Electricity load forecasting method, device, electronic device and storage medium

Abstract

The present invention discloses a method, device, electronic device, and storage medium for predicting power load, belonging to the technical field of power load prediction. The method comprises: obtaining the actual load at the current moment; searching a database for first Gaussian mixture distribution data corresponding to the current moment, wherein the database stores a correspondence between moments within a preset period and Gaussian mixture distribution data, wherein the Gaussian mixture distribution data is obtained by fitting historical load data at the same moment in different periods; searching the database for second Gaussian mixture distribution data corresponding to the next moment; and predicting the power load at the next moment based on the actual load, the first Gaussian mixture distribution data, and the second Gaussian mixture distribution data.

Description

Technical Field

[0001] The present invention relates to the technical field of electricity load forecasting, and in particular to an electricity load forecasting method, device, electronic equipment and storage medium. Background Art

[0002] Power load forecasting is a crucial foundation for the core business of power retail companies. Specifically, key business modules such as contract signing strategies, quotation strategies, and transaction strategies are all based on power load forecasting results.

[0003] Related technologies use a periodic time naive forecasting method when forecasting electricity load. This method uses the observed value at the corresponding moment in the previous time period as the forecast value. For example, if the electricity load needs to be forecasted at 10:00 AM on Monday, the actual electricity load at 10:00 AM on Monday of the previous week is used as the forecast value.

[0004] This approach is suitable for scenarios where the power load changes regularly. However, in scenarios where the power load changes less regularly or the power load fluctuates significantly, the prediction errors obtained by this approach are large, making it difficult to accurately predict future power loads. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the defect of inaccurate power load prediction in the prior art and to provide a power load prediction method, device, electronic device and storage medium.

[0006] The present invention solves the above technical problems through the following technical solutions:

[0007] In a first aspect, the present invention provides a method for predicting electricity load, the method comprising:

[0008] Get the actual load at the current moment;

[0009] searching a database for first Gaussian mixture distribution data corresponding to the current moment, wherein the database stores a correspondence between moments in a preset period and Gaussian mixture distribution data, the Gaussian mixture distribution data being obtained by fitting historical load data at the same moment in different periods;

[0010] Searching the database for second Gaussian mixture distribution data corresponding to the next moment;

[0011] The power load at the next moment is predicted based on the actual load, the first Gaussian mixture distribution data, and the second Gaussian mixture distribution data.

[0012] Optionally, predicting the power load at the next moment according to the actual load, the first Gaussian mixture distribution data, and the second Gaussian mixture distribution data includes:

[0013] determining the standard score at the current moment according to the actual load and the first Gaussian mixture distribution data;

[0014] The power load at the next moment is predicted based on the standard score and the second Gaussian mixture distribution data.

[0015] Optionally, the first Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean, and standard deviation of each Gaussian distribution; and the standard score matrix at the current moment is determined according to the following formula:

[0016]

[0017] Where k is the number of Gaussian distributions, i is an integer between [0, k], is the standard score matrix at the current moment, is the standard score of the i-th Gaussian distribution corresponding to the current moment, is the weight of the i-th Gaussian distribution corresponding to the current moment, is the mean of the i-th Gaussian distribution corresponding to the current moment, is the standard deviation of the i-th Gaussian distribution corresponding to the current moment.

[0018] Optionally, the second Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean and standard deviation of each Gaussian distribution;

[0019] The power load at the next moment is predicted according to the following formula:

[0020]

[0021] Where k is the number of Gaussian distributions, i is an integer between [0, k], and P t+1 is the power load at the next moment, is the weight of the i-th Gaussian distribution corresponding to the next moment in the historical period, is the mean of the i-th Gaussian distribution corresponding to the next moment in the historical period, is the standard deviation of the i-th Gaussian distribution corresponding to the next moment in the historical period.

[0022] Optionally, the method further includes:

[0023] Divide the historical monitoring time into a number of preset periods according to the preset time, and obtain the historical load data corresponding to each moment in each of the preset periods;

[0024] Performing Gaussian mixture distribution fitting on the historical load data corresponding to the same moment, and extracting Gaussian mixture distribution data corresponding to the moment based on the fitting result;

[0025] The correspondence between the time moments and the Gaussian mixture distribution data is stored to form the database.

