Peak shaving auxiliary service optimization decision-making method based on fairness index

Abstract

The invention discloses a fairness index-based peak regulation auxiliary service optimization decision-making method, and aims to solve the problem of unfair cost allocation caused by supply and demand mismatch, insufficient flexibility, carbon emission and other multi-dimensional risks in a new energy power system. A new fairness index is provided and is used for quantifying the deviation between the actual allocation amount and the theoretical allocation amount of the participants, and the difference between the system risk and the individual preference is comprehensively considered. A combined weighting method is introduced, a risk weight is determined through an analytic hierarchy process and an anti-entropy weight method, a risk-weighted Shapley value cost sharing model is constructed, and dynamic adjustment of individual risk responsibility is realized. And further establishing a peak regulation auxiliary service scheduling optimization model, and optimizing a peak regulation scheme by taking minimization of the total cost of the system as a target and comprehensively considering system balance, standby and unit scheduling constraint conditions. According to the method, the fairness and scientificity of cost allocation are effectively improved, and important technical support is provided for efficient operation of a new energy high-proportion access power system.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a peak load auxiliary service optimization decision method based on fairness indicators. Background Art

[0002] Driven by the "dual carbon" goals, the power system is accelerating its transformation to a new form dominated by new energy. This change has brought about the challenge of significantly increased uncertainty on the supply and demand sides, which is manifested in multi-dimensional system risks such as supply and demand mismatch, lack of flexibility and carbon emissions. Traditional risk assessment methods mainly target random uncertainties such as component failures and focus on a single load shedding risk. They cannot fully characterize the multi-source risk characteristics of the new energy system, and it is difficult to reasonably allocate risk responsibilities. Therefore, in order to ensure the safety of system operation and the efficiency of resource allocation, new theories and tools are urgently needed.

[0003] The introduction of fairness indicators aims to solve the unfairness problems existing in the current risk assessment and sharing mechanisms. Traditional methods are mostly based on the principles of marginal contribution and equalization, and fail to give sufficient consideration to the actual risk contribution and preferences of individuals. This may lead to low cost sharing for high-risk responsible entities or excessive burden on low-risk responsible entities, which will not only weaken the behavioral incentives of participants, but also make it difficult to reflect the risk efficiency of the overall system.

[0004] The new fairness index quantifies system risks and individual preference differences, and establishes an evaluation method for the comprehensive contribution of individuals to the overall system risk. This indicator can not only evaluate the rationality of risk distribution, but also provide a quantitative basis for system planning and operation optimization. It can be applied to specific scenarios such as auxiliary service optimization and scheduling. By dynamically adjusting the allocation ratio, the system cost allocation can better reflect the actual source of risk and individual differences. It is of great significance in the new energy power system, and can enhance the willingness of various entities to cooperate, optimize resource allocation, and improve the overall system operation efficiency and safety. Summary of the invention

[0005] This paper proposes a fairness index that takes risk factors into account, which aims to comprehensively consider systemic risk and individual risk preferences, evaluate the fairness of distribution by quantifying the deviation between the actual and theoretical allocations of individuals, and fully reflect the individual's contribution to systemic risks such as supply-demand mismatch, carbon emissions and lack of flexibility by combining the conditional value-at-risk (cVaR) model. Based on the risk-weighted Shapley value method, the cost allocation rules are optimized to achieve dynamic adjustment of individual risk responsibilities.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A peak load auxiliary service optimization decision method based on fairness index includes the following steps:

[0008] Define a fairness indicator to measure the deviation between the actual share of participants and the theoretical share;

[0009] Introduce risk weights into fairness indicators and set risk factors to describe the impact of systemic risk and individual risk preferences on the allocation results;

[0010] Based on the risk weights, a risk-weighted cost allocation model based on the original Shapley value method is constructed;

[0011] Optimize the cost sharing model based on fairness indicators and dynamically adjust the sharing rules to improve the fairness of distribution;

[0012] Based on the optimized cost-sharing model, a peak-shaving auxiliary service scheduling optimization model caused by the grid connection of new energy is constructed.

