Microgrid operation method and system based on carbon flow tracking and multi-target collaborative optimization

Abstract

The invention relates to a microgrid operation method and system based on carbon flow tracking and multi-target collaborative optimization, and the method comprises the following steps: S1, carrying out the high-precision, real-time and synchronous data collection and processing, and providing complete and reliable data input for carbon flow tracking and optimal scheduling; s2, according to the data acquired in the S1, realizing carbon flow dynamic tracking through a directed graph, quantifying the carbon emission intensity of each node in the microgrid, introducing a carbon label technology, tracking the carbon emission source of each kilowatt-hour, and providing a data basis for optimization; s3, by taking minimization of carbon emission and operation cost as targets, a long-term scheduling plan is generated for the equipment, and multi-target collaborative optimization is realized; and S4, further optimizing the energy storage system on the basis of realizing the multi-target collaborative optimization in the step S3 so as to improve the demand response speed. According to the invention, the carbon emission reduction rate and the demand response speed during the operation of the industrial micro-grid can be effectively improved.

Description

[0001] The present invention relates to the technical field of power systems, and in particular to an industrial microgrid operation optimization method and system based on carbon flow tracking and multi-objective collaborative optimization. Background Art

[0002] With the deepening of the "dual carbon" goals, industrial microgrids, as an important carrier on the energy consumption side, play a key role in achieving clean energy consumption and improving energy efficiency. However, the existing industrial microgrid operation system still faces severe challenges in the coordination of carbon emission control and demand response. Traditional microgrids lack the ability to track carbon flows in a refined manner, making it difficult to accurately measure electricity carbon emissions. Existing carbon accounting mostly uses regional average carbon emission factors, which cannot reflect the real-time carbon flow distribution under the dynamic access of distributed energy, resulting in unclear definition of carbon responsibility and restricting the formulation of low-carbon scheduling strategies. In addition, the coordinated optimization of energy storage systems and demand response is insufficient. Especially under the peak-valley electricity price mechanism, the existing system often gives priority to economic efficiency and ignores carbon emission reduction targets, resulting in high-carbon standby units participating in peak regulation, which weakens the benefits of new energy consumption. Summary of the invention

[0003] In response to the above problems, the present invention provides a microgrid operation method and system based on carbon flow tracking and multi-objective collaborative optimization, which effectively improves the carbon emission reduction rate and demand response speed of industrial microgrids during operation through dynamic carbon flow tracking and priority calling based on demand response.

[0004] The technical solution of the present invention is: a microgrid operation method based on carbon flow tracking and multi-objective collaborative optimization, comprising the following steps:

[0005] S1. High-precision, real-time synchronous data collection and processing provide complete and reliable data input for carbon flow tracking and optimized scheduling.

[0006] S11. Collect environmental data (photovoltaic, wind power), load data (adjustable load demand, fixed load demand), market data (time-of-use electricity price, carbon trading price) and required equipment parameters (charging and discharging efficiency of energy storage system, rated capacity and output-fuel consumption curve of diesel generator, etc.).

[0007] S12, processing the data collected in S11. Perform outlier detection on the collected data and remove outlier data, and then use the LSTM neural network to perform 72-hour rolling forecast to fill in the missing data.

[0008] S13. Align, fuse and store data from different sources to provide data input for subsequent operations. Use timestamp interpolation to unify data with different sampling frequencies to 1-minute granularity, perform confidence-weighted fusion on data from different sources, obtain fused data and store it.

[0009] S2. Based on the data collected in S1, realize the dynamic tracking of carbon flow through a directed graph, quantify the carbon emission intensity of each node (such as photovoltaic, energy storage, diesel generator, etc.) in the microgrid, introduce the carbon label technology, track the carbon emission source of each degree of electricity, and provide a data basis for optimization.

[0010] S21. Construct a directed graph to realize the tracking and tracing of the carbon flow path. The nodes of this directed graph respectively correspond to the data of photovoltaic, wind power, diesel generator, energy storage system, power grid and load collected in S11, and the weight of the edges of this directed graph is the power value;

[0011] S22. According to the weighted calculation of the upstream carbon flow, define the carbon intensity propagation equation to calculate the carbon intensity of different nodes;

[0012] In step S22,

[0013] Use the carbon intensity propagation equation to predict the carbon intensity CI j (t) of node j at time t. The formula is as follows:

[0014]

[0015] In the formula, up(j) represents the set of all upstream nodes of node j, CI i (t) represents the carbon intensity of upstream node i at time t, P ij (t) represents the power between node i and node j at time t, down(j) represents the set of all downstream nodes of node j, P ik (t) represents the power between node i and node k at time t, represents the carbon emission of node j.

