criterion can be used in combination with business logic to correct abnormal values. After filling the missing values and correcting the abnormal values, the industry expansion installation data can also be logarithmically transformed to eliminate the right-skewed distribution.
[0031] S120, determine a holiday window based on the historical power consumption data and the holiday data.
[0032] The holiday window includes a pre-holiday sub-window, a holiday sub-window and a post-holiday sub-window. The holiday window includes a window start date and a window end date.
[0033] The pre-holiday sub-window is D1 days before the first day of the holiday to the day before the first day of the holiday; the holiday sub-window is the entire holiday; and the post-holiday sub-window is one day after the end of the holiday to D2 days after the end of the holiday. D1 is the number of days in the pre-holiday sub-window, and D2 is the number of days in the post-holiday sub-window. Assuming T0 is the first day of the holiday and N is the number of days of the holiday, the pre-holiday sub-window can be represented as: [T0-D1, T0-1], the holiday sub-window can be represented as: [T0, T0+N-1], and the post-holiday sub-window can be represented as: [T0+N, T0+N-1+D2].
[0034] In this embodiment, in view of the complexity of the holiday system, two types of differentiated effect window mechanisms can be constructed, and a double calendar engine is used to realize high-precision mapping of double calendar dates. The holiday window can include a first calendar holiday window and a second calendar holiday window. The first calendar can be understood as the Gregorian calendar, and the second calendar can be understood as the lunar calendar. The Gregorian calendar holidays can include New Year's Day, May Day, National Day, etc.; and the lunar calendar holidays can include the Spring Festival, the Mid-Autumn Festival, etc.
[0035] Optionally, the way of determining the holiday window based on the historical power consumption data and the holiday data can be: for the first calendar holiday window, determining the pre-holiday sub-window and the post-holiday sub-window based on the power consumption fluctuation rate, and determining the holiday sub-window based on the holiday data.
[0036] wherein the power consumption fluctuation rate is determined based on the mean and the standard deviation of the historical power consumption data, and the formula can be expressed as CV = σ / μ, CV is the power consumption fluctuation rate, σ is the power consumption standard deviation, and μ is the power consumption mean. Specifically, the process of determining the pre-holiday sub-window and the post-holiday sub-window can be: starting from T0-1 and scanning forward to find the first date that satisfies CV(d)≤a set value (for example: 0.25) as the window start date, and taking the window start date to the day before the holiday start date as the pre-holiday sub-window. Starting from T0+N and scanning backward to find the first date that satisfies CV(d)≤0.25 as the window end date, and taking T0+N to the window end date as the post-holiday sub-window.
[0037] Optionally, the way of determining the holiday window based on the historical power consumption data and the holiday data can be: for the second calendar holiday window, determining the pre-holiday sub-window based on the first preset number of days and determining the post-holiday sub-window based on the second preset number of days, and determining the holiday sub-window based on the holiday data.
[0038] wherein the first preset number of days and the second preset number of days are the same or different. The first preset number of days can be understood as D1 described above, and the second preset number of days can be understood as D2 described above.
[0039] In this embodiment, the window offset can also be determined according to the proportion of power consumption of each industry in the target area. Assuming that Δ is the window offset, the calculation formula can be expressed as: , wherein P ind is the proportion of industrial power consumption, P com is the proportion of commercial power consumption, E ind is the industrial elasticity coefficient, and E com is the commercial elasticity coefficient. The industrial elasticity coefficient is determined by the ratio of the total power consumption growth rate to the industrial output value growth rate, and the commercial elasticity coefficient is determined by the ratio of the total power consumption growth rate to the commercial activity index growth rate. is the weight, which is estimated by ridge regression. After obtaining the window offset, the pre-holiday sub-window can be expressed as: [T0-D1+Δ, T0-1], and the post-holiday sub-window can be expressed as: [T0+N, T0+N-1+D2+Δ].
[0040] In the application scenario, extreme temperature also affects the power consumption (caused by refrigeration or heating), so when extreme temperature occurs during the holiday, the holiday window needs to be adjusted. The extreme temperature can be understood as the maximum temperature being greater than the first threshold or less than the second threshold, the first threshold can be understood as the high temperature threshold, and the second threshold can be understood as the low temperature threshold. The first threshold and the second threshold are determined based on historical power consumption data and historical temperature data.
