Transformer substation carbon emission calculation method based on machine learning

Through machine learning-based methods, the access points are subdivided and the working parameters of each time step are collected, the substation parameter matrix and carbon emission matrix are constructed, and the carbon emissions of the substation are predicted, which solves the failure to consider the power access points in traditional methods. The problem of load changes improves the prediction accuracy of carbon emissions.

CN119090063APending Publication Date: 2024-12-06MARKETING SERVICE CENT OF STATE GRID HEILONGJIANG ELECTRIC POWER CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202411132892.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The traditional substation carbon emission calculation method fails to effectively consider the power load changes at the power access point, resulting in low prediction accuracy.

Method used

Using a machine learning-based method, by constructing the substation parameter matrix and carbon emission matrix, using hidden layers and classifier models to predict the power loss value and carbon emissions, subdividing the access points and collecting working parameters of each time step, considering daily operation and maintain carbon emissions.

Benefits of technology

It improves the accuracy of the prediction of carbon emissions in substations, takes into account the workload and carbon emission impacts of each access point, and provides a more accurate comprehensive carbon emission calculation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119090063A_ABST
    Figure CN119090063A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of transformer substation carbon emission calculation, and discloses a transformer substation carbon emission calculation method based on machine learning, and the method comprises the following steps: S101, constructing a transformer substation parameter matrix A; s102, inputting the transformer substation parameter matrix A into a first prediction model, wherein an output value represents a predicted electric quantity loss value in a period M; and S103, obtaining the carbon emission generated by each access point in the operation process of the transformer substation in the time period T, and calculating the predicted operation carbon emission of the transformer substation in the period M. According to the method, the working parameters of each access point in each time step are collected, and the carbon emission of the transformer substation in the future time period T is predicted according to the historical working parameter information of each access point. The influence of the workload on the carbon emission of each access point of the transformer substation is considered, and the prediction calculation result is more accurate.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of substation carbon emission calculation, and more specifically, to a substation carbon emission calculation method based on machine learning. Background Art

[0002] The carbon emissions of traditional substations are calculated by combining the fuel consumption method with the electricity consumption method. The fuel consumption method is to obtain the fuel carbon emissions by multiplying the fuel consumption (such as coal, natural gas, diesel, etc.) of the substation by the fuel carbon emission coefficient corresponding to different fuels. The electricity consumption method is to obtain the electricity carbon emissions by multiplying the electricity consumption of the substation by the corresponding electricity carbon emission coefficient. The sum of the two is the carbon emissions of the substation.

[0003] The publication number is CN1 15564138A, and the patent name is a Chinese patent for a method, device and equipment for predicting carbon emissions from a substation. It discloses: the carbon emissions of a substation are obtained by taking a weighted sum of the service life of the substation, the operating energy consumption of electrical equipment, the maintenance probability and the replacement probability.

[0004] Although the above scheme takes into account the maintenance and replacement of electrical equipment in the substation, it does not take into account the impact of the power load at the power access point of the substation on the substation, such as the increase in power consumption due to air conditioning in summer and the increase in power consumption due to heating in winter, resulting in low accuracy in the prediction calculation. Summary of the invention

[0005] The present invention provides a method for calculating carbon emissions of a substation based on machine learning, which solves the technical problems in the above-mentioned background technology.

[0006] The present invention provides a method for calculating carbon emissions of a substation based on machine learning, comprising the following steps:

[0007] Step S101, moving forward a time period T from the current time, collecting the working parameters of the substation in the time period T, and constructing the substation parameter matrix A, which is expressed as:

[0008] in, It represents the working parameters of the first access point of the substation from the 1st time step to the nth time step; represents the working parameters of the substation from the 1st access point to the mth access point at the 1st time step; Indicates the working parameters of the first access point to the mth access point of the substation from the first time step to the nth time step;

[0009] in, Respectively represent the maximum current parameter, minimum current parameter, current loss, maximum voltage parameter and minimum voltage parameter of the mth access point of the substation at the nth time step and the rainfall change parameter, light intensity change parameter and temperature change parameter corresponding to each access point;

[0010] Step S102, the substation parameter matrix A is input into the first prediction model, and the output value represents the predicted power loss value within a period M, where M represents the predicted time period;

[0011] Step S103, obtaining the carbon emissions generated by each access point during the operation of the substation in the time period T, and calculating the predicted operating carbon emissions of the substation in the period M;

[0012] Step S104, calculating the predicted comprehensive carbon emissions of the substation within the period M.