[0016] S23. Use the carbon flow path directed graph in S21 and the carbon intensity calculated in S22 to label the carbon flow with carbon labels, which is convenient for tracing the carbon emissions. Use the blockchain-based evidence storage system to associate a carbon label with each kWh of electric energy. When the electric energy flows through the energy storage system, the carbon label is dynamically updated according to the principle of last-in, first-out;

[0017] S24. Construct a carbon flow state space model, use the real-time detection data collected in S11 to correct the predicted value of the carbon intensity calculated in S22 in real time, and correct the generated carbon label.

[0018] S3. Based on the load data and market data collected in S1, with the goal of minimizing carbon emissions and operating costs, generate a long-term scheduling plan for the equipment to achieve multi-objective collaborative optimization.

[0019] S31. Comprehensively consider the operating cost, carbon emission cost and carbon peak penalty to construct the objective function of the microgrid operation optimization model.

[0020] S32. Set the constraint conditions for the operation of the microgrid, mainly including power balance constraints, equipment operation constraints, and carbon flow constraints.

[0021] S33. Use the Benders decomposition algorithm for multi-objective decomposition. Regard the optimization of the objective function as the main problem, regard the carbon flow constraint as the sub-problem, verify the carbon flow feasibility, and simultaneously satisfy the objective function and constraint conditions of the microgrid operation optimization.

[0022] S4. On the basis of achieving multi-objective collaborative optimization in S3, further optimize the energy storage system to improve the demand response speed. Store the excess electric energy during the peak period of renewable energy generation, and then, during the peak demand or insufficient power generation, preferentially call low-carbon resources to avoid the start of high-carbon standby power supplies and simultaneously meet the demand response.

[0023] S41. Calculate the carbon potential difference between the charging stage and the discharging stage.

[0024] S42. Define the charge-discharge threshold rule according to the carbon potential difference calculated in S41. The charging condition is when there is excess renewable energy, and the discharging condition is during the peak load or low-carbon demand.

[0025] S43. Define the calling order of different energy devices. While meeting the demand response, preferentially call low-carbon resources. Set the energy storage discharge as the first priority, set the adjustable load D flex,k as the second priority, and set the diesel generator P DG as the third priority.

[0026] The microgrid operation system based on carbon flow tracing and multi-objective collaborative optimization includes:

[0027] A preprocessing module for data acquisition and processing;

[0028] A tracing module for realizing dynamic carbon flow tracing of the processed data through a directed graph, quantifying each node in the microgrid; introducing carbon label technology to trace the carbon emission source of each degree of electricity;

[0029] A generation module for generating a scheduling plan for the equipment with the goal of minimizing carbon emissions and operating costs;

[0030] An optimization module for optimizing the energy storage system to improve the demand response speed.

[0031] In the operation of the present invention, first, through carbon flow dynamic tracking, the carbon emission intensity of each node in the microgrid is quantified, and the carbon label technology is introduced to trace the carbon emission source of each degree of electricity. Secondly, in order to generate a long-term scheduling plan for the equipment, with the goal of minimizing carbon emissions and operating costs, multi-objective collaborative optimization is achieved. In addition, from the perspective of optimizing the energy storage system, the redundant electric energy is stored during the peak period of renewable energy generation, and then during the peak demand period or when the power generation is insufficient, the low-carbon resources are preferentially called, the start of the high-carbon standby power supply is avoided, and the demand response is satisfied at the same time.

[0032] The present invention has carried out an innovative design for the operation of the industrial microgrid, effectively improving the carbon emission reduction rate and the demand response speed during the operation of the industrial microgrid. Brief Description of the Drawings

[0033] Figure 1 is a flowchart of the method of the present invention;

[0034] Figure 2 is a schematic diagram of carbon flow tracking designed by the present invention (including current and virtual carbon flow);

[0035] Figure 3 is a schematic diagram of multi-objective collaborative optimization designed by the present invention;

[0036] Figure 4 is a schematic diagram of demand response-carbon emission coupling control designed by the present invention. Detailed Embodiment

[0037] The following will describe the specific embodiments of the present invention in conjunction with the drawings, so that those skilled in the art can better understand the present invention.

