Electricity consumption gap prediction method and device
By combining gray prediction and neural network prediction, the problem of inaccurate prediction of electricity consumption gaps in the power market is solved, and a more accurate and suitable power consumption gap prediction is achieved in a nonlinear environment.
Patent Information
- Application Number
- CN202411796999.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-09
- Publication Date
- 2025-04-01
AI Technical Summary
The existing power consumption prediction methods are inaccurate in the power market, especially in a highly nonlinear environment, and it is difficult to capture the medium- and long-term trends of the power market and the entire spot process dynamics.
By obtaining the historical monthly electricity consumption gap data and historical daily electricity consumption gap data of the target load, a combination of gray prediction and neural network prediction is adopted. The specific steps include: making gray predictions based on historical monthly electricity consumption gap data to obtain the first predicted value; making neural network predictions based on historical daily electricity consumption gap data to obtain the second predicted value; and then correcting these predicted values to make them more suitable for the nonlinear environment of the power market, thereby improving prediction accuracy.
By combining gray prediction and neural network prediction, the power consumption gap in the power market can be predicted more accurately, improving the accuracy and applicability of the prediction results, and is suitable for the highly nonlinear environment of the power market.
Smart Images
Figure CN119741048A_ABST
Abstract
Description
Technical Field
[0001] The present application belongs to the technical field of electricity consumption gap prediction, and in particular, relates to an electricity consumption gap prediction method and device. Background Art
[0002] With the continuous growth of global energy demand and the diversification of the development of the power market, both medium- and long-term and spot transactions in the province can be completed through centralized bidding. Since the power market in the province has always been in a state of supply exceeding demand, the power generation side reports quantity and quotation in the spot market, and the power consumption side only reports quantity but not quotation. Therefore, predicting the power consumption gap of power buyers in the market has become a crucial issue. In the power trading market, the power consumption gap on the user side mainly includes medium- and long-term power and spot power. Among them, the medium- and long-term power gap is supplemented by macro power regulation; and the remaining spot power gap is supplemented by the spot market. Accurate prediction results can not only help participants in the power market to effectively allocate resources, but also have a far-reaching impact on the stable operation of the power system and the efficient use of energy.
[0003] Most existing electricity consumption forecasting methods rely mainly on time series analysis, such as autoregressive moving average (ARMA), autoregressive integrated moving average (ARIMA) and other models. These models perform well when dealing with linear relationships, but their forecasting accuracy is often limited in a highly nonlinear environment such as the electricity market. Although some machine learning-based methods can capture nonlinear relationships, they often lack comprehensive consideration of the medium- and long-term trends in the electricity market and the dynamics of the entire spot process. Summary of the invention
[0004] In order to overcome the technical problem of inaccurate prediction of electricity consumption gap in the power market existing in the related art, the embodiments of the present application provide a method and device for predicting electricity consumption gap.
[0005] This application is implemented through the following technical solutions:
[0006] In a first aspect, an embodiment of the present application provides a method for predicting a power consumption gap, comprising:
[0007] Obtain the historical monthly electricity consumption gap data and historical daily electricity consumption gap data of the target load.
[0008] Grey prediction is performed based on the historical monthly electricity consumption gap data to obtain the first prediction value; neural network prediction is performed based on the historical daily electricity consumption gap data to obtain the second prediction value.
[0009] The second prediction value is corrected based on the first prediction value to obtain the prediction value of the daily electricity consumption gap data of the target load in the target time period; the first prediction value is corrected based on the second prediction value to obtain the prediction value of the monthly electricity consumption gap data of the target load in the target time period.
[0010] In combination with the first aspect, in some possible implementations, the second prediction value is corrected based on the first prediction value to obtain a prediction value of the daily power consumption gap data of the target load in the target time period, including:
[0011] The first formula is used to calculate the predicted value of the daily electricity consumption gap data of the target load in the target time period.
