Novel power system robust optimization scheduling method based on node carbon potential response

Abstract

The invention relates to the field of power system scheduling, in particular to a novel power system robust optimization scheduling method based on node carbon potential response. Aiming at the defects that in the prior art, carbon emission responsibility allocation is unreasonable, a demand response mechanism is imperfect, and a scheduling model comprehensively considering economical efficiency and low-carbon performance is lacked, a carbon emission flow model is constructed, node carbon potential is calculated, and responsibility is allocated; a carbon emission responsibility allocation mechanism is established, load node carbon emission is subdivided and calculated, and a quota is determined; developing a dual low-carbon demand response model which comprises a price type model and an excitation type model; and constructing a source-load collaborative low-carbon optimal scheduling model which comprises an upper-layer economic scheduling model and a lower-layer low-carbon scheduling model. Through cooperative work and iterative optimization of the models, fairness allocation of carbon emission responsibility, improvement of demand response effectiveness, dual optimization of economy and low-carbon performance of the electric power system and improvement of stability are realized, and the method is suitable for various energy power generation and electric power systems with different types of loads.

Description

Technical Field

[0001] The present invention relates to the field of power system dispatching, and in particular to a novel power system robust optimization dispatching method based on node carbon potential response. Background Art

[0002] As the world pays more attention to environmental protection and sustainable development, low-carbon operation of power systems has become a research hotspot. Traditional power system dispatching methods mainly focus on the economy and reliability of the system, and often ignore the impact of carbon emissions. In the existing technology, there are the following defects:

[0003] The allocation of carbon emission responsibility is unreasonable. The existing methods lack a detailed division of carbon emission responsibility and do not fully consider the differences in energy efficiency and carbon potential of different load nodes. This results in some high-carbon emission loads being unable to be effectively managed, while some low-carbon emission loads may bear too much carbon emission responsibility, making it impossible to achieve fair allocation of carbon emission responsibility. For example, industrial loads and residential loads have great differences in energy efficiency and carbon emission characteristics, but traditional methods may use a unified standard to calculate carbon emission responsibility, which cannot accurately reflect the actual situation.

[0004] The demand response mechanism is imperfect. Traditional demand response strategies are often relatively simple and cannot fully mobilize users' enthusiasm for participating in low-carbon scheduling. For example, through simple time-of-use electricity price adjustments, users may not be very sensitive to changes in electricity prices, and cannot effectively guide users to adjust their electricity consumption behavior during low-carbon emission periods. Moreover, the existing demand response mechanism lacks targeted incentives for load aggregates, and cannot be flexibly adjusted according to the characteristics of load aggregates and carbon emissions, making it difficult to achieve effective low-carbon scheduling on the load side.

[0005] There is a lack of dispatch models that comprehensively consider economy and low carbon. Most existing dispatch models either focus on economy and ignore the requirements of low carbon emissions; or overemphasize low carbon, resulting in excessively high system operating costs. No effective dispatch model can achieve the goal of low carbon emissions while ensuring the economy of the system. For example, some dispatch methods aimed at low carbon may over-rely on renewable energy generation, which may lead to reduced system reliability and increased power generation costs when renewable energy is unstable.

[0006] The novel robust optimization scheduling method for power system based on node carbon potential response proposed in the present invention improves the defects of the above-mentioned prior art. Summary of the invention

[0007] In view of the above deficiencies in the prior art, the object of the present invention is to achieve comprehensive optimization of the economy, low carbon and stability of the power system.

[0008] In order to achieve the above objectives, the present invention adopts the following technical solutions:

[0009] A new robust optimization dispatching method for power system based on node carbon potential response includes the following steps:

[0010] S1. Construct a carbon emission flow model for the power system, construct a carbon emission flow model for the power system, and track the carbon emission flow in the power system;

[0011] S2. Establish a carbon emission responsibility sharing mechanism. Based on the node carbon potential obtained in S1 and the actual electricity consumption of each load node, calculate the carbon emissions corresponding to each load node, and determine the initial carbon emission quota of each load aggregate according to its purchased electricity and node carbon potential; use the initial carbon emission quota of each load aggregate to design a ladder carbon trading model, and use different carbon trading prices to encourage load aggregates to optimize their electricity consumption behavior within the carbon emission quota;

[0012] S3. Develop a dual low-carbon demand response model. Based on the carbon emission responsibility sharing mechanism established in S2, a dual low-carbon demand response model is constructed;

[0013] S4. Construct a low-carbon optimization dispatch model for power system source-load coordination. Based on the dual low-carbon demand response model developed in S3, construct a low-carbon optimization dispatch model for power system source-load coordination.

[0014] S5. Realize source-load coordinated low-carbon scheduling. Based on the source-load coordinated low-carbon optimization scheduling model constructed in S4, realize source-load coordinated low-carbon scheduling.

[0015] Furthermore, in S1, the method for constructing a carbon emission flow model for a power system includes:

[0016] S11. Define the carbon emission flow index. In the power system, the determination of the carbon emission flow index is the key to quantifying carbon emissions. Different fuels have different carbon emission characteristics. Taking the i-th fuel as an example, its carbon emission coefficient Ci,f represents the carbon emission generated by the combustion of a unit mass of fuel. The mass flow rate of the i-th fuel used by the power generation equipment j during operation is mi,f, which reflects the fuel consumption rate. At the same time, the operating efficiency ηj of the power generation equipment has an important impact on carbon emissions. If the equipment efficiency is low, more fuel will be consumed in order to generate the same amount of electricity, resulting in more carbon emissions. Therefore, the carbon emission flow index Ej is calculated by summing the carbon emissions generated by the use of various fuels and considering the influence of equipment efficiency. The formula is:

[0017]