[0013] Furthermore, M, T, m and n are all custom parameters, which are determined according to the actual number of access points of the substation.

[0014] Furthermore, the first prediction model includes n columns of hidden layers, each hidden layer includes m hidden units, wherein the hidden unit in the i-th row and j-th column inputs the working parameters of the i-th row and j-th column of the substation parameter matrix A, and the lower m-th hidden unit of the n-th column of the hidden layer outputs to a classifier, and the classification space of the classifier represents the power loss value of the substation within the prediction time period M, wherein 1≤i≤n, 1≤j≤m.

[0015] Furthermore, the calculation formula of the first prediction model includes:

[0016] Update gate of the vth hidden unit in the uth hidden layer The calculation formula is as follows:

[0017]

[0018] Where W z , U z and b z They represent the update gate of the vth hidden unit in the uth hidden layer. corresponding first weight parameter, second weight parameter and bias parameter;

[0019] The reset gate of the vth hidden unit in the uth hidden layer The calculation formula is as follows:

[0020]

[0021] Among them, W r , U r and b rRepresents the reset gate of the vth hidden unit in the uth hidden layer corresponding first weight parameter, second weight parameter and bias parameter;

[0022] The hidden state of the vth hidden unit in the uth hidden layer The calculation formula is as follows:

[0023]

[0024] Where W f , U f and b f Respectively represent the current memory content of the vth hidden unit in the uth hidden layer corresponding first weight parameter, second weight parameter and bias parameter;

[0025] Candidate hidden state of the vth hidden unit in the uth hidden layer The calculation formula is as follows:

[0026]

[0027] In the above formula, Represents the input of the current time step; h u-1 represents the hidden state of the previous time step; All are assigned a value of 0; σ represents the sigmoid activation function; tanh represents the hyperbolic tangent activation function; ⊙ represents the dot product.

[0028] Furthermore, the predicted power loss value corresponding to the training sample used to train the first prediction model is determined by the actually detected power loss value.

[0029] Furthermore, the calculation formula for the carbon emissions corresponding to the predicted power loss value of the substation in period M is as follows:

[0030] Q = α q;

[0031] Where Q represents the carbon emissions corresponding to the predicted electricity loss value of the substation within period M; α represents the carbon emission coefficient of the region where the substation is located; and q represents the predicted electricity loss value of the substation within M.

[0032] Furthermore, the steps for calculating the predicted operating carbon emissions of the substation in period M are as follows:

[0033] Step S201, constructing a carbon emission matrix B according to the information collected in step S103, the carbon emission matrix B is expressed as:

[0034] in, represents the set of carbon emissions of the first access point of the substation from the first time period to the nth time period; represents the set of carbon emissions from the first access point to the mth access point of the substation in the first time period; represents the set of carbon emissions from the 1st access point to the mth access point of the substation from the 1st time step to the nth time step;

[0035] in, They represent the operation carbon emission, maintenance carbon emission and fault carbon emission of the first access point of the substation at the mth time step respectively;

[0036] Step S202: input the carbon emission matrix B into the second prediction model, and the output value represents the predicted carbon emission of substation operation and maintenance within the period M.

[0037] Furthermore, the calculation formula of the second prediction model is the same as that of the first prediction model.

[0038] Furthermore, the predicted carbon emissions from operation and maintenance of the substation corresponding to the training samples used to train the second prediction model are determined by actually detected carbon emissions.