[0013] Further preferably, the fairness index is defined as follows:

[0014]

[0015] In the formula, F i The fairness index of participant i, A i is the actual share of participant i, E i is the theoretical share of participant i calculated based on risk weight.

[0016] Further preferably, the risk weight is expressed as:

[0017] E i =W i ·T

[0018]

[0019] In the formula, E i is the systematic risk of participant i, W i is the risk weight of participant i; T is the total system allocation value; R i Contribution to the risk factor of participant i;

[0020] Systematic risks include supply-demand mismatch risk, carbon emission risk and insufficient flexibility risk, which can be expressed as follows:

[0021] R v =∑ s∈S p s s s,v

[0022]

[0023] R Eu =ΔR u,t ·p{R u,t <Pn,t+1 -P n,t}

[0024] R Ed =ΔR d,t ·p{R d,t <P n,t -P n,t+1}

[0025] In the formula, R v is the risk of supply-demand mismatch; v represents the current operation mode of the system; s represents the system state; S is the system state set; p s is the probability of state s appearing; s s,v Indicates the severity of state s under the current operation mode; R C is the carbon emission risk; C N is the carbon capture capacity of the system; C E is the system carbon emissions; R Eu The flexibility to increase the adjustment is insufficient; R Ed To adjust the flexibility down to the expected lack of flexibility; R u,t is the increase capacity available to the system at time t; R d,t is the down-regulation capacity available to the system at time t; P n,t is the net load of the system at time t; P n,t+1 is the net load of the system at time t+1; ΔR u,t is the change in the system's increased capacity at time t; ΔR d,t is the change in the system's capacity reduction at time t; p{·} represents the probability of the event occurring;

[0026] The individual risk heterogeneity of participants is expressed by individual conditional risk value, which is expressed as:

[0027]

[0028] Where cVaR i is the individual conditional risk value of participant i, η i is the risk threshold of participant i; i is the risk aversion parameter of participant i, χ i ∈[0,1), reflecting the attitude of individual participants towards uncertainty, the larger the value, the more risk-averse they are; u s is the amount of scene loss that exceeds the threshold in state s.

[0029] Further optimization is performed to construct various risk value indicators for different participants for system risk and individual conditional risk values, and the comprehensive risk indicator is determined using the combined weighting method, which is specifically expressed as:

[0030]

[0031] In the formula, R v,i is the supply-demand mismatch risk of participant i; P s,i is the gap of participant i in the supply-demand mismatch; N is the total number of participants, R C,i is the carbon emission risk of participant i; C i is the carbon emissions of participant i; C t is the total carbon emission of the system; R f,i is the risk of insufficient flexibility of participant i; ΔL i is the flexibility response deficit of participant i; ΔL t is the total flexibility response shortfall of the system;

[0032] R i =W 1s R v,i +W 2s R C,i +W 3s R f,i +W 4s QUR i

[0033] W 1s +W 2s +W 3s +W 4s =1

[0034] In the formula, R i is the risk factor contribution of participant i; W 1s , W 2s , W 3s , W 4s They are the risk weights of supply-demand mismatch risk, carbon emission risk, insufficient flexibility risk and individual conditional risk value determined by the combination weighting method.

[0035] Further preferably, the risk-weighted cost allocation model based on the original Shapley value method is constructed based on the risk weight, including: obtaining the risk-weighted marginal cost of the target participant using the risk weight, and correcting the original Shapley value based on the risk-weighted marginal cost, which is specifically expressed as follows:

[0036] h i (C)=W i ·[C(S U )-C(S U -{i})]

[0037]

[0038] In the formula, h i (C) is the risk-weighted marginal cost of participant i, S Uare all alliances that the target participant i participates in; C is the cost function; C(S U ) indicates alliance S U The total cost value to be shared; C(S U -{i}) represents the alliance S U The cost value that should be shared by participants other than i in the calculation; C(S U )-C(S U -{i}) represents the target participant i’s participation in the alliance S U The marginal contribution of i is the risk weight of participant i; N represents the total number of participants; φ i (C) is the risk-weighted peak-shaving cost caused by participant i.