[0018] S12. Calculate the node carbon potential. The calculation of the node carbon potential needs to consider the inflow and outflow of various energy sources. The charging power of the energy storage device at node k is Pc,k, and the discharge power is Pd,k. The corresponding carbon potential when the energy storage device is charged is The corresponding carbon potential during discharge is This is because the energy sources and destinations of energy storage devices are different during the charging and discharging process, and the carbon potential is also different. Charging may come from a power source with high carbon potential, and discharging may supply a load with low carbon potential. At the same time, the equivalent carbon emissions of the distributed power supply at node k are E DG ,k, equivalent carbon emissions E DG ,k takes into account the implicit carbon emissions of distributed power equipment manufacturing and maintenance; there are n branches connected to node k, and the carbon emission flow of the lth branch is Ekl; the node carbon potential The calculation formula is:

[0019]

[0020] The meaning of the formula is that the node carbon potential It is equal to the sum of the carbon emission flows of all connected branches, plus the equivalent carbon emissions of distributed power generation, plus the increase in carbon potential caused by charging of energy storage equipment. Subtract the carbon potential taken away by the energy storage device during discharge

[0021] S13. Construct an active flux matrix, which describes the flow path and size of power in the power system. With the operation of the power system, the network topology may change, and the dynamic characteristics of the load will also affect the power flow. At time t, the connection matrix corresponding to the network topology is A(t), which reflects the connection relationship between nodes. The load vector is L(t), which represents the load demand of each node. The generator output vector is G(t), which represents the output power of each generator. The active flux matrix P(t) is obtained by multiplying the network connection matrix with the difference between the generator output and the load vector. The formula is:

[0022] P(t) = A(t)(G(t) - L(t));

[0023] The formula is expressed as follows: If a transmission line fails at a certain moment, the connection matrix A(t) will change accordingly, thus affecting the active flux matrix P(t), and thus affecting the flow of power in the system;

[0024] S14. Apply the carbon flow tracking method to apportion the carbon emission responsibility. In order to fairly apportion the carbon emission responsibility from the power generation side to the load side, it is necessary to calculate based on the total power generation and carbon emissions of the system; the total power generation of the system is Gt, the total carbon emissions are Et, and the power consumption of load node m is Um; since the carbon emission responsibility should be related to the power consumption, the carbon emissions Em allocated to the load node m is proportional to its power consumption, and the proportionality coefficient is the ratio of the total carbon emissions of the system to the total power generation, and the calculation formula is:

[0025]

[0026] Furthermore, the step of establishing a carbon emission responsibility sharing mechanism in S2 includes:

[0027] S21. Calculate the carbon emissions of the load nodes. In order to accurately calculate the carbon emissions, the load nodes need to be subdivided. The load node n is subdivided into p parts, each of which contains its own energy utilization efficiency and carbon potential. For each subdivided part, its carbon emission calculation needs to consider carbon potential, electricity consumption and energy utilization efficiency. First, the carbon emissions of each subdivided part are calculated as:

[0028] Then add up the carbon emissions of all subdivided parts to get the carbon emissions Em of the entire load node n. The calculation formula is:

[0029] Among them, ηn,q is the energy efficiency of the qth part, is the carbon potential corresponding to the energy utilization efficiency ηn,q of the qth part, Un,q is the electricity consumption corresponding to the energy utilization efficiency ηn,q of the qth part;

[0030] S22. Determine the initial carbon emission quota of the load aggregate. When determining the initial carbon emission quota of the load aggregate, it is necessary to comprehensively consider multiple factors, including the purchased electricity volume, node carbon potential and industry attribute adjustment coefficient of the load aggregate; among them, the purchased electricity volume and node carbon potential are the basic considerations, and the industry attribute adjustment coefficient is determined based on the historical data of energy utilization efficiency, etc., to reflect the differences between different industries; the initial carbon emission quota of the load aggregate is jointly determined by the purchased electricity volume, node carbon potential and industry attribute adjustment coefficient, and the calculation formula is:

[0031]

[0032] Where Cr is the initial carbon emission quota of load aggregate r; Qr is the electricity purchase amount of load aggregate r; is the node carbon potential; αr is the industry attribute adjustment coefficient;

[0033] S23. Design a ladder carbon trading model. Use the initial carbon emission quota of each load aggregate to design a ladder carbon trading model. Divide the carbon trading price into s ladders, and the price of the i-th ladder is pi. When the carbon emission Er of the load aggregate r is within the i-th ladder range, the load-side carbon trading cost Tr is calculated according to the price of the ladder. The calculation formula is: Tr = Er × pi.

[0034] At the same time, the carbon credit trading mechanism is introduced. The carbon credit obtained by the load aggregate r is Cr,r, and the carbon credit price is pc. Then its carbon credit trading income is:

[0035] Rr=Cr,r×pc;

[0036] The introduction of a carbon credit trading mechanism is used to encourage load aggregates to obtain additional economic benefits by taking emission reduction measures, thereby more actively participating in carbon emission control.

[0037] Furthermore, in S3, the step of developing a dual low-carbon demand response model includes:

[0038] S31. Construct a price-based demand response model. The price-based demand response model guides users' electricity consumption behavior by adjusting electricity prices to achieve the goal of low carbon emissions. The price-based demand response model is expressed as:

[0039] Cu(t)=Uu(t)×p(t)×λ(t);

[0040]

[0041] Uu(t) is the user's electricity consumption; p(t) is the time-of-use electricity price; Cu(t) is the user's electricity cost; Esys(t) is the current carbon emission; β(t) is the proportion of renewable energy access; L(t) is the total system load; a, b, c are constants determined based on the historical data and characteristics of the power system; λ(t) is the dynamic carbon emission factor;

[0042] S32. Construct an incentive-based demand response model. The incentive-based demand response model is a demand response model based on the node carbon potential. The incentive-based demand response model is expressed as:

[0043]

[0044] Uv is the electricity consumption of load polymer v; Cv is the electricity cost of load polymer v; is the node carbon potential; u is the power consumption plan adjustment coefficient of load aggregate v; k1, k2, k3 are constants determined according to historical data and power system characteristics;

[0045] The carbon reward and penalty mechanism is used to encourage load aggregation to optimize power consumption plans. When the node carbon potential is high, the node carbon potential When the node carbon potential is low, u is also large, which increases the electricity cost Cv, thereby prompting the load aggregate to reduce electricity consumption. If is small, u is also small, the electricity cost Cv is relatively low, and the load aggregation may reasonably increase electricity consumption.