[0041] Specifically, after determining the holiday window based on historical power consumption data and holiday data, the following steps are included: if the temperature of any day in the holiday window meets the following conditions: greater than the first threshold or less than the second threshold, then determine the window adjustment amount according to the temperature data and the reference temperature; determine the power consumption proportion of each industry in the target area based on the historical power consumption data; adjust the holiday window based on the power consumption proportion of each industry and the window adjustment amount.
[0042] Among them, the industrial power consumption includes industrial power consumption, commercial power consumption and residential power consumption. The first threshold is greater than the second threshold. The window adjustment amount can be understood as the number of days of window delay or extraction. The calculation formula of the window adjustment amount according to the temperature data and the reference temperature can be expressed as: , wherein, is the regional sensitivity coefficient, which reflects the sensitivity of regional power consumption to temperature, which can be determined by the proportion of power consumption change caused by unit temperature change and the average power consumption, T is the maximum temperature of the day, is the reference temperature, which can be obtained by analyzing the lowest point of the historical temperature-power curve.
[0043] In this embodiment, when the power consumption proportion of a certain industry exceeds a certain value (such as 50%), it can be considered that the target area is dominated by that industry. Specifically, the way to adjust the holiday window based on the power consumption proportion of each industry and the window adjustment amount can be: for an industrial dominant area, if the temperature of any day in the holiday window is greater than the first threshold, the holiday window is delayed based on the window adjustment amount; if the temperature of any day in the holiday window is less than the second threshold, the holiday window is advanced based on the window adjustment amount. For commercial or residential dominant areas, if the temperature of any day in the holiday window is greater than the first threshold, the holiday window is advanced based on the window adjustment amount; if the temperature of any day in the holiday window is less than the second threshold, the holiday window is delayed based on the window adjustment amount.
[0044] S130, determine the holiday effect value based on the holiday window and the historical power consumption data.
[0045] The holiday effect value is used to reflect the influence of the holiday on the power consumption. In this embodiment, for the three sub-windows of the holiday window, different functions can be used to determine the power consumption in the corresponding sub-window.
[0046] Specifically, the manner of determining the holiday effect value based on the holiday window and the historical power consumption data can be: determining the benchmark power consumption of the pre-holiday sub-window starting day and the benchmark power consumption of the post-holiday sub-window starting day based on the historical power consumption data as the first benchmark power consumption and the second benchmark power consumption; determining the power consumption of each day in the pre-holiday sub-window based on the first benchmark power consumption and the first function as the pre-holiday power consumption; determining the power consumption of each day in the post-holiday sub-window based on the second benchmark power consumption and the second function as the post-holiday power consumption; determining the power consumption of each day in the post-holiday sub-window based on the average power consumption during the holiday in the historical power consumption data as the mid-holiday power consumption; and determining the holiday effect value of each day in the holiday window based on the pre-holiday power consumption, the mid-holiday power consumption and the post-holiday power consumption.
[0047] The benchmark power consumption of the pre-holiday sub-window starting day and the post-holiday sub-window starting day can be determined based on the mean or median of the historical power consumption on the same date.
[0048] The first function can be a linear growth function, reflecting the gradual increase of power consumption during the holiday preparation period. The calculation formula (i.e., the first function) of the pre-holiday power consumption can be expressed as: , wherein, is the first benchmark power consumption, t1 is the first day of the pre-holiday sub-window, and t is a certain day in the pre-holiday sub-window. is the pre-holiday daily growth rate, which is determined by linear regression fitting the slope of the historical pre-holiday power consumption sequence.
[0049] For the mid-holiday sub-window, the power consumption of each day in the mid-holiday sub-window can be determined based on the average power consumption during the holiday in the historical power consumption data, and the formula is: , is the average power consumption during the holiday in the historical power consumption data.
[0050] The second function can be an exponential decay function, reflecting the gradual recovery of power consumption after the end of the holiday. The calculation formula (i.e., the second function) of the post-holiday power consumption can be expressed as: , wherein, represents the second benchmark power consumption, t is a certain day in the post-holiday sub-window, and t3 is the first day of the post-holiday sub-window. is the decay rate, which is determined by nonlinear least squares fitting of the exponential curve based on the historical power consumption data.
[0051] Specifically, the way of determining the holiday effect value of each day in the holiday window based on the pre-holiday effect application power, the mid-holiday effect application power and the post-holiday effect application power can be: averaging the pre-holiday effect application power, averaging the post-holiday effect application power, and finally weighting and summing the average pre-holiday effect application power, the mid-holiday effect application power and the average post-holiday effect application power to obtain the holiday effect value. The calculation formula can be represented as: , wherein, , respectively, represent the average pre-holiday effect application power, the mid-holiday effect application power and the average post-holiday effect application power, , respectively, represent the weights, which can be obtained by Akaike Information Criterion (AIC) optimization.
