Two-stage power supply centralized optimization configuration modeling method considering uncertainty and inertia

Abstract

The invention relates to a two-stage power supply centralized optimization configuration modeling method considering uncertainty and inertia, and the method comprises the following steps: S1, constructing a two-stage optimization objective function by taking the sum of the minimum total investment cost and the operation cost and the simultaneous subtraction of the maximum new energy consumption capability as a comprehensive objective; s2, executing first-stage investment-pre-scheduling joint optimization, and establishing a first-stage constraint; s3, implementing second-stage real-time rescheduling optimization, and establishing a second-stage constraint; s4, performing collaborative optimization on the first-stage constraint and the second-stage constraint to obtain a general mathematical model; and S5, solving the general mathematical model to obtain a final optimization model. According to the method, the power supply investment cost and the new energy power generation consumption capacity are jointly optimized, the new energy consumption capacity under a certain power supply structure can be quantified, and clear guidance is provided for new energy power generation investment.

Description

Technical Field

[0001] The present invention relates to the technical field of centralized optimal configuration of power system power sources, and particularly relates to a two-stage modeling method for centralized optimal configuration of power sources considering uncertainty and inertia. Background Art

[0002] With the improvement of new energy penetration rate and the promotion of carbon neutrality goals, the power system faces dual challenges of low-inertia stability and new energy uncertainty: on the one hand, new energy units are decoupled through power electronic devices, resulting in the lack of system rotational inertia. Traditional power source planning models ignore the potential of virtual inertia support of wind power and are difficult to cope with the risk of frequency instability; on the other hand, existing uncertainty modeling methods have inherent defects - stochastic optimization relies on accurate probability distributions while actual data is scarce, robust optimization is overly conservative and damages economic efficiency, and the two-stage coupling model is forced to simplify key elements such as unit start-stop and network constraints due to the curse of dimensionality, resulting in the deviation of the planning scheme from actual requirements. Current research mostly focuses on inertia regulation at the operation level, but does not incorporate inertia adequacy constraints into long-term power source investment decisions. At the same time, there is a lack of an effective method to directly quantify the dynamic relationship between the power source structure and the new energy consumption capacity. Existing curtailment penalty or risk constraint mechanisms rely on repetitive sensitivity analysis and are difficult to achieve the coordinated optimization of safety, economy, and consumption capacity, restricting the low-carbon transformation process of the new power system. Therefore, it is very necessary to design a new power source configuration model to overcome the above problems.

[0003] It can be understood that the above statements only provide background art related to the present invention and do not necessarily constitute prior art. Summary of the Invention

[0004] The purpose of the present invention is to provide a two-stage modeling method for centralized optimal configuration of power sources considering uncertainty and inertia, which jointly optimizes the power source investment cost and the new energy consumption capacity, can quantify the new energy consumption capacity under a certain power source structure, and provides clear guidance for new energy power generation investment.

[0005] To achieve the above purpose, the present invention provides a two-stage modeling method for centralized optimal configuration of power sources considering uncertainty and inertia, specifically including the following steps:

[0006] S1. Construct a two-stage optimization objective function with the comprehensive goal of minimizing the sum of the total investment cost and the operation cost, while subtracting the maximized new energy consumption capacity;

[0007] S2. Execute the first-stage investment-pre-scheduling joint optimization and establish the first-stage constraints;

[0008] S3. Implement the second-stage real-time re-scheduling optimization and establish the second-stage constraints;

[0009] S4. Co - optimize the constraints in the first stage and the second stage to obtain a general mathematical model;

[0010] S5. Solve the above - mentioned general mathematical model to obtain the final optimization model.

[0011] Optionally, in S1, the calculation formula of the objective function is specifically:

[0012]

[0013] where min represents the minimization function; g represents a thermal power unit; G S represents the set of planned thermal power units to be built, G E represents the set of existing thermal power units; w represents wind power; W represents the set of planned wind power to be built; t is the time period; T represents the set of time periods; and respectively represent the annualized investment costs of thermal power unit g and wind power w; δ represents the weight coefficient; the function represents the operating cost function of thermal power unit g; represents the installed capacity of the planned thermal power unit to be built; represents the installed capacity of the planned wind power w to be built; and respectively represent the pre - dispatch power outputs of the planned and existing thermal power units at time period t based on the wind power prediction value; a represents the probability that the system completely absorbs the wind power output; represents the investment cost of the thermal power unit; represents the investment cost of the wind power; represents the operating costs of wind power and thermal power; δa represents the probability of maximizing the system's absorption of new energy under the preset weight coefficient.

[0014] Optionally, in S2, the first - stage constraints include: power investment constraint, minimum green power ratio constraint, first - stage power balance constraint, first - stage line power flow constraint, first - stage spinning reserve constraint, first - stage dispatch power output constraint and its coupling constraint with spinning reserve, first - stage thermal power ramp - up constraint, and inertia - frequency change rate constraint.

[0015] Optionally, the power investment constraint is shown as the following formula and includes:

[0016]

[0017]

[0018] where β g represents whether to build a new thermal power unit. When β g = 1, it means construction; when β g = 0, it means non - construction, G SDenote the set of thermal power units to be built; w represents wind power, and W represents the set of wind power to be built. Denote the installed capacity to be built for wind power w. Denote the upper bound of the newly built capacity for wind power w. Denote the existing installed capacity of wind power w. Denote the total installed capacity of wind power w. Is the maximum output of the thermal power units to be built.

[0019] The green power ratio constraint is shown as the following formula:

[0020]

[0021] Among them, the set N represents the node set, ΔT is the scheduling interval; γ is the minimum requirement for the ratio of new energy power generation to total power consumption; d tn Is the power load demand at node n in time period t. Is the predicted output of wind power w in time period t.