[0038] As Figures 1-4 shown, the present invention provides a microgrid operation method based on carbon flow tracking and multi-objective collaborative optimization, which includes the following steps:

[0039] S1. High-precision and real-time synchronous data acquisition and processing provide complete and reliable data input for carbon flow tracking and optimal scheduling. The specific implementation steps are as follows:

[0040] S11. Collect environmental data, load data, market data, and the required equipment parameters.

[0041] Among them, the environmental data mainly includes the output of renewable energy, such as the photovoltaic output power P PV (t) and the wind power output power P WT (t), and the carbon emission factor η, such as the carbon emission factor η DG of the diesel generator and the carbon emission factor η grid of the power grid.

[0042] The load data mainly includes the adjustable load demand D flex,k (t), which includes K types of load, such as interruptible load (e.g., air conditioner) and shiftable load (e.g., electric boiler). k represents the load type number (k = 1, 2, …, K). At the same time, it also includes the fixed load demand D fix (t).

[0043] The market data mainly includes the time-of-use electricity price π grid (t) obtained from the power trading platform, and the carbon trading price obtained from the carbon emissions exchange

[0044] The equipment parameters mainly include the charging efficiency of the energy storage system discharge efficiency rated capacity It also includes the output-fuel consumption curve P DG (t) = f(F fuel (t)) of the diesel generator.

[0045] S12. Process the data collected in S11. Detect outliers in the collected data and remove the outlier data. Then use the LSTM neural network model for 72-hour rolling prediction to fill in the missing data. Use the existing data at time t to predict the missing data D(t + Δt) at the future Δt moment. The expression is as follows:

[0046] D(t + Δt) = LSTM[P re (t), D load (t), π grid (t)] + ε(t)

[0047] where P re (t) and D load (t) represent the values of renewable energy and load respectively, and ε(t) represents the Kalman filter correction term.

[0048] S13. Align, fuse and store the data from different sources to provide data input for subsequent operations. First, use the timestamp interpolation method to unify the data X(t) with different sampling frequencies to a 1-minute granularity, and obtain the aligned data X align (t) at time t. The formula is as follows:

[0049]

[0050] where N represents the number of sampling points of the original data within the Δt time window, i represents the i-th sampling point, and X(t i ) represents the data of the i-th sampling point at time t.

[0051] Then define a dynamic weight ω i, perform confidence-weighted fusion on data X from different sources i (t), obtain the fused data X(t) and store it:

[0052]

[0053] In S1, multi-source heterogeneous data is collected, and the collected data is subjected to abnormal data elimination and missing data filling. At the same time, the multi-source heterogeneous data is aligned and fused. Compared with the prior art that usually relies on a single data source (such as a SCADA system) and has a lag in data update; the present invention provides rich and reliable data sources, which can significantly reduce the risk of wrong decisions.

[0054] S2. According to the data collected in S1, realize the dynamic tracking of carbon flow through a directed graph, quantify the carbon emission intensity of each node (such as photovoltaic, energy storage, diesel generator, etc.) in the microgrid, introduce carbon label technology, track the carbon emission source of each degree of electricity, and provide a data basis for optimization. The specific implementation steps are as follows:

[0055] S21. Construct a directed graph to realize the tracking and tracing of the carbon flow path. The nodes of this directed graph respectively correspond to different data collected in S11, including photovoltaic, wind power, diesel generator, energy storage system, grid and load data. The nodes of this directed graph can be expressed as V = {PV, WT, DG, ESS, Grid, Load}. The edges of this directed graph are expressed as E = {(i, j)|node i delivers electric energy to node j}, and the weight of the edge represents the power value P ij (t), for example, P PV→Load (t) represents the power of photovoltaic direct supply to the load. Use an adjacency matrix A(t) of dimension |V|×|V| to describe the connection relationship between nodes, and its definition is as follows:

[0056]

[0057] Among them, A ij (t) represents the element value of the i-th row and j-th column in the adjacency matrix A(t), and its meaning is that node i delivers electric energy to node j at time t.

