Electricity-carbon collaborative optimization method and system based on carbon emission flow theory

Abstract

The invention discloses an electricity-carbon collaborative optimization method and system based on a carbon emission flow theory, and belongs to the technical field of optimization of power systems and energy management. The electricity-carbon collaborative optimization method based on the carbon emission flow theory comprises the following steps: constructing a regional power distribution network architecture comprising load nodes and power generation nodes, and establishing a power model of the load nodes; establishing a carbon emission flow model according to the power flow model; constructing an electricity-carbon collaborative optimization model based on the power model and the carbon emission flow model of the load node; predicting actual photovoltaic and wind power generation and load data in the power distribution network by using an LSTM network, collecting charging demand information of a charging station, and performing parameter assignment in the power transmission-carbon collaborative optimization model; and solving the electricity-carbon collaborative optimization model by using a solver, determining the optimal power flow of each load per hour, and calculating the carbon emission intensity and the carbon emission rate to obtain a final scheduling scheme.

Description

Technical Field

[0001] The present invention belongs to the technical field of optimization of power systems and energy management, and specifically relates to an electricity-carbon collaborative optimization method and system based on carbon emission flow theory. Background Art

[0002] The distribution network is a bridge connecting the transmission network and end users. The carbon emission intensity of the main grid where it is located will change due to different operating modes. Even if the power structure of the main grid is known, when the operating state of the main grid changes, the carbon emissions per unit of electricity consumption (i.e., carbon potential) of the access node of the distribution system may also change accordingly. With the continuous optimization of the energy structure, new energy information subjects represented by new energy vehicles, data centers, electrolysis hydrogen production, cluster photovoltaic wind power generation, cloud energy storage, etc. have emerged. These generators with different energy efficiency, emission reduction potential and cost, as well as new subjects with different load properties, have greatly reshaped the power flow distribution of the power system and the carbon emission distribution between branches. For example, wind power generation and photovoltaic generators can effectively reduce system carbon emissions and optimize carbon flow distribution. Since the carbon emission flow is regarded as a virtual network flow that depends on the existence of active power flow in the power grid, the concept of carbon emission flow can be introduced to capture the changes in the carbon potential of system nodes caused by changes in the main grid operation mode or active power flow, thereby optimizing the operation strategy and state of the distribution network and realizing the operation of the entire system in a low-carbon and economic direction.

[0003] However, the coordinated electricity and carbon flow management within the solar-hydrogen-storage-charging multi-energy system has not been fully studied. Therefore, an electricity-carbon collaborative optimization method based on carbon emission flow theory is proposed. Summary of the invention

[0004] In view of the deficiencies in the prior art, the purpose of the present invention is to provide an electric-carbon collaborative optimization method and system based on carbon emission flow theory, which solves the problems in the prior art.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] The electricity-carbon synergistic optimization method based on carbon emission flow theory includes the following steps:

[0007] Construct a regional distribution network architecture including load nodes and power generation nodes, and establish a power model for load nodes;

[0008] According to the power flow model, a carbon emission flow model is established;

[0009] Based on the power model of the load node and the carbon emission flow model, an electricity-carbon collaborative optimization model is constructed;

[0010] The LSTM network is used to predict the actual photovoltaic and wind power generation and load data in the distribution network, collect the charging demand information of the charging station, and assign parameters in the power-carbon collaborative optimization model;

[0011] The solver is used to solve the electricity-carbon collaborative optimization model, determine the optimal power flow of each load every hour, calculate the carbon emission intensity and carbon emission rate, and obtain the final scheduling plan.

[0012] Furthermore, the load nodes include: electrolysis hydrogen production equipment, charging piles, commercial loads, and civil loads; the power generation nodes include: cluster photovoltaic and wind power generation, coal-fired units, and natural gas power generation units.

[0013] Furthermore, the carbon emission flow model is:

[0014]

[0015] In the formula, e Ni is the carbon potential of node i, I + is the set of branches with flow flowing into node i, ρ s is the carbon flow density of branch s, P Bs is the active power of branch s, P Nii is the node active flux matrix; R all is the sum of carbon flow rates of power system nodes; N is the total number of nodes in the distribution network; R i is the carbon flow rate of node i; Input active power for the generator set at node i; is the carbon emission intensity of the generator set at node i.