[0046] Furthermore, in S4, the step of constructing a power system source-load coordinated low-carbon optimization dispatch model includes:

[0047] S41. Establish an upper-level economic dispatch model. The upper-level economic dispatch model aims to optimize the economic efficiency of system operation and needs to comprehensively consider multiple cost factors. The objective function J1 of the upper-level economic dispatch model is:

[0048] J1=Cg+Cw+Ct+Caux;

[0049]

[0050] Pg=ΔPi+Pi

[0051] Cw=W×cw;

[0052]

[0053] Cg is the cost of power generation; Pg is the actual output of the power generation equipment; ΔPi is the output adjustment of the power generation equipment; Pi is the current output of the power generation equipment; Cw is the penalty cost for wind and solar power abandonment; Ct is the carbon trading cost; Caux is the ancillary service cost; cg is the power generation cost coefficient; W is the amount of wind and solar power abandonment; cw is the penalty coefficient, which is used to punish the wind and solar power abandonment caused by various reasons during the power generation process to encourage the full use of renewable energy; Eg is the carbon emission flow indicator of the power generation equipment; χ is the completion ratio of the low-carbon target; γ is the proportion of renewable energy access; δ is the carbon emission adjustment coefficient on the power generation side;

[0054] S42. Establish a lower-level low-carbon dispatch model. The lower-level low-carbon dispatch model takes the minimum system low-carbon dispatch cost and load carbon emission responsibility cost as the goal. J2 is:

[0055] J2=Cl+Ce+Cim;

[0056] Cim is the implementation cost of low-carbon scheduling measures; Cl is the low-carbon scheduling cost, which is calculated based on the relevant operating costs of low-carbon scheduling; Ce is the load carbon emission responsibility cost, which is calculated based on the load carbon emissions and related cost coefficients.

[0057] Furthermore, in S5, the step of constructing a power system source-load coordinated low-carbon optimization dispatch model includes:

[0058] S51. The upper model adjusts the unit output plan and finds a set of values ​​of the output adjustment of the power generation equipment ΔPi, so that the objective function J1 of the upper economic dispatch model is minimized and satisfies:

[0059] ΔPi≤riΔt;

[0060] Pmin,i≤Pi+ΔPi≤Pmax,i;

[0061] And try to reduce the number of unit starts and stops;

[0062] Among them, ri is the climbing rate; Δt is the time interval; Pmin,i is the minimum stable output; Pmax,i is the maximum stable output;

[0063] The load curve L1 after the response is passed to the lower model, wherein the load curve L1 is obtained based on the unit output adjustment and the supply and demand balance principle of the power system; when the unit output changes, the load curve will change accordingly according to the supply and demand balance principle of the power system;

[0064] S52. The lower model guides the flexible load to change the power consumption plan, and finds a set of values ​​of the power consumption plan adjustment ΔUj of the flexible load within the time interval Δt, so that the objective function J2 of the lower low-carbon scheduling model is minimized, and at the same time satisfies:

[0065] hj(t)ΔUj≤Δt;

[0066]

[0067] ΔUj-Uj|≤ΔUmax,j

[0068] ΔUmax,j is the maximum adjustment range; flexible load response characteristic function hj(t); kj is a coefficient determined according to the importance;

[0069] The load curve L2 corresponding to the adjusted flexible load power consumption plan is obtained based on the new power consumption distribution formed in each time period after the flexible load is adjusted;

[0070] S53. Iterative optimization. In each iteration, the upper model adjusts the unit output plan according to the power generation cost and carbon trading cost, and transmits the response load curve L1 to the lower model; the lower model guides the flexible load to change the power consumption plan according to the node carbon potential and load curve, and feeds back the response load curve L2 to the upper model;

[0071] During the iteration process, determine whether the convergence condition is met; if it is met and When S≤ε and D≤ε, the iteration ends;

[0072] Among them, S is the stability index; ε is the comprehensive convergence standard; D is the load satisfaction index; and are the objective function values ​​of the upper and lower models at the nth iteration respectively; and are the objective function values ​​of the upper and lower models at the n-1th iteration, respectively;

[0073] During the iteration process, the convergence criterion ε is adjusted according to the actual situation.

[0074] In summary, due to the adoption of the above technical solution, the beneficial technical effects of the invention are:

[0075] To achieve fair sharing of carbon emission responsibilities, we build a carbon emission flow model for the power system, define carbon emission flow indicators using measured or simulated data, calculate the node carbon potential, and apply the carbon flow tracking method to share the carbon emission responsibility on the power generation side to the load side. On this basis, we establish a carbon emission responsibility sharing mechanism, calculate the carbon emissions of the load nodes in detail, and determine the initial carbon emission quota based on the power purchase amount and node carbon potential of the load aggregate.

[0076] This refined carbon emission responsibility allocation method fully considers the energy utilization efficiency and carbon potential differences of different load nodes, and can accurately allocate carbon emission responsibility to each load node and load aggregate, avoiding the problem of unreasonable carbon emission responsibility allocation and achieving fair allocation of carbon emission responsibility.

[0077] To improve the effectiveness of demand response, a dual low-carbon demand response model was developed, including a price-based demand response model based on dynamic carbon emission factors and an incentive-based demand response model based on node carbon potential. The price-based demand response model guides users to increase electricity consumption during low-carbon emission periods through time-of-use electricity prices, while the incentive-based demand response model encourages load aggregates to optimize electricity consumption plans through a carbon reward and penalty mechanism.