[0026] In a second aspect, the present invention provides an electricity load forecasting device, the device comprising:

[0027] The first acquisition module is used to obtain the actual load at the current moment;

[0028] a first search module, configured to search a database for first Gaussian mixture distribution data corresponding to the current moment, wherein the database stores a correspondence between moments in a preset period and Gaussian mixture distribution data, wherein the Gaussian mixture distribution data is obtained by fitting historical load data at the same moment in different periods;

[0029] A second search module is used to search the database for second Gaussian mixture distribution data corresponding to the next moment;

[0030] A prediction module is used to predict the power load at the next moment based on the actual load, the first Gaussian mixture distribution data and the second Gaussian mixture distribution data.

[0031] Optionally, the prediction module includes:

[0032] a determining unit, configured to determine the standard score at the current moment according to the actual load and the first Gaussian mixture distribution data;

[0033] A prediction unit is used to predict the power load at the next moment based on the standard score and the second Gaussian mixture distribution data.

[0034] Optionally, the first Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean, and standard deviation of each Gaussian distribution; and the determination unit determines the standard score matrix at the current moment according to the following formula:

[0035]

[0036] Where k is the number of Gaussian distributions, i is an integer between [0, k], is the standard score matrix at the current moment, is the standard score of the i-th Gaussian distribution corresponding to the current moment, is the weight of the i-th Gaussian distribution corresponding to the current moment, is the mean of the i-th Gaussian distribution corresponding to the current moment, is the standard deviation of the i-th Gaussian distribution corresponding to the current moment.

[0037] Optionally, the second Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean and standard deviation of each Gaussian distribution;

[0038] The prediction unit predicts the power load at the next moment according to the following formula:

[0039]

[0040] Where k is the number of Gaussian distributions, i is an integer between [0, k], and P t+1 is the power load at the next moment, is the weight of the i-th Gaussian distribution corresponding to the next moment in the historical period, is the mean of the i-th Gaussian distribution corresponding to the next moment in the historical period, is the standard deviation of the i-th Gaussian distribution corresponding to the next moment in the historical period.

[0041] Optionally, the device further comprises:

[0042] The second acquisition module is used to divide the historical monitoring time into a plurality of preset periods according to the preset time, and obtain the historical load data corresponding to each moment in each of the preset periods;

[0043] an extraction module, configured to perform Gaussian mixture distribution fitting on the historical load data corresponding to the same moment, and extract the Gaussian mixture distribution data corresponding to the moment according to the fitting result;

[0044] A storage module is used to store the corresponding relationship between the time and the Gaussian mixture distribution data to form the database.

[0045] In a third aspect, the present invention provides 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 computer program, the power load forecasting method as described in the first aspect above is implemented.

[0046] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the power load forecasting method as described in the first aspect above.

[0047] The positive progress effect of the present invention is:

[0048] The power load forecasting method provided by the present invention extracts Gaussian mixture distribution data from historical load data. This data, combined with the current actual load, predicts the power load at the next moment. This overall method reflects the influence of historical patterns and current data on the forecast results, achieving high prediction accuracy and efficiency, overcoming the significant prediction discrepancies found in related technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a flow chart of a method for predicting power load according to an exemplary embodiment;

[0050] Figure 2 is a flowchart of step S104 according to an exemplary embodiment;

[0051] Figure 3 is a flow chart of a method for predicting power load according to another exemplary embodiment;

[0052] Figure 4 is a schematic diagram showing changes in BIC and AIC according to an exemplary embodiment;

[0053] Figure 5A and Figure 5B is a schematic diagram of Gaussian mixture distribution fitting according to different exemplary embodiments;

[0054] Figure 6 is a block diagram of a power load prediction device according to an exemplary embodiment;

[0055] Figure 7 is a block diagram of a power load prediction device according to an exemplary embodiment;

[0056] Figure 8 is a block diagram of a power load prediction device according to an exemplary embodiment;

[0057] Figure 9 This is a schematic structural diagram of an electronic device provided in this embodiment. DETAILED DESCRIPTION

[0058] The present invention is further described below by way of examples, but the present invention is not limited to the scope of the examples.