[0016] Furthermore, the objective function of the electricity-carbon collaborative optimization model is:

[0017]

[0018] In the formula, p t is the load power, is the residential load power at time t, is the power generated by renewable energy, is the power of hydrogen production by electrolysis, Power for electric vehicle charging stations The power of centralized energy storage equipment; For electricity price, Output of each generator set, c k is the power generation cost of unit k, is the carbon emission intensity of the kth generator set, c c is the price per ton of CO2 emitted; is the output matrix vector of the generator set.

[0019] Furthermore, the constraints of the electricity-carbon collaborative optimization model include:

[0020]

[0021] In the formula, They are the maximum power constraints for residential loads, renewable energy generation, water electrolysis hydrogen production, electric vehicle charging piles, and energy storage devices; They are the minimum power constraints for electric vehicle charging piles, residential loads, and water electrolysis to produce hydrogen; and are the minimum and maximum output power constraints of the generator sets respectively; and is the climbing constraint of the generator set.

[0022] Furthermore, the solver adopts Gurobi Optimizer.

[0023] The electricity-carbon coordinated optimization system based on carbon emission flow theory includes:

[0024] Load node power model building module: builds a regional distribution network architecture including load nodes and power generation nodes, and establishes a power model for load nodes;

[0025] Carbon emission flow model building module: build a carbon emission flow model based on the power flow model;

[0026] Collaborative optimization model building module: builds an electricity-carbon collaborative optimization model based on the power model and carbon emission flow model of the load node;

[0027] Parameter prediction module: Use the LSTM network to predict the actual photovoltaic, wind power generation and load data in the distribution network, collect the charging demand information of the charging station, and assign parameters in the power-carbon collaborative optimization model;

[0028] And, the model solving module: use the solver to solve the electricity-carbon collaborative optimization model, determine the optimal power flow of each load every hour, and calculate the carbon emission intensity and carbon emission rate to obtain the final scheduling plan.

[0029] A computer storage medium stores a readable program, which can execute the above-mentioned electricity-carbon collaborative optimization method based on carbon emission flow theory when the program is run.

[0030] An electronic device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0031] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the above-mentioned electricity-carbon collaborative optimization method based on carbon emission flow theory.

[0032] A computer program product includes computer instructions, wherein the computer instructions instruct a computing device to execute operations corresponding to the above-mentioned electricity-carbon collaborative optimization method based on carbon emission flow theory.

[0033] Beneficial effects of the present invention:

[0034] 1. From the perspective of electricity-carbon coupling, the present invention realizes the optimal power distribution between various loads and generator sets by analyzing the active power flow distribution in the source-grid-load-storage and distribution networks.

[0035] 2. This study analyzes the spatiotemporal distribution patterns of carbon emission transfers within the system, quantifies the impact of emission- and non-emission generators, and the impact of electricity and carbon flows on specific loads (such as charging stations or hydrogen electrolyzers). BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0037] Figure 1 is a schematic diagram of the electric-carbon coupled distribution network of the present invention;

[0038] Figure 2 is a schematic diagram of an IEEE-14bus distribution network system calculation example in Embodiment 2 of the present invention;

[0039] Figure 3 This is a statistical diagram of the 24-hour power and carbon flow rate of the charging pile in Example 2 of the present invention;

[0040] Figure 4 It is a statistical diagram of power output and carbon flow rate of units (G1-G5) in Example 2 of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] Example 1

[0043] The electricity-carbon synergistic optimization method based on carbon emission flow theory includes the following steps:

[0044] S1, construct a regional distribution network architecture including load nodes and power generation nodes, and establish a power model of load nodes;

[0045] like Figure 1 The regional distribution network shown in Figure 1 is considered as an active distribution network (ADN), which includes load nodes such as electrolysis hydrogen production equipment, charging piles, commercial / civilian, and power generation nodes such as cluster photovoltaic and wind power generation, coal-fired units, and natural gas generators. It is assumed that each ADN has a control center to control the energy flow between different loads and collect power information of loads and generators.

