Inter-provincial tie line power flow prediction and value evaluation method based on data driving

Abstract

The invention discloses an inter-provincial tie line power flow prediction and value evaluation method based on data driving, and the method comprises the steps: obtaining thermal power generation, new energy power generation, load, DC channel delivery and inter-provincial tie line power flow data of each province, and aggregating the province into a region, i.e., integrating a plurality of nodes in the province into a single region node; according to the obtained data, the provincial injection power after new energy power generation is considered is solved; according to the obtained provincial injection power data and inter-provincial tie line power flow data, a multiple linear regression equation set with a constant term being zero is established, and a regional transfer factor is obtained through fitting; taking the goodness of fit as an index, selecting a region transfer factor corresponding to the time period with the highest fitting precision, and substituting the region transfer factor into subsequent analysis; and establishing a simulated power grid dispatching optimization model P1 by taking the minimum power generation cost as an optimization target according to the regional transfer factor corresponding to the time period with the highest fitting precision, relaxing the capacity constraint of the tie line, constructing a Lagrange function, and calculating the marginal value of the capacity of the tie line according to the optimization model P1.

Description

Technical Field

[0001] The present invention relates to the technical field of power system operation and dispatch automation, and particularly to a data-driven method for predicting and evaluating the value of inter-provincial tie-line power flow. Background Art

[0002] In the context of new energy consumption issues and challenges in coping with peak load curtailment, inter-provincial tie-lines play an important role. As a key channel for transmitting new energy between provinces, tie-lines enable the utilization of surplus new energy generation and promote the integration of intermittent resources in the power grid. Through inter-provincial tie-lines, the power system can more efficiently balance and allocate renewable energy and flexible resources, reduce the need for curtailment, and improve the overall grid stability. Similarly, tie-lines can also be used to send electricity from areas rich in power generation resources to peak load regions, reducing peak load curtailment. Therefore, how to quantify the value of inter-provincial tie-lines for planning tie-line construction, so as to maximize the utility of inter-provincial tie-lines in reducing power generation costs, improving new energy consumption, and reducing load shedding has become a new problem in the field of cross-provincial power grid dispatch.

[0003] Traditional tie-line power flow prediction requires the use of Power Transfer Distribution Factors (PTDF). The calculation of PTDF in traditional nodal network systems may encounter limitations in some markets, which stem from the challenges of obtaining accurate and timely data, combined with concerns about information protection, because node-specific relevant data and internal parameters of tie-lines are often difficult to obtain, and this calculation does not consider emerging new energy generation.

[0004] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present invention, and thus may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] Aiming at the deficiencies in the prior art, the purpose of the present invention is to provide a data-driven method for predicting and evaluating the value of inter-provincial tie-line power flow, aggregating the node systems within a province into a regional system with the province as a unit, using the least squares method to fit the Zonal Shift Factor, and using the zonal shift factor to estimate the magnitude of the tie-line power flow. Based on the duality theory, Lagrange multipliers are introduced to quantify the marginal value of tie-line capacity in terms of reducing power generation costs, improving new energy consumption, and reducing load shedding.

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

[0007] A data-driven method for predicting and evaluating the value of inter-provincial tie-line power flow according to the present invention includes:

[0008] Step 1: Obtain the data of thermal power generation, new energy power generation, load, DC channel power transmission, and inter-provincial tie-line power flow in each province, and aggregate the provinces into regions, that is, integrate multiple nodes within a province into a single regional node;

[0009] Step 2: According to the data obtained in Step 1, solve the provincial injection power considering new energy power generation;

[0010] Step 3: According to the provincial injection power and the inter-provincial tie-line power flow data obtained in Step 1, establish a multiple linear regression equation system with a constant term of zero, and fit to obtain the regional transfer factor;

[0011] Step 4: Using the goodness of fit as an index, select the regional transfer factor corresponding to the time period with the highest fitting accuracy and substitute it into the subsequent analysis;

[0012] Step 5: According to the regional transfer factor corresponding to the time period with the highest fitting accuracy obtained in Step 4, with the minimum power generation cost as the optimization goal, establish a simulated power grid dispatching optimization model P1;

[0013] Step 6: According to the optimization model P1, relax the tie-line capacity constraint, construct the Lagrangian function, and calculate the marginal value of the tie-line capacity.

[0014] In the method described above, the calculation method of the regional injection power is: Determine the objective function as:

[0015] P zinj =P co +P re -P sd -P 1d (1)

[0016] where P co and P re are thermal power generation and new energy power generation respectively, P sd and P 1d are the DC channel power transmission and the load respectively, and P zinj is the provincial injection power considering new energy power generation.

[0017] In the method described above, the fitting of the regional transfer factor includes,

[0018] 1) Establish the relationship between the provincial injection power and the inter-provincial tie-line power flow

[0019] P f1 =ZP zinj (2)

[0020] where is the inter-provincial tie-line power flow, L is the number of inter-provincial tie-lines, is the provincial injection power, K is the number of provinces, is the regional transfer factor matrix,

[0021] 2) Substitute the sample data at N times into equation (2), and use the least squares method to solve the regional transfer factor matrix, and construct a loss function with a single inter-provincial tie line.

[0022]

[0023] Where is the theoretical power flow calculated using the fitted regional transfer factor, is the actual power flow of the sample. The goal of the least squares method is to obtain a set of parameter values that minimize the loss function value.

[0024] 3) Use the matrix method of the least squares method to solve the multiple linear regression. Since the sum of the injected powers of each province is zero, set province 1 as the balancing province, and only need to solve k - 1 parameters (without considering the constant term).