[0039] Further optimization is performed by quantitatively evaluating the cost sharing model based on fairness indicators, and dynamically adjusting the sharing rules to improve the fairness of the allocation, which is specifically expressed as follows:

[0040]

[0041] In the formula, C A The actual allocation amount for the current target participants; is the theoretical share of participant i calculated based on the modified Shapley value; when F i =1, the allocation is completely fair; when F i >1, the actual allocation is higher than the theoretical allocation, and the cost allocation needs to be reduced; when F i When <1, the actual allocation is lower than the theoretical allocation, and the cost allocation needs to be increased; η is the adjustment coefficient, which is used to control the sensitivity of the fairness index to the allocation adjustment; is the dynamically adjusted share of the target participant, and C∑ is the objective function of the peak load auxiliary service scheduling optimization model caused by the grid connection of new energy.

[0042] Further preferably, the peak load ancillary service scheduling optimization model caused by the new energy grid connection takes the lowest total peak load cost of the system as the goal, which is expressed as follows:

[0043] Its objective function is:

[0044]

[0045] In the formula, G c is the collection of thermal power units; Ω D It is a collection of deep peak-shaving gears; is the bid for deep peak load regulation of unit j in segment o; is the peak load regulation amount won by thermal power unit j in the oth deep peak load regulation period t; j,tis a 0-1 variable indicating whether thermal power unit j is turned on in period t, 1 indicates start, otherwise 0; is the startup cost of thermal power unit j in period t; z j,t is a 0-1 variable indicating whether thermal power unit j is shut down in period t, 1 indicates shutdown, otherwise 0; is the shutdown cost of thermal power unit j in period t; V LE Cut costs per unit of energy; is the reduction amount of renewable energy v in period t, G v for a collection of renewable energy sources;

[0046] The constraints are:

[0047] System balance constraints:

[0048]

[0049] Where P j,t is the output of thermal power unit j in period t; P v,t is the output of renewable energy v in period t; D k,t is the load demand of load k in time period t; J is the set of loads;

[0050] System backup constraints:

[0051]

[0052] In the formula, u j,t is a 0-1 variable indicating the operating status of thermal power unit j. The value 1 indicates that thermal power unit j is in the on state during period t, and the value 0 indicates that it is in the off state. and are respectively the positive and negative reserve requirements of the system in time period t; and are the maximum and minimum technical outputs of thermal power unit j respectively;

[0053] Constraints on thermal power units:

[0054]

[0055] y j,t -z j,t =u j,t -u j,t-1 ,j∈G c ,t∈T

[0056] y j,t +z j,t ≤1,j∈G c ,t∈T

[0057]

[0058] In the formula, is the minimum technical output of basic peak load regulation of thermal power unit j in time period t; is the maximum capacity of thermal power unit j in deep peak regulation at stage o; and They are respectively the upward and downward climbing speed constraints of thermal power unit j; and are the minimum continuous operation and shutdown time of thermal power unit j respectively; u j,τ is the operating status of thermal power unit j at time τ;

[0059] Renewable energy output constraints:

[0060]

[0061] In the formula, is the predicted output of renewable energy ν in time period t.