[0046] Construct the power model of the distribution network load node. The specific process is as follows:

[0047] The distribution network has N nodes, among which K nodes have generator injection, and the residential load power at time t is defined as Power generation from renewable energy sources (light, wind, etc.) Hydrogen electrolysis power Electric vehicle charging station power And the power of centralized energy storage equipment At time t, for the electricity price Consider different loads The maximum power constraint is:

[0048]

[0049] In the formula, They are the maximum power constraints for residential loads, renewable energy generation, water electrolysis hydrogen production, electric vehicle charging piles, and energy storage devices.

[0050] Based on the neural network machine learning model to predict the hourly electricity consumption and power generation in the system, it can effectively solve the prediction of long time series. The predicted power of each load is:

[0051]

[0052] In the formula, They are the power forecast values ​​of residential load, renewable energy generation, and electric vehicle charging piles, respectively.

[0053] The optimal power is input with the predicted value of the load, where the charging station can use the V2G function of the electric vehicle to achieve peak load shifting according to the demand response signal. A positive value of indicates energy storage, while a negative value indicates discharging electricity to the grid. Electricity consumption and electricity prices follow a logarithmic function, and all loads in the distribution network are:

[0054]

[0055] Output of generator set G Including conventional power plants and renewable energy Right now:

[0056]

[0057] For the sake of uniformity, the output of each generator set is expressed as The power generation cost of unit k is c k , k∈K, without distinguishing the types of generator sets.

[0058] S2, establishing a carbon emission flow (CEF) model based on the power flow model;

[0059] The core of power system flow calculation is to determine the operating state of the entire system according to specific operating conditions and network structure. In the power grid, the flow is mainly constrained by the grid structure, system parameters and boundary conditions. Corresponding to the flow calculation, the carbon emission flow calculation of the power system aims to quantitatively analyze the dynamic state of the carbon emission flow in the power system according to the flow distribution, so as to reveal the source and destination of carbon emissions in the power system. Specifically, the power system flow calculation determines the system state through operating conditions and network structure, and is constrained by the grid structure and parameters. The carbon emission flow calculation analyzes the flow state of carbon emissions based on the flow distribution. The carbon emission flow is closely related to the flow, but it is also affected by the carbon emission characteristics of the generator set and has unique flow properties. Both are affected by factors such as network topology and line impedance, but the carbon emission flow is also affected by the carbon emission intensity of the unit.

[0060] First, we introduce the basic concepts of power flow calculation and power balance equation. For each node i, the power balance equation can be expressed as:

[0061] P i =P Gi +P Li ,Q i =Q Gi +Q Li , (1)

[0062] Among them, P i and Q i is the active and reactive power of node i; P Gi and Q Gi is the power generation of node i; P Li and Q Li is the load power of node j.

[0063] The node voltage can be expressed by the relationship between the node current and impedance:

[0064]

[0065] Among them, V i is the voltage at node i, I ij is the current from node i to node j, Z ij is the impedance between node i and node j.

[0066] In power flow calculations, active and reactive power can be calculated using the relationship between node voltage and current:

[0067] P=V·I * =V·(G+jB)·V * , (3)

[0068] Q=V·I * =V·(G-jG)·V * , (4)

[0069] Where G and B are the real and imaginary parts of the node's admittance matrix, respectively.

[0070] The power flow calculation is usually solved by iterative methods such as the Newton-Raphson method or the Gauss-Seidel method until the set convergence conditions are met. Since the energy consumption and carbon emissions in the power system are mainly related to the active output of the power source and are slightly affected by the reactive output of the generator set, the carbon emission flow (CEF) is mainly affected by the active power flow in the system. The following is the basic calculation method for carbon emission flow.