[0025]

[0026] J(b) = 1 / 2(P zinj b - P fl ) T (P zinj b - P fl )

[0027]

[0028] b = ((P zinj ) T P zinj ) -1 (P zinj ) T P fl (3)

[0029] Where P zinj ∈ R N×K-1 is the injected power of each province for N samples, b ∈ R K-1 is the parameter array, corresponding to k - 1 regions, is the estimated power flow array for N samples, P fl ∈ R N is the actual power flow array for N samples, J(b) is defined as the loss function, is the partial derivative of the loss function with respect to the array b, and after sorting, the solution of the parameter is b = ((P zinj ) T P zinj ) -1 (P zinj ) T P fl ,

[0030] 4) Calculate the parameter array b of L tie lines, combine them by rows, and finally add a column of zeros to the first column to obtain the area transfer factor matrix.

[0031] In the method described above, according to the goodness of fit R 2 Selecting the area transfer factor corresponding to the time period with the highest fitting accuracy includes:

[0032] 1) The goodness of fit R 2 Is used as an index to measure the goodness of fit in regression analysis

[0033]

[0034] Where n is the number of samples, y i Is the true value of the sample, Is the fitted value of the sample, Is the arithmetic mean of the true values of the samples; Represents the sum of squared errors after exact fitting, Represents the sum of squared errors after rough fitting using the sample mean, R 2 The value range is 0 - 1. The closer it is to 1, the closer the sum of squared errors after fitting is to 0, that is, the higher the fitting accuracy.

[0035] 2) Divide the samples into multiple sample groups with different time spans according to the time scale,

[0036] 3) Perform least squares fitting on multiple sample groups respectively to obtain the corresponding goodness of fit R 2 Value,

[0037] 4) Compare the goodness of fit R 2 Values corresponding to multiple sample groups, and select the group of samples with the highest R 2 To obtain the area transfer factor under the corresponding time scale, and substitute it into subsequent model analysis and power flow prediction.

[0038] In the method described above, establishing the simulated power grid dispatching optimization model P1 includes:

[0039] 1) Determine the objective function as

[0040]

[0041] Where, P k，t , Are the outputs of thermal power units, wind power curtailment, photovoltaic power curtailment, and load shedding in the k-th province at time t respectively. f() is the thermal power generation cost. B, D, and c are the wind power curtailment penalty, photovoltaic power curtailment penalty, and load shedding penalty respectively. T is the total number of time periods, and K is the total number of provinces.

[0042] 2) Introduce the tie line power flow constraint

[0043] Pflow =ZP inj (6)

[0044] where P flow ∈R T×L is the tie-line power flow matrix, L is the number of tie-lines, Z is the area transfer factor obtained in step 4, and P inj ∈R T×K is the provincial injection power matrix. This constraint reflects the relationship between tie-line power flow and provincial injection power.

[0045] 3) Introduce the provincial power balance constraint

[0046]

[0047] where P k，w，t is the wind power output of the k-th province at time t, P k，w，t is the photovoltaic power output of the k-th province at time t, P l，k，t is the electricity quantity transmitted to the k-th province through the l-th tie-line at time t, P k，d，t is the load of the k-th province at time t. Here, the power output of the DC channel is regarded as part of the load. is the set of all time periods, is the set of all provinces. This constraint indicates that the load of any province at any time is equal to the sum of the thermal power unit output, photovoltaic unit output, wind power unit output, the electricity quantity transmitted to this province through all tie-lines and the load shedding of this province.

[0048] 4) Introduce the tie-line capacity constraint

[0049]

[0050] where P l，max is the capacity of the l-th tie-line, P l，t is the power flow on the l-th tie-line at time t, is the set of all tie-lines. This constraint indicates that the positive and negative power flows on any tie-line at any time cannot exceed the capacity of the tie-line.

[0051] 5) Introduce the thermal power unit output constraint

[0052]

[0053] where, is the lower limit of the thermal power unit output of the k-th province at time t, is the upper limit of the thermal power unit output of the k-th province at time t. This constraint indicates that the value of the thermal power unit output of any province at any time is between its upper and lower limits.

[0054] 6) Introduce the wind power unit output constraint

[0055]

[0056] wherein, is the lower limit of the output of wind turbines in the k-th province at time t, is the upper limit of the output of wind turbines in the k-th province at time t. This constraint indicates that the output value of wind turbines in any province at any time is between its upper and lower limits.

[0057] 7) Introduce the output constraint of photovoltaic power generation units

[0058]

[0059] wherein, is the lower limit of the output of photovoltaic power generation units in the k-th province at time t, is the upper limit of the output of photovoltaic power generation units in the k-th province at time t. This constraint indicates that the output value of photovoltaic power generation units in any province at any time is between its upper and lower limits.

[0060] 8) Introduce the wind curtailment constraint

[0061]

[0062] This constraint indicates that the wind curtailment energy in any province at any time is equal to the maximum value of the output of wind turbines in this province minus the actual output value of wind turbines in this province.

[0063] 9) Introduce the photovoltaic curtailment constraint

[0064]

[0065] This constraint indicates that the photovoltaic curtailment energy in any province at any time is equal to the maximum value of the output of photovoltaic power generation units in this province minus the actual output value of photovoltaic power generation units in this province.