[0039] The beneficial effects of the present invention are as follows: the substation is subdivided into m access points, and the working parameters of each access point at each time step are collected, and the carbon emissions of the substation in the future time period T are predicted based on the historical working parameter information of each access point, and the carbon emissions of the daily operation and maintenance of the substation are also taken into account. Compared with the common carbon emission prediction and calculation method, the impact of the workload on the carbon emissions of each access point of the substation is taken into account, and the prediction and calculation results are more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is the overall flow chart of the present invention;

[0041] Figure 2 It is a flowchart of calculating the predicted operating carbon emissions of a substation within a period M according to the present invention. DETAILED DESCRIPTION

[0042] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0043] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in one or more embodiments of the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in one or more embodiments of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0044] like Figure 1-Figure 2 As shown, a method for calculating carbon emissions of a substation based on machine learning includes the following steps:

[0045] Step S101, moving forward a time period T from the current time, collecting the working parameters of the substation in the time period T, and constructing the substation parameter matrix A, which is expressed as:

[0046] in, It represents the working parameters of the first access point of the substation from the 1st time step to the nth time step; represents the working parameters of the substation from the 1st access point to the mth access point at the 1st time step; Indicates the working parameters of the first access point to the mth access point of the substation from the first time step to the nth time step;

[0047] in, Respectively represent the maximum current parameter, minimum current parameter, current loss, maximum voltage parameter and minimum voltage parameter of the mth access point of the substation at the nth time step and the rainfall change parameter, light intensity change parameter and temperature change parameter corresponding to each access point;

[0048] The maximum current parameter, minimum current parameter, current consumption, maximum voltage parameter and minimum voltage parameter of the substation can represent the working load of the substation;

[0049] The rainfall variation parameters, light intensity variation parameters and temperature variation parameters corresponding to the substation access point can reflect the climate conditions of the area corresponding to the access point;

[0050] Step S102, the substation parameter matrix A is input into the first prediction model, and the output value represents the predicted power loss value within a period M, where M represents the predicted time period;

[0051] Step S103, obtaining the carbon emissions generated by each access point during the operation of the substation in the time period T, and calculating the predicted operating carbon emissions of the substation in the period M;

[0052] Step S104, calculating the predicted comprehensive carbon emissions of the substation within the period M.

[0053] In one embodiment of the present invention, M, T, m and n are all custom parameters, wherein they are determined according to the actual number of access points of the substation. Preferably, T>M, which can ensure that the collected working parameters of the substation are richer and improve the accuracy of predicting the power loss value within the period M.

[0054] In one embodiment of the present invention, the first prediction model includes n columns of hidden layers, each hidden layer includes m hidden units, wherein the hidden unit in the i-th row and j-th column inputs the working parameters of the i-th row and j-th column of the substation parameter matrix A, and the lower m-th hidden unit in the n-th column of the hidden layer outputs to a classifier, and the classification space of the classifier represents the power loss value of the substation within the prediction time period M, wherein 1≤i≤n, 1≤j≤m.

[0055] In one embodiment of the present invention, the calculation formula of the first prediction model includes:

[0056] Update gate of the vth hidden unit in the uth hidden layer The calculation formula is as follows:

[0057]

[0058] Where W z , U z and b z They represent the update gate of the vth hidden unit in the uth hidden layer. corresponding first weight parameter, second weight parameter and bias parameter;

[0059] The reset gate of the vth hidden unit in the uth hidden layer The calculation formula is as follows:

[0060]

[0061] Among them, W r , U r and b r Represents the reset gate of the vth hidden unit in the uth hidden layer corresponding first weight parameter, second weight parameter and bias parameter;

[0062] The hidden state of the vth hidden unit in the uth hidden layer The calculation formula is as follows:

[0063]

[0064] Where W f , U f and b f Respectively represent the current memory content of the vth hidden unit in the uth hidden layer corresponding first weight parameter, second weight parameter and bias parameter;

[0065] Candidate hidden state of the vth hidden unit in the uth hidden layer The calculation formula is as follows:

[0066]

[0067] In the above formula, Represents the input of the current time step; h u-1 represents the hidden state of the previous time step; All are assigned a value of 0; σ represents the sigmoid activation function; tanh represents the hyperbolic tangent activation function; ⊙ represents the dot product.