[0071] Assume that the distribution network has N nodes, K of which have generator injection and M of which are loads. According to the above power flow calculation formula, the N-order branch power flow distribution matrix can be obtained, which can be expressed as P B =(P Bij ) N×N It means that this matrix can describe the active power flow distribution of the power system and give the boundary conditions of the carbon emission flow distribution from the power network level. Because this matrix contains both the topological structure information of the power network and the distribution information of the system dynamic active power flow. If there is a branch connecting node i and node j (i, j = 1, 2, .., N), and the positive active power flow from node i to node j through this branch is p, then P Bij =p,P Bji =p; if the active power flow p flowing through the branch is a reverse flow, then vice versa. The unit injection distribution matrix is ​​a K×N matrix, using P G =(P Gkj ) K×N Indicates. GIt can describe the relationship between all generators and power nodes and the active power injected into the system, and also describe the boundary conditions of the carbon emission flow generated by the system's heavy generators. The load distribution matrix is ​​a K×N matrix, represented by P L =(P Lmj ) K×N According to the above matrix, we can get the node active flux matrix, which is an N-order diagonal matrix, P N =(P Nij ) N×N For node i, P Nii =P Bs +P Gi , where s is the branch with flow flowing into the node. B and P G It is known that P can be directly generated N matrix.

[0072] When the carbon emission intensity of the kth generator set is Then the carbon emission intensity vector of the generator set is When the carbon potential of the i-th node is e Ni When the node carbon potential vector After obtaining the node carbon potential vector, the carbon flow rate of each branch can be further obtained, which is defined as an N-order matrix, R B =(R Bij ) N×N , which is similar to the calculation principle of active power flow. In particular, all diagonal elements are 0. According to the matrix defined previously, R B =P B diag(E N ).

[0073] After calculating the node carbon potential, the carbon emission intensity of the node load is equal to the carbon potential of the node. Combined with the load distribution matrix, the carbon flow rate corresponding to all loads can be obtained. The physical meaning is the carbon emissions generated per unit time by the power supply node load on the power generation side. The load carbon flow rate vector can be expressed as: From the above analysis, we can get that R L =P L E N .

[0074] According to the definition of node carbon potential, the carbon potential e of node i in the system can be obtained as Ni is the carbon emission on the power generation side corresponding to the unit electricity consumed by node i, which is used to characterize the carbon emission intensity of the node. Ni The expression is as follows:

[0075]

[0076] Where: I + is the set of branches with flow flowing into node i, ρ s is the carbon flow density of branch s, which is defined as the ratio of the carbon flow rate of any branch of the power system to the active power flow: P Bs is the active power of branch s, P Nii is the node active flux matrix; at the same time, the carbon potential calculation formula of all nodes in the whole system is: E N =(P N -P B T )-1 P G T E G .

[0077] The total carbon flow rate of the system is the sum of the carbon flow rates of all nodes in the power system, and the expression is as follows:

[0078]

[0079] Where: R all is the sum of carbon flow rates of power system nodes; N is the total number of nodes in the distribution network; R i is the carbon flow rate of node i; Input active power for the generator set at node i; is the carbon emission intensity of the generator set at node i.

[0080] The establishment of a theoretical system of carbon emission flows can make it possible to infer the carbon emissions on the load side of the system from the power emitted by the power generation side and the power grid flow, so that the carbon emission responsibility originally calculated on the power generation side can be transferred to the load side for calculation, thereby achieving more effective carbon emission responsibility sharing on the load side.

[0081] S3, based on the power model of the load nodes in S1 and the carbon emission flow model in S2, to build an electricity-carbon collaborative optimization model;

[0082] Under the premise of ensuring the safe operation of ADN, based on historical data of energy load and power generation data While ensuring the safety constraints of the current model, the output of each generator set and the power p of each load are optimized with the goal of minimizing the overall operating cost of the system. t .

[0083] The electricity-carbon synergistic optimization model is constructed, and its objective function is:

[0084]

[0085] The total operating cost is the sum of the total electricity cost and the total carbon emission cost in the T period; where p t is the load power, is the residential load power at time t, is the power generated by renewable energy, is the power of hydrogen production by electrolysis, Power for electric vehicle charging stations The power of centralized energy storage equipment; For electricity price, Output of each generator set, c k is the power generation cost of unit k, is the carbon emission intensity of the kth generator set, c c is the price per ton of CO2 emitted; is the power vector of the generator set at time t.