[0022] The first-stage power balance constraint is shown as the following formula:

[0023]

[0024] The line power flow constraint is shown as the following formula:

[0025]

[0026] Among them, Denote the transmission capacity of line l, A nl Denote the power transfer distribution factor of node n with respect to line l, G S (n) denotes the set of thermal power units to be built at node n, G E (n) denotes the set of existing thermal power units at node n, W(n) denotes the set of wind power to be built at node n, and L denotes the line set

[0027] The first-stage spinning reserve constraint is shown as the following formula, including:

[0028]

[0029] Among them, Denote the system spinning reserve demand in time period t; And Respectively denote the 10-minute spinning reserve coefficients of the thermal power units to be built and existing ones; And Respectively denote the spinning reserves of the thermal power units to be built and existing ones in the pre-scheduling time period t; Denote the maximum output of the existing thermal power unit g;

[0030] The first-stage scheduling output constraint and its coupling constraint with the spinning reserve are shown as follows, including:

[0031]

[0032] Among them, is the typical output coefficient of wind power w at time t, represents the total installed capacity of wind power w; represents the maximum output of the existing thermal power units;

[0033] The first-stage thermal power ramp-up constraint is shown as follows, including:

[0034]

[0035] Among them, and respectively represent the downward and upward ramp coefficients of the to-be-built thermal power units; and respectively represent the downward and upward ramp coefficients of the existing thermal power units;

[0036] The inertia-frequency change rate constraint is shown as follows:

[0037]

[0038] Among them and respectively represent the inertia of the to-be-built thermal power units, wind power w and existing thermal power units, represents the upper limit of the frequency change rate, f0 represents the reference frequency of the system, ΔP Sys represents the unbalanced power of the system, H Sys represents the total inertia of the system.

[0039] Optionally, in S3, the second stage is the rescheduling stage, which is a scheduling process for coping with the wind power prediction error. Its constraint is the operation constraint after the wind power uncertainty appears and involves the random variable ξ tw , representing the prediction error of the output of wind power at time t.

[0040] Among them, the second-stage constraints include: operation state constraints, new energy consumption range constraints, and distributionally robust joint chance constraints; the operation state constraints include: second-stage power balance constraints, second-stage line power flow constraints, second-stage spinning reserve constraints, second-stage scheduling output constraints and their coupling constraints with the spinning reserve, second-stage thermal power ramp-up constraints and rescheduling ramp-up constraints.

[0041] Optionally, the second-stage power balance constraint is shown as follows:

[0042]

[0043] Among them, and respectively represent the actual dispatching outputs of the thermal power units to be built and existing during period t; the random variable ξ tw , represents the prediction error of the output of wind power w during period t; the random variable ξ t , represents the prediction error of all wind power during period t.

[0044] The second-stage line power flow constraint is shown as follows:

[0045]

[0046] The second-stage spinning reserve constraint is shown as follows, including:

[0047]

[0048] Among them, and respectively represent the spinning reserves of the thermal power units to be built and existing during period t of the rescheduling stage;

[0049] The second-stage dispatching output constraint and its coupling constraint with the spinning reserve are shown as follows, including:

[0050]

[0051] Among them, and respectively represent the spinning reserves of the thermal power unit g to be built and existing during period t of the rescheduling stage;

[0052] The second-stage thermal power ramp constraint is shown as follows, including:

[0053]

[0054] The rescheduling ramp constraint is shown as follows, including:

[0055]

[0056] Among them, and respectively represent the downward and upward ramp coefficients of the thermal power units to be built during the rescheduling stage; and respectively represent the downward and upward ramp coefficients of the existing thermal power units during the rescheduling stage.

[0057] Optionally, the new energy consumption range constraint is shown as follows:

[0058]

[0059] Among them, and respectively represent the upper and lower bounds of the wind power w consumption range in time period t.

[0060] Optionally, the establishment of the distributionally robust joint chance constraint specifically includes the following steps:

[0061] J1. Establish a fuzzy set

[0062]

[0063] Among them, represents the probability distribution of the prediction error of wind power based on historical data; ξ tw represents the prediction error of the output of the wind turbine; μ tw represents the mean value of the prediction error of the output of the wind turbine; represents the variance of the prediction error of the output of the wind turbine; represents the expectation of the probability distribution ;

[0064] J2. Model the wind power consumption capacity and quantify the probability that the wind power is consumed by the system, as shown in the following formula:

[0065]

[0066] a0 ≤ a ≤ 1 (36)

[0067] Among them, a represents the minimum value of the probability that the prediction error of the wind power falls within the system's consumable range;

[0068] represents the upper and lower bounds of the prediction error of all wind power in time period t; a0 represents the lower bound constraint of the minimum value a of the prediction error of the wind power falling within the system's consumable range.

[0069] Optionally, in the above-mentioned S4, the general mathematical model established is:

[0070]

[0071] x ∈ X (38)

[0072]

[0073] Qξ L ≤ q, Rξ U ≤ r (41)

[0074] Among them, Equation (37) represents the objective function, Equation (38) represents the first-stage constraint, Equation (39) represents the operating state constraint in the second-stage constraint, Equation (40) represents the distributionally robust joint chance constraint in the second-stage constraint, and Equation (41) represents the constraint on the upper and lower limits of the accommodation range; the variable x represents the decision variable in the first-stage constraint; X represents the set of x; the vector K represents the coefficient in the objective function; T(x) represents the set of functions of the variable x, W represents the set of parameters related to the function of the random variable ξ, and J represents the set of parameters related to the random variable ξ. ξ L and ξ U respectively represent the sets of the lower and upper limits of the random variable ξ; a represents the wind power accommodation capacity; δ represents the weight coefficient; the vector y(ξ) represents the operating state decision variable in the second-stage constraint; Q and q represent the parameters related to the lower limit of the random variable ξ; R and r represent the parameters related to the upper limit of the random variable ξ.

[0075] Optionally, in S5, the general mathematical model is solved according to the affine decision rule, Bonferroni approximation, Gaussian inequality, and second-order cone relaxation to obtain the final optimization model:

[0076] Objective function:

[0077]

[0078] Second-stage power balance constraint:

[0079]

[0080] Second-stage line power flow constraint:

[0081]

[0082] Second-stage spinning reserve constraint:

[0083]

[0084]

[0085] Second-stage dispatch output constraint and its coupling constraint with spinning reserve:

[0086]

[0087] Second-stage thermal power ramp constraint:

[0088]

[0089] Rescheduling ramp constraint:

[0090]

[0091] Auxiliary variable-related constraints, including: E tw and and

[0092]

[0093] And other constraints remain unchanged.

[0094] In summary, compared with the prior art, a two-stage centralized optimization configuration modeling method for power sources considering uncertainty and inertia provided by the present invention has at least the following beneficial effects:

[0095] (1) The present invention simultaneously considers the requirements of frequency stability and reliability. In terms of frequency stability, the system inertia is modeled, including traditional inertia and virtual inertia, as well as corresponding frequency constraints; in terms of reliability, uncertainty is described by a fuzzy set, and the accommodation capacity of new energy is modeled in the form of a distributionally robust joint chance constraint.