[0052] Optionally, after determining the holiday effect value based on the holiday window and the historical power consumption data, the following steps are further included: determining the power consumption proportion of each industry in the target area based on the historical power consumption data; and adjusting the holiday effect value based on the power consumption proportion of each industry.
[0053] Among them, the industrial power consumption includes industrial power consumption, commercial power consumption and residential power consumption. Specifically, the calculation formula for adjusting the holiday effect value based on the power consumption proportion of each industry is: , wherein, is the weight factor of the i-th industry, which is determined by historical power consumption proportion normalization, is the power consumption activity coefficient of the i-th industry on the t-th day, reflecting the expansion speed of industrial power consumption.
[0054] Optionally, the holiday effect value can also be adjusted according to the temperature, and the calculation formula can be represented as: , wherein, , respectively, are the high-temperature correction coefficient and the low-temperature correction coefficient, , respectively, represent the high-temperature threshold and the low-temperature threshold.
[0055] Optionally, the way of determining the holiday effect value based on the holiday window and the historical power consumption data can be: when two or more holiday windows overlap, the maximum holiday effect value is taken as the holiday effect value of the overlapping part.
[0056] In this embodiment, when the same date is in the overlapping area of multiple holiday windows, the maximum holiday effect value is taken as the holiday effect value of the day.
[0057] S140, based on the X13 algorithm, the historical power consumption data, the exogenous variable data and the holiday data are processed to obtain the historical trend item, the historical seasonal holiday effect item and the historical residual random item.
[0058] The exogenous variable data includes holiday effect value, industry expansion and installation quantity data, and temperature data. The historical trend item, the historical seasonal holiday effect item, and the historical residual random item are sequentially represented. The X13 algorithm (full name: X13-ARIMA-SEATS) is a time series seasonal adjustment and decomposition tool, and the core principle is to separate the trend item, the seasonal holiday effect item, and the residual random item from the historical power consumption sequence.
[0059] The X13 algorithm includes a moving average filter. When the filter interval of the moving average filter includes a holiday window, the window length of the moving average filter is adjusted to the cumulative result of the original length and the holiday window length. For example, assuming that the original length of the moving average filter is L, and the holiday window is △, when the filter interval of the moving average filter includes the holiday window, the window length of the moving average filter is adjusted to L+△.
[0060] Specifically, the process of processing the historical power consumption data, the exogenous variable data, and the holiday data based on the X13 algorithm can be: extracting a long-term trend by using a filter, introducing the industry expansion and installation quantity data as auxiliary data at the time, extracting the historical trend item through a multivariate regression model; extracting the seasonal factor based on the improved X11 algorithm, and then correcting based on the holiday effect value, the temperature data, and the holiday data to obtain the seasonal holiday effect item; and separating the historical residual item through triggering operation.
[0061] Optionally, after obtaining the historical seasonal holiday effect item, the Bootstrap method can be used to correct the abnormal values in the historical seasonal holiday effect item. The specific principle is referred to the existing Bootstrap algorithm, which is not limited here.
[0062] S150, predicting the power consumption based on the historical trend item, the historical seasonal holiday effect item, and the historical residual random item.
[0063] Specifically, the way of predicting the power consumption based on the historical trend item, the historical seasonal holiday effect item, and the historical residual random item can be: determining a predicted trend item based on the historical trend item and the industry expansion and installation quantity data; determining a predicted seasonal effect item based on the historical seasonal holiday effect item; determining a predicted residual random item based on the historical residual random item; and fusing the predicted trend item, the predicted seasonal effect item, and the predicted residual random item to obtain the predicted power consumption.
[0064] In this embodiment, the time series regression model is used to obtain the predicted trend item T from the historical trend item and the industry expansion and installation quantity data. The historical seasonal holiday effect item is compared with the historical seasonal factor library to match the typical seasonal pattern, so as to obtain the predicted seasonal holiday effect item S. The residual statistical correction method (such as moving average or exponential smoothing) is used to process the historical residual random item to obtain the predicted residual random item I.
[0065] Specifically, the way of fusing the predicted trend item, the predicted seasonal effect item and the predicted residual random item can be: multiplying the predicted trend item, the predicted seasonal effect item and the predicted residual random item to obtain the predicted power. The formula can be expressed as: Y = T x S x I.