[0066] Establish the power dispatching model P1 of the simulated northwest power grid

[0067]

[0068] In the described method, constructing the Lagrangian function to calculate the marginal value of the tie-line capacity includes:

[0069] 1) Relax the tie-line capacity constraint Construct the Lagrangian function with the minimum generation cost, maximum new energy consumption, and minimum load shedding as the objectives respectively

[0070]

[0071] wherein, L 1 is used to calculate the marginal value of the tie-line capacity based on reducing the generation cost, L2 The marginal value of tie-line capacity for calculating based on improving the accommodation of new energy, L 3 The marginal value of tie-line capacity for calculating based on reducing load shedding

[0072] 2) Solve the dual problem corresponding to the Lagrangian function to obtain the marginal value of tie-line capacity

[0073]

[0074]

[0075]

[0076] where λ 1 + μ 1 is the marginal value of tie-line capacity based on reducing power generation cost, λ 2 + μ 2 is the marginal value of tie-line capacity based on improving the accommodation of new energy, λ 3 + μ 3 is the marginal value of tie-line capacity based on reducing load shedding

[0077] Beneficial effects

[0078] The present invention obtains the data of provincial thermal power generation, new energy generation, DC external transmission, load and inter-provincial tie-line power flow, and aggregates provinces into regions, that is, integrates multiple nodes within a province into a single regional node, uses the least squares method to fit and obtain the Zone Shift Factor (ZSF) matrix, and then uses the ZSF matrix for power flow prediction. Compared with the traditional Power Transfer Distribution Factor (PTDF) prediction method that requires knowing the specific parameters of tie-lines and detailed node injection power data, the data required by this method is easier to obtain, the calculation is simpler, and it has more practical significance. The Zone Shift Factor (ZSF) is a simplified PTDF matrix used to approximately calculate the power flow between different regions in the power grid. Since ZSF only depends on publicly available and ex-post inter-regional power flow data as input, market operators prefer to use ZSF because this type of data is easier to obtain than detailed node information. Moreover, even in the absence of accurate tie-line parameters, the ZSF-based method can achieve reasonable accuracy. Similarly, this method is also applicable to various regional types including built-in node systems at the municipal and county levels. By aggregating the node systems within a region into a single regional-level node, this method can be used for power flow prediction and analysis. In this method, the goodness of fit R 2Sample selection also ensures that the final calculated regional transfer factor matrix has high accuracy. The introduction of Lagrangian function to obtain the marginal value of interconnection line capacity based on different aspects can evaluate the impact of interconnection lines on power system operation from multiple dimensions, including more effective balancing and allocation of renewable energy and flexible resources in different regions, thereby reducing wind curtailment demand and improving overall grid stability; transmitting electricity from areas with abundant power generation resources to peak load areas, thereby achieving the purpose of peak shaving; and using power generation resources in regions with lower power generation costs as much as possible to reduce the overall operating cost of the power grid. This method has broad application prospects in the fields of inter-regional power flow prediction and interconnection line value assessment, including: regional power flow prediction, interconnection line construction planning, and interconnection line investment return prediction.

[0079] The above description is only an overview of the technical solution of the present invention. In order to make the technical means of the present invention clearer and to enable those skilled in the art to implement it according to the contents of the specification, and to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described below by way of example. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0081] By reading the detailed description of the preferred specific embodiments below, various other advantages and benefits of the present invention will become clear to those of ordinary skill in the art. The drawings in the specification are only for the purpose of illustrating the preferred embodiments and are not considered to be limitations of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative work. Moreover, the same reference numerals are used to represent the same components throughout the drawings.

[0082] In the attached picture:

[0083] Figure 1 It is a flow chart of the steps of fitting the inter-provincial regional transfer factor (ZSF) of the present invention;

[0084] Figure 2 It is a flow chart of the steps of calculating the marginal value of the tie line according to the present invention;

[0085] Figure 3 It is a topological structure diagram of the contact lines of the five northwestern provinces (regions) in the embodiment of the present invention;

[0086] Figure 4 It is a line graph in an embodiment of the present invention that uses the region transfer factor obtained by fitting to calculate the estimated power flow of a certain tie line in the current same time period and compares it with the actual power flow;

[0087] Figure 5 It is a line graph in an embodiment of the present invention that uses the region transfer factor obtained by fitting to calculate the estimated power flow of a certain tie line in a future time period and compares it with the actual power flow;

[0088] Figure 6 It is a schematic diagram in an embodiment of the present invention of the marginal value of the capacity of a certain tie line measured based on reducing the power generation cost;

[0089] Figure 7 It is a schematic diagram in an embodiment of the present invention of the marginal value of the capacity of a certain tie line measured based on improving the accommodation of new energy;

[0090] Figure 8 It is a schematic diagram in an embodiment of the present invention of the marginal value of the capacity of a certain tie line measured based on reducing the load shedding.

[0091] The present invention will be further explained below in conjunction with the accompanying drawings and embodiments. Specific Embodiments

[0092] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0093] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0094] It should be noted that: Similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0095] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc. are based on the orientation or positional relationships shown in the drawings. These are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention.

[0096] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, the meaning of "a plurality" is two or more unless otherwise specifically defined.

[0097] In the present invention, unless otherwise clearly specified and defined, the terms such as "mounted", "connected", "connected to", "fixed" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or integrated; it may be directly connected or indirectly connected through an intermediate medium, and it may be the communication inside two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0098] In the present invention, unless otherwise clearly specified and defined, the first feature being "above" or "below" the second feature may include the direct contact between the first and second features, or may include the situation where the first and second features are not in direct contact but in contact through other features between them. Moreover, the first feature being "above", "over" and "on" the second feature includes that the first feature is directly above and obliquely above the second feature, or merely indicates that the horizontal height of the first feature is higher than that of the second feature. The first feature being "below", "under" and "beneath" the second feature includes that the first feature is directly below and obliquely below the second feature, or merely indicates that the horizontal height of the first feature is lower than that of the second feature.