[0062] Beneficial effects of the present invention: The present invention introduces a fairness index that takes into account risk factors, comprehensively considers multi-dimensional system risks such as supply and demand mismatch, carbon emissions and insufficient flexibility, and solves the problem that the traditional sharing mechanism is difficult to fully reflect the multi-source risk characteristics of the new energy power system. The risk weight is combined with the conditional value at risk (cVaR) through a combined weighting method. This method can quantify the actual contribution of individuals to system risks, and dynamically adjust the sharing ratio to reflect individual risk preference differences, significantly improving the scientificity and fairness of the distribution. Based on the improved Shapley value method, the present invention optimizes the cost sharing rules so that the sharing results not only reflect the individual marginal contribution, but also reasonably distribute the system risk responsibility, especially in the cost sharing of peak-shaving auxiliary services caused by the grid connection of new energy, which is highly applicable. The dynamically optimized fairness index ensures that the sharing process is more transparent and reasonable, effectively enhances the willingness of the participating entities to cooperate, and provides important support for the efficient operation and resource allocation of the new energy power system, and has a strong promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0064] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0065] Reference Figure 1 A peak load auxiliary service optimization decision method based on fairness index of the present invention comprises the following steps:

[0066] S1: Define a fairness indicator to measure the deviation between the actual share of participants and the theoretical share;

[0067] S2: Introduce risk weights into fairness indicators and set risk factors to describe the impact of systemic risk and individual risk preferences on the allocation results;

[0068] S3: Construct a risk-weighted cost allocation model based on the original Shapley value method based on risk weights;

[0069] S4: Optimize the cost sharing model based on fairness indicators, and dynamically adjust the sharing rules to improve the fairness of distribution; S5: Based on the optimized cost sharing model, construct a peak-shaving auxiliary service scheduling optimization model caused by the grid connection of new energy.

[0070] Specifically, the fairness index is defined as follows:

[0071]

[0072] In the formula, F i The fairness index of participant i, A i is the actual share of participant i, E i is the theoretical share of participant i calculated based on risk weights, when F i =1, indicating that the allocation is completely fair; when F i >1, it means the apportionment is too high; when F i When <1, it means the apportionment amount is low.

[0073] Specifically, the S2 includes:

[0074] S21: Risk weights are introduced based on fairness indicators and expressed as:

[0075] E i =W i ·T

[0076]

[0077] In the formula, E i is the systematic risk of participant i, W i is the risk weight of participant i; T is the total system allocation value; R i Contribution to the risk factor of participant i;

[0078] S22: The calculation of risk factors takes into account systemic risk factors and personal risk preferences, mainly including:

[0079] The risk of supply-demand mismatch is expressed as: R v =∑ s∈S p s s s,v

[0080] In the formula, R v is the risk of supply-demand mismatch; v represents the current operation mode of the system; s represents the system state; S is the system state set; p s is the probability of state s appearing; s s,v Indicates the severity of state s under the current operation mode;

[0081] Carbon emission risk is expressed as:

[0082] In the formula, R C is the carbon emission risk; C N is the carbon capture capacity of the system; C E is the system carbon emissions, R C The higher the value, the lower the carbon emission of the system to the environment. C The lower it is, the higher the system's carbon footprint to the environment.

[0083] The risk of insufficient flexibility is expressed as:

[0084] R Eu =ΔR u,t ·p{R u,t <P n,t+1 -P n,t}

[0085] R Ed =ΔR d,t ·p{R d,t <P n,t -P n,t+1}

[0086] In the formula, R Eu The flexibility to increase the adjustment is insufficient; R Ed To adjust the flexibility down to the expected lack of flexibility; R u,t is the increase capacity available to the system at time t; R d,t is the down-regulation capacity available to the system at time t; P n,t is the net load of the system at time t; P n,t+1 is the net load of the system at time t+1; ΔR u,t is the change in the system's increased capacity at time t; ΔR d,t is the change in the system's capacity reduction at time t; p{·} represents the probability of the event occurring.

[0087] The individual risk heterogeneity of participants is expressed by individual conditional risk value, which is expressed as:

[0088]

[0089] Where cVaRi is the individual conditional risk value of participant i, η i is the risk threshold of participant i; i is the risk aversion parameter of participant i, χ i ∈[0,1), reflecting the attitude of individual participants towards uncertainty, the larger the value, the more risk-averse they are; u s is the amount of scene loss that exceeds the threshold in state s.