[0068] In one embodiment of the present invention, the predicted electricity loss value corresponding to the training sample used to train the first prediction model is determined by the actual detected electricity loss value. When training the first prediction model, the historical working parameters of the substation are input into the first prediction model, and then the actual electricity loss value is compared with the electricity loss value predicted by the first prediction model. If the error exceeds the set error threshold, the parameters are adjusted and retrained until the error is within the set error threshold. The training of the first prediction model is terminated and the error threshold is set as a custom parameter.

[0069] In one embodiment of the present invention, the calculation formula for the carbon emissions corresponding to the predicted power loss value of the substation in period M is as follows:

[0070] Q = α q;

[0071] Where Q represents the carbon emissions corresponding to the predicted electricity loss value of the substation within period M; α represents the carbon emission coefficient of the region where the substation is located; and q represents the predicted electricity loss value of the substation within M.

[0072] In one embodiment of the present invention, the steps of calculating the predicted operating carbon emissions of the substation in period M are as follows:

[0073] Step S201, constructing a carbon emission matrix B according to the information collected in step S103, the carbon emission matrix B is expressed as:

[0074] in, represents the set of carbon emissions of the first access point of the substation from the first time period to the nth time period; represents the set of carbon emissions from the first access point to the mth access point of the substation in the first time period; represents the set of carbon emissions from the 1st access point to the mth access point of the substation from the 1st time step to the nth time step;

[0075] in, They represent the operating carbon emissions, maintenance carbon emissions, and fault carbon emissions of the first access point of the substation at the mth time step. During the operation of the substation, control personnel and maintenance personnel are required to go to the substation for operation, which will generate indirect carbon emissions. In addition, when faults occur at different access points, maintenance personnel are required to go to the substation for repair. In this way, the carbon emissions of different access points will be different, which can better reflect the impact of different loads of each access point on carbon emissions.

[0076] Step S202: input the carbon emission matrix B into the second prediction model, and the output value represents the predicted carbon emission of substation operation and maintenance within the period M.

[0077] In one embodiment of the present invention, the calculation formula of the second prediction model is the same as that of the first prediction model.

[0078] In one embodiment of the present invention, the predicted carbon emissions from substation operation and maintenance corresponding to the training samples used to train the second prediction model are determined by actually detected carbon emissions, using the same method as the training method of the first prediction model.

[0079] In one embodiment of the present invention, after the first prediction model and the second prediction model have completed their predictions, the carbon emissions corresponding to the predicted electricity loss value within period M and the carbon emissions of the predicted substation operation and maintenance within period M are weighted and summed to obtain the predicted comprehensive carbon emissions of the substation within period M.

[0080] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A method for calculating carbon emissions from a substation based on machine learning, characterized in that: The steps include: Step S101, moving forward a time period T from the current time, collecting the working parameters of the substation in the time period T, and constructing the substation parameter matrix A, which is expressed as: in, It represents the working parameters of the first access point of the substation from the 1st time step to the nth time step; represents the working parameters of the substation from the 1st access point to the mth access point at the 1st time step; Indicates the working parameters of the first access point to the mth access point of the substation from the first time step to the nth time step; in, Respectively represent the maximum current parameter, minimum current parameter, current loss, maximum voltage parameter and minimum voltage parameter of the mth access point of the substation at the nth time step and the rainfall change parameter, light intensity change parameter and temperature change parameter corresponding to each access point; Step S102, the substation parameter matrix A is input into the first prediction model, and the output value represents the predicted power loss value within a period M, where M represents the predicted time period; Step S103, obtaining the carbon emissions generated by each access point during the operation of the substation in the time period T, and calculating the predicted operating carbon emissions of the substation in the period M; Step S104, calculating the predicted comprehensive carbon emissions of the substation within the period M.