[0058] S22. According to the weighted calculation of the upstream carbon flow, define the carbon intensity propagation equation, calculate the carbon intensity of different nodes, and use the grid carbon emission factor η grid (t) (dynamically updated over time), the carbon emission factor η of the diesel generator DG Obtain the carbon intensity of the grid and the diesel generator. For example, the carbon intensity of grid power purchase at time t can be expressed as CI Grid (t) = η grid (t), and the carbon intensity of the diesel generator is CI DG = η DG, the carbon intensity of renewable energy is 0, that is, the carbon intensity CI of photovoltaic PV = 0, and the carbon intensity CI of wind power WT = 0. Use the carbon intensity propagation equation to predict the carbon intensity CI of node j at time t j (t), and the formula is as follows:

[0059]

[0060] In the formula, up(j) represents the set of all upstream nodes of node j, and CI i (t) represents the carbon intensity of upstream node i at time t, and P ij (t) represents the power between node i and node j at time t. down(j) represents the set of all downstream nodes of node j, and P ik (t) represents the power between node i and node k at time t. Represents the carbon emission of node j.

[0061] S23. Use the carbon flow path digraph in S21 and the carbon intensity calculated in S22 to label the carbon flow with carbon tags, which is convenient for tracing carbon emissions. Use the blockchain-based evidence storage system to associate a carbon tag with each kWh of electric energy e represents the e-th unit of electric energy. Among them, t e represents the time stamp of electric energy production, Path e represents the sequence of nodes passed through, represents the cumulative carbon intensity, and Source e represents the number of the carbon emission power supply equipment.

[0062] When the electric energy flows through the energy storage system, the carbon tag is dynamically updated according to the principle of last in first out. During the charging stage of the energy storage system, its carbon intensity at time t is equal to the carbon intensity CI of the power grid at time t Grid (t). During the discharging stage, take the carbon tag of the nearest charging batch to avoid the aliasing of carbon intensity.

[0063] S24. Build a carbon flow state space model, use real-time detection data to correct the predicted value of the carbon intensity calculated in S22 in real time, and correct the generated carbon tags. The carbon flow state space model is a dynamic mathematical model used to describe the law of the carbon emission intensity of each node (such as power source, energy storage, load) in the microgrid changing with time.

[0064] The predicted value of the carbon intensity calculated in S22 itself may have errors, and the carbon intensity of the node will also change dynamically with time and sudden changes in load demand or adjustments of energy storage charge and discharge strategies, etc. Therefore, it is necessary to use real-time data to correct the carbon intensity of the node in real time.

[0065] After calibrating the carbon intensity estimation values of relevant nodes using real-time data, trace back and correct the carbon labels generated in S23.

[0066] The calibrated carbon intensity and carbon labels will be used as a reference for minimizing carbon emissions in S3 to adjust the scheduling plan for microgrid operation.

[0067] First, define the carbon flow state vector CI(t) = [CI1(t), CI2(t), …, CI |V| (t)] T , and thus construct a discrete-time state equation to calculate the carbon flow state vector CI(t + 1) at the next moment of time t, with the formula as follows:

[0068] CI(t + 1) = A d ·CI(t) + B d ·u(t) + w(t)

[0069] Among them, A d represents the state transition equation, which is calculated based on the adjacency matrix A(t) and the power distribution ratio; B d represents the input matrix, characterizing the regulation effect of energy storage charging and discharging on the carbon flow; u(t) represents the power when the energy storage system is charging or discharging, and w(t) represents the process noise.

[0070] Next, uniformly represent the real-time detection data such as the photovoltaic output power P PV (t) and the wind power output power P WT (t) collected in S11 as Z(t) = [P PV (t), P DG (t), …] T , and use it to update the calculated value of the carbon intensity:

[0071] CI corr (t) = CI pred (t) + K(t)(Z(t) - H·CI pred (t))

[0072] Among them, CI corr (t) represents the calibrated carbon intensity, CI pred (t) represents the predicted value of the carbon intensity of the node calculated in S22, H is the observation matrix, used to identify the position of the measurable node; K(t) is the Kalman gain matrix.

[0073] In S2, a dynamic carbon flow network is constructed using a directed graph. The carbon label technology is adopted to record the carbon flow path, and the state space model is used to dynamically correct the estimated value of carbon intensity based on real-time data. Compared with the existing technology that usually adopts the average carbon emission factor and ignores the dynamic carbon intensity change of the power grid, the present invention can reflect the real-time power distribution and carbon flow propagation. The carbon label technology enables the carbon intensity of the load node to be traced back to the specific power supply equipment, which is convenient for controlling carbon emissions.