[0090] S23: Based on systemic risk and individual behavior differences, construct risk value indicators for different participants:

[0091] The supply-demand mismatch risk of participant i is expressed as:

[0092]

[0093] In the formula, R v,i is the supply-demand mismatch risk of participant i; P s,i is the gap of participant i in the supply and demand mismatch; N is the total number of participants.

[0094] The carbon emission risk of participant i is expressed as:

[0095]

[0096] In the formula, R C,i is the carbon emission risk of participant i; C i is the carbon emissions of participant i; C t is the total carbon emissions of the system.

[0097] The risk of insufficient flexibility of participant i is expressed as:

[0098]

[0099] In the formula, R f,i is the risk of insufficient flexibility of participant i; ΔL i is the flexibility response deficit of participant i; ΔL t It is the total flexibility response shortfall of the system.

[0100] S24: Risk factor contribution of participant i, expressed as:

[0101] R i =W 1s R v,i +W 2s R C,i +W 3s R f,i +W 4s QUR i

[0102] W 1s +W 2s +W 3s +W 4s =1

[0103] In the formula, R i is the risk factor contribution of participant i; W 1s , W 2s , W 3s , W 4s They are the risk weights of supply-demand mismatch risk, carbon emission risk, insufficient flexibility risk and individual conditional risk value determined by the combination weighting method.

[0104] The weight coefficient is determined by the combined weighting method based on the hierarchical analysis method and the anti-entropy weight method. The evaluation process is as follows:

[0105] 1) Determine the subjective weight w by analytic hierarchy process i

[0106] Construct comparison judgment matrix R = (a ij ) m×n .a ij The elements of the matrix

[0107] Calculate the attribute judgment matrix, normalize the comparison judgment matrix, and convert it into a weight matrix so that the sum of all elements is 1. The calculation formula is expressed as:

[0108]

[0109] Where K is an integer greater than or equal to 2, b ij is the normalized element;

[0110] Determine the subjective weight value, expressed as:

[0111]

[0112] Where W 1(i) is the subjective weight value of the ith indicator.

[0113] 2) Anti-entropy weight method to determine objective weights

[0114] Data normalization, let:

[0115]

[0116] In the formula, v ij is the original value of market entity j on the ith indicator; r ij is the standardized value of market entity j on the ith indicator.

[0117] Calculate the anti-entropy, expressed as:

[0118]

[0119] Among them, u i is anti-entropy, n is the number of market entities; f ij is the weight of market entity j on the ith indicator;

[0120] Determine the objective weight value, expressed as:

[0121]

[0122] Where W 2i is the objective weight value of the ith indicator, and m is the number of indicators.

[0123] 3) Determine the combined weight of indicators using the variance minimization principle:

[0124]

[0125] Among them, μ k is the anti-entropy of the kth index, k = 1 or 2, W k is the objective weight value prepared for the kth time; W k The transpose of

[0126] It is transformed into a system of linear equations by differential properties and first-order derivative conditions:

[0127]

[0128] Solving the equation can obtain μ1, μ2, and further obtain the normalized combination weight:

[0129]

[0130] Among them, W is is the normalized combined weight.

[0131] Further, the S3 includes:

[0132] S31: Based on the total risk-weighted cost borne by all participants N and the total risk-weighted cost borne by all participants except the target participant, the risk-weighted marginal cost of the target participant is obtained, which is expressed as:

[0133] h i (C)=W i ·[C(S U )-C(S U -{i})]

[0134] In the formula, h i(C) is the risk-weighted marginal cost of participant i, S U are all alliances that the target participant i participates in; C is the cost function; C(S U ) indicates alliance S U The total cost value to be shared; C(S U -{i}) represents the alliance S U The cost value that should be shared by participants other than i in the calculation; C(S U )-C(S U -{i}) represents the target participant i’s participation in the alliance S U The marginal contribution of i is the risk weight of participant i.