[0096] (2) The model proposed by the present invention jointly optimizes the power source investment cost and the accommodation capacity of new energy power generation, can quantify the accommodation capacity of new energy under a certain power source structure, and provides clear guidance for new energy power generation investment.

[0097] (3) The general mathematical model provided by the present invention is difficult to solve, so affine decision rules, second-order cone reconstruction, and dual techniques are adopted to obtain an approximately equivalent model.

[0098] (4) Through experimental verification, the present invention obtains the effectiveness and scalability of the proposed method, and proves the importance of considering inertia response requirements and virtual inertia in the power source investment problem. Description of the Drawings

[0099] Figure 1 is the overall model framework diagram of the two-stage centralized optimization configuration modeling method for power sources considering uncertainty and inertia of the present invention;

[0100] Figure 2 is the topological diagram of the 9-node system and existing and to-be-built thermal power units in an embodiment of the present invention;

[0101] Figure 3 is the schematic diagram of the trade-off relationship between investment cost and accommodation range in the 9-node system in an embodiment of the present invention;

[0102] Figure 4 is the relationship between cost and accommodation probability under different d' and γ in the 9-node system in an embodiment of the present invention;

[0103] Figure 5 Schematic diagram of the system inertia level considering and not considering inertia constraints in an embodiment of the present invention;

[0104] Figure 6 Schematic diagram of the new energy accommodation range obtained by the deterministic model and the proposed model in an embodiment of the present invention under γ = 45%, γ = 55%, and γ = 65% respectively. Detailed implementation manners

[0105] The following further elaborates on the present invention by Figure 1 - Figure 6 detailed description of a preferred specific embodiment in conjunction with the attached

[0106] It should be noted that the attached drawings are in a very simplified form and use non-precise scales, only for conveniently and clearly assisting in explaining the purpose of the embodiment mode of the present invention, and are not used to limit the limiting conditions for the implementation of the present invention. Therefore, they do not have technical substantial significance. Any modification of the structure, change of the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope covered by the technical content disclosed by the present invention.

[0107] It should be noted that in the present invention, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or sequence between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements clearly listed, but also includes other elements not clearly listed, or further includes elements inherent to such process, method, article or device.

[0108] As Figure 1 shown, the present invention provides a two-stage power centralized optimization configuration modeling method considering uncertainty and inertia. Through two-stage collaborative decision-making of investment-pre-scheduling joint optimization and real-time re-scheduling optimization, the new energy accommodation capacity is maximized, which specifically includes the following steps:

[0109] S1. Construct a two-stage optimization objective function with the comprehensive goal of minimizing the sum of the total investment cost and the operating cost, and at the same time subtracting the maximized new energy accommodation capacity;

[0110] S2. Execute the first-stage investment-pre-scheduling joint optimization and establish the first-stage constraints;

[0111] S3. Implement the second-stage real-time re-scheduling optimization and establish the second-stage constraints;

[0112] S4. Co - optimize the first - stage constraints and the second - stage constraints to obtain a general mathematical model;

[0113] S5. Solve the above - mentioned general mathematical model to obtain the final optimization model.

[0114] Further, in the above - mentioned S1, the calculation formula of the objective function is specifically:

[0115]

[0116] where, min represents the minimization function; g represents a thermal power unit; G S represents the set of thermal power units to be built, G E represents the set of existing thermal power units; w represents wind power; W represents the set of wind power to be built; t is the time period; T represents the set of time periods; and respectively represent the annualized investment cost of thermal power unit g and wind power w (unit: $ / MW - year); δ represents the weight coefficient; the function represents the operating cost function of thermal power unit g; represents the installed capacity of the thermal power unit to be built (unit: MW); represents the installed capacity of the wind power w to be built (unit: MW); and respectively represent the pre - dispatch output of the thermal power units to be built and existing ones at time period t based on the wind power prediction value (unit: MW); a represents the probability that the system completely absorbs the wind power output; represents the investment cost of the thermal power unit; represents the investment cost of the wind power; represents the operating cost of the wind power and thermal power; δa represents the probability of maximizing the absorption of new energy under the preset weight coefficient.

[0117] Further, in the above - mentioned S2, the first stage is the pre - dispatch stage, and the first - stage constraints established in this stage include: power investment constraints, minimum green power ratio constraints, first - stage power balance constraints, first - stage line power flow constraints, first - stage spinning reserve constraints, first - stage dispatch output constraints and their coupling constraints with spinning reserve, first - stage thermal power ramp - up constraints and inertia - frequency change rate constraints.

[0118] Specifically, the power investment constraints are shown as follows and include:

[0119]

[0120] where, equation (2) represents whether to build a new thermal power unit; β g represents whether to build a new thermal power unit. When β g = 1, it means construction, and when β gWhen it is 0, it means no construction.

[0121] Among them, Equation (3) represents the new capacity limit of wind power, and Equation (4) represents the total capacity of wind power; represents the upper bound of the new capacity of wind power w (unit: MW), represents the existing installed capacity of wind power w (unit: MW), represents the total installed capacity of wind power w (unit: MW).

[0122] Among them, Equation (5) represents the installed capacity of the to-be-built thermal power unit; is the maximum output of the to-be-built thermal power unit (unit: MW).

[0123] Specifically, the minimum green power ratio constraint is shown as follows:

[0124]

[0125] Among them, Equation (6) represents the share requirement of new energy power generation in the total power generation; the set N represents the node set, ΔT is the scheduling interval; γ is the minimum requirement for the ratio of new energy power generation in the total power consumption; d tn is the power load demand of node n at time t (unit: MW); is the predicted output of wind power w at time t (unit: MW).

[0126] Specifically, the first-stage power balance constraint is shown as follows:

[0127]

[0128] Among them, Equation (7) represents the system power balance constraint for the predicted power value of wind power in the pre-scheduling stage.

[0129] Specifically, the first-stage line power flow constraint is shown as follows:

[0130]

[0131] Among them, represents the transmission capacity of line l (unit: MW), A nl represents the power transfer distribution factor of node n with respect to line l, G S (n) represents the set of to-be-built thermal power units at node n, G E (n) represents the set of existing thermal power units at node n, W(n) represents the set of to-be-built wind power at node n, and L represents the set of lines.