[0099] To enable those skilled in the art to better understand the technical solutions of the present invention, the following will Figures 1 to 8 further introduce the present invention in detail, and each drawing does not constitute a limitation to the embodiments of the present invention.

[0100] In one embodiment, as Figures 1 to 8 shown, the present disclosure provides a data-driven method for inter-provincial tie-line power flow prediction and value evaluation, including the following steps:

[0101] Step 1: Obtain the data of thermal power generation (considering hydroelectric power generation as part of thermal power generation), new energy power generation (including wind power generation and photovoltaic power generation), load, DC channel power transmission, and inter-provincial tie-line power flow in each province, and aggregate the provinces into regions, that is, integrate multiple nodes within the province into a single regional node.

[0102] Step 2: According to the data obtained in Step 1, solve the provincial injection power considering new energy power generation.

[0103] Step 3: Based on the provincial injection power and the inter-provincial tie-line power flow data obtained in Step 1, establish a multiple linear regression equation system with a zero constant term, and fit to obtain the area transfer factor.

[0104] Step 4: Using the goodness of fit as an index, select the area transfer factor corresponding to the time period with the highest fitting accuracy and substitute it into the subsequent analysis.

[0105] Step 5: Based on the area transfer factor corresponding to the time period with the highest fitting accuracy obtained in Step 4, establish a simulated power grid dispatching optimization model P1 with the minimum power generation cost as the optimization goal.

[0106] Step 6: According to the optimization model P1, relax the tie-line capacity constraint, construct the Lagrangian function, and calculate the marginal value of the tie-line capacity.

[0107] In the preferred implementation manner of the described method, the calculation method of the provincial injection power is: Determine the objective function as:

[0108] P zinj =P co +P re -P sd -P ld (1)

[0109] where P co and P re are thermal power generation and new energy power generation respectively, P sd and P ld are the DC channel power transmission and the load respectively, and P zinj is the provincial injection power considering new energy power generation.

[0110] In the preferred implementation manner of the described method, the fitting of the area transfer factor includes

[0111] 1) Establish the relationship between the provincial injection power and the inter-provincial tie-line power flow

[0112] P fl =ZP zinj (2)

[0113] where For the power flow of the inter-provincial tie line, L is the number of inter-provincial tie lines, For the power injection of the province, K is the number of provinces, For the regional transfer factor matrix,

[0114] 2) Substitute the sample data at N times into the relational expression (2), and use the least squares method to solve the regional transfer factor matrix. A loss function is constructed with a single inter-regional tie line

[0115]

[0116] where is the theoretical power flow calculated using the fitted regional transfer factor, is the actual power flow of the sample. The goal of the least squares method is to obtain a set of parameter values that minimize the loss function value,

[0117] 3) Use the matrix method of the least squares method to solve the multiple linear regression. Since the sum of the power injections of each province is zero, set province 1 as the balancing province, and only need to solve k - 1 parameters (excluding the constant term)

[0118]

[0119] J(b) = 1 / 2(P zinj b - P fl ) T (P zinj b - P fl )

[0120]

[0121] b = ((P zinj ) T P zinj ) -1 (P zinj ) T P fl (3)

[0122] where P zinj ∈R N×K-1 is the power injection of the province for N samples, b ∈ R K-1 is the parameter array, corresponding to k - 1 provinces, is the estimated power flow array for N samples, P fl ∈R N is the actual power flow array for N samples, J(b) is defined as the loss function, is the partial derivative of the loss function with respect to the array b. After sorting, the solution of the parameter is b = ((P zinj ) T P zinj ) -1 (P zinj )T P fl ,

[0123] 4) Calculate the parameter array b of L tie lines, combine them row by row, and finally add a zero column to the first column, corresponding to the balancing province 1, to obtain the regional transfer factor matrix.

[0124] In the preferred embodiment of the method described, according to the goodness of fit R 2 Selecting the regional transfer factors corresponding to the time period with the highest fitting accuracy includes:

[0125] 1) The goodness of fit R 2 Used as an index to measure the goodness of fit in regression analysis

[0126]

[0127] where n is the number of samples, y i is the true value of the sample, is the fitted value of the sample, is the arithmetic mean of the true values of the samples; represents the sum of squared errors after precise fitting, represents the sum of squared errors after rough fitting using the sample mean, R 2 The value range is 0 - 1. The closer it is to 1, the closer the sum of squared errors after fitting is to 0, that is, the higher the fitting accuracy.

[0128] 2) Divide the samples into multiple sample groups with different time spans according to the time scale.

[0129] 3) Perform least squares fitting on multiple sample groups respectively to obtain the corresponding goodness of fit R 2 value,

[0130] 4) Compare the goodness of fit R 2 values corresponding to multiple sample groups, and select the group of samples with the highest R 2 to obtain the regional transfer factors under the corresponding time scale, and substitute them into subsequent model analysis and power flow prediction.

[0131] In the preferred embodiment of the method described, establishing the simulated power grid dispatching optimization model P1 includes:

[0132] 1) Determine the objective function as

[0133]

[0134] where, P k，t , They are the thermal power unit output, wind power curtailment, photovoltaic power curtailment, and load shedding in the k-th province at time t, respectively. f() is the thermal power generation cost. B, D, and C are the wind power curtailment penalty, photovoltaic power curtailment penalty, and load shedding penalty, respectively. T is the total number of time periods, and K is the total number of provinces.

[0135] 2) Introduce the tie-line power flow constraint

[0136] P flow =ZP inj (6)

[0137] Among them, P flow ∈R T×L is the tie-line power flow matrix, L is the number of tie-lines, Z obtained in step 4 is the area transfer factor, and P inj ∈R T×K is the provincial injection power matrix. This constraint reflects the relationship between tie-line power flow and provincial injection power.