[0078] The two demand response models work together to fully mobilize the enthusiasm of users and load aggregators to participate in low-carbon scheduling and improve the effectiveness of demand response. Users can flexibly adjust their electricity consumption behavior according to time-of-use electricity prices and carbon reward and penalty mechanisms, and load aggregators can also optimize their electricity consumption plans according to their own carbon emissions and incentives, effectively guiding low-carbon scheduling on the load side.

[0079] To achieve dual optimization of the economy and low carbon of the power system, a source-load coordinated low-carbon optimization dispatch model for the power system was constructed, including an upper-level economic dispatch model and a lower-level low-carbon dispatch model. The upper-level economic dispatch model aims to optimize the economic efficiency of system operation, taking into account the power generation cost, the penalty cost of wind and solar power abandonment, and the carbon trading cost; the lower-level low-carbon dispatch model aims to minimize the system low-carbon dispatch cost and the load carbon emission responsibility cost.

[0080] Through the collaborative work of the two models, the goal of low carbon emissions is achieved while ensuring the economy of the system. On the one hand, the upper economic dispatch model reduces the power generation cost and carbon trading cost by reasonably adjusting the unit output plan, thereby improving the economy of the system; on the other hand, the lower low carbon dispatch model reduces the load carbon emission responsibility cost and the system low carbon dispatch cost by guiding the flexible load to change the power consumption plan, thereby achieving the requirements of low carbon emissions. Ultimately, the dual optimization of the economy and low carbon of the power system is achieved.

[0081] Improve the stability of the power system. In the iterative optimization process, the unit output plan and flexible load power consumption plan are continuously adjusted until the system meets the dual requirements of economy and low carbon. This iterative optimization method can enable the system to continuously adapt to load changes and energy supply during operation, thus improving the stability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 A method flow chart of a novel robust optimization dispatching method for power systems based on node carbon potential response;

[0083] Figure 2 A flow chart of the method for constructing a carbon emission flow model for a power system;

[0084] Figure 3 A flow chart of the method for establishing a carbon emission responsibility allocation mechanism;

[0085] Figure 4 A methodological flow chart for developing a dual low-carbon demand response model;

[0086] Figure 5 A flow chart of the method for constructing a source-load coordinated low-carbon optimization dispatch model for the power system;

[0087] Figure 6 Flowchart of the method for constructing a source-load coordinated low-carbon optimization dispatch model for power systems. DETAILED DESCRIPTION

[0088] In order to make the purpose, technical solution and advantages of the invention more clear, the invention is further described in detail below in combination with the embodiments. It should be understood that the specific embodiments described here are only used to explain the invention and are not used to limit the invention.

[0089] like Figure 1 As shown, the novel robust optimization dispatching method of power system based on node carbon potential response includes the following steps:

[0090] S1. Construct a carbon emission flow model for the power system, define carbon emission flow indicators through measured or simulated data, then use the defined carbon emission flow indicators to calculate the node carbon potential, construct an active flux matrix, and finally apply the carbon flow tracking method to allocate the carbon emission responsibility from the power generation side to the load side;

[0091] S2. Establish a carbon emission responsibility sharing mechanism. Based on the node carbon potential obtained in S1 and the actual electricity consumption of each load node, calculate the carbon emissions corresponding to each load node, and determine the initial carbon emission quota of each load aggregate according to its purchased electricity and node carbon potential. Use the initial carbon emission quota of each load aggregate to design a ladder carbon trading model, and use different carbon trading prices to encourage load aggregates to optimize electricity consumption within the carbon emission quota, thereby achieving fair sharing of carbon emission responsibilities and providing an incentive mechanism for the demand response model.

[0092] S3. Develop a dual low-carbon demand response model. On the basis of the carbon emission responsibility sharing mechanism established in S2, a price-based demand response model based on dynamic carbon emission factors is constructed to guide users to increase electricity consumption during low-carbon emission periods through time-of-use electricity prices. Secondly, an incentive-based demand response model based on node carbon potential is constructed to encourage load aggregates to optimize electricity consumption plans through a carbon reward and penalty mechanism. The price-based demand response model and the incentive-based demand response model based on node carbon potential realize low-carbon dispatch on the load side, providing a demand-side response strategy for the power system source-load coordinated low-carbon optimization dispatch model.

[0093] S4. Construct a low-carbon optimization dispatch model for power system source-load coordination. Based on the dual low-carbon demand response model developed in S3, construct a low-carbon optimization dispatch model for power system source-load coordination. First, establish an upper-level economic dispatch model with the goal of optimizing the economic efficiency of system operation, taking into account the power generation cost, the penalty cost of wind and solar power abandonment, and the carbon trading cost. Then, establish a lower-level low-carbon dispatch model with the goal of minimizing the system low-carbon dispatch cost and the load carbon emission responsibility cost. Through the collaborative work of the two models, the dual optimization of the economy and low carbon of the power system is achieved.

[0094] S5. Realize source-load coordinated low-carbon dispatch. Based on the source-load coordinated low-carbon optimization dispatch model constructed in S4, realize source-load coordinated low-carbon dispatch. First, the upper-level model adjusts the unit output plan according to the power generation cost and carbon trading cost, and passes the responded load curve to the lower-level model. Then, the lower-level model guides the flexible load to change the power consumption plan through incentive-based demand response according to the node carbon potential and load curve, and feeds back the responded load curve to the upper-level model. Through iterative optimization, source-load coordinated low-carbon dispatch is realized until the dual requirements of system economy and low carbon are met, ensuring the efficient and environmentally friendly operation of the power system.