[0135] S32: Based on the risk-weighted marginal cost of the target participant, we obtain: The cost-sharing Shapley value modified by the risk factor is expressed as

[0136]

[0137] Where N is the total number of participants; φ i (C) is the risk-weighted peak-shaving cost caused by participant i.

[0138] S33: Construct a risk-weighted cost allocation model based on the Shapley value model modified by risk factors.

[0139] The cost sharing amount of the target participant is expressed as:

[0140]

[0141] C∑ is the objective function of the peak load ancillary service scheduling optimization model caused by the grid connection of new energy.

[0142] Further, the S4 includes:

[0143] S41: Based on the above fairness indicators, the cost sharing model is quantitatively evaluated and expressed as follows:

[0144]

[0145] In the formula, C A The actual allocation amount for the current target participants; is the theoretical allocation based on the modified Shapley value. i =1, the allocation is completely fair; when F i >1, the actual allocation is higher than the theoretical allocation, and the cost allocation needs to be reduced; when F i When <1, the actual apportionment amount is lower than the theoretical apportionment amount, and the cost apportionment amount needs to be increased.

[0146] S42: Evaluate the cost sharing model based on fairness indicators and dynamically adjust the sharing rules to improve the fairness of distribution, which can be expressed as:

[0147]

[0148] In the formula, η is the adjustment coefficient, which is used to control the sensitivity of the fairness index to the allocation adjustment; The dynamically adjusted share amount for the target participant.

[0149] Further, the S5 includes:

[0150]

[0151] In the formula, G c is the collection of thermal power units; Ω D It is a collection of deep peak-shaving gears; is the bid for deep peak load regulation of unit j in segment o; is the peak load regulation amount won by thermal power unit j in the oth deep peak load regulation period t; j,t is a 0-1 variable indicating whether thermal power unit j is turned on in period t, 1 indicates start, otherwise 0; is the startup cost of thermal power unit j in period t; z j,t is a 0-1 variable indicating whether thermal power unit j is shut down in period t, 1 indicates shutdown, otherwise 0; is the shutdown cost of thermal power unit j in period t; V LE Cut costs per unit of energy; is the reduction amount of renewable energy v in period t, G v for the collection of renewable energy sources;

[0152] The constraints are:

[0153] System balance constraints:

[0154]

[0155] Where P j,t is the output of thermal power unit j in period t; P ν,t is the output of renewable energy ν in period t; D k,t is the load demand of load k in time period t; K is the set of loads

[0156] System backup constraints:

[0157]

[0158] In the formula, u j,tis a 0-1 variable of the operating state of thermal power unit j, where 1 indicates that thermal power unit j is in the on state during period t, and 0 indicates that it is in the off state; and are respectively the positive and negative reserve requirements of the system in time period t; and are the maximum and minimum technical outputs of thermal power unit j respectively;

[0159] Constraints on thermal power units:

[0160]

[0161] y j,t -z j,t =u j,t -u j,t-1 ,j∈G c ,t∈T

[0162] y j,t +z j,t ≤1,j∈G c ,t∈T

[0163]

[0164] In the formula, is the minimum technical output of basic peak load regulation of thermal power unit j in time period t; is the maximum capacity of thermal power unit j in deep peak regulation at stage o; and They are respectively the upward and downward climbing speed constraints of thermal power unit j; and are the minimum continuous operation and shutdown time of thermal power unit j respectively; u j,τ is the operating status of thermal power unit j at time τ;

[0165] Renewable energy output constraints:

[0166]

[0167] In the formula, is the predicted output of renewable energy ν in time period t.

[0168] Furthermore, based on the cost sharing model described in S4, the peak load auxiliary service costs caused by the introduction of new energy units are shared, and based on the fairness index described in S1, the sharing results are quantitatively evaluated and dynamically adjusted to achieve fairness.