[0138] 3) Introduce the provincial power balance constraint

[0139]

[0140] Among them, P k，w，t is the wind power output in the k-th province at time t, P k，v，t is the photovoltaic power output in the k-th province at time t, P l，k，t is the electricity transmitted from the l-th tie-line to the k-th province at time t, P k，d，t is the load in the k-th province at time t. Here, the power transmitted through the DC channel is regarded as part of the load. is the set of all time periods, is the set of all provinces. This constraint indicates that the load in any province at any time is equal to the sum of the thermal power unit output, photovoltaic power unit output, wind power unit output, the electricity transmitted from all tie-lines to this province, and the load shedding in this province.

[0141] 4) Introduce the tie-line capacity constraint

[0142]

[0143] Among them, P l，max is the capacity of the l-th tie-line, and P l，t is the power flow on the l-th tie-line at time t. is the set of all tie-lines. This constraint indicates that the positive and negative power flows on any tie-line at any time cannot exceed the capacity of the tie-line.

[0144] 5) Introduce the thermal power unit output constraint

[0145]

[0146] Among them, is the lower limit of the output of the thermal power units in the k-th province at time t, is the upper limit of the output of the thermal power units in the k-th province at time t. This constraint indicates that the output value of the thermal power units in any province at any time is between its upper and lower limits.

[0147] 6) Introduce the output constraint of wind turbines

[0148]

[0149] Among them, is the lower limit of the output of the wind turbines in the k-th province at time t, is the upper limit of the output of the wind turbines in the k-th province at time t. This constraint indicates that the output value of the wind turbines in any province at any time is between its upper and lower limits.

[0150] 7) Introduce the output constraint of photovoltaic generators

[0151]

[0152] Among them, is the lower limit of the output of the photovoltaic generators in the k-th province at time t, is the upper limit of the output of the photovoltaic generators in the k-th province at time t. This constraint indicates that the output value of the photovoltaic generators in any province at any time is between its upper and lower limits.

[0153] 8) Introduce the wind curtailment constraint

[0154]

[0155] This constraint indicates that the wind curtailment energy in any province at any time is equal to the maximum output of the wind turbines in this province minus the actual output of the wind turbines in this province.

[0156] 9) Introduce the photovoltaic curtailment constraint

[0157]

[0158] This constraint indicates that the photovoltaic curtailment energy in any province at any time is equal to the maximum output of the photovoltaic generators in this region minus the actual output of the photovoltaic generators in this province.

[0159] Establish the simulation of the northwest power grid power dispatch model P1,

[0160]

[0161] In the preferred embodiment of the described method, constructing the Lagrangian function to calculate the marginal value of the tie-line capacity includes:

[0162] 1) Relax the tie-line capacity constraint Construct the Lagrangian function with the objectives of minimizing the power generation cost, maximizing the consumption of new energy, and minimizing the load shedding respectively.

[0163]

[0164] Wherein, L 1 is used to calculate the marginal value of the tie-line capacity based on reducing the power generation cost, and L 2 is used to calculate the marginal value of the tie-line capacity based on increasing the consumption of new energy, and L 3 is used to calculate the marginal value of the tie-line capacity based on reducing the load shedding.

[0165] 2) Solve the dual problem corresponding to the Lagrangian function to obtain the marginal value of the tie-line capacity.

[0166]

[0167]

[0168]

[0169] Wherein, λ 1 + μ 1 is the marginal value of the tie-line capacity based on reducing the power generation cost, λ 2 + μ 2 is the marginal value of the tie-line capacity based on increasing the consumption of new energy, and λ 3 + μ 3 is the marginal value of the tie-line capacity based on reducing the load shedding.

[0170] In one embodiment, the method has a flowchart as Figure 1 as Figure 2 shown, and includes the following steps:

[0171] Step 1: Obtain the thermal power generation (considering hydropower generation as part of thermal power generation), new energy generation (including wind power generation and photovoltaic power generation), load, DC channel power transmission, and inter-provincial tie-line power flow data of each province, as well as the schematic diagrams of provinces and tie-lines. At the same time, aggregate the provinces into regions, that is, integrate multiple nodes within the province into a single regional node.

[0172] Step 2: According to the thermal power generation, new energy generation, load, and DC channel power transmission data of each province obtained in Step 1, define the grid injection power at the regional level considering new energy generation on the premise of ignoring network losses:

[0173] P zinj = P co + P re - P sd - P 1d(1)

[0174] where P co , P re are thermal power generation and new energy power generation respectively, and P sd , P 1d are the DC channel transmission power and load respectively, and P zinj is the grid injection power at the regional level considering new energy power generation. It should be noted that if the DC channel transmission power P sd and the new energy power generation P re are removed from the above formula, it becomes the formula for calculating the grid injection power at the node level of the traditional node network, and there is no DC channel between this node network and the nodes outside the network.

[0175] Step 3: According to the provincial injection power data obtained in Step 2 and the inter-provincial tie-line power flow data obtained in Step 1, establish a relationship between the two that includes the regional transfer factor

[0176] P fl =ZP zinj (2)

[0177] where is the inter-provincial tie-line power flow (L is the number of inter-provincial tie-lines), is the provincial injection power (K is the number of provinces), is the regional transfer factor matrix. Similar to the definition of the traditional PTDF matrix, the regional transfer factor ZSF is essentially the coefficient of L groups corresponding to K regions in the multiple linear equation system between the regional injection power and the inter-provincial tie-line power flow. Therefore, the regional transfer factor matrix can be obtained by fitting through multiple linear regression analysis of a large amount of historical data. In particular, this equation system does not contain a constant term. And without considering network losses, the sum of the provincial injection powers is zero. To ensure the linear independence of each dependent variable (i.e., the provincial injection powers), Province 1 can be set as the balanced province and not substituted into the fitting process. Correspondingly, the coefficients (the first column) corresponding to Region 1 in the regional transfer factor matrix are all zero.

