A Method and System for Carbon Emission Governance in a Distribution Network with Flexible Load Participation

Abstract

The present invention belongs to the technical field of power systems, and particularly relates to a method and system for carbon emission governance of a distribution network involving flexible loads. The method includes constructing a flexible load prediction model for users; constructing a total load prediction model for users, and inputting the flexible load prediction data of users into the total load prediction model of users to obtain the total load prediction data of users; calculating the three-phase unbalance loss in the distribution area based on the total load prediction data of users, and calculating the carbon emission index in the distribution area based on the total load prediction data of users and the energy output prediction data; constructing a bi-objective optimization model with the minimum three-phase unbalance loss in the distribution area and the minimum carbon emission index in the distribution area as the objectives, and obtaining the optimal flexible load data of users by solving the bi-objective optimization model. The present invention matches the flexible load with the output of distributed power sources, reduces the peak load and the grid connection loss, and at the same time reduces the unbalance degree and the transformer line loss by optimizing the three-phase load.

Description

Technical Field

[0001] The present invention belongs to the technical field of power systems, and particularly relates to a method and system for carbon emission governance of a distribution network involving flexible loads. Background Art

[0002] Integrated energy systems can promote the consumption of renewable energy, achieve the optimal utilization of resources, and improve the utilization rate of integrated energy, which is an important measure to achieve energy conservation and emission reduction. However, the access of distributed photovoltaic has complicated the current situation of network losses in the distribution network. The penetration rate of distributed photovoltaic in the distribution network is rapidly increasing, and the line loss level and influencing factors of the low-voltage distribution network are the key points and difficulties of line loss management, and its line loss rate directly involves the economic interests of both power supply and power consumption parties.

[0003] At present, effective methods for energy conservation and loss reduction of distribution networks with sources have been proposed at home and abroad. The main methods include: energy-saving transformation of lines, using reactive power compensation devices, reducing the power supply range of distribution areas, and achieving loss reduction. This method calculates the power supply area and reasonably distributes the transformation funds according to network losses, nests the calculation of line losses of a single feeder under the power supply sub-area, forms a hierarchical optimization model of power supply sub-area - feeder line losses, and generates alternative transformation plans. However, this plan is based on the division of distributed power supply sub-areas, and as the photovoltaic penetration rate increases, the sub-area transformation plan changes accordingly.

[0004] Optimizing distributed power sources, considering the uncertainty of voltage and source-load, and from the perspective of the operation of distribution networks with sources, achieving the effect of reducing network losses by optimizing the location and capacity of distributed power sources; however, there are similar problems to line transformation, and the installation location, available capacity, and actual output efficiency of power sources are restricted by geography, meteorology, user needs, etc. Summary of the Invention

[0005] To solve the deficiencies of the prior art, the present invention provides a method and system for carbon emission governance of a distribution network involving flexible loads, which realizes local consumption by matching the flexible loads with the output of distributed power sources, reduces peak loads and grid connection losses, and at the same time reduces the unbalance degree and transformer line losses by optimizing three-phase loads.

[0006] To solve the deficiencies of the prior art, the technical solution provided by the present invention is as follows:

[0007] A method for carbon emission governance of a distribution network involving flexible loads includes:

[0008] Constructing a flexible load prediction model for users to obtain flexible load prediction data of users;

[0009] Constructing a total load prediction model for users, and inputting the flexible load prediction data of users into the total load prediction model of users to obtain total load prediction data of users;

[0010] Calculate the three-phase unbalance loss in the distribution area based on the total load prediction data of users, and calculate the carbon emission index in the distribution area based on the total load prediction data of users and the energy output prediction data;

[0011] Construct a bi-objective optimization model with the minimum three-phase unbalance loss in the distribution area and the minimum carbon emission index in the distribution area as the objectives, and obtain the optimal data of the flexible load of users by solving the bi-objective optimization model.