[0132] Specifically, the first-stage spinning reserve constraint is shown as follows, including:

[0133]

[0134] Among them, Equation (9) represents the system spinning reserve requirement constraint in the pre-scheduling stage, and Equations (10) and (11) respectively represent the spinning reserve ranges of the to-be-built and existing thermal power units in the pre-scheduling stage; among them, represents the system spinning reserve requirement at time period t; and respectively represent the 10-minute spinning reserve coefficients of the to-be-built and existing thermal power units; and respectively represent the spinning reserves of the to-be-built and existing thermal power units at time period t in the pre-scheduling stage (unit: MW); represents the maximum output of the existing thermal power unit g (unit: MW).

[0135] Specifically, the first-stage dispatching output constraint and its coupling constraint with the spinning reserve are shown as follows, including:

[0136]

[0137] Among them, Equation (12) represents the predicted output of wind power in the pre-scheduling stage, that is, the typical output coefficient multiplied by the capacity of wind power; is the typical output coefficient of wind power w at time period t, equal to the ratio of its typical output to the rated capacity, represents the total installed capacity of wind power w.

[0138] Among them, Equations (13) and (14) respectively represent the upper limits of the output and the provision of spinning reserve of the to-be-built and existing thermal power units in the pre-scheduling stage, and Equations (15) and (16) respectively represent that the dispatching outputs of the to-be-built and existing thermal power units in the pre-scheduling stage are all non-negative; represents the maximum output of the existing thermal power unit (unit: MW).

[0139] Specifically, the first-stage thermal power ramp-up constraint is shown as follows, including:

[0140]

[0141] Among them, Equations (17) and (18) respectively represent the ramp-up constraints of the to-be-built and existing thermal power units; and respectively represent the downward and upward ramp-up coefficients of the to-be-built thermal power unit; and respectively represent the downward and upward ramp-up coefficients of the existing thermal power unit.

[0142] Specifically, the inertia-frequency change rate constraint is shown as follows:

[0143]

[0144] Among them, Equation (19) represents the inertia constraint related to the rate of change of frequency; and respectively represent the inertia (unit: s) of the to-be-built thermal power unit, wind power w, and the existing thermal power unit, represents the upper limit of the rate of change of frequency (unit: Hz / s), f0 represents the base frequency of the system (unit: Hz), ΔP Sys represents the unbalanced power of the system (unit: p.u.), H Sys represents the total inertia of the system (unit: s).

[0145] Furthermore, in S3, the second stage is the rescheduling stage, which is a scheduling process for coping with wind power prediction errors. Its constraints are the operation constraints after the wind power uncertainty appears and involve the random variable ξ tw , which represents the prediction error of the output of wind power w at time t (unit: MW), that is, the difference between the actual output and the predicted output. It can be understood that during this process, some operation variables, such as the actual output of thermal power, can be regarded as a function of the prediction errors of all wind power at time t (unit: MW).

[0146] Furthermore, the second-stage constraints established in the second stage include: operation state constraints, new energy consumption range constraints, and distributionally robust joint chance constraints.

[0147] Even further, the operation state constraints include: the second-stage power balance constraint, the second-stage line power flow constraint, the second-stage spinning reserve constraint, the second-stage dispatch output constraint and its coupling constraint with spinning reserve, the second-stage thermal power ramp constraint, and the rescheduling ramp constraint.

[0148] Specifically, the second-stage power balance constraint is shown as the following formula:

[0149]

[0150] Among them, Equation (20) represents the system power balance constraint for the predicted wind power value in the rescheduling stage; and respectively represent the actual dispatch outputs (unit: MW) of the to-be-built and existing thermal power units g at time t, that is, the outputs in the rescheduling stage, and this output is related to the prediction error of wind power; the random variable ξ tw , which represents the prediction error of the output of wind power w at time t (unit: MW), that is, the difference between the actual output and the predicted output; the random variable ξ t , which represents the prediction errors of all wind power at time t (unit: MW).

[0151] Specifically, the second-stage line power flow constraint is shown as the following formula:

[0152]

[0153] Among them, Equation (21) represents the line power flow constraint based on the power transfer distribution factor in the rescheduling stage.

[0154] Specifically, the second-stage spinning reserve constraint is shown as follows and includes:

[0155]

[0156] Among them, Equation (22) represents the spinning reserve demand constraint in the rescheduling stage, and Equations (23) and (24) respectively represent the spinning reserve ranges of the to-be-built and existing thermal power units in the rescheduling stage; and respectively represent the spinning reserve (unit: MW) of the to-be-built and existing thermal power units at time t in the rescheduling stage.

[0157] Specifically, the second-stage dispatch output constraint and its coupling constraint with the spinning reserve are shown as follows and include:

[0158]

[0159] Among them, Equations (25) and (26) respectively represent the upper limits of the output and the spinning reserve provided by the to-be-built thermal power units in the rescheduling stage, and Equations (27) and (28) respectively represent that the dispatch outputs of the to-be-built and existing thermal power units in the rescheduling stage are all non-negative; and respectively represent the spinning reserve of the to-be-built and existing thermal power units at time t in the rescheduling stage.

[0160] Specifically, the second-stage thermal power ramp-up constraint is shown as follows and includes:

[0161]

[0162] Among them, Equations (29) and (30) respectively represent the to-be-built and existing thermal power ramp-up constraints in the rescheduling stage.

[0163] Specifically, the rescheduling ramp-up constraint is shown as follows and includes:

[0164]

[0165] Among them, Equations (31) and (32) respectively represent the to-be-built and existing thermal power ramp-up constraints in the rescheduling stage; and respectively represent the downward and upward ramp-up coefficients of the to-be-built thermal power units in the rescheduling stage; and respectively represent the downward and upward ramp-up coefficients of the existing thermal power units in the rescheduling stage.