[0086] The constraints of the electricity-carbon collaborative optimization model include:

[0087]

[0088] Among them, constraint (7) ensures power balance at any time;

[0089]

[0090] Constraints (8)-(10) define the load The minimum and maximum ranges of

[0091]

[0092] Constraint (11) limits the energy storage system The charge and discharge power,

[0093]

[0094] Constraint (12) limits the output power of the generator set.

[0095]

[0096] Constraint (13) is the ramp constraint of the generator set. For simplicity, the model does not consider the excess cost due to voltage and line flow.

[0097] S4, using LSTM network to predict actual photovoltaic, wind power generation and load data in the distribution network, collect charging demand information of charging stations, and send it to the electricity-carbon collaborative optimization model constructed in S3 for parameter assignment;

[0098] It predicts actual photovoltaic and wind power generation and base load data in the distribution network, and then collects charging demand information from charging stations. After that, it determines the optimal power flow for each load every hour and calculates the carbon flow, including carbon emission intensity and carbon emission rate. The goal is to optimize the management of power flow and power generation units in this power-carbon system to minimize power and carbon costs.

[0099] First, a recurrent neural network is used to predict the hourly power generation of photovoltaic (PV) and wind power as well as base load data, which are trained based on 24 months of real-time series data from 2020 to 2021 provided by the State Grid Corporation of China. These data are recorded every 0.5 hours and include power load and environmental information, making them suitable for time series analysis, especially in the scenario of predicting power load based on meteorological conditions. The model is based on a long short-term memory (LSTM) network. In order to reduce the impact of the data value range on the final accuracy of the network, the input data needs to be normalized to the interval of [0.5,1]. Min-Max normalization is used in this embodiment. In terms of neural network parameter configuration, the Adam optimizer is used, the batch size is set to 10, and the learning rate is 0.0001. The network is trained for a total of 2000 cycles. Specifically, it adopts a single-layer architecture with 10 LSTM units, using the ReLU activation function. In addition, 85% of the dataset is used to train the model, and the remaining 15% is used for testing. The implementation of LSTM uses PyTorch.

[0100] S5, uses the solver to solve the electricity-carbon collaborative optimization model, determines the optimal power flow of each load every hour, and calculates the carbon emission intensity and carbon emission rate to obtain the final scheduling plan.

[0101] It first predicts actual PV and wind power generation and baseload data. Subsequently, it collects charging demand information from charging stations. Next, it determines the optimal power flow for various loads every hour and calculates carbon flow, including CEF intensity and carbon emission rate. The goal is to effectively manage power flows and unit generation in this power-carbon system to minimize power and carbon costs over multiple time steps.

[0102] To ensure the optimality of the solution, the optimization model was coded in Python and solved using Gurobi Optimizer. All experiments were conducted on a desktop computer equipped with an Intel(R) Core(TM) i5-8250U CPU (1.60 GHz) and 16 GB of memory.

[0103] Based on similar inventive concepts, an embodiment of the present invention also provides a computer storage medium storing a readable program, which can execute the above-mentioned electricity-carbon collaborative optimization method based on carbon emission flow theory when the program is running.

[0104] Based on similar inventive concepts, an embodiment of the present invention provides an electronic device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0105] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the above-mentioned electricity-carbon collaborative optimization method based on carbon emission flow theory.

[0106] Based on similar inventive concepts, an embodiment of the present invention also provides a computer program product, including computer instructions, which instruct a computing device to perform operations corresponding to the above-mentioned electricity-carbon collaborative optimization method based on carbon emission flow theory.

[0107] Example 2

[0108] In this embodiment, the electricity-carbon collaborative optimization method based on the carbon emission flow theory mentioned in Example 1 is verified by simulation;

[0109] like Figure 2 As shown in the figure, the system is verified in the modified IEEE-14bus distribution network system. This example does not consider network loss, and the system power flow distribution is calculated using the DC power flow method. The output of each unit and the load of each user (in MW) in the system are calculated by the collaborative optimization framework, and the steady-state active power flow distribution of the system is calculated by the simulation software.