[0059] In a first aspect, an embodiment of the present invention provides a method for predicting electricity load. Figure 1 is a flow chart showing a method for predicting power load according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0060] Step S101: Obtain the actual load at the current moment.

[0061] The actual load at the current moment refers to the sum of the power consumed by various electrical equipment at the current moment, and the standard unit is KW.

[0062] Step S102: searching for first Gaussian mixture distribution data corresponding to the current moment in a database. The database stores a correspondence between moments in a preset period and Gaussian mixture distribution data. The Gaussian mixture distribution data is obtained by fitting historical load data at the same moment in different periods.

[0063] When establishing the database, a preset period is divided into multiple moments at preset time intervals. Historical load data includes the actual power load corresponding to each moment in different preset periods. As an example, a week is used as the preset period and divided into 672 moments at 15-minute intervals. Historical load data includes data from multiple preset periods. For example, historical load data includes the actual power load at 672 moments in the first week, the actual power load at 672 moments in the second week, and so on and so forth.

[0064] The actual power load corresponding to the same time in different preset periods is used as the target data group. The data in the target data group is fitted to obtain a Gaussian mixture distribution model for the target data group, and then the Gaussian mixture distribution data corresponding to the time is extracted. As an example, once again using a preset period of one week, the actual power load corresponding to 10:00 a.m. on Monday from week 1 to week n is used as the target data group. The model is fitted and data extracted from the target data group to obtain the Gaussian mixture distribution data corresponding to 10:00 a.m. on Monday.

[0065] The duration of the preset period and the time interval for dividing the moments are not specifically limited and can be flexibly set as needed. For example, a preset period of one month and a time interval of half an hour are used for dividing the moments.

[0066] In the database, the Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean, and standard deviation of each Gaussian distribution. Accordingly, the first Gaussian mixture distribution data obtained from the database in step S102 accurately reflects the current power load variation pattern in the historical data.

[0067] Step S103: Search the database for the second Gaussian mixture distribution data corresponding to the next moment.

[0068] For example, with one week as a preset cycle and every 15 minutes as a time interval, the current moment is 10:00 a.m. on Monday, and the next moment is 10:15 a.m. on Monday. Step S103 specifically searches the database for the second Gaussian mixture distribution data corresponding to 10:15 a.m. on Monday. It can be understood that the second Gaussian mixture distribution data obtained in step S103 reflects the change pattern of power load in the historical data at the next moment. Moreover, steps S102 and S103 can efficiently and quickly obtain the first Gaussian mixture distribution data and the second Gaussian mixture distribution data by searching the database.

[0069] Step S104: predicting the power load at the next moment according to the actual load, the first Gaussian mixture distribution data, and the second Gaussian mixture distribution data.

[0070] The first Gaussian mixture distribution data accurately reflects the distribution pattern of the power load corresponding to the current moment in the historical data, and the impact of the power load at the current moment on the power load at the next moment is reflected through the actual load at the current moment and the first Gaussian mixture distribution data. The second Gaussian mixture distribution data accurately reflects the distribution pattern of the power load corresponding to the next moment in the historical data, and the impact of the historical distribution pattern of the power load on the power load at the next moment is reflected through the second Gaussian mixture distribution data. In other words, the prediction method provided by the embodiment of the present invention comprehensively considers the power load at the current moment and the historical power load distribution pattern, thereby improving prediction accuracy.

[0071] In one example, Figure 2 FIG. 1 is a flow chart of step S104 according to an exemplary embodiment. Figure 2 As shown, step S104 specifically includes:

[0072] Step S1041: Determine the standard score at the current moment based on the actual load and the first Gaussian mixture distribution data.