[0012] Preferably, the construction of the flexible load prediction model of users to obtain the flexible load prediction data of users includes,

[0013]

[0014] Among them, is the flexible load prediction data of the i-th user at time t; i = 1, 2, 3, …, M, where M is the total number of users in the distribution area; T(t) represents the predicted outdoor temperature value at time t, represents the preset temperature value of the air conditioner of the i-th user at time t; θ i represents the air conditioner energy efficiency coefficient of the i-th user, δ i represents the electric thermal resistance of the air conditioner of the i-th user;

[0015] Satisfy the following constraint conditions:

[0016]

[0017] Among them, is the indoor temperature of the i-th user at time t + Δt; is the indoor temperature of the i-th user at time t; is the air conditioner operation indicator of the i-th user at time t; β i is the air conditioner cooling parameter of the i-th user; is the maximum air conditioner load of the i-th user;

[0018] Among them, β i The calculation formula is as follows:

[0019]

[0020] Among them, C i represents the air conditioner capacitance of the i-th user;

[0021] Among them, Satisfy:

[0022]

[0023] Among them, ε i(t) is the bandwidth, representing the difference between the indoor temperature of the i-th user at time t and the air conditioner set temperature value. The difference.

[0024] Preferably, the total load prediction model of the user is constructed, and the flexible load prediction data of the user is input into the total load prediction model of the user to obtain the total load prediction data of the user, including

[0025] Taking every 15 minutes as an interval, extracting the daily total load historical data of the user, and performing denoising and normalization processing on the daily total load historical data;

[0026] Using the K-means algorithm to perform clustering analysis on the daily total load historical data of the user, and classifying the user into four types of users by using the SSE clustering effect evaluation index;

[0027] Based on the non-intrusive load decomposition method, decomposing the daily total load historical data of the user into flexible load historical data and basic load historical data;

[0028] Taking historical electricity prices, flexible load historical data, basic load historical data, and historical meteorological data as feature quantities, and taking the daily total load historical data of the user as the result, training the deep learning LSTM network model for four types of users respectively;

[0029] Inputting the predicted electricity price, flexible load prediction data, basic load prediction data, and meteorological prediction data into the deep learning LSTM network model to obtain the total load prediction data of the user.

[0030] Preferably, calculating the three-phase unbalance loss in the distribution area based on the total load prediction data of the user, including

[0031]

[0032] Among them, F loss (t) represents the total sum of the three-phase unbalance losses of all transformers in the distribution area at time t, and ρ j (t) represents the three-phase negative sequence unbalance degree of the j-th transformer at time t; R a is the first line resistance; R b is the second line resistance; R c is the third line resistance; j = 1, 2, 3,..., N, and N is the total number of transformers in the distribution area;

[0033] ρ j (t) The calculation formula is as follows:

[0034]

[0035] Among them, L j (t) represents the first parameter of the three-phase unbalance index, and the calculation formula is as follows:

[0036]

[0037] Among them, a j (t), b j (t), c j (t) are respectively the effective values of the fundamental wave components of the three-phase currents of the j-th transformer at the t-th moment, and a j (t), b j (t), c j (t) are respectively equal to the normalized values of the sum of the total load prediction data of the users connected to the first line, the sum of the total load prediction data of the users connected to the second line, and the sum of the total load prediction data of the users connected to the third line of the j-th transformer at the t-th moment.

[0038] Preferably, calculating the carbon emission index in the distribution area based on the total load prediction data and energy output prediction data of users includes,

[0039]

[0040] Among them, CI(t) represents the carbon emission index of the distribution area at the t-th moment, is the power emission factor of the transformer in the distribution area at the t-th moment; F i (t) is the total load prediction data of the i-th user at the t-th moment; i = 1, 2, 3,..., M, where M is the total number of users in the distribution area;

[0041] The calculation formula of

[0042]

[0043] Among them, K m (t) represents the proportion of the predicted output data of the m-th type of energy at the t-th moment in the total predicted output data of the distribution area, m = 1, 2, 3,..., Q, where Q is the total number of energy types; E m represents the carbon emission factor of the m-th type of energy.