[0110] Assume that node 6 is a water electrolysis hydrogen production device, and nodes 10 and 12 are distributed electric vehicle charging stations (DC fast charging stations). Given the carbon emission intensity vectors (in gCO2 / kW·h) of all generator sets G, assume that G1 is a coal-fired unit with the highest carbon emission intensity; G2 and G4 are distributed photovoltaic and wind turbines with a carbon emission intensity of 0; G3 and G5 are coal-fired units and gas units with carbon complements, with medium carbon emission intensity. G As shown below:

[0111] E G =[900,0,600,0,500].

[0112] According to the calculation formula in Example 1, the active flux and carbon potential of all nodes at a certain time can be calculated (the following table takes the data at 12:00 pm), as shown in Table 1.

[0113] Table 1 Node active flux and carbon potential

[0114]

[0115] According to Table 1, for nodes 1, 8, 11, and 12, the flows flowing into these nodes only come from a certain generator set or a node connected to it. Therefore, the carbon potential of these nodes is consistent with the carbon potential of the generator set or node that provides active power flow upstream. At this moment, the injected power of units G1-G5 at nodes 1, 2, 3, 6, and 8 is 55MW, 18MW, 40MW, 10MW, and 30MW, respectively.

[0116] Furthermore, the carbon flow rate distribution matrix R of the branch can be calculated B , the carbon flow rates of all branches are obtained. Table 2 lists the branch carbon flow rates at a certain time (data taken at 12:00 pm) as a reference.

[0117] Table 2 Branch active power flow and carbon flow rate

[0118]

[0119] Among them, the line with the largest flow is node 7 to 9 (37.14MW), and the line with the densest branch carbon flow density is from node 1 to 2 and from 1 to 5, which is 900.0gCO2 / kWh, showing the intensity of carbon injection of coal-fired units and the impact on the entire carbon emission flow. The highest carbon flow rate appears in nodes 7 to 9 (19.87tCO2 / h) and nodes 1 to 2 (27.41t CO2 / h). Due to the existence of fragmented carbon flow and high carbon flow density, optimizing load management or introducing renewable energy on high carbon emission branches (such as 1 to 2, 1 to 5, 7 to 9) can effectively reduce carbon emissions and reduce the carbon flow density and carbon flow rate of these branches.

[0120] Further, the carbon flow rate vector and distribution matrix R of the outflow node are calculated L , so that the carbon flow rate injected by all loads and units at a certain moment (the table below takes the data at 12:00 pm) can be obtained, as shown in Table 3.

[0121] Table 3 Load and unit carbon flow rate

[0122]

[0123] As can be seen from Table 3, ignoring the network loss, the carbon flow rate injected into the system is equal to the sum of the carbon flow rate outflowing from the system. The above table shows the carbon flow data after the electricity-carbon coordinated optimization at a certain moment. The change process of carbon potential at different nodes in the distribution network system at different time periods can also be obtained. Node 3 not only has the highest load, but also has the highest load carbon flow rate (30.81tCO2 / h), which is consistent with its high load. Nodes 1, 3 and 5 have unit injection carbon flow rates of 49.5, 24.0 and 15.0t CO2 / h, respectively. For example, although the load of node 1 is 0, the unit injection carbon flow rate is 49.5tCO2 / h. Overall, the total load is 155.72MW, the total load carbon flow rate is 89.65t CO2 / h, and the total unit injection carbon flow rate is 88.5t CO2 / h. These data show that most of the carbon emissions in the system come from the load-related carbon flow rate, and the carbon flow rate injected by the unit also accounts for a considerable proportion.