[0073] The first Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean and standard deviation of each Gaussian distribution. In step S1041, the standard score matrix at the current moment is determined according to the following formula:

[0074]

[0075] Where k is the number of Gaussian distributions, i is an integer between [0, k], is the standard score matrix at the current moment, is the standard score of the i-th Gaussian distribution corresponding to the current moment, is the weight of the i-th Gaussian distribution corresponding to the current moment, is the mean of the i-th Gaussian distribution corresponding to the current moment, is the standard deviation of the i-th Gaussian distribution corresponding to the current moment.

[0076] Step S1042: predict the power load at the next moment based on the standard score and the second Gaussian mixture distribution data.

[0077] The second Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean and standard deviation of each Gaussian distribution. In step S1042, the power load at the next moment is predicted according to the following formula:

[0078]

[0079] Where k is the number of Gaussian distributions, i is an integer between [0, k], and P t+1 is the power load at the next moment, is the weight of the i-th Gaussian distribution corresponding to the next moment in the historical cycle, is the mean of the i-th Gaussian distribution corresponding to the next moment in the historical cycle, is the standard deviation of the i-th Gaussian distribution corresponding to the next moment in the historical cycle.

[0080] In summary, the power load forecasting method provided by the embodiments of the present invention uses Gaussian mixture distribution data extracted from historical load data and the current actual load to predict the power load at the next moment. This method achieves high accuracy and efficiency, overcoming the drawback of large prediction discrepancies in related technologies.

[0081] In one embodiment, the power load forecasting method provided by the embodiment of the present invention further includes establishing the database. Figure 3 FIG. 1 is a flow chart of a method for predicting power load according to another exemplary embodiment. Figure 3 As shown, the method further includes:

[0082] Step S301: Divide the historical monitoring duration into a number of preset periods according to the preset duration, and obtain the historical load data corresponding to each moment in each of the preset periods.

[0083] The historical monitoring duration can be divided into multiple preset periods based on the preset duration. For example, if the historical monitoring duration is 2 years and the preset duration is one week, 2 years can be divided into 2×52=104 preset periods. Furthermore, a period can be divided into multiple moments based on preset time intervals. For example, a week can be divided into 7×24×4=672 moments at 15-minute intervals. In this example, step S301 specifically obtains historical load data for 672 moments in the 104 preset periods.

[0084] In step S301, the historical data is preliminarily grouped, and the historical load array corresponding to the same time is regarded as a sample group. Then, in subsequent steps, Gaussian mixture distribution fitting is performed on the data of each sample group.

[0085] Step S302: Perform Gaussian mixture distribution fitting on the historical load data corresponding to the same moment, and extract the Gaussian mixture distribution data corresponding to the moment based on the fitting result.

[0086] The Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean, and variance of each Gaussian distribution. Taking the processing of historical load data corresponding to a moment as an example, step S302 specifically includes:

[0087] The first step is to extract the mean, standard deviation and weight of k Gaussian mixture distributions.

[0088] Where k is a positive integer representing the number of Gaussian distributions in the Gaussian mixture distribution model obtained by fitting the historical data corresponding to a moment. Specifically, the EM algorithm is used to extract the mean, standard deviation, and weight of k Gaussian distributions, and the maximum log-likelihood function of the fitted mixture Gaussian distribution model is calculated:

[0089]

[0090] Among them, k means that there are k Gaussian distributions in the Gaussian mixture distribution model; β i is the probability of the i-th Gaussian distribution being generated and β i ≥0; p(x j |μ i , C i ) is the probability density function of Gaussian distribution; μ i is the mean vector, C i is the covariance matrix, x j The sample is specifically the historical load data corresponding to a moment in the present invention.

[0091] The second step is to determine the number of Gaussian distributions.

[0092] Repeat the first step with different values of k. Furthermore, in the second step, the BIC and AIC indicators are introduced to determine the number of Gaussian distributions. Here, BIC = log(m)p-2(L) is determined based on the Bayesian Information Criterion, and AIC = 2p-2(L) is determined based on the Akaike Information Criterion. L is the value of the maximum logarithmic function, m is the number of samples, and p is the number of parameters in the Gaussian mixture distribution model with k Gaussian distributions, including weights, means, and variances.