[0066] The technical scheme of the embodiment, the historical power consumption data, the industry expansion and decoration data, the holiday data and the temperature data of the target area are acquired; the holiday window is determined based on the historical power consumption data and the holiday data; wherein the holiday window includes the pre-holiday sub-window, the holiday sub-window and the post-holiday sub-window; the holiday effect value is determined based on the holiday window and the historical power consumption data; the historical trend item, the historical seasonal holiday effect item and the historical residual random item are obtained by processing the historical power consumption data, the exogenous variable data and the holiday data based on the X13 algorithm; wherein the exogenous variable data includes the holiday effect value, the industry expansion and decoration data, and the temperature data; the power consumption is predicted based on the historical trend item, the historical seasonal holiday effect item and the historical residual random item. The power consumption prediction method provided by the embodiment determines the holiday effect value, and outputs the holiday effect value, the industry expansion and decoration data and the temperature data as exogenous variables to the X13 algorithm, so that the X13 algorithm splits the historical power consumption data into the historical trend item, the historical seasonal holiday effect item and the historical residual random item based on the exogenous variables to predict the power consumption. That is, the holiday data, the temperature data and the industry expansion and decoration data are considered when predicting the power consumption, which can improve the accuracy of the power consumption prediction.
[0067] Embodiment two
[0068] Figure 2 is a structural schematic diagram of a power consumption prediction device provided by the embodiment two of the present application, as Figure 2 shown, the device comprises:
[0069] The data acquisition module 210 is configured to acquire historical power consumption data, industry expansion and decoration data, holiday data and temperature data of a target area.
[0070] The holiday window determination module 220 is configured to determine a holiday window based on the historical power consumption data and the holiday data; wherein the holiday window includes a pre-holiday sub-window, a holiday sub-window and a post-holiday sub-window.
[0071] The holiday effect value determination module 230 is configured to determine a holiday effect value based on the holiday window and the historical power consumption data.
[0072] The data processing module 240 is configured to process the historical power consumption data, the exogenous variable data and the holiday data based on the X13 algorithm to obtain a historical trend item, a historical seasonal holiday effect item and a historical residual random item; the exogenous variable data includes holiday effect values, industry expansion installation quantity data and temperature data.
[0073] The power consumption prediction module 250 is configured to predict power consumption based on the historical trend item, the historical seasonal holiday effect item and the historical residual random item.
[0074] Optionally, the holiday window includes a first calendar holiday window and a second calendar holiday window; and the holiday window determination module 220 is further configured to:
[0075] For the first calendar holiday window, a pre-holiday sub-window and a post-holiday sub-window are determined based on a power consumption fluctuation rate, and a holiday sub-window is determined based on the holiday data; the power consumption fluctuation rate is determined based on a mean value and a standard deviation of the historical power consumption data.
[0076] For the second calendar holiday window, the pre-holiday sub-window is determined based on a first preset number of days, the post-holiday sub-window is determined based on a second preset number of days, and the holiday sub-window is determined based on the holiday data; the first preset number of days and the second preset number of days are the same or different.
[0077] Optionally, the method further includes a holiday window adjustment module configured to:
[0078] If the temperature of any day in the holiday window satisfies the following condition: greater than a first threshold value or less than a second threshold value, a window adjustment amount is determined according to the temperature data and a reference temperature; the first threshold value is greater than the second threshold value.
[0079] The historical power consumption data is used to determine a proportion of power consumption of each industry in the target region; the industry power consumption includes industrial power consumption, commercial power consumption and residential power consumption.
[0080] The holiday window is adjusted based on the proportion of power consumption of each industry and the window adjustment amount.
[0081] Optionally, the holiday effect value determination module 230 is further configured to:
[0082] The historical power consumption data is used to determine a reference power consumption on a starting day of the pre-holiday sub-window and a reference power consumption on a starting day of the post-holiday sub-window as a first reference power consumption and a second reference power consumption.
[0083] The first reference power consumption and a first function are used to determine a power consumption of each day in the pre-holiday sub-window as a pre-holiday effect power consumption.
[0084] The second reference power consumption and a second function are used to determine a power consumption of each day in the post-holiday sub-window to obtain a post-holiday effect power consumption.
[0085] determining the average power consumption during the holiday in the historical power consumption data, and obtaining the power consumption of each day in the holiday window;
[0086] determining the holiday effect value of each day in the holiday window based on the pre-holiday effective power consumption, the holiday effective power consumption and the post-holiday effective power consumption.