[0095] like Figure 2As shown, in S1, the method for constructing a carbon emission flow model for a power system includes:

[0096] S11. Define the carbon emission flow index. In the power system, the determination of the carbon emission flow index is the key to quantifying carbon emissions. Different fuels have different carbon emission characteristics. Taking the i-th fuel as an example, its carbon emission coefficient Ci,f represents the carbon emission generated by the combustion of a unit mass of fuel. The mass flow rate of the i-th fuel used by the power generation equipment j during operation is mi,f, which reflects the fuel consumption rate. At the same time, the operating efficiency ηj of the power generation equipment has an important impact on carbon emissions. If the equipment efficiency is low, more fuel will be consumed in order to generate the same amount of electricity, resulting in more carbon emissions. Therefore, the carbon emission flow index Ej is calculated by summing the carbon emissions generated by the use of various fuels and considering the influence of equipment efficiency. The formula is:

[0097]

[0098] S12. Calculate the node carbon potential. The calculation of the node carbon potential needs to consider the inflow and outflow of various energy sources. The charging power of the energy storage device at node k is Pc,k, and the discharge power is Pd,k. The corresponding carbon potential when the energy storage device is charged is The corresponding carbon potential during discharge is This is because the energy sources and destinations of energy storage devices are different during the charging and discharging process, and the carbon potential is also different. Charging may come from a power source with high carbon potential, and discharging may supply a load with low carbon potential. At the same time, the equivalent carbon emissions of the distributed power supply at node k are E DG ,k, equivalent carbon emissions E DG ,k takes into account the implicit carbon emissions of distributed power equipment manufacturing and maintenance; there are n branches connected to node k, and the carbon emission flow of the lth branch is Ekl; the node carbon potential The calculation formula is:

[0099]

[0100] The meaning of the formula is that the node carbon potential It is equal to the sum of the carbon emission flows of all connected branches, plus the equivalent carbon emissions of distributed power generation, plus the increase in carbon potential caused by charging of energy storage equipment. Subtract the carbon potential taken away by the energy storage device during discharge

[0101] S13. Construct an active flux matrix, which describes the flow path and size of power in the power system. With the operation of the power system, the network topology may change, such as line failure, new equipment access, etc. At the same time, the dynamic characteristics of the load will also affect the power flow. At time t, the connection matrix corresponding to the network topology is A(t), which reflects the connection relationship between nodes. The load vector is L(t), which represents the load demand of each node. The generator output vector is G(t), which represents the output power of each generator. The active flux matrix P(t) is obtained by multiplying the network connection matrix with the difference between the generator output and the load vector. The formula is:

[0102] P(t) = A(t)(G(t) - L(t));

[0103] The formula is expressed as follows: If a transmission line fails at a certain moment, the connection matrix A(t) will change accordingly, thus affecting the active flux matrix P(t), and thus affecting the flow of power in the system;

[0104] S14. Apply the carbon flow tracking method to apportion the carbon emission responsibility. In order to fairly apportion the carbon emission responsibility from the power generation side to the load side, it is necessary to calculate based on the total power generation and carbon emissions of the system; the total power generation of the system is Gt, the total carbon emissions are Et, and the power consumption of load node m is Um; since the carbon emission responsibility should be related to the power consumption, the carbon emissions Em allocated to the load node m is proportional to its power consumption, and the proportionality coefficient is the ratio of the total carbon emissions of the system to the total power generation, and the calculation formula is:

[0105]

[0106] like Figure 3 As shown, the steps of establishing the carbon emission responsibility sharing mechanism in S2 include:

[0107] S21. Calculate the carbon emissions of the load nodes. In order to accurately calculate the carbon emissions, the load nodes need to be subdivided. The load node n is subdivided into p parts, each of which contains its own energy utilization efficiency and carbon potential. For each subdivided part, its carbon emission calculation needs to consider carbon potential, electricity consumption and energy utilization efficiency. First, the carbon emissions of each subdivided part are calculated as:

[0108] Then add up the carbon emissions of all subdivided parts to get the carbon emissions Em of the entire load node n. The calculation formula is:

[0109] Among them, ηn,q is the energy efficiency of the qth part, is the carbon potential corresponding to the energy utilization efficiency ηn,q of the qth part, Un,q is the electricity consumption corresponding to the energy utilization efficiency ηn,q of the qth part;

[0110] S22. Determine the initial carbon emission quota of the load aggregate. When determining the initial carbon emission quota of the load aggregate, it is necessary to comprehensively consider multiple factors, including the purchased electricity volume, node carbon potential and industry attribute adjustment coefficient of the load aggregate; among them, the purchased electricity volume and node carbon potential are the basic considerations, and the industry attribute adjustment coefficient is determined based on the historical data of energy utilization efficiency, etc., to reflect the differences between different industries; the initial carbon emission quota of the load aggregate is jointly determined by the purchased electricity volume, node carbon potential and industry attribute adjustment coefficient, and the calculation formula is:

[0111]

[0112] Where Cr is the initial carbon emission quota of load aggregate r; Qr is the electricity purchase amount of load aggregate r; is the node carbon potential; αr is the industry attribute adjustment coefficient;

[0113] S23. Design a ladder carbon trading model. Use the initial carbon emission quota of each load aggregate to design a ladder carbon trading model. Divide the carbon trading price into s ladders, and the price of the i-th ladder is pi. When the carbon emission Er of the load aggregate r is within the i-th ladder range, the load-side carbon trading cost Tr is calculated according to the price of the ladder. The calculation formula is: Tr = Er × pi.

[0114] At the same time, the carbon credit trading mechanism is introduced. The carbon credit obtained by the load aggregate r is Cr,r, and the carbon credit price is pc. Then its carbon credit trading income is:

[0115] Rr=Cr,r×pc;

[0116] The introduction of a carbon credit trading mechanism can encourage load aggregates to obtain additional economic benefits by taking emission reduction measures, thereby participating more actively in carbon emission control.