[0044] A distribution network carbon emission governance system involving flexible loads includes a flexible load prediction module, a total load prediction module, a three-phase unbalance loss calculation module, a carbon emission index calculation module, and an optimization module;

[0045] The flexible load prediction module is used to construct a flexible load prediction model for users and obtain the flexible load prediction data of users;

[0046] The total load prediction module is used to construct a total load prediction model for users and input the flexible load prediction data of users into the total load prediction model of users to obtain the total load prediction data of users;

[0047] The three-phase unbalance loss calculation module is configured to calculate the three-phase unbalance loss in the distribution area based on the predicted data of the total user load;

[0048] The carbon emission index calculation module is configured to calculate the carbon emission index in the distribution area based on the predicted data of the total user load and the predicted data of the energy output;

[0049] The optimization module is configured to construct a bi-objective optimization model with the goals of minimizing the three-phase unbalance loss and the carbon emission index in the distribution area, and obtain the optimal flexible load data of the user by solving the bi-objective optimization model.

[0050] Preferably, the flexible load prediction module is configured to calculate the predicted flexible load data according to the following formula:

[0051]

[0052] where, is the predicted flexible load data of the i-th user at time t; i = 1, 2, 3, …, M, where M is the total number of users in the distribution area; T(t) represents the predicted outdoor temperature at time t, represents the preset temperature value of the air conditioner of the i-th user at time t; θ i represents the air conditioner energy efficiency coefficient of the i-th user, and δ i represents the electric thermal resistance of the air conditioner of the i-th user;

[0053] which satisfies the following constraint conditions:

[0054]

[0055] where, is the indoor temperature of the i-th user at time t+Δt; is the indoor temperature of the i-th user at time t; is the air conditioner operation indicator of the i-th user at time t; β i is the air conditioner refrigeration parameter of the i-th user; is the maximum air conditioner load of the i-th user;

[0056] where, β i is calculated according to the following formula:

[0057]

[0058] where, C i represents the air conditioner capacitance of the i-th user;

[0059] where, satisfies:

[0060]

[0061] Among them, ε i (t) is the bandwidth, representing the difference between the indoor temperature of the i-th user at time t and the air-conditioning set temperature value. The difference.

[0062] Preferably, the total load prediction module is used to

[0063] Extract the historical daily total load data of users at intervals of every 15 minutes, and perform denoising and normalization processing on the historical daily total load data;

[0064] Use the K-means algorithm to perform clustering analysis on the historical daily total load data of users, and divide users into four categories of users by using the SSE clustering effect evaluation index;

[0065] Based on the non-intrusive load decomposition method, decompose the historical daily total load data of users into flexible load historical data and basic load historical data;

[0066] Use historical electricity prices, flexible load historical data, basic load historical data, and historical meteorological data as feature quantities, and use the historical daily total load data of users as the result to train the deep learning LSTM network model for the four categories of users respectively;

[0067] Input the predicted electricity price, flexible load prediction data, basic load prediction data, and meteorological prediction data into the deep learning LSTM network model to obtain the total load prediction data of users.

[0068] Preferably, the three-phase unbalance loss calculation module is used to calculate the three-phase unbalance loss according to the following formula:

[0069]

[0070] Among them, F loss (t) represents the total three-phase unbalance loss of all transformers in the distribution area at time t, and ρ j (t) represents the three-phase negative sequence unbalance degree of the j-th transformer at time t; R a is the first line resistance; R b is the second line resistance; R c is the third line resistance; j = 1, 2, 3,..., N, and N is the total number of transformers in the distribution area;

[0071] ρ j (t) The calculation formula is as follows:

[0072]

[0073] Among them, L j(t) represents the first parameter of the three-phase unbalance index, and the calculation formula is as follows:

[0074]

[0075] Among them, a j (t), b j (t), c j (t) are respectively the effective values of the fundamental wave components of the three-phase currents of the jth transformer at the tth moment. a j (t), b j (t), c j (t) are respectively equal to the normalized values of the sum of the total load prediction data of the users connected to the first line, the sum of the total load prediction data of the users connected to the second line, and the sum of the total load prediction data of the users connected to the third line of the jth transformer at the tth moment.

[0076] Preferably, the carbon emission index calculation module is used to calculate the carbon emission index according to the following formula:

[0077]

[0078] Among them, CI(t) represents the carbon emission index of the distribution area at the tth moment, is the power emission factor of the transformer in the distribution area at the tth moment; F i (t) is the total load prediction data of the ith user at the tth moment; i = 1, 2, 3,..., M, where M is the total number of users in the distribution area;

[0079] The calculation formula of is as follows:

[0080]

[0081] Among them, K m (t) represents the proportion of the predicted output data of the mth type of energy at the tth moment in the total predicted output data of the distribution area, m = 1, 2, 3,..., Q, where Q is the total number of energy types; E m represents the carbon emission factor of the mth type of energy.