[0166] Furthermore, the new - energy accommodation range is constrained as follows:

[0167]

[0168] Wherein, and respectively represent the upper and lower bounds (unit: MW) of the accommodation range of wind power w in time period t, that is, the upper and lower bounds of the prediction error of wind power w that can be accommodated in time period t. This accommodation range is defined as when the prediction error of wind power w output in time period t is greater than and less than , the power system can fully accommodate the wind power generation at this time, and there will be no curtailment of wind power. It should be noted that after implementing the first - stage investment - pre - dispatch joint optimization and establishing the power - source investment constraint in the first - stage constraint, the accommodation range of wind power should also be given accordingly, that is and The larger the range, the higher the probability that wind power is fully accommodated, and the stronger the power system's ability to accommodate wind power.

[0169] Furthermore, the distribution - robust joint chance constraint, that is, modeling the uncertainty of wind power, specifically includes the following steps:

[0170] J1. Establish a fuzzy set Wherein, this fuzzy set is a set composed of a series of probability distributions with the same characteristics, and has the following two characteristics: one is that the prediction error ξ tw of the wind - turbine output has the same mean μ tw and variance The other is that the prediction error ξ tw of the wind - turbine output is unimodal with respect to the mean μ tw , that is, the probability density function of the prediction error ξ tw of the wind - turbine output is non - decreasing from 0 to the mean μ tw and non - increasing after the mean μ tw , as shown in the following formula:

[0171]

[0172] Wherein, represents the probability distribution of the prediction error of wind power based on historical data; represents the probability distribution expectation.

[0173] J2. Model the wind - power accommodation ability and quantify the probability that wind power is accommodated by the system, that is, the probability that the wind - power prediction error falls within the range, as shown in the following formula:

[0174]

[0175] a0 ≤ a ≤ 1 (36)

[0176] Among them, Equation (35) indicates that the minimum value of the probability that the prediction error of wind power falls within the system's consumable range is a; Equation (36) represents the upper and lower bound constraints of the minimum value a of this probability, that is, the upper and lower bound constraints of the minimum consumption probability; represents the upper and lower bounds of the prediction error of all wind power at time t; a0 represents the lower bound of the minimum probability that the prediction error of wind power falls within the system's consumable range, that is, it is required that the minimum consumption probability of new energy is not less than a0.

[0177] According to the above Equations (1) - (36), all constraints are simplified to obtain the general mathematical model of the present invention in S4:

[0178]

[0179] x ∈ X (38)

[0180]

[0181] Qξ L ≤ q, Rξ U ≤ r (41)

[0182] Among them, Equation (37) represents the objective function, Equation (38) represents the first-stage constraint, Equation (39) represents the operating state constraint in the second-stage constraint, Equation (40) represents the distributionally robust joint chance constraint in the second-stage constraint, and Equation (41) represents the constraint on the upper and lower limits of the consumption range; the variable x represents the decision variable in the first-stage constraint, that is, the power source investment plan and the pre-scheduling output of thermal power; X represents the set of x; the vector K represents the coefficient in the objective function; T(x) represents the set of functions of the variable x, W represents the set of parameters related to the function of the random variable ξ, and J represents the set of parameters related to the random variable ξ. ξ L and ξ U respectively represent the sets of the lower and upper limits of the random variable ξ; a represents the wind power consumption capacity; δ represents the weight coefficient used to adjust the trade-off relationship between the construction and operation total cost and the wind power consumption capacity; the vector y(ξ) represents the operating state decision variable in the second-stage constraint; Q and q represent the parameters related to the lower limit of the random variable ξ; R and r represent the parameters related to the upper limit of the random variable ξ.

[0183] The solution of the general mathematical model in S5 will be introduced below:

[0184] In this embodiment, the prediction error of the operating state variable in the second-stage constraint for wind power output follows an affine decision rule, and there is an affine relationship between the decision variable and the uncertainty factor, that is:

[0185] y(ξ) = Bξ + b (42)

[0186] where y is the decision variable, b is a constant vector, and B is a matrix.

[0187] The physical meaning of this mathematical assumption is the response behavior of the system after the uncertainty of wind power is revealed, simulating the adjustment response of AGC in the power system.

[0188] Taking the dispatching output of the planned thermal power unit at time t in actual dispatching as an example, it is specifically expressed as:

[0189]

[0190] where B tgw and b tg represent the response of the thermal power unit output to the prediction deviation of wind power output.

[0191] It can be understood that by using this affine decision rule, the optimization space of the original optimization problem is restricted within the affine function range, reducing the solution difficulty.

[0192] In this embodiment, based on Bonferroni approximation, Gaussian inequality, and second-order cone relaxation, the distributionally robust joint chance constraint is transformed, that is, transforming Equation (35) into the following second-order cone form:

[0193]

[0194] where g tw 、r tw and z tw are auxiliary variables, μ tw and σ tw are the mean and standard deviation of the random variable fuzzy set, respectively.

[0195] In this embodiment, it involves changes in the upper and lower limits of the accommodation range, specifically including the following:

[0196] First, define E := diag(ξ U - ξ L ), then the random variable ξ can be written as ξ L + Ev, v ∈ [0, e], where e is a vector of all 1s, then Equation (39) can be written as:

[0197]

[0198] Next, the quadratic terms BE and Bξ LReplace with S and s respectively to obtain:

[0199]

[0200] Then, the standard robust form of Equation (51) is:

[0201]

[0202] Finally, Equation (52) can be dualized to:

[0203] Re≤Jξ L -T(x)-Ws(53)

[0204] R≥WS-JE, R≥0(54)

[0205] Therefore, after the above transformation, the final optimization model in S5 of the present invention is as follows:

[0206] Objective function, Equation (55) corresponds to Equation (1):

[0207]

[0208] Power balance constraint in the second stage, Equations (56) and (57) correspond to Equation (20):

[0209]

[0210] Line power flow constraint in the second stage, Equations (58) and (59) correspond to Equation (21):

[0211]

[0212]

[0213] Reserve spinning constraint in the second stage, Equation (60) corresponds to Equation (22), Equations (61) and (62) correspond to Equation (23), and Equations (63) and (64) correspond to Equation (24):

[0214]

[0215] Dispatch output constraint in the second stage and its coupling constraint with reserve spinning, Equation (65) corresponds to Equation (25), Equation (66) corresponds to Equation (26), Equation (67) corresponds to Equation (27), and Equation (68) corresponds to Equation (28):

[0216]

[0217] Thermal power ramp rate constraint in the second stage, Equations (69) and (70) correspond to Equation (29), and Equations (71) and (72) correspond to Equation (30):

[0218]

[0219] The rescheduling ramp constraint. Equations (73) and (74) correspond to Equation (31), and Equations (75) and (76) correspond to Equation (32):

[0220]

[0221] Constraints related to auxiliary variables, including: E tw 、 and

[0222]

[0223] Other constraints, the first-stage constraints: Equations (2)-(19), the new energy consumption range constraint: Equation (33), the minimum and maximum consumption probability constraints: Equation (36), and the distributionally robust joint chance constraints: Equations (44)-(49).