[0117] like Figure 4 As shown, in S3, the steps of developing the dual low-carbon demand response model include:

[0118] S31. Construct a price-based demand response model. The price-based demand response model guides users' electricity consumption behavior by adjusting electricity prices to achieve the goal of low carbon emissions. The price-based demand response model is expressed as:

[0119] Cu(t)=Uu(t)×p(t)×λ(t);

[0120]

[0121] Uu(t) is the user's electricity consumption; p(t) is the time-of-use electricity price; Cu(t) is the user's electricity cost; Esys(t) is the current carbon emission; β(t) is the proportion of renewable energy access; L(t) is the total system load; a, b, c are constants determined based on the historical data and characteristics of the power system; λ(t) is the dynamic carbon emission factor;

[0122] The above model can be understood as guiding users to adjust their electricity consumption behavior according to carbon emissions in different time periods by adjusting the time-of-use electricity price and the dynamic carbon emission factor λ(t), such as increasing electricity consumption in time periods with lower dynamic carbon emission factors;

[0123] S32. Construct an incentive-based demand response model. The incentive-based demand response model is a demand response model based on the node carbon potential. The incentive-based demand response model is expressed as:

[0124]

[0125]

[0126] Uv is the electricity consumption of load polymer v; Cv is the electricity cost of load polymer v; is the node carbon potential; u is the power consumption plan adjustment coefficient of load aggregate v; k1, k2, k3 are constants determined according to historical data and power system characteristics;

[0127] The carbon reward and penalty mechanism is used to encourage load aggregation to optimize power consumption plans. When the node carbon potential is high, the node carbon potential When the node carbon potential is low, u is also large, which increases the electricity cost Cv, thereby prompting the load aggregate to reduce electricity consumption. If is small, u is also small, the electricity cost Cv is relatively low, and the load aggregation may reasonably increase electricity consumption.

[0128] like Figure 5 As shown, in S4, the steps of constructing a power system source-load coordinated low-carbon optimization dispatch model include:

[0129] S41. Establish an upper-level economic dispatch model. The upper-level economic dispatch model aims to optimize the economic efficiency of system operation and needs to comprehensively consider multiple cost factors. The objective function J1 of the upper-level economic dispatch model is:

[0130] J1=Cg+Cw+Ct+Caux;

[0131]

[0132] Pg=ΔPi+Pi

[0133] Cw=W×cw;

[0134]

[0135] Cg is the cost of power generation; Pg is the actual output of the power generation equipment; ΔPi is the output adjustment of the power generation equipment; Pi is the current output of the power generation equipment; Cw is the penalty cost for wind and solar power abandonment; Ct is the carbon trading cost; Caux is the cost of auxiliary services (such as the cost of auxiliary services such as frequency regulation and voltage regulation); cg is the power generation cost coefficient; W is the amount of wind and solar power abandonment; cw is the penalty coefficient, which is used to punish the wind and solar power abandonment caused by various reasons during the power generation process to encourage the full use of renewable energy; Eg is the carbon emission flow indicator of the power generation equipment; χ is the completion ratio of the low-carbon target; γ is the proportion of renewable energy access; δ is the carbon emission adjustment coefficient on the power generation side;

[0136] S42. Establish a lower-level low-carbon dispatch model. The lower-level low-carbon dispatch model takes the minimum system low-carbon dispatch cost and load carbon emission responsibility cost as the goal. J2 is:

[0137] J2=Cl+Ce+Cim;

[0138] Cim is the implementation cost of low-carbon dispatching measures; Cl is the low-carbon dispatching cost, which is calculated based on the relevant operating costs of low-carbon dispatching (such as communication costs and control equipment costs for demand response measures to guide flexible loads to change power consumption plans); Ce is the load carbon emission responsibility cost, which is calculated based on the load's carbon emissions and related cost coefficients.

[0139] like Figure 6 As shown, in S5, the steps of constructing a power system source-load coordinated low-carbon optimization dispatch model include:

[0140] S51. The upper model adjusts the unit output plan and finds a set of values ​​of the output adjustment of the power generation equipment ΔPi, so that the objective function J1 of the upper economic dispatch model is minimized and satisfies:

[0141] ΔPi≤riΔt;

[0142] Pmin,i≤Pi+ΔPi≤Pmax,i;

[0143] And try to reduce the number of unit starts and stops;

[0144] Where ri is the ramp rate; Δt is the time interval; Pmin,i is the minimum stable output; Pmax,i is the maximum stable output;

[0145] The load curve L1 after the response is passed to the lower model, where the load curve L1 is obtained based on the unit output adjustment and the supply and demand balance principle of the power system; when the unit output changes, according to the supply and demand balance principle of the power system, the load curve will change accordingly; for example, if the unit output increases, under the condition that other conditions remain unchanged, the load curve will show an overall upward trend, reflecting that the system can meet more load demands; conversely, if the unit output decreases, the load curve will decrease;

[0146] S52. The lower model guides the flexible load to change the power consumption plan, and finds a set of values ​​of the power consumption plan adjustment ΔUj of the flexible load within the time interval Δt, so that the objective function J2 of the lower low-carbon scheduling model is minimized, and at the same time satisfies:

[0147] hj(t)ΔUj≤Δt;

[0148]

[0149] |ΔUj-Uj|≤ΔUmax,j

[0150] ΔUmax,j is the maximum adjustment range; flexible load response characteristic function hj(t); kj is a coefficient determined according to the importance;

[0151] The load curve L2 corresponding to the adjusted flexible load power consumption plan is obtained based on the new power consumption distribution formed in each time period after the flexible load is adjusted;

[0152] S53. Iterative optimization. In each iteration, the upper model adjusts the unit output plan according to the power generation cost and carbon trading cost, and transmits the response load curve L1 to the lower model; the lower model guides the flexible load to change the power consumption plan according to the node carbon potential and load curve, and feeds back the response load curve L2 to the upper model;

[0153] During the iteration process, determine whether the convergence condition is met; if it is met and When S≤ε and D≤ε, the iteration ends;

[0154] Among them, S is the stability index; ε is the comprehensive convergence standard; D is the load satisfaction index; and are the objective function values ​​of the upper and lower models at the nth iteration respectively; and are the objective function values ​​of the upper and lower models at the n-1th iteration, respectively;

[0155] During the iteration process, the convergence criterion ε is adjusted according to the actual situation; for example, when the system is close to the optimal solution, ε can be appropriately tightened to improve the optimization accuracy; at the same time, in order to avoid local optimal solutions, multiple initial values ​​can be used for iteration, or combined with other optimization algorithms for hybrid optimization, for example, iterations can be performed under different initial unit output plans and flexible load power consumption plans, and then different results can be compared to select the optimal solution.