[0082] Advantages of the present invention:

[0083] The present invention proposes a two-layer planning model for a distribution network. In the upper-layer new energy consumption planning, the minimum comprehensive carbon emissions of the distribution network are taken as the optimization goal; in the lower-layer line loss governance planning, the minimum three-phase unbalance loss is taken as the optimization goal, and a dual-objective optimization algorithm is used for solving. Through the matching of flexible loads and distributed power generation outputs, in-situ consumption is realized, peak loads and grid connection losses are reduced. At the same time, by optimizing the three-phase loads, the unbalance degree and transformer line losses are reduced. Description of the Drawings

[0084] Figure 1 It is a flowchart of the method for managing carbon emissions in a distribution network with flexible loads provided by the present invention. Specific embodiments

[0085] The present invention will be further described below in conjunction with the embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention and cannot be used to limit the protection scope of the present invention.

[0086] An embodiment of the present invention provides a method for managing carbon emissions in a distribution network with flexible loads. Refer to Figure 1 , including

[0087] 1. Build a flexible load prediction model for users to obtain flexible load prediction data of users:

[0088] Define the flexible load of users as the air-conditioning load of users, establish an air-conditioning cooling model, and the air-conditioning load prediction model of users, that is, the flexible load prediction model, is as follows:

[0089]

[0090] Where is the air-conditioning load prediction data of the i-th user at time t, and is also the flexible load prediction data of the i-th user; i = 1, 2, 3,..., M, where M is the total number of users in the distribution area; T(t) represents the predicted outdoor temperature value at time t, represents the preset temperature value of the air conditioner of the i-th user at time t; θ i represents the air-conditioning energy efficiency coefficient of the i-th user, and δ i represents the electric resistance of the air conditioner of the i-th user;

[0091] Satisfy the following constraint conditions:

[0092]

[0093] Where is the indoor temperature of the i-th user at time t + Δt; is the indoor temperature of the i-th user at time t; is the air-conditioning operation indicator of the i-th user at time t; β i is the air-conditioning cooling parameter of the i-th user; is the maximum air-conditioning load of the i-th user;

[0094] Where β i The calculation formula is as follows:

[0095]

[0096] Where Ci Denote the air conditioner capacitor of the \(i\)th user;

[0097] Satisfy:

[0098]

[0099] where \(\varepsilon\) i (t) is the bandwidth, representing the difference between the indoor temperature of the \(i\)th user at time \(t\) and the air conditioner set temperature value of the difference; is 1, indicating that the air conditioner needs to run for cooling, is 0, indicating that the air conditioner needs to stop working.

[0100] 2. Construct a total load prediction model for users, and input the flexible load prediction data of users into the total load prediction model of users to obtain the total load prediction data of users:

[0101] Taking 15 minutes as an interval, extract the daily total load historical data of users, and perform denoising and normalization processing on the daily total load historical data;

[0102] Use the K-means algorithm to perform clustering analysis on the daily total load historical data of users, and use the SSE clustering effect evaluation index to determine that the optimal number of clusters is 4, and divide users into four types of users;

[0103] Based on the non-intrusive load decomposition method, according to the current wave characteristics of flexible loads, decompose the daily total load historical data of users into flexible load historical data and basic charge historical data;

[0104] Taking historical electricity prices, flexible load historical data, basic load historical data, and historical meteorological data as feature quantities, and taking the daily total load historical data of users as the result, train deep learning LSTM network models for four types of users respectively; the training set is one week of continuous feature data and 7*96 point load data, and the test set is the feature data of the day after one week and 96 point load data;

[0105] After the deep learning LSTM network model is trained, input the predicted electricity price, flexible charge prediction data, basic load prediction data, and meteorological prediction data into the deep learning LSTM network model to obtain the total load prediction data of users.

[0106] Among them, the predicted electricity price can be obtained by the unary linear regression method. The basic load prediction data is obtained by removing noise from the basic load historical data through the moving average filtering algorithm.