[0178] Taking a single tie-line as an example, obtain the relationship between its power flow and the provincial injection power

[0179]

[0180] where is the power flow of this tie-line at a certain moment, is the injection power of the i-th (i∈2, 3…K) province at the same moment, and b i is the coefficient corresponding to the i-th (i∈2, 3…K) province, that is, the parameter to be solved. There are sample data at T times. Use the least squares method to solve the multiple linear regression problem of Equation (3), and construct the objective function

[0181]

[0182] where is the theoretical power flow calculated using the fitting area transfer factor, is the actual power flow of the sample. The purpose of the least squares method is to obtain a set of b values that minimize the value of Equation (4). Here, the matrix method of the least squares method is used to solve:

[0183]

[0184] J(b) = 1 / 2(P zinj b - P fl ) T (P zinj b - P fl )

[0185]

[0186] b = ((P zinj ) T P zinj ) -1 (P zinj ) T P fl (5)

[0187] where P zinj ∈R N×K-1 is the area injection power of N samples, b ∈ R K-1 is the coefficient array, corresponding to k - 1 provinces, is the estimated power flow array of N samples, P fl ∈R N is the actual power flow array of N samples, J(b) is defined as the loss function, is the partial derivative of the loss function with respect to the array b, and after sorting, the solution of the coefficient is b = ((P zinj ) T P zinj ) -1 (P zinj ) T P fl . By calculating the coefficient array b of L tie lines and combining them row by row, and finally adding a zero column in the first column, corresponding to the balancing province 1, the area transfer factor matrix can be obtained.

[0188] In step 4, select the time period according to the goodness of fit R 2 :

[0189] To reduce the impact of the uncertainty of tie-line parameters on the regional transfer factor matrix and improve the accuracy of fitting, it is necessary to conduct accuracy tests and screenings on the obtained regional transfer factor matrix, and select the set of regional transfer factor matrices with the highest accuracy to substitute into subsequent model analysis and power flow prediction. The goodness of fit R 2 is often used as an indicator to measure the quality of fitting in linear regression problems

[0190]

[0191] where n is the number of samples, and y i is the true value of the sample, is the fitted value of the sample, is the arithmetic mean of the true values of the samples; represents the sum of squared errors after precise fitting, represents the sum of squared errors after rough fitting using the sample mean. Therefore, R 2 can be expressed as the degree of improvement in the overall error after this precise fitting compared to the rough fitting using the sample mean. R 2 ranges from 0 to 1. The closer it is to 1, the closer the sum of squared errors after fitting is to 0, that is, the higher the fitting accuracy.

[0192] The sample data is divided into N sample groups with different time scales according to the time scale size, and the least squares fitting is performed on the N sample groups respectively to obtain N goodness-of-fit R 2 values, and each group contains L (L is the number of tie lines) values. Compare the N goodness-of-fit R 2 values, and obtain a set of time spans with the highest comprehensive R 2 value, and select the corresponding regional transfer factor matrix and related data to substitute into subsequent model analysis and power flow prediction, which can ensure the accuracy of fitting and subsequent models.

[0193] Step 5: Based on the regional transfer factors obtained in Step 4, with the goal of minimizing the power generation cost, establish a simulation of the northwest power grid power dispatch optimization model P1;

[0194] The optimization goal of model P1 is to minimize the power generation cost, and the objective function can be described as

[0195]

[0196] where, P k，t , They are the power output of thermal power units, wind power curtailment, photovoltaic power curtailment, and load shedding in the k-th province at time t, respectively. f() is the power generation cost of thermal power units, which is generally a function related to time. In the test, it is set as a constant that does not change with time for simplified analysis. B, D, and C are the wind power curtailment penalty, photovoltaic power curtailment penalty, and load shedding penalty respectively. By adding penalty coefficients, the model is made to minimize wind power curtailment, photovoltaic power curtailment, and load shedding as much as possible, meeting the actual operation requirements of the power system. T is the total number of time periods. K is the total number of provinces.

[0197] The constraint conditions of the established model P1 include

[0198] Interconnection line power flow constraint:

[0199] P flow =ZP inj (6)

[0200] Among them, P flow ∈R T×L is the interconnection line power flow matrix, L is the number of interconnection lines, Z is the area transfer factor, which is obtained by fitting a large amount of historical power flow and provincial injection power data in step 4, and P inj ∈R T×K is the provincial injection power matrix. This constraint reflects the relationship between the interconnection line power flow and the provincial injection power.

[0201] Provincial power balance constraint:

[0202]

[0203] Among them, P k，w，t is the wind power output in the k-th province at time t, P k，w，t is the photovoltaic power output in the k-th province at time t, P l，k，t is the electricity transmitted from the l-th interconnection line to the k-th province at time t, P k，d，t is the load in the k-th province at time t. All the motors in each province are equivalent to a thermal power unit, a wind power unit, and a photovoltaic power unit. The DC external transmission channel data of each province is regarded as a part of the load of each province. The positive and negative of the interconnection line power flow are used to represent the direction of the energy transported on the interconnection line. is the set of all time periods, is the set of all provinces. This constraint indicates that the load in any province at any time is equal to the sum of the power output of the thermal power unit, the photovoltaic power unit, the wind power unit, the electricity transmitted from all interconnection lines to this province, and the load shedding in this province.