2. The method for calculating carbon emissions of a substation based on machine learning according to claim 1 is characterized in that: M, T, m and n are all custom parameters, which are determined according to the actual number of access points in the substation.

3. The method for calculating carbon emissions of a substation based on machine learning according to claim 1 is characterized in that: The first prediction model includes n columns of hidden layers, each hidden layer includes m hidden units, wherein the hidden unit in the i-th row and j-th column inputs the working parameters of the i-th row and j-th column of the substation parameter matrix A, and the lower m-th hidden unit of the n-th hidden layer outputs to a classifier, and the classification space of the classifier represents the power loss value of the substation within the prediction time period M, wherein 1≤i≤n, 1≤j≤m.

4. The method for calculating carbon emissions of a substation based on machine learning according to claim 3 is characterized in that: The calculation formula of the first prediction model includes: Update gate of the vth hidden unit in the uth hidden layer The calculation formula is as follows: Where W z , U z and b z They represent the update gate of the vth hidden unit in the uth hidden layer. corresponding first weight parameter, second weight parameter and bias parameter; The reset gate of the vth hidden unit in the uth hidden layer The calculation formula is as follows: Among them, W r , U r and b r Represents the reset gate of the vth hidden unit in the uth hidden layer corresponding first weight parameter, second weight parameter and bias parameter; The hidden state of the vth hidden unit in the uth hidden layer The calculation formula is as follows: Where W f , U f and b f Respectively represent the current memory content of the vth hidden unit in the uth hidden layer corresponding first weight parameter, second weight parameter and bias parameter; Candidate hidden state of the vth hidden unit in the uth hidden layer The calculation formula is as follows: In the above formula, 1≤u≤n, 1≤v≤n, Represents the input of the current time step; h u-1 represents the hidden state of the previous time step; All are assigned a value of 0; σ represents the sigmoid activation function; tanh represents the hyperbolic tangent activation function; ⊙ represents the dot product.

5. The method for calculating carbon emissions of a substation based on machine learning according to claim 4 is characterized in that: The predicted power loss value corresponding to the training sample used to train the first prediction model is determined by the actually detected power loss value.

6. The method for calculating carbon emissions of a substation based on machine learning according to claim 1 is characterized in that: The calculation formula for the carbon emissions corresponding to the predicted power loss value of the substation in period M is as follows: Q = α q; Where Q represents the carbon emissions corresponding to the predicted electricity loss value of the substation within period M; α represents the carbon emission coefficient of the region where the substation is located; and q represents the predicted electricity loss value of the substation within M.

7. The method for calculating carbon emissions of a substation based on machine learning according to claim 5 is characterized in that: The steps to calculate the predicted operating carbon emissions of a substation in period M are as follows: Step S201, constructing a carbon emission matrix B according to the information collected in step S103, the carbon emission matrix B is expressed as: in, represents the set of carbon emissions of the first access point of the substation from the first time period to the nth time period; represents the set of carbon emissions from the first access point to the mth access point of the substation in the first time period; represents the set of carbon emissions from the 1st access point to the mth access point of the substation from the 1st time step to the nth time step; in, They represent the operation carbon emission, maintenance carbon emission and fault carbon emission of the first access point of the substation at the mth time step respectively; Step S202: input the carbon emission matrix B into the second prediction model, and the output value represents the predicted carbon emission of substation operation and maintenance within the period M.

8. The method for calculating carbon emissions of a substation based on machine learning according to claim 7 is characterized in that: The calculation formula of the second prediction model is the same as that of the first prediction model.

9. The method for calculating carbon emissions of a substation based on machine learning according to claim 8, characterized in that: The predicted carbon emissions from operation and maintenance of the substation corresponding to the training samples used to train the second prediction model are determined by actually detected carbon emissions.

Citation Information

Patent Citations

  • Method, device and equipment for predicting carbon emission of transformer substation

    CN115564138A