[0107] 3. Calculate the three-phase unbalance loss in the distribution area based on the total load prediction data of users:

[0108]

[0109] Among them, F loss (t) represents the total three-phase unbalance loss of all transformers in the distribution area at time t, and ρ j (t) represents the three-phase negative-sequence unbalance degree of the j-th transformer at time t; R a is the resistance of the first line; R b is the resistance of the second line, and R c is the resistance of the third line, a j (t), b j (t), c j (t) are respectively the effective values of the fundamental wave components of the three-phase currents of the j-th transformer at time t, a j (t), b j (t), c j (t) are respectively the normalized values of the sum of the total load prediction data of the users connected to the first line, the sum of the total load prediction data of the users connected to the second line, and the sum of the total load prediction data of the users connected to the third line of the j-th transformer at time t, j = 1, 2, 3,..., N, where N is the total number of transformers in the distribution area;

[0110] The intermediate parameters are calculated as follows:

[0111]

[0112] Among them, L j (t) represents the first parameter of the three-phase unbalance index, and the calculation formula is as follows:

[0113]

[0114] 4. Calculate the carbon emission index in the distribution area based on the total load prediction data and energy output prediction data of users:

[0115]

[0116] Among them, CI(t) represents the carbon emission index of the distribution area at time t, is the power emission factor of the distribution area transformer at time t; F i (t) is the total load prediction data of the i-th user at time t;

[0117]

[0118] Among them, K m (t) represents the proportion of the predicted output data of the m-th type of energy at time t in the total predicted output data of the distribution area, m = 1, 2, 3,..., Q, where Q is the total number of energy types; E mIt represents the carbon emission factor of the m-th type of energy. The power emission factors of different energy types are as follows: for general power, it is 0.4872; for coal power generation, it is 1.205; for natural gas, it is 0.4515; for large hydropower, it is 0.0035; for photovoltaic power generation, it is 0.0704; for wind power generation, it is 0.024.

[0119] Among them, taking the case where there is only photovoltaic output and general power in the distribution area as an example to illustrate the method for obtaining the carbon emission index. Usually, the photovoltaic output is not sufficient to supply all users, and general power is needed to make up for it. First, obtain the predicted data of photovoltaic output. From the predicted data of the total load of users in the distribution area, the predicted data of the total output in the distribution area can be obtained. Combining the predicted data of photovoltaic output, the predicted data of the output of general power can be obtained, and thus the power emission factor of the distribution area transformer can be obtained.

[0120] Use the existing photovoltaic output prediction model to obtain the predicted data of photovoltaic output, which will not be elaborated here.

[0121] 5. Construct a bi-objective optimization model based on the three-phase unbalance index of users and the carbon emission index in the distribution area. Solve the optimization problem through a heuristic optimization algorithm combined with the distribution network power flow simulation software to obtain the best data of the flexible load of users from the bi-objective optimization model:

[0122] The objective functions of the bi-objective optimization model are as follows:

[0123] Minimize F loss (t)

[0124] Minimize CI(t)

[0125] Finally, obtain the best data of the air-conditioning load (flexible load) of the i-th user at time t after optimization by solving the objective functions of the bi-objective optimization model

[0126] The distribution network power flow simulation software can be PSCAD, DIGSILENT, or PSSE.

[0127] The embodiment of the present invention also provides a distribution network carbon emission governance system involving flexible loads, including a flexible load prediction module, a total load prediction module, a three-phase unbalance loss calculation module, a carbon emission index calculation module, and an optimization module;

[0128] The flexible load prediction module is used to construct a flexible load prediction model for users and obtain the predicted data of the flexible load of users;

[0129] The total load prediction module is used to construct a total load prediction model for users and input the predicted data of the flexible load of users into the total load prediction model of users to obtain the predicted data of the total load of users;

[0130] The three-phase unbalance loss calculation module is used to calculate the three-phase unbalance loss in the distribution area based on the user's total load prediction data;

[0131] The carbon emission index calculation module is used to calculate the carbon emission index in the distribution area based on the user's total load prediction data and energy output prediction data;

[0132] The optimization module is used to construct a bi-objective optimization model with the minimum three-phase unbalance loss in the distribution area and the minimum carbon emission index in the distribution area as the objectives, and obtain the optimal flexible load data of the user by solving the bi-objective optimization model.