[0074] S3. Based on the load data and market data collected in S11, with the goal of minimizing carbon emissions and operating costs, a long-term scheduling plan is generated for the equipment to achieve multi-objective collaborative optimization. The specific implementation steps are as follows:

[0075] S31. Comprehensively consider the operating cost, carbon emission cost, and carbon peak penalty to construct the objective function of the microgrid operation optimization model. The expression of the objective function is as follows:

[0076]

[0077] Among them, represents the operating cost at time t, and π grid (t) represents the time-of-use electricity price at time t, c DG , c ESS respectively represent the unit power generation cost of the diesel generator and the energy storage cycle loss cost coefficient. P Grid (t), P DG (t), respectively represent the power grid power, diesel generator power, and charging and discharging power of the energy storage system at time t.

[0078] represents the carbon emission cost at time t, represents the carbon trading price at time t, CI Grid (t), CI DG represent the carbon intensities of the power grid and the diesel generator at time t. CI peak (t) represents the peak value of the overall carbon emission intensity of the microgrid at time t. α, β, γ represent the dynamic adjustment coefficients. T represents the total operation time of the microgrid.

[0079] S32. Set the constraint conditions for the microgrid operation.

[0080] (1) Power balance constraint:

[0081]

[0082] Among them, D flex,k (t) represents the adjustable load demand, and D fix (t) represents the fixed load demand.

[0083] In addition, the constraints on adjustable loads are as follows:

[0084]

[0085] Among them, D flex,k (t) represents the demand power of the k-th type of adjustable load at time t, represents the maximum power of the adjustable load, represents the total power of the k-th type of adjustable load.

[0086] (2) Equipment operation constraints

[0087] Operating constraints of diesel generators:

[0088]

[0089] Among them, represent the lower and upper limits of the output of the diesel generator respectively, represents the ramp rate limit of the diesel generator.

[0090] Operating constraints of the energy storage system:

[0091]

[0092] 0.2 ≤ SOC(t) ≤ 0.9

[0093]

[0094] Among them, represents the state of charge of the energy storage at time t, represents the rated capacity of the energy storage, E ESS (t) represents the capacity of the energy storage at time t, represents the charging and discharging efficiency of the energy storage. represent the charging and discharging power of the energy storage system at time t respectively, represent the maximum charging and discharging powers of the energy storage system respectively.

[0095] (3) Carbon flow constraints

[0096] Carbon intensity limit for each node:

[0097] CI load (t) ≤ CI target

[0098] Among them, CI load (t) represents the carbon emission intensity of the load node, calculated according to the propagation equation in S22. CI target represents the target threshold of the carbon emission intensity.

[0099] Carbon intensity of the energy storage system:

[0100]

[0101] Among them, Τ ch represents the set of all charging periods within the current discharge cycle. represents the carbon intensity during the charging period τ when the energy storage is charging, represents the charging capacity of the energy storage system during the charging period τ.

[0102] S3 uses the Benders decomposition algorithm for multi-objective decomposition, simultaneously satisfying the objective function and constraint conditions of the microgrid operation optimization. Regarding the optimization of the objective function as the main problem, it is expressed as follows:

[0103] min[αC op +βC carbon s.t. Equipment operation constraints. Among them, C op represents the operation cost of the power grid, and C carbon represents the carbon emission cost. Regarding the carbon flow constraint as the sub-problem, verify the carbon flow feasibility and verify CI load (t) ≤ CI target . If the constraint is violated, return to the main problem.

[0104] In S3, with the goal of minimizing both carbon emissions and operation costs, the operation strategy of the microgrid is adjusted in a timely manner by combining the carbon intensity of the nodes and the real-time carbon price. Compared with the single-objective optimization in the prior art, such as only considering the operation cost or carbon emissions and being unable to respond to the fluctuations of the real-time carbon price; through multi-objective collaborative optimization, the present invention can reduce carbon emissions and operation costs while ensuring the normal operation of the microgrid.