[0169] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A peak load auxiliary service optimization decision method based on fairness index, characterized in that: The following steps are involved: Define a fairness indicator to measure the deviation between the actual share of participants and the theoretical share; Introduce risk weights into fairness indicators and set risk factors to describe the impact of systemic risk and individual risk preferences on the allocation results; Based on the risk weights, a risk-weighted cost allocation model based on the original Shapley value method is constructed; Optimize the cost sharing model based on fairness indicators and dynamically adjust the sharing rules to improve the fairness of distribution; Based on the optimized cost-sharing model, a peak-shaving auxiliary service scheduling optimization model caused by the grid connection of new energy is constructed.

2. The peak load auxiliary service optimization decision method based on fairness index according to claim 1 is characterized in that: The fairness index is defined as follows: In the formula, F i The fairness index of participant i, A i is the actual share of participant i, E i is the theoretical share of participant i calculated based on risk weight.

3. The peak load auxiliary service optimization decision method based on fairness index according to claim 1 is characterized in that: The risk weight is expressed as: E i =W i ·T In the formula, E i is the systematic risk of participant i, W i is the risk weight of participant i; T is the total allocation value of the system; R i Contribution to the risk factor of participant i; Systematic risks include supply-demand mismatch risk, carbon emission risk and insufficient flexibility risk, which can be expressed as: R v =∑ s∈S p s s s,v R Eu =ΔR u,t ·p{R u,t <P n,t+1 -P n,t } R Ed =ΔR d,t ·p{R d,t <P n,t -P n,t+1 } In the formula, R v is the risk of supply-demand mismatch; v represents the current operation mode of the system; s represents the system state; S is the system state set; p s is the probability of state s appearing; s s,v Indicates the severity of state s under the current operation mode; R C is the carbon emission risk; C N is the carbon capture capacity of the system; C E is the system carbon emissions; R Eu The flexibility to increase the adjustment is insufficient; R Ed To adjust the flexibility down to the expected lack of flexibility; R u,t is the increase capacity available to the system at time t; R d,t is the down-regulation capacity available to the system at time t; P n,t is the net load of the system at time t; P n,t+1 is the net load of the system at time t+1; ΔR u,t is the change in the system's increased capacity at time t; ΔR d,t is the change in the system's capacity reduction at time t; p{·} represents the probability of an event occurring; The individual risk heterogeneity of participants is expressed by individual conditional risk value, which is expressed as: Where cVaR i is the individual conditional risk value of participant i, η i is the risk threshold of participant i; χ i is the risk aversion parameter of participant i, χ i ∈[0,1), reflecting the attitude of individual participants towards uncertainty, the larger the value, the more risk-averse they are; u s is the amount of scene loss that exceeds the threshold in state s.

4. The peak load auxiliary service optimization decision method based on fairness index according to claim 3 is characterized in that: For systemic risk and individual conditional risk values, various risk value indicators for different participants are constructed, and the comprehensive risk indicator is determined using the combined weighting method, which is specifically expressed as: In the formula, R v,i is the supply-demand mismatch risk of participant i; P s,i is the gap of participant i in the supply-demand mismatch; N is the total number of participants, R C,i is the carbon emission risk of participant i; C i is the carbon emissions of participant i; C t is the total carbon emission of the system; R f,i is the risk of insufficient flexibility of participant i; ΔL i is the flexibility response deficit of participant i; ΔL t is the total flexibility response shortfall of the system; R i =W 1s R v,i +W 2s R C,i +W 3s R f,i +W 4s cVaR i IN 1s +W 2s +W 3s +W 4s =1 In the formula, R i Contribution to the risk factor of participant i; W 1s , W 2s , W 3s , W 4s They are the risk weights of supply-demand mismatch risk, carbon emission risk, insufficient flexibility risk and individual conditional risk value determined by the combination weighting method.