[0133] The flexible load prediction module is used to calculate the flexible load prediction data according to the following formula:

[0134]

[0135] where, is the flexible load prediction data of the i-th user at time t; i = 1, 2, 3,..., M, where M is the total number of users in the distribution area; T(t) represents the predicted outdoor temperature value at time t, represents the preset temperature value of the air conditioner of the i-th user at time t; θ i represents the air conditioner energy efficiency coefficient of the i-th user, and δ i represents the electric heating resistance of the air conditioner of the i-th user;

[0136] satisfies the following constraint conditions:

[0137]

[0138] where, is the indoor temperature of the i-th user at time t + Δt; is the indoor temperature of the i-th user at time t; is the air conditioner operation indicator of the i-th user at time t; β i is the air conditioner cooling parameter of the i-th user; is the maximum air conditioner load of the i-th user;

[0139] where, β i The calculation formula is as follows:

[0140]

[0141] where, C i represents the air conditioner capacitor of the i-th user;

[0142] where, satisfies:

[0143]

[0144] Among them, ε i (t) is the bandwidth, representing the difference between the indoor temperature of the i-th user at time t and the air conditioner set temperature value. The difference value.

[0145] The total load prediction module is used to

[0146] Extract the daily total load historical data of users at intervals of every 15 minutes, and perform denoising and normalization processing on the daily total load historical data;

[0147] Adopt the K-means algorithm to perform clustering analysis on the daily total load historical data of users, and divide users into four categories of users by using the SSE clustering effect evaluation index;

[0148] Based on the non-intrusive load decomposition method, decompose the daily total load historical data of users into flexible load historical data and basic load historical data;

[0149] Taking historical electricity prices, flexible load historical data, basic load historical data, and historical meteorological data as feature quantities, and taking the daily total load historical data of users as the result, train the deep learning LSTM network model for four categories of users respectively;

[0150] Input the predicted electricity price, flexible load prediction data, basic load prediction data, and meteorological prediction data into the deep learning LSTM network model to obtain the total load prediction data of users.

[0151] The three-phase unbalance loss calculation module is used to calculate the three-phase unbalance loss according to the following formula:

[0152]

[0153] Among them, F loss (t) represents the total three-phase unbalance loss of all transformers in the distribution area at time t, and ρ j (t) represents the three-phase negative sequence unbalance degree of the j-th transformer at time t; R a is the first line resistance; R b is the second line resistance; R c is the third line resistance; j = 1, 2, 3,..., N, and N is the total number of transformers in the distribution area;

[0154] ρ j (t) The calculation formula is as follows:

[0155]

[0156] Among them, L j (t) represents the first parameter of the three-phase unbalance index, and the calculation formula is as follows:

[0157]

[0158] Among them, a j (t), b j (t), c j (t) are respectively the effective values of the fundamental wave components of the three-phase currents of the j-th transformer at the t-th moment. a j (t), b j (t), c j (t) are respectively equal to the normalized values of the sums of the total load prediction data of the users connected to the first line, the second line, and the third line of the j-th transformer at the t-th moment.

[0159] The carbon emission index calculation module is used to calculate the carbon emission index according to the following formula:

[0160]

[0161] Among them, CI(t) represents the carbon emission index of the distribution area at the t-th moment. is the power emission factor of the transformer in the distribution area at the t-th moment; F i (t) is the total load prediction data of the i-th user at the t-th moment; i = 1, 2, 3,..., M, and M is the total number of users in the distribution area.

[0162] The calculation formula of is as follows:

[0163]

[0164] Among them, K m (t) represents the proportion of the output prediction data of the m-th type of energy at the t-th moment in the total output prediction data of the distribution area, m = 1, 2, 3,..., Q, and Q is the total number of energy types; E m represents the carbon emission factor of the m-th type of energy.

[0165] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0166] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a means for implementing the functions specified in one or more flows and / or blocks. Figure 1 in one or more flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more blocks.

[0167] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction means that implements the functions specified in one or more flows and / or blocks. Figure 1 in one or more flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more blocks.

[0168] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more flows and / or blocks. Figure 1 in one or more flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more blocks.

[0169] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Those of ordinary skill in the art, under the inspiration of the present invention and without departing from the spirit and scope of the present invention as defined by the claims, can also make many forms, and all of these fall within the protection scope of the present invention.