[0012] The first formula is:
[0013]
[0014] in, It represents the predicted value of the daily electricity consumption gap data of the target load in the target time period. L xh represents the second predicted value, μ L Indicates the confidence of grey prediction, its value is between 0 and 1, L zc represents the first predicted value, H zc Represents the time span of grey prediction.
[0015] In combination with the first aspect, in some possible implementations, the first prediction value is corrected based on the second prediction value to obtain a prediction value of the monthly power consumption gap data of the target load in the target time period, including:
[0016] The second formula is used to calculate the predicted value of the monthly electricity consumption gap data of the target load in the target time period.
[0017] The second formula is:
[0018]
[0019] in, It represents the predicted value of the monthly electricity consumption gap data of the target load in the target time period. L zc represents the first predicted value, λ L Indicates the confidence of the neural network prediction, its value is between 0 and 1, L xh represents the second predicted value, H xh Represents the time span for the neural network prediction.
[0020] In combination with the first aspect, in some possible implementations, obtaining historical monthly power consumption gap data and historical daily power consumption gap data of the target load includes:
[0021] The original data of the historical monthly electricity consumption gap data is obtained, which is recorded as the first original data, and the original data of the historical daily electricity consumption gap data is obtained, which is recorded as the second original data.
[0022] The first original data is cleaned and normalized, and the CEEMDAN algorithm is used to decompose and reconstruct the first original data to obtain the historical monthly electricity consumption gap data.
[0023] The second original data is cleaned and normalized, and the CEEMDAN algorithm is used to decompose and reconstruct the second original data to obtain the historical daily electricity consumption gap data.
[0024] In combination with the first aspect, in some possible implementations, the first original data is decomposed and reconstructed using the CEEMDAN algorithm to obtain historical monthly electricity consumption gap data, including:
[0025] Gaussian white noise is introduced into the first original data to obtain the first original data after the noise is introduced.
[0026] The first original data after the introduction of noise is subjected to empirical mode decomposition to obtain residual components.
[0027] If the number of extreme value points of the residual component is greater than or equal to 2, Gaussian white noise is added to the residual component, and the empirical mode decomposition is performed again to determine the number of extreme value points.
[0028] If the number of extreme points of the residual component is less than 2, the decomposition and reconstruction of the first original data is completed to obtain the historical monthly electricity consumption gap data.
[0029] In combination with the first aspect, in some possible implementations, the first original data after the noise is introduced is:
[0030] xi(t)=x(t)+cλi(t)
[0031] Among them, λ i (t) represents the Gaussian white noise introduced for the i-th time, c represents the Gaussian white noise coefficient, x(t) represents the first original data, x i (t) represents the first original data after the noise is introduced for the i-th time.
[0032] The third formula is used to calculate the residual component.
[0033] The third formula is:
[0034] r1(t)=x i (t)-I1(t)
[0035]
[0036] Where r1(t) represents the first residual component, N represents the total number of components obtained by the first empirical mode decomposition, represents the i-th component after the first empirical mode decomposition.
[0037] The historical monthly electricity consumption gap data is:
[0038]
[0039] Among them, x1 represents the historical monthly electricity consumption gap data, M represents the number of empirical mode decomposition, and R represents the final residual component.
[0040] In combination with the first aspect, in some possible implementations, gray prediction is performed based on historical monthly electricity consumption gap data to obtain a first prediction value, including:
[0041] The historical monthly electricity consumption gap data is processed by first-order accumulation to obtain the first processed data.
[0042] A differential equation is established based on the first processed data, and a constant factor is solved.
[0043] A first prediction equation is established based on the constant factor, and a first prediction value is obtained according to the first prediction equation.
[0044] In combination with the first aspect, in some possible implementations, the historical monthly electricity consumption gap data is subjected to first-order accumulation processing to obtain first processed data, including:
[0045] The fourth formula is used to calculate the first processed data.