[0093] Figure 4FIG. 1 is a schematic diagram showing changes in BIC and AIC according to an exemplary embodiment. Figure 4 As shown in Figure 2, the k value when AIC and BIC do not decrease significantly is used as the number of Gaussian distributions finally determined.

[0094] Continue to refer to Figure 3 , after step S302, step S303 is performed, specifically as follows:

[0095] Step S303: Store the correspondence between the time and the Gaussian mixture distribution data to form the database. Optionally, the database is implemented as a load distribution dictionary, with the Gaussian mixture distribution data as the value and the time as the key. The load distribution dictionary stores the correspondence between the value and the key. In one example, the structure of the load distribution dictionary is as follows:

[0096]

[0097]

[0098] Through steps S301 to S303 , Gaussian mixture distribution fitting is performed on the historical load data corresponding to a moment. Figure 5A and Figure 5B Schematic diagrams of Gaussian mixture distribution fitting according to different exemplary embodiments are shown. Figure 5A and Figure 5B As shown, in this embodiment, the optimal number of Gaussian distributions in the Gaussian mixture distribution model is two. By constructing a Gaussian mixture distribution model, the distribution of historical load data can be more accurately fitted, thereby improving the accuracy of the prediction method implemented based on this database. Furthermore, the database formed in this way can quickly determine the corresponding Gaussian mixture distribution data at a given time, improving the efficiency of the prediction method.

[0099] In a second aspect, an embodiment of the present invention provides an electricity load forecasting device. Figure 6 is a block diagram of a power load prediction device according to an exemplary embodiment. Figure 6 As shown, the device includes:

[0100] A first acquisition module 610 is used to obtain the actual load at the current moment;

[0101] A first search module 620 is configured to search a database for first Gaussian mixture distribution data corresponding to the current moment. The database stores a correspondence between moments in a preset period and Gaussian mixture distribution data. The Gaussian mixture distribution data is obtained by fitting historical load data at the same moment in different periods.

[0102] A second search module 630 is used to search the database for second Gaussian mixture distribution data corresponding to the next moment;

[0103] The prediction module 640 is used to predict the power load at the next moment based on the actual load, the first Gaussian mixture distribution data and the second Gaussian mixture distribution data.

[0104] In one embodiment, Figure 7 is a block diagram of a power load prediction device according to an exemplary embodiment. Figure 7 As shown, the prediction module 640 includes:

[0105] A determination unit 641 is configured to determine a standard score at a current moment based on the actual load and the first Gaussian mixture distribution data;

[0106] The prediction unit 642 is used to predict the power load at the next moment based on the standard score and the second Gaussian mixture distribution data.

[0107] In one embodiment, the first Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean, and standard deviation of each Gaussian distribution; and the determination unit determines the standard score matrix at the current moment according to the following formula:

[0108]

[0109] Where k is the number of Gaussian distributions, i is an integer between [0, k], is the standard score matrix at the current moment, is the standard score of the i-th Gaussian distribution corresponding to the current moment, is the weight of the i-th Gaussian distribution corresponding to the current moment, is the mean of the i-th Gaussian distribution corresponding to the current moment, is the standard deviation of the i-th Gaussian distribution corresponding to the current moment.

[0110] In one embodiment, the second Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean, and standard deviation of each Gaussian distribution;

[0111] The prediction unit predicts the power load at the next moment according to the following formula:

[0112]

[0113] Where k is the number of Gaussian distributions, i is an integer between [0, k], and P t+1 is the power load at the next moment, is the weight of the i-th Gaussian distribution corresponding to the next moment in the historical cycle, is the mean of the i-th Gaussian distribution corresponding to the next moment in the historical cycle, is the standard deviation of the i-th Gaussian distribution corresponding to the next moment in the historical cycle.