[0170] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art of this industry should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and all of these changes and improvements fall within the scope of the present invention claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for carbon emission governance in a distribution network involving flexible loads, characterized in that, including, building a flexible load prediction model for users, Among them, is the flexible load prediction data of the i-th user at time t; i = 1, 2, 3, …, M, where M is the total number of users in the distribution area; T(t) represents the predicted outdoor temperature value at time t, represents the preset temperature value of the air conditioner of the i-th user at time t; θ i represents the air conditioner energy efficiency coefficient of the i-th user, and δ i represents the electric heating resistance of the air conditioner of the i-th user; to obtain the flexible load prediction data of users; building a total load prediction model for users, inputting the flexible load prediction data of users into the total load prediction model of users to obtain the total load prediction data of users, and inputting the predicted electricity price, flexible load prediction data, basic load prediction data, and meteorological prediction data into the deep learning LSTM network model to obtain the total load prediction data of users; calculating the three-phase unbalance loss in the distribution area based on the total load prediction data of users, Among them, F loss (t) represents the total three-phase unbalance loss of all transformers in the distribution area at time t, and ρ j (t) represents the three-phase negative-sequence unbalance degree of the j-th transformer at time t; R a is the first line resistance; R b is the second line resistance; R c is the third line resistance; j = 1, 2, 3, …, N, where N is the total number of transformers in the distribution area; ρ j (t) is calculated as follows: Among them, L j (t) represents the first parameter of the three-phase unbalance index, and the calculation formula is as follows: Among them, a j (t), b j (t), c j (t) are respectively the effective values of the fundamental wave components of the three-phase currents of the j-th transformer at the t-th moment. a j (t), b j (t), c j (t) are respectively equal to the normalized values of the sums of the total load prediction data of the users connected to the first line, the second line, and the third line of the j-th transformer at the t-th moment; calculating the carbon emission index in the distribution area based on the total load prediction data of users and the energy output prediction data, Among them, CI(t) represents the carbon emission index of the distribution area at time t, is the power emission factor of the transformer in the distribution area at time t; F i (t) is the total load prediction data of the i-th user at time t; i = 1, 2, 3, …, M, where M is the total number of users in the distribution area; building a bi-objective optimization model with the minimum three-phase unbalance loss in the distribution area and the minimum carbon emission index in the distribution area as the objectives, and obtaining the optimal flexible load data of users by solving the bi-objective optimization model.

2. The method for carbon emission governance of a distribution network involving flexible loads according to claim 1, characterized in that The satisfies the following constraint conditions: Among them, is the indoor temperature of the i-th user at time t + Δt; is the indoor temperature of the i-th user at time t; is the air conditioner operation indicator of the i-th user at time t; β i is the air conditioner cooling parameter of the i-th user; is the maximum air conditioner load of the i-th user; Among them, β i The calculation formula is as follows: Among them, C i represents the air conditioner capacitor of the i-th user; Among them, Satisfy: Among them, ε i (t) is the bandwidth, representing the difference between the indoor temperature of the i-th user at time t and the air conditioner set temperature value .

3. A method for carbon emission governance in a distribution network with flexible loads involved, as claimed in claim 1, wherein The building of the total load prediction model for users includes, extracting the daily total load historical data of users at 15-minute intervals, and denoising and normalizing the daily total load historical data; using the K-means algorithm to perform clustering analysis on the daily total load historical data of users, and classifying users into four categories by using the SSE clustering effect evaluation index; based on the non-intrusive load decomposition method, decomposing the daily total load historical data of users into flexible load historical data and basic load historical data; using the historical electricity price, flexible load historical data, basic load historical data, and historical meteorological data as feature quantities, and the daily total load historical data of users as the result, training the deep learning LSTM network model for the four categories of users respectively.

4. A method for carbon emission governance of a distribution network involving flexible loads according to claim 1, characterized in that, The said has the following calculation formula: Among them, K m (t) represents the proportion of the predicted output data of the m-th type of energy at time t in the total predicted output data of the distribution area, where m = 1, 2, 3, …, Q, and Q is the total number of energy types; E m represents the carbon emission factor of the m-th type of energy.