[0046] The fourth formula is:
[0047]
[0048] Among them, x (1) ={x (1) (1),x (1) (2),x (1) (3),...,x (1) (n)} represents the first processed data, x (0) ={x (0) (1),x (0) (2),x (0) (3),...,x (0) (n)} represents the historical monthly electricity consumption gap data.
[0049] The differential equation is:
[0050]
[0051] The constant factor in the differential equation is:
[0052]
[0053] The first prediction equation is:
[0054]
[0055] In combination with the first aspect, in some possible implementations, performing a neural network prediction based on historical daily electricity consumption gap data to obtain a second prediction value includes:
[0056] Based on the historical daily electricity consumption gap data, an LSTM neural network prediction is performed to obtain a second prediction value; the LSTM neural network includes an input gate, an output gate and a forget gate.
[0057] In a second aspect, an embodiment of the present application provides a device for predicting a power consumption gap, including:
[0058] The data acquisition module is used to obtain the historical monthly electricity consumption gap data and the historical daily electricity consumption gap data of the target load.
[0059] The data prediction module is used to perform grey prediction based on the historical monthly electricity consumption gap data to obtain a first prediction value; and to perform neural network prediction based on the historical daily electricity consumption gap data to obtain a second prediction value.
[0060] The data correction module is used to correct the second prediction value based on the first prediction value to obtain the prediction value of the daily electricity consumption gap data of the target load in the target time period; and to correct the first prediction value based on the second prediction value to obtain the prediction value of the monthly electricity consumption gap data of the target load in the target time period.
[0061] It can be understood that the beneficial effects of the second aspect mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here.
[0062] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:
[0063] The present application uses historical monthly electricity consumption gap data and historical daily electricity consumption gap data for prediction, respectively, to obtain a first prediction value (obtained based on the historical monthly electricity consumption gap data prediction) and a second prediction value (obtained based on the historical daily electricity consumption gap data prediction). After obtaining the first prediction value and the second prediction value, the second prediction value is used to correct the first prediction value, and the first prediction value is used to correct the second prediction value, so that the two corrected values both incorporate the characteristics of the historical monthly electricity consumption gap data and the characteristics of the historical daily electricity consumption gap data, which are more suitable for the highly nonlinear environment of the electricity market, and make the prediction results more accurate.
[0064] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present specification. BRIEF DESCRIPTION OF THE DRAWINGS
[0065] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0066] Figure 1 It is a flowchart of a method for predicting electricity consumption gap provided by an embodiment of the present application;
[0067] Figure 2 This is a schematic diagram of the LSTM neural network structure provided by an embodiment of the present application;
[0068] Figure 3 This is a schematic diagram of the idea of a method for predicting electricity consumption gap provided in one embodiment of the present application;
[0069] Figure 4 It is a structural schematic diagram of a power consumption gap prediction device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0070] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.
[0071] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.
[0072] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.
[0073] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.
[0074] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.
[0075] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.
[0076] The present application provides a method for predicting power consumption shortfall. Figure 1 is a schematic diagram of the process of the power consumption gap prediction method provided in an embodiment of the present application, referring to Figure 1 The detailed description of the electricity consumption gap prediction method is as follows:
[0077] Step 101, obtaining historical monthly power consumption gap data and historical daily power consumption gap data of the target load.
[0078] Exemplarily, the historical monthly electricity consumption gap data of the target load is the medium- and long-term electricity consumption in the user-side electricity consumption gap, and the historical daily electricity consumption gap data of the target load is the spot electricity consumption in the user-side electricity consumption gap.
[0079] Exemplarily, step 101 may include:
[0080] The original data of the historical monthly electricity consumption gap data is obtained, which is recorded as the first original data, and the original data of the historical daily electricity consumption gap data is obtained, which is recorded as the second original data.
[0081] The first original data is cleaned and normalized, and the CEEMDAN algorithm is used to decompose and reconstruct the first original data to obtain the historical monthly electricity consumption gap data.
[0082] The second original data is cleaned and normalized, and the CEEMDAN algorithm is used to decompose and reconstruct the second original data to obtain the historical daily electricity consumption gap data.