[0114] In one embodiment, Figure 8 is a block diagram of a power load prediction device according to an exemplary embodiment. Figure 8 As shown, the device also includes:

[0115] The second acquisition module 650 is configured to divide the historical monitoring duration into a plurality of preset periods according to a preset duration, and acquire the historical load data corresponding to each moment in each of the preset periods;

[0116] An extraction module 660 is configured to perform Gaussian mixture distribution fitting on the historical load data corresponding to the same moment, and extract Gaussian mixture distribution data corresponding to the moment based on the fitting result;

[0117] The storage module 670 is used to store the correspondence between the time and the Gaussian mixture distribution data to form the database.

[0118] In a third aspect, an embodiment of the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the power load forecasting method described in the first aspect.

[0119] Figure 9 This is a schematic diagram of the structure of an electronic device provided in this embodiment. The electronic device includes at least one processor and a memory in communication with the at least one processor. 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 load forecasting method of Example 1. The electronic device provided in this embodiment can be a personal computer, such as a desktop computer, an all-in-one computer, a laptop computer, a tablet computer, etc., and can also be a terminal device such as a mobile phone, a wearable device, or a PDA. Figure 9 The electronic device 3 shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0120] The components of the electronic device 3 may include, but are not limited to: the at least one processor 4 mentioned above, the at least one memory 5 mentioned above, and a bus 6 connecting different system components (including the memory 5 and the processor 4).

[0121] The bus 6 includes a data bus, an address bus, and a control bus.

[0122] The memory 5 may include a volatile memory, such as a random access memory (RAM) 51 and / or a cache memory 52 , and may further include a read-only memory (ROM) 53 .

[0123] The memory 5 may also include a program / utility 55 having a set (at least one) of program modules 54, such program modules 54 including but not limited to: an operating system, one or more application programs, other program modules and program data, each of which or some combination may include an implementation of a network environment.

[0124] The processor 4 executes various functional applications and data processing, such as the above-mentioned power load forecasting method, by running computer programs stored in the memory 5 .

[0125] The electronic device 3 can also communicate with one or more external devices 7 (e.g., keyboard, pointing device, etc.). Such communication can be performed through an input / output (I / O) interface 8. Furthermore, the electronic device 3 can also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 9. Figure 9 As shown, the network adapter 9 communicates with other modules of the electronic device 3 via the bus 6. It should be understood that although Figure 9 Not shown, other hardware and / or software modules may be used in conjunction with the electronic device 3, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems.

[0126] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of the present invention, the features and functions of two or more units / modules described above may be embodied in a single unit / module. Conversely, the features and functions of a single unit / module described above may be further divided and embodied by multiple units / modules.

[0127] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method for predicting power load as described in the first aspect is implemented.

[0128] The readable storage medium may include, but is not limited to, a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0129] In a possible implementation, the present invention can also be implemented in the form of a program product, which includes program code. When the program product runs on an electronic device, the program code is used to enable the electronic device to execute the power load forecasting method provided in the first aspect.

[0130] The program code for executing the present invention may be written in any combination of one or more programming languages, and the program code may be executed entirely on the electronic device, partially on the electronic device, as an independent software package, partially on the electronic device and partially on a remote device, or entirely on the remote device.

[0131] Although specific embodiments of the present invention have been described above, those skilled in the art will appreciate that these are merely illustrative and that the scope of the present invention is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, and such changes and modifications are intended to fall within the scope of the present invention.