5. A distribution network carbon emission governance system involving flexible loads, characterized in that, including a flexible load prediction module, a total load prediction module, a three-phase unbalance loss calculation module, a carbon emission index calculation module, and an optimization module; The flexible load prediction module is used to build a flexible load prediction model for users, Among them, is the flexible load prediction data of the i-th user at time t; i = 1, 2, 3, …, M, where M is the total number of users in the distribution area; T(t) represents the predicted outdoor temperature value at time t, represents the preset temperature value of the air conditioner of the i-th user at time t; θ i represents the air conditioner energy efficiency coefficient of the i-th user, and δ i represents the electric heating resistance of the air conditioner of the i-th user; to obtain the flexible load prediction data of users; The total load prediction module is used to build a total load prediction model for users, inputting the flexible load prediction data of users into the total load prediction model of users to obtain the total load prediction data of users, and inputting the predicted electricity price, flexible load prediction data, basic load prediction data, and meteorological prediction data into the deep learning LSTM network model to obtain the total load prediction data of users; The three-phase unbalance loss calculation module is used to calculate the three-phase unbalance loss in the distribution area based on the total load prediction data of users, Among them, F loss (t) represents the total three-phase unbalance loss of all transformers in the distribution area at time t, and ρ j (t) represents the three-phase negative-sequence unbalance degree of the j-th transformer at time t; R a is the first line resistance; R b is the second line resistance; R c is the third line resistance; j = 1, 2, 3, …, N, where N is the total number of transformers in the distribution area; ρ j (t) is calculated as follows: Among them, L j (t) represents the first parameter of the three-phase unbalance index, and the calculation formula is as follows: Among them, a j (t), b j (t), c j (t) are respectively the effective values of the fundamental wave components of the three-phase currents of the j-th transformer at the t-th moment. a j (t), b j (t), c j (t) are respectively equal to the normalized values of the sum of the total load prediction data of the users connected to the first line, the sum of the total load prediction data of the users connected to the second line, and the sum of the total load prediction data of the users connected to the third line of the j-th transformer at the t-th moment; The carbon emission index calculation module is used to calculate the carbon emission index in the distribution area based on the total load prediction data of users and the energy output prediction data, Among them, CI(t) represents the carbon emission index of the distribution area at time t, is the power emission factor of the transformer in the distribution area at time t; F i (t) is the total load prediction data of the i-th user at time t; i = 1, 2, 3, …, M, where M is the total number of users in the distribution area; The optimization module is used to build a bi-objective optimization model with the minimum three-phase unbalance loss in the distribution area and the minimum carbon emission index in the distribution area as the objectives, and obtaining the optimal flexible load data of users by solving the bi-objective optimization model.

6. The carbon emission governance system of a distribution network involving flexible loads according to claim 5, characterized in that The said satisfies the following constraint conditions: Among them, is the indoor temperature of the i-th user at time t + Δt; is the indoor temperature of the i-th user at time t; is the air conditioner operation indicator of the i-th user at time t; β i is the air conditioner cooling parameter of the i-th user; is the maximum air conditioner load of the i-th user; Among them, β i The calculation formula is as follows: Among them, C i represents the air conditioner capacitor of the i-th user; Among them, Satisfy: Among them, ε i (t) is the bandwidth, representing the difference between the indoor temperature of the i-th user at time t and the air-conditioning set temperature value .

7. A distribution network carbon emission governance system involving flexible loads according to claim 5, characterized in that, The total load prediction module is used to, extract the daily total load historical data of users at 15-minute intervals, and denoise and normalize the daily total load historical data; using the K-means algorithm to perform clustering analysis on the daily total load historical data of users, and classifying users into four categories by using the SSE clustering effect evaluation index; Based on the non-intrusive load decomposition method, the historical data of the user's total daily load is decomposed into the historical data of flexible load and the historical data of basic load.

8. A flexible load-involved distribution network carbon emission governance system according to claim 5, characterized in that The said has the following calculation formula: Among them, K m (t) represents the proportion of the predicted output data of the m-th type of energy at time t in the total predicted output data of the distribution area, where m = 1, 2, 3, …, Q, and Q is the total number of energy types; E m represents the carbon emission factor of the m-th type of energy.

Images