[0224] Furthermore, in the preferred embodiment of the present invention, the 9-node system and the IEEE 118-node system are respectively selected as the test systems. Both are programmed and solved based on the required solver on the simulation platform. The relative optimal gap tolerance is set to 0.01%. All numerical simulations are carried out on the central processing unit of the same model. Assuming that the minimum probability of the power system fully consuming wind power is 2 / 3, one day in each of the four seasons of spring, summer, autumn, and winter is selected as the typical day (T = 96h). When calculating the operating cost, it needs to be converted to the annual operating cost. It should also be noted that this embodiment gives the numerical results of the above model under different parameter sensitivities. The adjusted parameters include: different green power ratio requirements γ, different load levels d', and the weight coefficient δ between the total cost and the wind power consumption capacity. It can be understood that the numerical results obtained from the simulation can be divided into two categories: the power supply configuration scheme and the range of wind power output that the system can fully consume. Therefore, the numerical results can be analyzed from two aspects: one is to analyze the impact of the inertia constraint on the power investment decision, and the other is to compare the investment schemes with and without virtual inertia support provided by wind power.

[0225] First, a set of wind power prediction error distributions is generated through a Gaussian distribution, with its mean set to 0 and the variance increasing with time within each day: starting from 10% of the installed capacity and increasing in steps of 0.1%. Then, these data are divided into two parts: the calibration part and the out-of-sample test part. The data in the calibration part are used to calibrate the mean and variance of the fuzzy sets. Next, the proposed model is solved to obtain the optimal wind power accommodation range and the optimal investment plan. The data in the out-of-sample test part are used to calculate the wind power accommodation probability and evaluate the out-of-sample performance of the proposed model. Finally, 1000 out-of-sample tests are conducted to verify the effectiveness of the proposed method.

[0226] In addition, the ratio of revenue to cost is used to quantify the profitability of the planning scheme, which is calculated as: the utility obtained from implementing the investment decision divided by the project investment cost. The utility comes from two aspects: the total value of the load shedding losses avoided by building new power sources, and the carbon emission reduction value of wind power replacing thermal power. Among them, VOLL (value of lost load) is set to 9000 $ / MWh, and the carbon emission reduction benefit is set to 10 $ / MWh (i.e., environmental benefit).

[0227] In the 9-node system, the system consists of 13 transmission lines, and the capacity of each line is 2500 MW. There are two thermal power units with rated capacities of 800 MW and 600 MW respectively, located at node 7 and node 3. The inertia of the existing thermal power units is 7 s. The system topology diagram and the parameters of the power sources to be built are as Figure 2 shown in and Table 1. In the figure, W represents the wind turbines to be built, the white G represents the existing thermal power units, and the orange G represents the thermal power units to be built. Figure 2 The percentage value next to each node in represents the percentage of the load at that node in the total system load. The minimum and maximum outputs of each thermal power unit are set to 20% and 100% of its rated capacity respectively. The investment-pre-scheduling interval in the first stage is 60 minutes, the re-scheduling interval in the second stage is 10 minutes, and the spinning reserve time scale is 10 minutes. The ramping capacity of the thermal power unit per minute is set to 2% of its rated capacity. In terms of the fault parameters, the active power deficit is set to the capacity of the largest unit in the system, the upper limit of the frequency change rate is 0.4 Hz / s, and the base frequency is 50 Hz. Thus, it can be calculated from Equation (19) that the minimum system inertia is 6.25 s

[0228] Table 1 Parameters of the power sources to be built

[0229]

[0230] (1) Analyze the trade-off between investment cost and wind power accommodation capacity

[0231] Furthermore, Table 2 lists the investment plans and corresponding investment costs under different γ and δ. Among them, in Table 2, the number before the parentheses in the third column represents the installation node location, and the number in the parentheses represents the capacity to be installed.

[0232] Table 2 Investment Plans for the 9-Node System

[0233]

[0234]

[0235] According to Table 2, when γ is fixed and δ increases, more thermal power units need to be built, or thermal power units with a higher inertia level need to be built, resulting in an increase in investment costs. Increasing δ means that the system operator hopes that the power system can accommodate more wind power, which requires more flexible power generation resources, such as stronger ramping capabilities and more sufficient inertia support, thus more thermal power units need to be built.

[0236] When δ is fixed and γ increases, more wind power will be built, resulting in an increase in investment costs. This is because a larger γ means that the system requires a higher new energy penetration rate, leading to more wind power installations. At the same time, since wind power can provide a certain amount of virtual inertia support, the installation of thermal power units can be appropriately reduced.

[0237] In addition, when γ is fixed and δ increases, or when δ is fixed and γ increases, the benefit-cost ratio of the system decreases. This is because compared with thermal power units, the reliable capacity of wind power is relatively low, that is, for the same 1MW of installed capacity, the available capacity of wind power is lower than that of thermal power units. This shows that to achieve a higher new energy power generation penetration rate, some economic benefits need to be sacrificed as a compromise for environmental benefits.

[0238] Figure 3 Shows the trade-off relationship between the wind power accommodation range and investment costs (at this time γ = 45%), that is, the wind power accommodation range under different δ. The space between the solid line and the dotted line of the same color represents the interval of wind power generation that the system can fully accommodate at a certain δ. The results show that when δ increases, that is, when the investment cost increases, the wind power accommodation range increases.

[0239] Furthermore, by gradually increasing δ under different d' and γ, out-of-sample tests are conducted to form the Pareto frontier of the total cost and accommodation probability. Among them, the total cost refers to the sum of the investment cost and the operating cost. As Figure 4 shown, increasing d' and γ will increase the total cost. Increasing δ can improve the new energy accommodation probability, but the total investment and operating costs will also increase. At the same time, at the same cost, the higher γ is, the lower the accommodation probability is. At high γ, higher investment costs and a larger δ are required to maintain the new energy accommodation probability.