[0083] Exemplarily, in order to ensure the subsequent processing efficiency and accuracy of the data, after the original data is obtained, the data is cleaned and normalized.
[0084] Exemplarily, the first original data is decomposed and reconstructed using the CEEMDAN algorithm to obtain historical monthly electricity consumption gap data, which may include:
[0085] Gaussian white noise is introduced into the first original data to obtain the first original data after the noise is introduced.
[0086] The first original data after the introduction of noise is subjected to empirical mode decomposition to obtain residual components.
[0087] If the number of extreme value points of the residual component is greater than or equal to 2, Gaussian white noise is added to the residual component, and the empirical mode decomposition is performed again to determine the number of extreme value points.
[0088] If the number of extreme points of the residual component is less than 2, the decomposition and reconstruction of the first original data is completed to obtain the historical monthly electricity consumption gap data.
[0089] For example, in order to better analyze the characteristics of each data, the CEEMDAN algorithm is used to decompose and reconstruct the original data to obtain historical monthly electricity consumption gap data and historical daily electricity consumption gap data.
[0090] Exemplarily, the first original data after the introduction of noise may be:
[0091] xi(t)=x(t)+cλi(t)
[0092] Among them, λ i (t) represents the Gaussian white noise introduced for the i-th time, c represents the Gaussian white noise coefficient, x(t) represents the first original data, x i (t) represents the first original data after the noise is introduced for the i-th time.
[0093] The third formula is used to calculate the residual component.
[0094] The third formula can be:
[0095] r1(t)=x i (t)-I1(t)
[0096]
[0097] Where r1(t) represents the first residual component, N represents the total number of components obtained by the first empirical mode decomposition, represents the i-th component after the first empirical mode decomposition.
[0098] The historical monthly electricity consumption gap data can be:
[0099]
[0100] Among them, x1 represents the historical monthly electricity consumption gap data, M represents the number of empirical mode decomposition, and R represents the final residual component.
[0101] Exemplarily, the second original data is decomposed and reconstructed using the CEEMDAN algorithm to obtain historical daily electricity consumption gap data, which may include:
[0102] Gaussian white noise is introduced into the second original data to obtain the second original data after the noise is introduced.
[0103] The second original data after the introduction of noise is subjected to empirical mode decomposition to obtain the residual component.
[0104] If the number of extreme value points of the residual component is greater than or equal to 2, Gaussian white noise is added to the residual component, and the empirical mode decomposition is performed again to determine the number of extreme value points.
[0105] If the number of extreme points of the residual component is less than 2, the decomposition and reconstruction of the second original data is completed to obtain the historical daily electricity consumption gap data.
[0106] Specifically, the specific process of decomposing and reconstructing the second original data using the CEEMDAN algorithm to obtain the historical daily electricity consumption gap data can be similar to the specific process of decomposing and reconstructing the first original data using the CEEMDAN algorithm to obtain the historical monthly electricity consumption gap data, except that the decomposed and reconstructed data are different.
[0107] Step 102, performing grey prediction based on historical monthly electricity consumption gap data to obtain a first prediction value; performing neural network prediction based on historical daily electricity consumption gap data to obtain a second prediction value.
[0108] Exemplarily, grey prediction is performed based on historical monthly electricity consumption gap data to obtain a first prediction value, including:
[0109] The historical monthly electricity consumption gap data is processed by first-order accumulation to obtain the first processed data.
[0110] A differential equation is established based on the first processed data, and a constant factor is solved.
[0111] A first prediction equation is established based on the constant factor, and a first prediction value is obtained according to the first prediction equation.
[0112] Exemplarily, performing first-order accumulation processing on the historical monthly electricity consumption gap data to obtain first processed data may include:
[0113] The fourth formula is used to calculate the first processed data.