5. The peak load auxiliary service optimization decision method based on fairness index according to claim 1 is characterized in that: The risk-weighted cost allocation model based on the original Shapley value method is constructed based on the risk weight, including: obtaining the risk-weighted marginal cost of the target participant by using the risk weight, and correcting the original Shapley value based on the risk-weighted marginal cost, which is specifically expressed as follows: h i (C)=W i ·[C(S U )-C(S U -{i})] In the formula, h i (C) is the risk-weighted marginal cost of participant i, S U are all alliances that the target participant i participates in; C is the cost function; C(S U ) indicates alliance S U The total cost value to be shared; C(S U -{i}) represents the alliance S U The cost value that should be shared by participants other than i in the calculation; C(S U )-C(S U -{i}) represents the target participant i’s participation in the alliance S U The marginal contribution of i is the risk weight of participant i; N represents the total number of participants; φ i (C) is the risk-weighted peak-shaving cost caused by participant i.

6. The peak load auxiliary service optimization decision method based on fairness index according to claim 1 is characterized in that: The cost sharing model is quantitatively evaluated based on fairness indicators, and the sharing rules are dynamically adjusted to improve the fairness of distribution. The specific expression is: In the formula, C A The actual allocation amount for the current target participants; is the theoretical share of participant i calculated based on the modified Shapley value; when F i =1, the allocation is completely fair; when F i >1, the actual allocation is higher than the theoretical allocation, and the cost allocation needs to be reduced; when F i When <1, the actual allocation is lower than the theoretical allocation, and the cost allocation needs to be increased; η is the adjustment coefficient, which is used to control the sensitivity of the fairness index to the allocation adjustment; is the dynamically adjusted share of the target participant, and C∑ is the objective function of the peak load auxiliary service scheduling optimization model caused by the grid connection of new energy.

7. The peak load auxiliary service optimization decision method based on fairness index according to claim 1, characterized in that: The peak load ancillary service scheduling optimization model caused by the new energy grid connection aims to minimize the total peak load cost of the system, which is expressed as follows: Its objective function is: In the formula, G c is the collection of thermal power units; Ω D It is a collection of deep peak-shaving gears; is the bid for deep peak load regulation of unit j in segment o; is the peak load regulation amount won by thermal power unit j in the oth deep peak load regulation period t; j,t is a 0-1 variable indicating whether thermal power unit j is turned on in period t, 1 indicates start, otherwise 0; is the startup cost of thermal power unit j in period t; z j,t is a 0-1 variable indicating whether thermal power unit j is shut down in period t, 1 indicates shutdown, otherwise 0; is the shutdown cost of thermal power unit j in period t; V LE Cut costs per unit of energy; is the reduction amount of renewable energy v in period t, G v for the collection of renewable energy sources; The constraints are: System balance constraints: Where P j,t is the output of thermal power unit j in period t; P ν,t is the output of renewable energy ν in period t; D k,t is the load demand of load k in time period t; J is the set of loads; System backup constraints: In the formula, u j,t is a 0-1 variable indicating the operating status of thermal power unit j. The value 1 indicates that thermal power unit j is in the on state during period t, and the value 0 indicates that it is in the off state. and are respectively the positive and negative reserve requirements of the system in time period t; and are the maximum and minimum technical outputs of thermal power unit j respectively; Constraints on thermal power units: yes j,t -z j,t =u j,t -u j,t-1 ,j∈G c ,t∈T y j,t +z j,t ≤1,j∈G c ,t∈T In the formula, is the minimum technical output of basic peak load regulation of thermal power unit j in time period t; is the maximum capacity of thermal power unit j in deep peak regulation at stage o; and They are respectively the upward and downward climbing speed constraints of thermal power unit j; and are the minimum continuous operation and shutdown time of thermal power unit j respectively; u j,τ is the operating status of thermal power unit j at time τ; Renewable energy output constraints: In the formula, is the predicted output of renewable energy ν in time period t.

Images