[0124] In Table 3, node 7 is the active power of hydrogen produced by water electrolysis, which is 6.51MW. Nodes 10 and 12 are two electric vehicle charging stations, which are 6MW and 2MW. Figure 3 The power and carbon flow rate of typical node electric vehicle charging stations (10 and 12) within 24 hours of a day are shown in the figure: the power change is relatively stable from late night to early morning, and the output power increases slowly until the morning traffic rush hour. It reaches a peak of 6.44MW around 12 noon, indicating that the charging demand is the largest during this period. For charging station 1, the corresponding carbon flow rate reaches a maximum value of 3.31tCO2 / h, which is particularly significant at noon and early afternoon, indicating that high-power charging during these periods may cause greater carbon emissions. When operating at low load, the carbon flow rate decreases accordingly. Different electric vehicle charging stations have large differences in power output and carbon emissions due to their different locations and the habits of the users they serve. By adjusting the injected power of the non-uniform generator set, the carbon emissions of the charging station can be effectively reduced through the zero-emission unit, and the contribution to reducing the carbon flow rate can be achieved through the V2G function.

[0125] When studying the output changes of the generating units and their related carbon emission rates, we analyzed the 24-hour operation data of the G1-G5 units ( Figure 4) shows that the coal-fired unit maintains a relatively stable output throughout the cycle, with the output power fluctuating between 30.16MW and 68.49MW. In particular, it reaches its maximum value in the 16th hour, showing its relative stability and reliability during peak load periods. The output of the G2 photovoltaic generator set is subject to sunshine conditions, gradually rising from 6:00 to 18:00, and reaching its maximum value from 12:00 to 13:00, reflecting the significant impact of the diurnal cycle on its power generation capacity. The fluctuation of the G3 wind turbine generator set is relatively small compared to photovoltaics, and it can provide clean energy for the system with an average daily output power of 13.5MW. G3 and G5 units provide the main load support at peak operating times (such as 9:00 and 10:00), while weakening at night such as 23:00 to 24:00. Through analysis, it can be concluded that the carbon emission rate of the G1 coal-fired unit gradually increases within 24 hours, with the peak occurring in the period consistent with its peak power generation, but the overall fluctuation rhythm is slow. The carbon flow rate of the G5 unit shows an increasing trend at night and during the peak power generation period, showing the load regulation ability of the gas power plant.

[0126] Example 3

[0127] In this embodiment, an electricity-carbon collaborative optimization system based on carbon emission flow theory is proposed, which specifically includes:

[0128] Load node power model building module: builds a regional distribution network architecture including load nodes and power generation nodes, and establishes a power model for load nodes;

[0129] Carbon emission flow model building module: build a carbon emission flow model based on the power flow model;

[0130] Collaborative optimization model building module: builds an electricity-carbon collaborative optimization model based on the power model and carbon emission flow model of the load node;

[0131] Parameter prediction module: Use the LSTM network to predict the actual photovoltaic, wind power generation and load data in the distribution network, collect the charging demand information of the charging station, and assign parameters in the power-carbon collaborative optimization model;

[0132] And, the model solving module: use the solver to solve the electricity-carbon collaborative optimization model, determine the optimal power flow of each load every hour, and calculate the carbon emission intensity and carbon emission rate to obtain the final scheduling plan.

[0133] The collaborative optimization method of the present invention may be implemented in hardware, firmware, or as software or computer code that may be stored in a recording medium (such as a CD ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code that is originally stored in a remote recording medium or a non-temporary machine-readable medium downloaded over a network and will be stored in a local recording medium, so that the method described herein may be stored in such software processing on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an ASIC or FPGA). It is understood that a computer, processor, microprocessor controller, or programmable hardware includes a storage component (e.g., RAM, ROM, flash memory, etc.) that can store or receive software or computer code, and when the software or computer code is accessed and executed by a computer, processor, or hardware, the method described herein is implemented. In addition, when a general-purpose computer accesses the code for implementing the method shown herein, the execution of the code converts the general-purpose computer into a dedicated computer for executing the method shown herein.