[0087] Optionally, the method further comprises:
[0088] determining the power consumption proportion of each industry of the target area based on the historical power consumption data; wherein, the industry power consumption comprises industrial power consumption, commercial power consumption and residential power consumption;
[0089] adjusting the holiday effect value based on the power consumption proportion of each industry.
[0090] Optionally, the holiday effect value determining module 230 is further configured to:
[0091] when two or more holiday windows overlap, taking the maximum holiday effect value as the holiday effect value of the overlapping part.
[0092] Optionally, the X13 algorithm comprises a moving average filter; when the filter interval of the moving average filter comprises the holiday window, the window length of the moving average filter is adjusted to the cumulative result of the original length and the holiday window length.
[0093] Optionally, the power consumption prediction module 250 is further configured to:
[0094] determining the predicted trend item based on the historical trend item and the industry expansion and installation data;
[0095] determining the predicted seasonal effect item based on the historical seasonal holiday effect item;
[0096] determining the predicted residual random item based on the historical residual random item;
[0097] fusing the predicted trend item, the predicted seasonal effect item and the predicted residual random item to obtain the predicted power consumption.
[0098] The device can execute the method provided by all the foregoing embodiments of the application, and has the corresponding function modules and beneficial effects of executing the foregoing method. Technical details not described in detail in the embodiment can be referred to the method provided by all the foregoing embodiments of the application.
[0099] Embodiment Three
[0100] Figure 3A structural diagram of an electronic device 10 that can be used to implement embodiments of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices (e.g., headsets, glasses, watches, etc.), and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present application described and / or claimed in this document.
[0101] As shown in Figure 3 The electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., connected to the at least one processor 11 in communication, where the memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer programs stored in the read-only memory (ROM) 12 or loaded into the random access memory (RAM) 13 from the storage unit 18. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.
[0102] A plurality of components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc., an output unit 17, such as various types of displays, speakers, etc., a storage unit 18, such as a magnetic disk, an optical disk, etc., and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunications networks.
[0103] The processor 11 can be various general and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 performs various methods and processes described above, such as the power prediction method.
[0104] In some embodiments, the power prediction method can be implemented as a computer program tangibly embodied in a computer readable storage medium, e.g., storage unit 18. In some embodiments, parts or all of the computer program can be loaded and / or installed onto electronic device 10 via, e.g., ROM 12 and / or communication unit 19. When the computer program is loaded onto RAM 13 and executed by processor 11, one or more steps of the power prediction method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the power prediction method by way of other means, e.g., with the aid of firmware.
[0105] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, specially designed application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.
[0106] Computer programs used to implement the methods of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor of the machine, implements the functions / acts specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine and partially on a remote machine or entirely on a remote machine or server.
[0107] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0108] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.
[0109] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.
[0110] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. Servers can be cloud servers, also known as cloud computing servers or cloud hosts, which are a host product in the cloud computing service system to solve the defects of great management difficulty and weak business scalability in traditional physical hosts and VPS services.
[0111] The embodiment of the present application further provides a computer program product comprising a computer program which, when executed by a processor, implements the power prediction method provided in any embodiment of the present application.
[0112] In implementing the computer program product, the computer program code for performing the operations of the present application can be written in one or more programming languages or combinations of languages including object oriented programming languages such as Java, Smalltalk, C++ or conventional procedural programming languages such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).
[0113] It should be understood that various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the spirit and scope of the present application. For example, the steps recited in the present application can be performed in parallel, in series, or in a different order, and the present application is not limited in this regard.
[0114] The specific embodiments described above are not intended to limit the scope of the present application. Those skilled in the art will understand that various modifications, combinations, sub-combinations and alternatives can be made to the specific embodiments without departing from the spirit and principles of the present application. Any further modifications, equivalents or alternatives within the spirit and principles of the present application are to be considered as falling within the scope of the present application.
Claims
1. A power prediction method, characterized by, The method comprises: obtaining historical power consumption data, industry expansion data, holiday data and temperature data of a target area; determining a holiday window based on the historical power consumption data and the holiday data; wherein the holiday window comprises a pre-holiday sub-window, a holiday sub-window and a post-holiday sub-window; determining a holiday effect value based on the holiday window and the historical power consumption data; processing the historical power consumption data, exogenous variable data and the holiday data based on an X13 algorithm to obtain a historical trend item, a historical seasonal holiday effect item and a historical residual random item; wherein the exogenous variable data comprises the holiday effect value, the industry expansion data and the temperature data; predicting power consumption based on the historical trend item, the historical seasonal holiday effect item and the historical residual random item.