[0114] The fourth formula can be:
[0115]
[0116] Among them, x (1) ={x (1) (1),x (1) (2),x (1) (3),...,x (1) (n)} represents the first processed data, x (0) ={x (0) (1),x (0) (2),x (0) (3),...,x (0) (n)} represents the historical monthly electricity consumption gap data.
[0117] The differential equation can be:
[0118]
[0119] The constant factors in differential equations can be:
[0120]
[0121] The first prediction equation can be:
[0122]
[0123] Exemplarily, performing a neural network prediction based on historical daily electricity consumption gap data to obtain a second prediction value may include:
[0124] Based on the historical daily electricity consumption gap data, an LSTM neural network prediction is performed to obtain a second prediction value; the LSTM neural network includes an input gate, an output gate and a forget gate.
[0125] like Figure 2 As shown in the figure, the detailed structure and functions of the LSTM neural network include:
[0126] Input gate: This layer receives a series of input variables (historical daily electricity consumption gap data).
[0127] i t =σ(W i [x t ;h t-1 ]+b i )
[0128] Forget gate: This gate controls the degree of forgetting of information about the state of the memory unit at the previous moment, and its output determines which information should be discarded.
[0129] f t =σ(W f [x t;h t-1 ]+b f )
[0130] Update unit: By deploying multiple neurons and using the activation function, nonlinear transformation and state update of data are achieved.
[0131] C t =f t C t-1 +i t tanh(W C [h t-1 ;x t ]+b C )
[0132] Output layer: Finally, this layer is responsible for outputting the prediction result (second prediction value) after being processed by the LSTM network.
[0133] o t =σ(W o [h t-1 ;x t ]+b o )
[0134] h t =o t tanh(C t )
[0135]
[0136] f t =σ(W f ·(h t-1 +x t )+b f )
[0137] y t =σ(W t h t +b y )
[0138] In this structure, the weight matrix W f , W i and W o and the bias term b f 、b i and b o Together, they affect the operation of each gating unit and neuron. The activation function σ and tanh are used for gating logic and state update respectively, and the final predicted value y t is determined by these complex interactions.
[0139] Step 103, based on the first prediction value, correct the second prediction value to obtain the prediction value of the daily electricity consumption gap data of the target load in the target time period; based on the second prediction value, correct the first prediction value to obtain the prediction value of the monthly electricity consumption gap data of the target load in the target time period.
[0140] Exemplarily, correcting the second prediction value based on the first prediction value to obtain the prediction value of the daily power consumption gap data of the target load in the target time period may include:
[0141] The first formula is used to calculate the predicted value of the daily electricity consumption gap data of the target load in the target time period.
[0142] The first formula can be:
[0143]
[0144] in, It represents the predicted value of the daily electricity consumption gap data of the target load in the target time period. L xh represents the second predicted value, μ L Indicates the confidence of grey prediction, its value is between 0 and 1, L zc represents the first predicted value, H zc Represents the time span of grey prediction.
[0145] Exemplarily, correcting the first prediction value based on the second prediction value to obtain the prediction value of the monthly power consumption gap data of the target load in the target time period may include:
[0146] The second formula is used to calculate the predicted value of the monthly electricity consumption gap data of the target load in the target time period.
[0147] The second formula can be:
[0148]
[0149] in, It represents the predicted value of the monthly electricity consumption gap data of the target load in the target time period. L zc represents the first predicted value, λ L Indicates the confidence of the neural network prediction, its value is between 0 and 1, L xh represents the second predicted value, H xh Represents the time span for the neural network prediction.
[0150] In order to facilitate the understanding of this scheme, Figure 3In some embodiments, the complete specific logic of this solution is: after obtaining the historical monthly electricity consumption gap data and the historical daily electricity consumption gap data, data preprocessing (data cleaning, data normalization, data decomposition and reconstruction) is first performed, and then predictions are performed separately to obtain respective prediction results, and finally the prediction results are corrected and the corrected prediction results are output.