[0156] The above description is a preferred embodiment of the invention and is not intended to limit the invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the invention should be included in the protection scope of the invention.

Claims

1. A novel robust optimization scheduling method for power systems based on node carbon potential response, characterized in that: The following steps are involved: S1. Construct a carbon emission flow model for the power system, construct a carbon emission flow model for the power system, and track the carbon emission flow in the power system; S2. Establish a carbon emission responsibility sharing mechanism. Based on the node carbon potential obtained in S1 and the actual electricity consumption of each load node, calculate the carbon emissions corresponding to each load node, and determine the initial carbon emission quota of each load aggregate according to its purchased electricity and node carbon potential; use the initial carbon emission quota of each load aggregate to design a ladder carbon trading model, and use different carbon trading prices to encourage load aggregates to optimize their electricity consumption behavior within the carbon emission quota; S3. Develop a dual low-carbon demand response model. Based on the carbon emission responsibility sharing mechanism established in S2, a dual low-carbon demand response model is constructed; S4. Construct a low-carbon optimization dispatch model for power system source-load coordination. Based on the dual low-carbon demand response model developed in S3, construct a low-carbon optimization dispatch model for power system source-load coordination. S5. Realize source-load coordinated low-carbon scheduling. Based on the source-load coordinated low-carbon optimization scheduling model constructed in S4, realize source-load coordinated low-carbon scheduling.

2. According to a novel robust optimization scheduling method for power system based on node carbon potential response according to claim 1, it is characterized in that: In S1, the method for constructing a carbon emission flow model for a power system includes: S11. Define the carbon emission flow index. In the power system, the determination of the carbon emission flow index is the key to quantifying carbon emissions. Different fuels have different carbon emission characteristics. The carbon emission flow index is expressed as: Ci,f is the carbon emission coefficient, which indicates the carbon emission generated by the combustion of a unit mass of fuel; mi,f is the mass flow rate of the i-th fuel used by the power generation equipment j during operation, and mi,f reflects the fuel consumption rate; ηj is the operating efficiency of the power generation equipment; S12. Calculate the node carbon potential. The calculation of the node carbon potential needs to consider the inflow and outflow of various energy sources. The calculation formula of the node carbon potential is: is the node carbon potential; Pc,k is the charging power of the energy storage device at node k, and Pd,k is the discharging power of the energy storage device at node k; is the carbon potential corresponding to the energy storage device at node k when it is charged; is the carbon potential corresponding to the discharge of the energy storage device at node k; E DG ,k is the equivalent carbon emission of distributed generation at node k; Ekl is the carbon emission flow of the lth branch; S13. Construct an active flux matrix. The active flux matrix describes the flow path and size of power in the power system. The active flux matrix is ​​obtained by multiplying the network connection matrix with the difference between the generator output and the load vector. The formula is: P(t) = A(t)(G(t) - L(t)); P(t) is the active flux matrix; A(t) is the connection matrix corresponding to the network topology at time t; L(t) is the load vector, which represents the load demand of each node; G(t) is the generator output vector, which represents the output power of each generator; S14. Apply the carbon flow tracking method to apportion the carbon emission responsibility. In order to fairly apportion the carbon emission responsibility from the power generation side to the load side, it is necessary to calculate based on the total power generation and carbon emissions of the system. The calculation formula is: Em is the carbon emissions allocated to load node m; Gt is the total power generation of the system; Et is the total carbon emissions; Um is the electricity consumption of load node m.

3. According to a novel robust optimization scheduling method for power system based on node carbon potential response according to claim 1, it is characterized in that: The steps of establishing a carbon emission responsibility sharing mechanism in S2 include: S21. Calculate the carbon emissions of the load nodes. In order to accurately calculate the carbon emissions, the load nodes need to be subdivided. The load node n is subdivided into p parts, each of which contains its own energy utilization efficiency and carbon potential. For each subdivided part, its carbon emission calculation needs to consider carbon potential, electricity consumption and energy utilization efficiency. First, the carbon emissions of each subdivided part are calculated as: Then add up the carbon emissions of all subdivided parts to get the carbon emissions Em of the entire load node n. The calculation formula is: Among them, ηn,q is the energy efficiency of the qth part, is the carbon potential corresponding to the energy utilization efficiency ηn,q of the qth part, Un,q is the electricity consumption corresponding to the energy utilization efficiency ηn,q of the qth part; S22. Determine the initial carbon emission quota of the load aggregate. When determining the initial carbon emission quota of the load aggregate, multiple factors need to be considered comprehensively, including the purchased electricity volume, node carbon potential and industry attribute adjustment coefficient of the load aggregate; among them, the purchased electricity volume and node carbon potential are the basic considerations, and the industry attribute adjustment coefficient is determined based on the historical data of energy utilization efficiency to reflect the differences between different industries; the initial carbon emission quota of the load aggregate is jointly determined by the purchased electricity volume, node carbon potential and industry attribute adjustment coefficient, and the calculation formula is: Where Cr is the initial carbon emission quota of load aggregate r; Qr is the electricity purchase amount of load aggregate r; is the node carbon potential; αr is the industry attribute adjustment coefficient; S23. Design a ladder carbon trading model. Use the initial carbon emission quota of each load aggregate to design a ladder carbon trading model. Divide the carbon trading price into s ladders, and the price of the i-th ladder is pi. When the carbon emission Er of the load aggregate r is within the i-th ladder range, the load-side carbon trading cost Tr is calculated according to the price of the ladder. The calculation formula is: Tr = Er × pi. At the same time, the carbon credit trading mechanism is introduced. The carbon credit obtained by the load aggregate r is Cr,r, and the carbon credit price is pc. Then its carbon credit trading income is: Rr=Cr,r×pc; The introduction of a carbon credit trading mechanism is used to encourage load aggregates to obtain additional economic benefits by taking emission reduction measures, thereby more actively participating in carbon emission control.