[0240] (2) Analyze the impact of considering inertia constraints on the system inertia level

[0241] Under different values of γ (set from 42.5% to 100%), the system inertia levels of the optimal power source structures with and without inertia constraints in the above model are as Figure 5 shown. When γ is lower than 59.4%, regardless of whether inertia constraints are considered, the system inertia level is greater than the minimum inertia requirement. However, when γ increases from 59.4% upwards, the system inertia obtained from the model without inertia constraints starts to be lower than the minimum inertia required by the system, which will affect the frequency stability of the system. On the contrary, if inertia constraints are considered, the model proposed in this chapter can ensure that the system inertia is always greater than the minimum inertia requirement of the system.

[0242] (3) Analyze the impact of considering virtual inertia support on the power source structure

[0243] The following analyzes the investment results with and without virtual inertia support. When ignoring the virtual inertia provided by wind power in the proposed model, the investment results under different values of γ with δ = 1 are obtained, as shown in Table 3.

[0244] Table 3 Investment plans for the 9 - node system (without virtual inertia support, δ = 1)

[0245]

[0246] Compared with Table 2 considering virtual inertia support, it can be seen from Table 3 that if virtual inertia support is ignored, in scenarios with the same wind power penetration rate, more thermal power units need to be built to provide sufficient system inertia, and these thermal power units are characterized by high inertia constants and high investment costs. When γ = 45%, the investment cost considering virtual inertia is 35.5% lower than that without considering virtual inertia. Moreover, in the case of γ = 45%, the benefit - cost ratio considering virtual inertia increases by 54.02% compared with that without considering virtual inertia. In addition, without considering virtual inertia, the maximum wind power generation share is 49%. This is because restricted by the system inertia requirement, for wind turbines that cannot provide virtual inertia support, if the new - energy generation share γ in the system exceeds 49%, the system will face insufficient inertia.

[0247] (4) Consider the impact of wind power generation uncertainty on the power source structure

[0248] Table 4 lists the investment results of the deterministic power source optimization configuration model under different conditions (without considering wind power generation uncertainty). Comparing Table 4 and Table 2, it can be seen that ignoring the uncertainty of wind power generation reduces the investment cost and improves the benefit - cost ratio. However, Figure 6It can be seen that ignoring the uncertainty of wind power generation reduces the accommodation range of wind power. For example, at t = 15h, the distance between the two red lines is smaller than the distance between the two blue lines.

[0249] Table 4 Investment Plan for 9-Node System (Deterministic Model)

[0250]

[0251] In the IEEE 118-node system, this 118-node system includes 19 generators, 35 synchronous condensers, 186 lines, 9 transformers, and 91 loads. Two candidate wind powers are located at Node 32 and Node 88 respectively. Table 5 lists the investment plan, the corresponding investment cost, the benefit-cost ratio, and the calculation time. It can be seen from this that the proposed model is effective in solving large-scale problems.

[0252] Table 1 Investment Plan for 118-Node System (δ = 1)

[0253]

[0254] In summary, for the two-stage centralized optimization configuration modeling method of power sources considering uncertainty and inertia in the present invention, the requirements of frequency stability and reliability are considered simultaneously first; in terms of the requirements of frequency stability, the system inertia is modeled, including traditional inertia and virtual inertia, as well as the corresponding frequency constraints. In terms of the requirements of reliability, the uncertainty is described by a fuzzy set, and the accommodation capacity of new energy is modeled in the form of a distributionally robust joint chance constraint, so that the proposed model jointly optimizes the power source investment cost and the accommodation capacity of new energy power generation, can quantify the accommodation capacity of new energy under a certain power source structure, and provides clear guidance for new energy power generation investment; secondly, the model established by the method of the present invention is a two-stage distributionally robust optimization model, and the upper and lower bounds of its random variables are variables rather than parameters, which is difficult to solve. Therefore, the present invention obtains a conservative approximate equivalent model by adopting an affine decision rule, second-order cone reconstruction, and dual techniques. This model is transformed into a single-stage robust optimization model to achieve the final solution.

[0255] Although the content of the present invention has been introduced in detail through the above preferred embodiments, it should be recognized that the above description should not be considered as a limitation of the present invention. After those skilled in the art have read the above content, various modifications and substitutions to the present invention will be obvious. Therefore, the protection scope of the present invention should be defined by the appended claims.

Claims

1. A two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia, characterized in that: The specific steps include: S1. Construct a two-stage optimization objective function with the comprehensive goal of minimizing the sum of total investment cost and operating cost and maximizing the new energy absorption capacity; S2, perform the first stage investment-pre-scheduling joint optimization and establish the first stage constraints; S3. Implement the second stage real-time rescheduling optimization and establish the second stage constraints; S4, collaboratively optimize the constraints of the first stage and the constraints of the second stage to obtain a general mathematical model; S5. Solve the above general mathematical model to obtain the final optimization model.

2. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 1 is characterized in that: In the above S1, the calculation formula of the objective function is specifically: Among them, min represents the minimization function; g represents the thermal power unit; G S Represents the set of thermal power units to be built, G E represents the set of existing thermal power units; w represents wind power; W represents the set of wind power units to be built; t represents the time period; T represents the set of time periods; and Respectively represent the annualized investment costs of thermal power units g and wind power w; δ represents the weight coefficient; function represents the operating cost function of thermal power unit g; Indicates the installed capacity of the thermal power unit to be built; represents the installed capacity of wind power w to be built; and They represent the pre-dispatched output of the thermal power units to be built and existing in period t based on the wind power forecast value; a represents the probability that the system can fully absorb the wind power output; represents the investment cost of thermal power units; represents the investment cost of wind power; Represents the operating cost of wind power and thermal power; δa represents the probability of maximizing the system's absorption of new energy under the preset weight coefficient.

3. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 2 is characterized in that: In the S2, the first-stage constraints include: power source investment constraints, minimum green electricity ratio constraints, first-stage power balance constraints, first-stage line flow constraints, first-stage rotating reserve constraints, first-stage dispatching output constraints and their coupling constraints with rotating reserve, first-stage thermal power ramping constraints and inertia-frequency change rate constraints.

4. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 3 is characterized in that: The power investment constraint is as shown in the following formula, including: Among them, β g Indicates whether to build a new thermal power unit. g =1 means construction, β g =0 means no construction, G S represents the set of thermal power units to be built; w represents wind power, and W represents the set of wind power units to be built. represents the installed capacity of wind power w to be built, represents the upper limit of the new capacity of wind power w, represents the existing installed capacity of wind power w, represents the total installed capacity of wind power w; is the maximum output of the thermal power unit to be built; The green electricity ratio constraint is as follows: Among them, the set N represents the node set, ΔT is the scheduling interval; γ is the minimum requirement for the ratio of renewable energy power generation to total power consumption; d tn is the power load demand of node n in time period t; is the predicted output of wind power w in time period t; The first stage power balance constraint is as follows: The line power flow constraint is as follows: in, represents the transmission capacity of line l, A nl represents the power transfer distribution factor of node n relative to line l, G S (n) represents the set of thermal power units to be built at node n, G E (n) represents the set of existing thermal power units at node n, W(n) represents the set of wind power units to be built at node n, and L represents the set of lines; the first-stage spinning reserve constraint is as follows, including: in, represents the system spinning reserve demand for time period t; and They represent the 10-minute spinning reserve coefficients of the thermal power units to be built and the existing ones respectively; and They represent the spinning reserves of the thermal power units to be built and the existing ones in the pre-dispatch phase period t respectively; Indicates the maximum output of the existing thermal power unit g; The first stage dispatch output constraint and its coupling constraint with spinning reserve are shown in the following formula, including: in, is the typical output coefficient of wind power w in period t, represents the total installed capacity of wind power w; Indicates the maximum output of existing thermal power units; The thermal power ramping constraint in the first stage is as follows: in, and They represent the downward and upward climbing coefficients of the thermal power unit to be built respectively; and They represent the downward and upward climbing coefficients of existing thermal power units respectively; The inertia-frequency rate of change constraint is as follows: in and They represent the inertia of the thermal power unit to be built, wind power w and the existing thermal power unit, represents the upper limit of the frequency change rate, f0 represents the base frequency of the system, ΔP Sys Represents the unbalanced power of the system, H Sys Represents the total inertia of the system.

5. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 4 is characterized in that: In S3, the second stage is the rescheduling stage, which is a scheduling process for dealing with wind power forecast errors. Its constraints are the operating constraints after the uncertainty of wind power emerges, involving random variables ξ tw , represents the prediction error of wind power output in time period t; Among them, the second stage constraints include: operating state constraints, new energy consumption scope constraints and distributed blue rod joint opportunity constraints; The operating state constraints include: second-stage power balance constraints, second-stage line flow constraints, second-stage spinning reserve constraints, second-stage dispatching output constraints and their coupling constraints with spinning reserve, second-stage thermal power ramping constraints and re-dispatching ramping constraints.

6. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 5 is characterized in that: The second stage power balance constraint is as follows: in, and They represent the actual dispatch output of the thermal power units to be built and the existing thermal power units in time period t respectively; the random variable ξ tw , represents the forecast error of wind power w in period t; random variable ξ t , represents the prediction error of all wind power in period t; The second stage line flow constraint is as follows: The second spinning reserve constraint in the stage is as shown in the following formula, including: in, and They represent the spinning reserves of the thermal power units to be built and the existing ones in the re-dispatch phase period t respectively; The second stage dispatch output constraint and its coupling constraint with the spinning reserve are shown in the following formula, including: in, and They represent the spinning reserve of the thermal power plant g to be built and the existing thermal power plant g in the re-dispatch phase period t respectively; The second stage thermal power ramp constraint is as follows: The rescheduling ramp constraint is as follows: in, and They represent the downward and upward ramping coefficients of the re-dispatch phase of the thermal power units to be built; and They respectively represent the downward and upward ramping coefficients of the re-dispatch phase of existing thermal power units.

7. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 6 is characterized in that: The new energy consumption range constraint is as follows: in, and They respectively represent the upper and lower limits of the wind power w consumption range in time period t.

8. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 7 is characterized in that: The establishment of the distributed robust joint opportunity constraint specifically includes the following steps: J1. Establish fuzzy sets in, represents the probability distribution of wind power forecast error based on historical data; ξ tw Represents the prediction error of wind turbine output; μ tw represents the mean prediction error of wind turbine output; represents the prediction error variance of wind turbine output; Represents probability distribution expectations; J2. Model the wind power absorption capacity and quantify the probability of wind power being absorbed by the system, as shown in the following formula: a0≤a≤1(36) Where a represents the minimum value of the probability that the wind power prediction error falls within the system's acceptable range; It represents the upper and lower bounds of the prediction error of all wind power in time period t; a0 represents the lower limit constraint of the minimum value a of the wind power prediction error that falls within the system's absorptive range.

9. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 8, characterized in that: In the above-mentioned S4, the general mathematical model established is: x∈X(38) Qξ L ≤q,Rξ U ≤r(41) Among them, formula (37) represents the objective function, formula (38) represents the first-stage constraints, formula (39) represents the operating state constraints in the second-stage constraints, formula (40) represents the distributed robust joint chance constraints of the second-stage constraints, and formula (41) represents the constraints on the upper and lower limits of the absorption range; the variable x represents the decision variable in the first-stage constraints; X represents the set of x; the vector K represents the coefficient in the objective function; T(x) represents the set of functions of the variable x, W represents the set of parameters related to the function of the random variable ξ, and J represents the set of parameters related to the random variable ξ. ξ L and U They represent the set of lower and upper limits of the random variable ξ respectively; a represents the wind power absorption capacity; δ represents the weight coefficient; the vector y(ξ) represents the operating state decision variable in the second stage constraint; Q and q represent the parameters related to the lower limit of the random variable ξ; R and r represent the parameters related to the upper limit of the random variable ξ.

10. The two-stage power centralized optimization configuration modeling method taking into account uncertainty and inertia as claimed in claim 9, characterized in that: In S5, the general mathematical model is solved according to the affine decision rule, Bonferroni approximation, Gaussian inequality, and second-order cone relaxation to obtain the final optimization model: Objective function: Second stage power balance constraints: The second stage line flow constraints: Second stage spinning reserve constraints: The second stage dispatch output constraints and their coupling constraints with spinning reserve: Second stage thermal power ramp constraints: Rescheduling ramp constraints: Auxiliary variable related constraints, including: E tw , and And other constraints remain unchanged.

Images