Claims

1. A method for predicting power load, characterized in that: The method comprises: Get the actual load at the current moment; searching a database for first Gaussian mixture distribution data corresponding to the current moment, wherein the database stores a correspondence between moments in a preset period and Gaussian mixture distribution data, the Gaussian mixture distribution data being obtained by fitting historical load data at the same moment in different periods; Searching the database for second Gaussian mixture distribution data corresponding to the next moment; Predicting the power load at the next moment according to the actual load, the first Gaussian mixture distribution data, and the second Gaussian mixture distribution data; The predicting the power load at the next moment according to the actual load, the first Gaussian mixture distribution data, and the second Gaussian mixture distribution data includes: determining the standard score at the current moment according to the actual load and the first Gaussian mixture distribution data; Predicting the power load at the next moment according to the standard score and the second Gaussian mixture distribution data; The first Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean, and standard deviation of each Gaussian distribution; the standard score matrix at the current moment is determined according to the following formula: Where k is the number of Gaussian distributions, i is an integer between [0, k], is the standard score matrix at the current moment, is the standard score of the i-th Gaussian distribution corresponding to the current moment, is the weight of the i-th Gaussian distribution corresponding to the current moment, is the mean of the i-th Gaussian distribution corresponding to the current moment, is the standard deviation of the i-th Gaussian distribution corresponding to the current moment, is the power load at the current moment; The second Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean and standard deviation of each Gaussian distribution; The power load at the next moment is predicted according to the following formula: Where k is the number of Gaussian distributions, i is an integer between [0, k], is the power load at the next moment, is the weight of the i-th Gaussian distribution corresponding to the next moment in the historical cycle, is the mean of the i-th Gaussian distribution corresponding to the next moment in the historical period, is the standard deviation of the i-th Gaussian distribution corresponding to the next moment in the historical period.

2. The power load forecasting method according to claim 1, characterized in that: The method further comprises: Divide the historical monitoring time into a number of preset periods according to the preset time, and obtain the historical load data corresponding to each moment in each of the preset periods; Performing Gaussian mixture distribution fitting on the historical load data corresponding to the same moment, and extracting Gaussian mixture distribution data corresponding to the moment based on the fitting result; The correspondence between the time moments and the Gaussian mixture distribution data is stored to form the database.

3. A power load forecasting device, characterized in that: The device comprises: The first acquisition module is used to obtain the actual load at the current moment; a first search module, configured to search a database for first Gaussian mixture distribution data corresponding to the current moment, wherein the database stores a correspondence between moments in a preset period and Gaussian mixture distribution data, wherein the Gaussian mixture distribution data is obtained by fitting historical load data at the same moment in different periods; A second search module is used to search the database for second Gaussian mixture distribution data corresponding to the next moment; a prediction module, configured to predict the power load at the next moment based on the actual load, the first Gaussian mixture distribution data, and the second Gaussian mixture distribution data; The prediction module includes: a determining unit, configured to determine the standard score at the current moment according to the actual load and the first Gaussian mixture distribution data; a prediction unit, configured to predict the power load at the next moment based on the standard score and the second Gaussian mixture distribution data; The first Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean, and standard deviation of each Gaussian distribution; the determination unit determines the standard score matrix at the current moment according to the following formula: Where k is the number of Gaussian distributions, i is an integer between [0, k], is the standard score matrix at the current moment, is the standard score of the i-th Gaussian distribution corresponding to the current moment, is the weight of the i-th Gaussian distribution corresponding to the current moment, is the mean of the i-th Gaussian distribution corresponding to the current moment, is the standard deviation of the i-th Gaussian distribution corresponding to the current moment, is the power load at the current moment; The second Gaussian mixture distribution data includes the number of Gaussian distributions, the weight, mean and standard deviation of each Gaussian distribution; The prediction unit predicts the power load at the next moment according to the following formula: Where k is the number of Gaussian distributions, i is an integer between [0, k], is the power load at the next moment, is the weight of the i-th Gaussian distribution corresponding to the next moment in the historical cycle, is the mean of the i-th Gaussian distribution corresponding to the next moment in the historical period, is the standard deviation of the i-th Gaussian distribution corresponding to the next moment in the historical period.

4. The power load forecasting device according to claim 3, characterized in that: The device further comprises: The second acquisition module is used to divide the historical monitoring time into a plurality of preset periods according to the preset time, and obtain the historical load data corresponding to each moment in each of the preset periods; an extraction module, configured to perform Gaussian mixture distribution fitting on the historical load data corresponding to the same moment, and extract the Gaussian mixture distribution data corresponding to the moment according to the fitting result; A storage module is used to store the corresponding relationship between the time and the Gaussian mixture distribution data to form the database.

5. 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 computer program, the power load forecasting method according to any one of claims 1 to 2 is implemented.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the power load forecasting method according to any one of claims 1 to 2 is implemented.

Images