4. According to a novel robust optimization scheduling method for power system based on node carbon potential response according to claim 1, it is characterized in that: In S3, the steps of developing a dual low-carbon demand response model include: S31. Construct a price-based demand response model. The price-based demand response model guides users' electricity consumption behavior by adjusting electricity prices to achieve the goal of low carbon emissions. The price-based demand response model is expressed as: Cu(t)=Uu(t)×p(t)×λ(t); Uu(t) is the user's electricity consumption; p(t) is the time-of-use electricity price; Cu(t) is the user's electricity cost; Esys(t) is the current carbon emission; β(t) is the proportion of renewable energy access; L(t) is the total system load; a, b, c are constants determined based on the historical data and characteristics of the power system; λ(t) is the dynamic carbon emission factor; S32. Construct an incentive-based demand response model. The incentive-based demand response model is a demand response model based on the node carbon potential. The incentive-based demand response model is expressed as: Uv is the electricity consumption of load polymer v; Cv is the electricity cost of load polymer v; is the node carbon potential; u is the power consumption plan adjustment coefficient of load aggregate v; k1, k2, k3 are constants determined according to historical data and power system characteristics; The carbon reward and penalty mechanism is used to encourage load aggregation to optimize power consumption plans. When the node carbon potential is high, the node carbon potential When the node carbon potential is low, u is also large, which increases the electricity cost Cv, thereby prompting the load aggregate to reduce electricity consumption. If is small, u is also small, the electricity cost Cv is relatively low, and the load aggregation may reasonably increase electricity consumption.

5. According to a novel robust optimization scheduling method for power system based on node carbon potential response according to claim 1, it is characterized in that: In S4, the steps of constructing a power system source-load coordinated low-carbon optimization dispatch model include: S41. Establish an upper-level economic dispatch model. The upper-level economic dispatch model aims to optimize the economic efficiency of system operation and needs to comprehensively consider multiple cost factors. The objective function J1 of the upper-level economic dispatch model is: J1=Cg+Cw+Ct+Caux; Pg=ΔPi+Pi Cw=W×cw; Cg is the cost of power generation; Pg is the actual output of the power generation equipment; ΔPi is the output adjustment of the power generation equipment; Pi is the current output of the power generation equipment; Cw is the penalty cost for wind and solar power abandonment; Ct is the carbon trading cost; Caux is the ancillary service cost; cg is the power generation cost coefficient; W is the amount of wind and solar power abandonment; cw is the penalty coefficient, which is used to punish the wind and solar power abandonment caused by various reasons during the power generation process to encourage the full use of renewable energy; Eg is the carbon emission flow indicator of the power generation equipment; χ is the completion ratio of the low-carbon target; γ is the proportion of renewable energy access; δ is the carbon emission adjustment coefficient on the power generation side; S42. Establish a lower-level low-carbon dispatch model. The lower-level low-carbon dispatch model takes the minimum system low-carbon dispatch cost and load carbon emission responsibility cost as the goal. J2 is: J2=Cl+Ce+Cim; Cim is the implementation cost of low-carbon scheduling measures; Cl is the low-carbon scheduling cost, which is calculated based on the relevant operating costs of low-carbon scheduling; Ce is the load carbon emission responsibility cost, which is calculated based on the load carbon emissions and related cost coefficients.

6. According to a novel robust optimization scheduling method for power system based on node carbon potential response according to claim 1, it is characterized in that: In S5, the steps of constructing a power system source-load coordinated low-carbon optimization dispatch model include: S51. The upper model adjusts the unit output plan and finds a set of values ​​of the output adjustment of the power generation equipment ΔPi, so that the objective function J1 of the upper economic dispatch model is minimized and satisfies: ΔPi≤riΔt; Pmin,i≤Pi+ΔPi≤Pmax,i; And try to reduce the number of unit starts and stops; Among them, ri is the climbing rate; Δt is the time interval; Pmin,i is the minimum stable output; Pmax,i is the maximum stable output; The load curve L1 after the response is passed to the lower model, wherein the load curve L1 is obtained based on the unit output adjustment and the supply and demand balance principle of the power system; when the unit output changes, the load curve will change accordingly according to the supply and demand balance principle of the power system; S52. The lower model guides the flexible load to change the power consumption plan, and finds a set of values ​​of the power consumption plan adjustment ΔUj of the flexible load within the time interval Δt, so that the objective function J2 of the lower low-carbon scheduling model is minimized, and at the same time satisfies: hj(t)ΔUj≤Δt; |ΔUj-Uj|≤ΔUmax,j ΔUmax,j is the maximum adjustment range; flexible load response characteristic function hj(t); kj is a coefficient determined according to the importance; The load curve L2 corresponding to the adjusted flexible load power consumption plan is obtained based on the new power consumption distribution formed in each time period after the flexible load is adjusted; S53. Iterative optimization. In each iteration, the upper model adjusts the unit output plan according to the power generation cost and carbon trading cost, and transmits the response load curve L1 to the lower model; the lower model guides the flexible load to change the power consumption plan according to the node carbon potential and load curve, and feeds back the response load curve L2 to the upper model; During the iteration process, determine whether the convergence condition is met; if it is met and When S≤ε and D≤ε, the iteration ends. Among them, S is the stability index; ε is the comprehensive convergence standard; D is the load satisfaction index; and are the objective function values ​​of the upper and lower models at the nth iteration respectively; and are the objective function values ​​of the upper and lower models at the n-1th iteration, respectively; During the iteration process, the convergence criterion ε is adjusted according to the actual situation.

Images