[0151] The above-mentioned electricity consumption gap prediction method uses the historical monthly electricity consumption gap data and the historical daily electricity consumption gap data for prediction, respectively, to obtain a first prediction value (obtained based on the historical monthly electricity consumption gap data prediction) and a second prediction value (obtained based on the historical daily electricity consumption gap data prediction). After obtaining the first prediction value and the second prediction value, the second prediction value is used to correct the first prediction value, and the first prediction value is used to correct the second prediction value, so that the two corrected values both incorporate the characteristics of the historical monthly electricity consumption gap data and the characteristics of the historical daily electricity consumption gap data, which are more suitable for the highly nonlinear environment of the electricity market, and make the prediction results more accurate.
[0152] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0153] Corresponding to the electricity consumption gap prediction method described in the above embodiment, Figure 2 A structural block diagram of the electricity consumption gap prediction device provided in an embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.
[0154] See also Figure 4 The power consumption gap prediction device in the embodiment of the present application may include:
[0155] The data acquisition module 401 is used to acquire the historical monthly power consumption gap data and the historical daily power consumption gap data of the target load.
[0156] The data prediction module 402 is used to perform grey prediction based on the historical monthly electricity consumption gap data to obtain a first prediction value; and perform neural network prediction based on the historical daily electricity consumption gap data to obtain a second prediction value.
[0157] The data correction module 403 is used to correct the second prediction value based on the first prediction value to obtain the prediction value of the daily electricity consumption gap data of the target load in the target time period; and correct the first prediction value based on the second prediction value to obtain the prediction value of the monthly electricity consumption gap data of the target load in the target time period.
[0158] Exemplarily, the data correction module 403 may be used to:
[0159] The first formula is used to calculate the predicted value of the daily electricity consumption gap data of the target load in the target time period.
[0160] The first formula can be:
[0161]
[0162] in, It represents the predicted value of the daily electricity consumption gap data of the target load in the target time period. L xh represents the second predicted value, μ L Indicates the confidence of grey prediction, its value is between 0 and 1, L zc represents the first predicted value, H zc Represents the time span of grey prediction.
[0163] Exemplarily, the data correction module 403 may be used to:
[0164] The second formula is used to calculate the predicted value of the monthly electricity consumption gap data of the target load in the target time period.
[0165] The second formula can be:
[0166]
[0167] in, It represents the predicted value of the monthly electricity consumption gap data of the target load in the target time period. L zc represents the first predicted value, λ L Indicates the confidence of the neural network prediction, its value is between 0 and 1, L xh represents the second predicted value, H xh Represents the time span for the neural network prediction.
[0168] Exemplarily, the data acquisition module 401 may be used to:
[0169] The original data of the historical monthly electricity consumption gap data is obtained, which is recorded as the first original data, and the original data of the historical daily electricity consumption gap data is obtained, which is recorded as the second original data.
[0170] The first original data is cleaned and normalized, and the CEEMDAN algorithm is used to decompose and reconstruct the first original data to obtain the historical monthly electricity consumption gap data.
[0171] The second original data is cleaned and normalized, and the CEEMDAN algorithm is used to decompose and reconstruct the second original data to obtain the historical daily electricity consumption gap data.
[0172] Exemplarily, the data acquisition module 401 may be used to:
[0173] Gaussian white noise is introduced into the first original data to obtain the first original data after the noise is introduced.
[0174] The first original data after the introduction of noise is subjected to empirical mode decomposition to obtain residual components.
[0175] If the number of extreme value points of the residual component is greater than or equal to 2, Gaussian white noise is added to the residual component, and the empirical mode decomposition is performed again to determine the number of extreme value points.
[0176] If the number of extreme points of the residual component is less than 2, the decomposition and reconstruction of the first original data is completed to obtain the historical monthly electricity consumption gap data.
[0177] Exemplarily, the first original data after the introduction of noise may be:
[0178] xi(t)=x(t)+cλi(t)
[0179] Among them, λ i (t) represents the Gaussian white noise introduced for the i-th time, c represents the Gaussian white noise coefficient, x(t) represents the first original data, x i (t) represents the first original data after the noise is introduced for the i-th time.