[0105] S4. On the basis of realizing multi-objective collaborative optimization in S3, further optimize the energy storage system to improve the demand response speed. Store the excess electric energy during the peak of renewable energy generation, and then, during the peak demand or insufficient power generation, preferentially call low-carbon resources (such as energy storage discharge, controllable load transfer) to avoid the start of high-carbon standby power sources and meet the demand response at the same time. The specific implementation steps are as follows:

[0106] S41. Calculate the carbon potential difference between the charging stage and the discharging stage. The carbon potential difference △CI ch (t) during the charging stage is calculated as follows:

[0107] ΔCI ch (t) = CI source (t) - CI ESS,avg

[0108] Among them, CI source (t) represents the real-time carbon intensity of the charging power source, and CI ESS,avg represents the historical average discharge carbon intensity of the energy storage.

[0109] Carbon potential difference △CI during the discharge stage dis (t) is calculated as follows:

[0110]

[0111] Where CI load (t) represents the current carbon intensity of the load node.

[0112] S42. Define the charge-discharge threshold rule according to the carbon potential difference calculated in S41. Rule 1: The charging condition is when renewable energy is in excess:

[0113]

[0114] Where δ high represents the adjustable upper limit of the carbon potential difference threshold, represents the power when renewable energy is in excess, and SOC(t) represents the state of charge of the energy storage at time t. When in excess, the energy storage system charges.

[0115] Rule 2: The discharging condition is during peak load or low carbon demand:

[0116]

[0117] Where δ low represents the adjustable lower limit of the carbon potential difference threshold, represents the peak value of the adjustable load demand.

[0118] S43. Define the calling order of different energy devices. While meeting the demand impact, preferentially call low-carbon resources. Set the energy storage discharge as the first priority, set the adjustable load D flex,k as the second priority, and set the diesel generator P DG as the third priority. The diesel generator P is enabled only when and only when SOC(t) < 0.25 and the load is non-adjustable DG . When the following conditions are met, starting the diesel generator is prohibited:

[0119]

[0120] In S4, on the basis of meeting the minimization of carbon emissions and operating costs, further optimize the energy storage system to improve the speed of demand response. Compared with the prior art where priority is given to scheduling low-carbon resources, some resource waiting time may be consumed and the demand response speed is ignored; the present invention uses the carbon potential difference to define the charge-discharge threshold conditions, preferentially calls the energy storage and adjustable load according to actual demands, and can implement the adjustment of the charge-discharge plan of the energy storage system, improving the demand response speed to the minute level.

[0121] In the operation of the present invention, through carbon flow dynamic tracking and priority call based on demand response, the problems of high carbon emissions, slow demand response, and difficulty in coordinating carbon emissions and demand response during the operation of industrial microgrids are solved, effectively improving the carbon emission reduction rate and demand response speed during the operation of industrial microgrids.

[0122] The present invention also provides a microgrid operation system based on carbon flow tracking and multi-objective collaborative optimization, including:

[0123] A preprocessing module for data acquisition and processing;

[0124] A tracking module for realizing carbon flow dynamic tracking of the processed data through a directed graph to quantify each node in the microgrid; introducing carbon label technology to track the carbon emission source of each degree of electricity;

[0125] A generation module for generating a scheduling plan for equipment with the goal of minimizing carbon emissions and operating costs;

[0126] An optimization module for optimizing the energy storage system to improve the demand response speed.

[0127] The present invention realizes real-time carbon flow update and improves the demand response speed. In addition, the present invention reduces the operating cost, while reducing the carbon emissions, further optimizing the operation of the microgrid. Taking the operation of a microgrid in a certain area within 24 hours as an example, the operation results of the proposed solution of the present invention and the traditional solution are compared, and the results are shown in Table 1. It can be seen from the table that compared with the traditional solution, the proposed solution of the present invention reduces the carbon emissions by 36.5%, reduces the carbon emission trading cost by 40.1%, and at the same time reduces the operating cost by 15%. In addition, the demand response speed is improved from the minute level to the second level, and the average daily response event number is increased by 10 times.

[0128] Table 1 Comparison of operation results between the proposed solution of the present invention and the traditional solution

[0129]

[0130] The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although specific embodiments are described in detail herein, those skilled in the art can still modify them or replace some technical features in an equivalent manner, and these changes do not deviate from the core idea and protection scope embodied in the embodiments of the present invention.