[0134] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. An electricity-carbon collaborative optimization method based on carbon emission flow theory, characterized in that: The following steps are involved: Construct a regional distribution network architecture including load nodes and power generation nodes, and establish a power model for load nodes; According to the power flow model, a carbon emission flow model is established; Based on the power model of the load node and the carbon emission flow model, an electricity-carbon collaborative optimization model is constructed; The LSTM network is used to predict the actual photovoltaic and wind power generation and load data in the distribution network, collect the charging demand information of the charging station, and assign parameters in the power-carbon collaborative optimization model; The solver is used to solve the electricity-carbon collaborative optimization model, determine the optimal power flow of each load every hour, calculate the carbon emission intensity and carbon emission rate, and obtain the final scheduling plan.

2. The electricity-carbon collaborative optimization method based on carbon emission flow theory according to claim 1 is characterized in that: The load nodes include: electrolysis hydrogen production equipment, charging piles, commercial loads, and civil loads; the power generation nodes include: cluster photovoltaic and wind power generation, coal-fired units, and natural gas power generation units.

3. The electricity-carbon collaborative optimization method based on carbon emission flow theory according to claim 1 is characterized in that: The carbon emission flow model is: In the formula, e Ni is the carbon potential of node i, I + is the set of branches with flow flowing into node i, ρ s is the carbon flow density of branch s, P Bs is the active power of branch s, P Nii is the node active flux matrix; R all is the sum of carbon flow rates of power system nodes; N is the total number of nodes in the distribution network; R i is the carbon flow rate of node i; Input active power for the generator set at node i; is the carbon emission intensity of the generator set at node i.

4. The electricity-carbon collaborative optimization method based on carbon emission flow theory according to claim 3 is characterized in that: The objective function of the electricity-carbon collaborative optimization model is: In the formula, p t is the load power, is the residential load power at time t, is the power generated by renewable energy, is the power of hydrogen production by electrolysis, Power for electric vehicle charging stations The power of centralized energy storage equipment; For electricity price, Output of each generator set, c k is the power generation cost of unit k, is the carbon emission intensity of the kth generator set, c c is the price per ton of CO2 emitted; is the output matrix vector of the generator set.

5. The electricity-carbon collaborative optimization method based on carbon emission flow theory according to claim 4 is characterized in that: The constraints of the electricity-carbon collaborative optimization model include: In the formula, They are the maximum power constraints for residential loads, renewable energy generation, water electrolysis hydrogen production, electric vehicle charging piles, and energy storage devices; They are the minimum power constraints for electric vehicle charging piles, residential loads, and water electrolysis to produce hydrogen; and are the minimum and maximum output power constraints of the generator sets respectively; and is the climbing constraint of the generator set.

6. The electricity-carbon collaborative optimization method based on carbon emission flow theory according to claim 1 is characterized in that: The solver adopts Gurobi Optimizer.

7. The electricity-carbon collaborative optimization system based on carbon emission flow theory is characterized by: include: Load node power model building module: builds a regional distribution network architecture including load nodes and power generation nodes, and establishes a power model for load nodes; Carbon emission flow model building module: build a carbon emission flow model based on the power flow model; Collaborative optimization model building module: builds an electricity-carbon collaborative optimization model based on the power model and carbon emission flow model of the load node; Parameter prediction module: Use the LSTM network to predict the actual photovoltaic, wind power generation and load data in the distribution network, collect the charging demand information of the charging station, and assign parameters in the power-carbon collaborative optimization model; And, the model solving module: use the solver to solve the electricity-carbon collaborative optimization model, determine the optimal power flow of each load every hour, and calculate the carbon emission intensity and carbon emission rate to obtain the final scheduling plan.

8. A computer storage medium storing a readable program, characterized in that: When the program is running, it can execute the electricity-carbon collaborative optimization method based on carbon emission flow theory as described in any one of claims 1-6.

9. An electronic device, characterized in that: include: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the electricity-carbon collaborative optimization method based on carbon emission flow theory as described in any one of claims 1-6.

10. A computer program product comprising computer instructions, characterized in that: The computer instructions instruct the computing device to perform operations corresponding to the electricity-carbon collaborative optimization method based on carbon emission flow theory as described in any one of claims 1-6.

Images