[0180] The third formula is used to calculate the residual component.
[0181] The third formula can be:
[0182] r1(t)=x i (t)-I1(t)
[0183]
[0184] Where r1(t) represents the first residual component, N represents the total number of components obtained by the first empirical mode decomposition, represents the i-th component after the first empirical mode decomposition.
[0185] The historical monthly electricity consumption gap data can be:
[0186]
[0187] Among them, x1 represents the historical monthly electricity consumption gap data, M represents the number of empirical mode decomposition, and R represents the final residual component.
[0188] Exemplarily, the data prediction module 402 may be used to:
[0189] The historical monthly electricity consumption gap data is processed by first-order accumulation to obtain the first processed data.
[0190] A differential equation is established based on the first processed data, and a constant factor is solved.
[0191] A first prediction equation is established based on the constant factor, and a first prediction value is obtained according to the first prediction equation.
[0192] Exemplarily, the data prediction module 402 may be used to:
[0193] The fourth formula is used to calculate the first processed data.
[0194] The fourth formula can be:
[0195]
[0196] Among them, x (1) ={x (1) (1),x (1) (2),x (1) (3),...,x (1) (n)} represents the first processed data, x (0) ={x (0) (1),x (0) (2),x (0) (3),...,x (0) (n)} represents the historical monthly electricity consumption gap data.
[0197] The differential equation can be:
[0198]
[0199] The constant factors in differential equations can be:
[0200]
[0201] The first prediction equation can be:
[0202]
[0203] Exemplarily, the data prediction module 402 may be used to:
[0204] Based on the historical daily electricity consumption gap data, an LSTM neural network prediction is performed to obtain a second prediction value; the LSTM neural network includes an input gate, an output gate and a forget gate.
[0205] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.
[0206] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.
[0207] Those of ordinary skill in the art will appreciate that the templates, units, and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0208] If the module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various electricity consumption gap prediction method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal and software distribution medium.
[0209] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.
Claims
1. A method for predicting electricity consumption gap, characterized in that: include: Obtain the historical monthly electricity consumption gap data and historical daily electricity consumption gap data of the target load; Perform grey prediction based on the historical monthly electricity consumption gap data to obtain a first prediction value; Performing neural network prediction based on the historical daily electricity consumption gap data to obtain a second prediction value; The second predicted value is corrected based on the first predicted value to obtain a predicted value of the daily power consumption gap data of the target load in the target time period; The first prediction value is corrected based on the second prediction value to obtain a prediction value of the monthly electricity consumption gap data of the target load in the target time period.
2. The method for predicting power consumption gap according to claim 1, characterized in that: The correcting the second predicted value based on the first predicted value to obtain a predicted value of the daily power consumption gap data of the target load in the target time period includes: The first formula is used to calculate the predicted value of the daily power consumption gap data of the target load in the target time period; The first formula is: in, represents the predicted value of the daily electricity consumption gap data of the target load in the target time period, L xh represents the second predicted value, μ L Indicates the confidence of grey prediction, its value is between 0 and 1, L zc represents the first predicted value, H zc Represents the time span of grey prediction.
3. The method for predicting power consumption gap according to claim 1, characterized in that: The step of correcting the first prediction value based on the second prediction value to obtain a prediction value of the monthly power consumption gap data of the target load in the target time period includes: The second formula is used to calculate the predicted value of the monthly electricity consumption gap data of the target load in the target time period; The second formula is: in, represents the predicted value of the monthly electricity consumption gap data of the target load in the target time period, L zc represents the first predicted value, λ L Indicates the confidence of the neural network prediction, its value is between 0 and 1, L xh represents the second predicted value, H xh Represents the time span for the neural network prediction.