Claims

1. A microgrid operation method based on carbon flow tracking and multi-objective collaborative optimization, characterized in that It includes the following steps: S1. Data collection and processing; S2. Implement carbon flow dynamic tracking for the processed data through a directed graph to quantify each node in the microgrid; introduce carbon label technology to track the carbon emission sources of each degree of electricity; S3. Based on S1 and S2, generate a scheduling plan for the equipment with the goal of minimizing carbon emissions and operating costs; S4. Based on S3, optimize the energy storage system to improve the demand response speed.

2. The microgrid operation method based on carbon flow tracking and multi-objective collaborative optimization according to claim 1, wherein Step S1 includes: S11. Collect environmental data, load data, market data, and equipment parameters; S12. Detect outliers in the collected data and remove the outlier data, and then use the LSTM neural network model for 72-hour rolling prediction to fill in the missing data; S13. Align, fuse, and store data from different sources.

3. According to the microgrid operation method based on carbon flow tracking and multi-objective collaborative optimization described in claim 2, wherein the environmental data includes photovoltaic output power, wind power output power, and carbon emission factors; the load data includes interruptible load, shiftable load, and fixed load demand; the market data includes time-of-use electricity price obtained from the power trading platform and carbon trading price obtained from the carbon emission exchange; the equipment parameters include the charging efficiency, discharging efficiency, and rated capacity of the energy storage system.

4. The microgrid operation method based on carbon flow tracking and multi-objective collaborative optimization according to claim 1, characterized in that Step S2 includes: S21. Construct a directed graph to realize the tracking and tracing of the carbon flow path; S22. Define the carbon intensity propagation equation to calculate the carbon intensity of different nodes; S23. Use the directed graph of the carbon flow path and the calculated carbon intensity to label the carbon flow with carbon labels; S24. Construct a carbon flow state space model, and use the collected real-time detection data to perform real-time correction on the predicted value of the calculated carbon intensity and correct the generated carbon labels.

5. The operating method of the microgrid based on carbon flow tracking and multi-objective collaborative optimization according to claim 1, characterized in that In step S21, the nodes of the directed graph respectively correspond to the data collected in S11, and the weight of the edge of the directed graph is the power value.

6. The microgrid operation method based on carbon flow tracking and multi-objective collaborative optimization according to claim 1, characterized in that In step S22, Predict the carbon intensity CI of node j at time t using the carbon intensity propagation equation, and the formula is as follows: j (t), and the formula is as follows: where up(j) represents the set of all upstream nodes of node j, and CI i (t) represents the carbon intensity of upstream node i at time t, and P ij (t) represents the power between node i and node j at time t, down(j) represents the set of all downstream nodes of node j, and P ik (t) represents the power between node i and node k at time t, represents the carbon emissions of node j.

7. The operating method of the microgrid based on carbon flow tracking and multi-objective collaborative optimization according to claim 1, characterized in that, Step S3 includes: S31. Comprehensively consider the operating cost, carbon emission cost, and carbon peak penalty to construct the objective function of the microgrid operation optimization model; S32. Set the constraint conditions for the microgrid operation, including power balance constraint, equipment operation constraint, and carbon flow constraint; S33. Adopt the Benders decomposition algorithm for multi-objective decomposition, regard the optimization of the objective function as the main problem, regard the carbon flow constraint as the sub-problem, verify the carbon flow feasibility, and simultaneously satisfy the objective function and constraint conditions of the microgrid operation optimization.

8. The operation method of the microgrid based on carbon flow tracking and multi-objective collaborative optimization according to claim 1, characterized in that Step S4 includes: S41. Calculate the carbon potential difference in the charging stage and the discharging stage; S42. Define the charge and discharge threshold rules according to the carbon potential difference calculated in S41; the charging condition is when there is excess renewable energy, and the discharging condition is during the load peak or low-carbon demand; S43. Define the call order of different energy devices. While meeting demand response, give priority to calling low-carbon resources; set energy storage discharge as the first priority, set the adjustable load D flex,k as the second priority, and set the diesel generator P DG as the third priority.

9. A microgrid operation system based on carbon flow tracking and multi-objective collaborative optimization, characterized in that It includes: A preprocessing module for data collection and processing; A tracking module for realizing carbon flow dynamic tracking for the processed data through a directed graph to quantify each node in the microgrid; introducing carbon label technology to track the carbon emission sources of each degree of electricity; A generation module for generating a scheduling plan for the equipment with the goal of minimizing carbon emissions and operating costs; Optimization module, used to optimize the energy storage system to improve the demand response speed.

Images