2. The method of claim 1, wherein, The holiday window comprises a first calendar holiday window and a second calendar holiday window; determining a holiday window based on the historical power consumption data and the holiday data comprises: for the first calendar holiday window, determining a pre-holiday sub-window and a post-holiday sub-window based on a power consumption fluctuation rate, and determining a holiday sub-window based on the holiday data; wherein the power consumption fluctuation rate is determined based on a mean value and a standard deviation of the historical power consumption data; for the second calendar holiday window, determining a pre-holiday sub-window based on a first preset number of days and determining a post-holiday sub-window based on a second preset number of days, and determining a holiday sub-window based on the holiday data; wherein the first preset number of days is the same as or different from the second preset number of days.
3. The method of claim 1, wherein, After determining the holiday window based on the historical power consumption data and the holiday data, the method further comprises: if the temperature on any day within the holiday window satisfies the following condition: greater than a first threshold value or less than a second threshold value, determining a window adjustment amount based on the temperature data and a reference temperature; wherein the first threshold value is greater than the second threshold value. determining a proportion of power consumption of each industry in the target area based on the historical power consumption data; wherein the industry power consumption comprises industrial power consumption, commercial power consumption and residential power consumption; adjusting the holiday window based on the proportion of power consumption of each industry and the window adjustment amount.
4. The method of claim 1, wherein, Determining a holiday effect value based on the holiday window and the historical power consumption data comprises: determining a reference power consumption on a starting day of the pre-holiday sub-window and a reference power consumption on a starting day of the post-holiday sub-window based on the historical power consumption data, as a first reference power consumption and a second reference power consumption; determining a power consumption on each day within the pre-holiday sub-window based on the first reference power consumption and a first function, as a pre-holiday effect power consumption; determining a power consumption on each day within the post-holiday sub-window based on the second reference power consumption and a second function, as a post-holiday effect power consumption; determining a power consumption on each day within the holiday sub-window based on an average power consumption during holidays in the historical power consumption data, as a holiday effect power consumption; determining a holiday effect value on each day within the holiday window based on the pre-holiday effect power consumption, the holiday effect power consumption and the post-holiday effect power consumption.
5. The method of claim 1, wherein, determining a holiday effect value based on the holiday window and the historical power consumption data, further comprising: determining an industry power consumption proportion of each industry in the target region based on the historical power consumption data; wherein the industry power consumption comprises industrial power consumption, commercial power consumption and residential power consumption; adjusting the holiday effect value based on the industry power consumption proportion of each industry.
6. The method of claim 1, wherein, determining a holiday effect value based on the holiday window and the historical power consumption data, comprising: when two or more holiday windows overlap, taking the maximum holiday effect value as the holiday effect value of the overlapping part.
7. The method of claim 1, wherein, the X13 algorithm comprises a moving average filter; when the filtering interval of the moving average filter comprises the holiday window, the window length of the moving average filter is adjusted to the cumulative result of the original length and the holiday window length.
8. The method of claim 1, wherein, the prediction of the power consumption based on the historical trend item, the historical seasonal holiday effect item and the historical residual random item, comprising: determining a predicted trend item based on the historical trend item and the industry expansion and installation data; determining a predicted seasonal effect item based on the historical seasonal holiday effect item; determining a predicted residual random item based on the historical residual random item; fusing the predicted trend item, the predicted seasonal effect item and the predicted residual random item to obtain the predicted power consumption.
9. An electric power prediction device characterized by comprising: comprising: a data acquisition module configured to acquire historical power consumption data, industry expansion and installation data, holiday data and temperature data of a target region; a holiday window determination module configured to determine a holiday window based on the historical power consumption data and the holiday data; wherein the holiday window comprises a pre-holiday sub-window, a holiday sub-window and a post-holiday sub-window; a holiday effect value determination module configured to determine a holiday effect value based on the holiday window and the historical power consumption data; a data processing module configured to process the historical power consumption data, exogenous variable data and the holiday data based on an X13 algorithm to obtain a historical trend item, a historical seasonal holiday effect item and a historical residual random item; wherein the exogenous variable data comprises the holiday effect value, the industry expansion and installation data and the temperature data; a power consumption prediction module configured to predict power consumption based on the historical trend item, the historical seasonal holiday effect item and the historical residual random item.
10. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected in communication with the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the power consumption prediction method in any one of claims 1-8.