4. The method for predicting power consumption gap according to claim 1, characterized in that: The obtaining of the historical monthly power consumption gap data and the historical daily power consumption gap data of the target load includes: Obtaining original data of historical monthly electricity consumption gap data, recorded as first original data, and obtaining original data of historical daily electricity consumption gap data, recorded as second original data; Cleaning and normalizing the first original data, and decomposing and reconstructing the first original data using the CEEMDAN algorithm to obtain the historical monthly electricity consumption gap data; The second original data is cleaned and normalized, and the CEEMDAN algorithm is used to decompose and reconstruct the second original data to obtain the historical daily electricity consumption gap data.
5. The method for predicting power consumption gap according to claim 4, characterized in that: The method of decomposing and reconstructing the first original data by using the CEEMDAN algorithm to obtain the historical monthly electricity consumption gap data includes: Introducing Gaussian white noise into the first original data to obtain the first original data after the noise is introduced; Performing empirical mode decomposition on the first original data after the noise is introduced to obtain a residual component; If the number of extreme value points of the residual component is greater than or equal to 2, then Gaussian white noise is added to the residual component, and empirical mode decomposition is performed again to determine the number of extreme value points; If the number of extreme points of the residual component is less than 2, the decomposition and reconstruction of the first original data is completed to obtain the historical monthly electricity consumption gap data.
6. The method for predicting power consumption gap according to claim 5, characterized in that: The first original data after the introduction of noise is: x i (t)=x(t)+cλ i (t) Among them, λ i (t) represents the Gaussian white noise introduced for the i-th time, c represents the Gaussian white noise coefficient, x(t) represents the first original data, x i (t) represents the first original data after the noise is introduced for the i-th time; Using the third formula, calculate the residual component; The third formula is: r1(t)=x i (t)-I1(t) Where r1(t) represents the first residual component, N represents the total number of components obtained by the first empirical mode decomposition, represents the i-th component after the first empirical mode decomposition; The historical monthly electricity consumption gap data are as follows: Wherein, x1 represents the historical monthly electricity consumption gap data, M represents the number of empirical mode decomposition, and R represents the final residual component.
7. The method for predicting power consumption gap according to claim 6, characterized in that: The grey prediction based on the historical monthly electricity consumption gap data to obtain a first prediction value includes: Performing first-order accumulation processing on the historical monthly electricity consumption gap data to obtain first processed data; Establishing a differential equation based on the first processed data and solving for a constant factor; A first prediction equation is established based on a constant factor, and the first prediction value is obtained according to the first prediction equation.
8. The method for predicting power consumption gap according to claim 7, characterized in that: The first order accumulation processing is performed on the historical monthly electricity consumption gap data to obtain first processed data, including: Using a fourth formula, calculating the first processed data; The fourth formula is: Among them, x (1) ={x (1) (1),x (1) (2),x (1) (3),...,x (1) (n)} represents the first processed data, x (0) ={x (0) (1),x (0) (2),x (0) (3),...,x (0) (n)} represents the historical monthly electricity consumption gap data; The differential equation is: The constant factor in the differential equation is: The first prediction equation is:
9. The method for predicting power consumption gap according to claim 1, characterized in that: The performing neural network prediction based on the historical daily electricity consumption gap data to obtain a second prediction value includes: An LSTM neural network prediction is performed based on the historical daily electricity consumption gap data to obtain a second prediction value; the LSTM neural network includes an input gate, an output gate and a forget gate.
10. A device for predicting power consumption shortfall, characterized in that: include: A data acquisition module is used to obtain historical monthly power consumption gap data and historical daily power consumption gap data of the target load; A data prediction module, configured to perform grey prediction based on the historical monthly electricity consumption gap data to obtain a first prediction value; Performing neural network prediction based on the historical daily electricity consumption gap data to obtain a second prediction value; A data correction module, configured to correct the second prediction value based on the first prediction value to obtain a prediction value of the daily power consumption gap data of the target load in a target time period; The first prediction value is corrected based on the second prediction value to obtain a prediction value of the monthly electricity consumption gap data of the target load in the target time period.