[0204] Interconnection line capacity constraint:

[0205]

[0206] Among them, P l，maxis the capacity of the l-th tie line, P l，t is the power flow on the l-th tie line at time t, is the set of all tie lines. This constraint indicates that the positive and negative power flows on any tie line at any time cannot exceed the capacity of the tie line.

[0207] Thermal power unit output constraint:

[0208]

[0209] where is the lower limit of the output of thermal power units in the k-th province at time t, is the upper limit of the output of thermal power units in the k-th province at time t. The upper and lower limits of the output of thermal power units are set according to the maximum and minimum values of the output of thermal power units in this area within a month. This constraint indicates that the output value of thermal power units in any province at any time is between its upper and lower limits.

[0210] Wind turbine unit output constraint:

[0211]

[0212] where is the lower limit of the output of wind turbine units in the k-th province at time t, is the upper limit of the output of wind turbine units in the k-th province at time t. The upper limit of the output of wind turbine units is obtained by dividing the wind power generation data of this province at this time by (1 - wind power abandonment rate). The lower limit of the output of wind turbine units is set to 0. This constraint indicates that the output value of wind turbine units in any province at any time is between its upper and lower limits.

[0213] Photovoltaic power unit output constraint:

[0214]

[0215] where is the lower limit of the output of photovoltaic power units in the k-th province at time t, is the upper limit of the output of photovoltaic power units in the k-th province at time t. The upper limit of the output of photovoltaic power units is obtained by dividing the photovoltaic power generation data of this province at this time by (1 - photovoltaic power abandonment rate). The lower limit of the output of photovoltaic power units is set to 0. This constraint indicates that the output value of photovoltaic power units in any province at any time is between its upper and lower limits.

[0216] Wind power abandonment constraint:

[0217]

[0218] This constraint indicates that the wind power abandonment in any province at any time is equal to the maximum value of the output of wind turbine units in this province minus the actual value of the output of wind turbine units in this province.

[0219] Photovoltaic power abandonment constraint:

[0220]

[0221] This constraint indicates that at any time, the curtailment of photovoltaic power in any province is equal to the maximum output of the photovoltaic power generation units in that province minus the actual output of the photovoltaic power generation units in that province.

[0222] In summary, the power dispatching model P1 of the Northwest Power Grid can be established.

[0223]

[0224] Step 6: According to the optimization model P1 established in Step 5, relax the tie-line capacity constraint, construct the Lagrangian function, and calculate the marginal value of the tie-line capacity;

[0225] First, relax the tie-line capacity constraint (8), and construct the Lagrangian function with the objectives of minimizing the generation cost, maximizing the new energy consumption, and minimizing the load shedding respectively

[0226]

[0227] Among them, L 1 is used to calculate the marginal value of the tie-line capacity based on reducing the generation cost, and L 2 is used to calculate the marginal value of the tie-line capacity based on increasing the new energy consumption, and L 3 is used to calculate the marginal value of the tie-line capacity based on reducing the load shedding.

[0228] After that, use the solver gurobi to solve the dual problem corresponding to the Lagrangian function. According to the duality theory, the value of the Lagrange multiplier corresponding to the tie-line constraint is the marginal value of the tie-line capacity.

[0229]

[0230] Among them, λ 1 +μ 1 is the marginal value of the tie-line capacity based on reducing the generation cost, λ 2 +μ 2 is the marginal value of the tie-line capacity based on increasing the new energy consumption, and λ 3 +μ 3 is the marginal value of the tie-line capacity based on reducing the load shedding.

[0231] Embodiment

[0232] To enable those skilled in the art to better understand the present invention, this embodiment uses the typical data of a certain month of the five provinces in the Northwest and 19 tie-lines in the region to verify the effectiveness of the proposed method of the present invention.

[0233] Figure 3 It shows the topological structure diagram of the connection lines among the five northwestern provinces. It can be seen from the figure that there are 17 nodes among the five northwestern provinces. If the traditional PTDF is used for power flow analysis at the node level, the required data is very large, and this type of data is relatively private and not easy to obtain. Therefore, in this embodiment, the five northwestern provinces are transformed into a regional system containing five regions, rather than a node system containing 17 nodes, because the relevant data at the regional level is more concise and easier to obtain.

[0234] First, process the relevant data of the five provinces, and calculate the regional injection power of the five provinces in this month according to the injection power formula at the regional level.

[0235] Select five connection lines for typical analysis. Divide the samples into five sample time groups of single day, five days, fifteen days, twenty days, and thirty-one days according to the time scale. Use the least squares method to fit the regional transfer factors respectively, and obtain five groups of goodness-of-fit R 2 As follows:

[0236] Table 1 Goodness-of-fit R of five connection lines under different time scale samples 2

[0237] <![CDATA[R 2 > Connection Line 5 Connection Line 9 Connection Line 14 Connection Line 16 Connection Line 19 Single day 0.9817 0.9803 0.8843 0.9846 0.9216 Five days 0.9652 0.9726 0.7617 0.9687 0.7731 Fifteen days 0.9573 0.9780 0.7219 0.9603 0.7112 Twenty days 0.9535 0.9741 0.7411 0.9686 0.7456 Thirty-one days 0.9440 0.9733 0.7406 0.9682 0.7825

[0238] It can be seen from the table that the goodness-of-fit R 2 decreases with the increase of the sample time scale, and the R 2 obtained within a single day is the highest. A reasonable explanation is that the uncertainty in the connection line increases with time, resulting in changes in the connection line parameters. To reduce the influence of this uncertainty caused by the time scale and improve the accuracy of fitting the regional transfer factors, select the samples within a single day for fitting the regional transfer factors, and obtain the regional transfer factor matrix and the corresponding R 2 As follows:

[0239] Table 2: Fitted regional transfer factors and goodness-of-fit R within a single day 2

[0240] Area 1 Area 2 Area 3 Area 4 Area 5 <![CDATA[R 2 > Connection Line 1 0 0.0225 -0.2935 0.0330 -0.2006 0.9573 Connection Line 2 0 0.1678 -0.1187 -0.0277 -0.1172 0.9381 Connection Line 3 0 -0.1634 -0.1588 0.0932 -0.1840 0.9793 Connection Line 4 0 -0.1626 -0.1583 0.0927 -0.1830 0.9793 Connection Line 5 0 -0.0634 0.1571 0.0359 0.0155 0.9817 Connection Line 6 0 -0.0632 0.1564 0.0356 0.0148 0.9820 Connection Line 7 0 -0.0632 0.1564 0.0356 0.0148 0.9820 Connection Line 8 0 -0.1923 0.1337 0.0303 0.1310 0.9388 Connection Line 9 0 -0.0172 -0.0066 0.0100 -0.2580 0.9803 Connection Line 10 0 -0.0180 -0.0074 0.0093 -0.2586 0.9794 Connection Line 11 0 0.1039 0.2019 0.0036 0.1404 0.9627 Connection Line 12 0 0.1045 0.2031 0.0038 0.1411 0.9627 Connection Line 13 0 0.0200 -0.2839 0.0312 -0.1939 0.9551 Connection Line 14 0 -0.0842 -0.0571 0.0028 0.2062 0.8843 Connection Line 15 0 -0.0830 -0.0562 0.0035 0.2079 0.8844 Connection Line 16 0 0.0169 0.0065 -0.0105 -0.2423 0.9846 Connection Line 17 0 0.0167 0.0065 -0.0102 -0.2422 0.9854 Connection Line 18 0 0.2793 0.0949 0.0134 0.0646 0.9216 Connection Line 19 0 0.2781 0.0941 0.0130 0.0635 0.9212

[0241] It can be seen from the table that the goodness-of-fit R 2 of the 17 connection lines except connection lines 14 and 15 is greater than 0.92. The R 2 corresponding to the highest connection line 17 reaches 0.9854. And the R 2 corresponding to connection lines 14 and 15 is also above 0.88. Excluding the influence of the internal uncertain factors of the connection lines, it can be considered that the fitted regional transfer factor matrix in this time has high accuracy and can be substituted into the subsequent model analysis and power flow prediction.

[0242] Figure 4 It shows a line graph of the same-day power flow estimation of Tie-line 17 using the region transfer factors obtained by fitting within a single day and comparing with the actual power flow. It can be visually seen from the graph that the fitting has a high accuracy this time.

[0243] Figure 5 It shows a line graph of the next-day power flow prediction of Tie-line 17 using the region transfer factors obtained by fitting within a single day and comparing with the actual power flow. It can be seen from the graph that this method also has a high accuracy in power flow prediction.

[0244] The region transfer factors obtained by fitting are used to calculate the marginal value of the tie-line. The traditional unit power generation cost, new energy abandonment penalty, and load shedding penalty used in this example are shown in the following table

[0245] Table 3: Specific parameter settings of the embodiment

[0246]

[0247] In this example, key attention is paid to the tightened Tie-line 8. Calculate the change in its marginal value based on reducing the power generation cost, increasing the new energy consumption, and reducing the load shedding within a day as the capacity of Tie-line 8 increases from 1000 Mw to 2400 Mw.

[0248] Figures 6 to 8 It shows the marginal value of the capacity of Tie-line 8 based on different aspects. The marginal value of the tie-line capacity is relatively high when the tie-line capacity is low. When the tie-line capacity increases, its marginal value gradually decreases. When it increases to a certain value, the marginal value of the tie-line capacity drops to 0, and at this time, increasing the capacity of the tie-line no longer generates value.

[0249] Although the embodiments of the present invention have been described above in conjunction with the accompanying drawings, the present invention is not limited to the above specific embodiments and application fields. The above specific embodiments are merely illustrative and guiding, rather than restrictive. Those of ordinary skill in the art can also make many forms under the inspiration of this specification and without departing from the scope protected by the claims of the present invention, and all of these fall within the scope of protection of the present invention.

Claims

1. A method for predicting and evaluating the value of inter-provincial tie-line power flow based on data-driven, characterized in that, it includes the following steps: Step 1: Obtain the data of thermal power generation, new energy power generation, load, DC channel power transmission, and inter-provincial tie-line power flow in each province, aggregate the provinces into regions, and integrate multiple nodes within the province into a single regional node; Step 2: According to the data obtained in Step 1, solve the provincial injection power considering new energy power generation; Step 3: According to the provincial injection power and the inter-provincial tie-line power flow data obtained in Step 1, establish a multiple linear regression equation system with a zero constant term, and fit to obtain the area transfer factor; Step 4: Using the goodness of fit as an index, select the area transfer factor corresponding to the time period with the highest fitting accuracy and substitute it into the subsequent analysis; Step 5: According to the area transfer factor corresponding to the time period with the highest fitting accuracy obtained in Step 4, with the minimum power generation cost as the optimization goal, establish a simulated power grid dispatching optimization model P1; Step 6: According to the optimization model P1, relax the tie-line capacity constraint, construct the Lagrangian function, and calculate the marginal value of the tie-line capacity.

2. The method according to claim 1, characterized in that, Preferably, the method for calculating the power saving injection is: determine the objective function as: P zinj = P co + P re - P sd - P ld (1) Among them, P co and P re are thermal power generation and new energy power generation respectively. P sd and P 1d are the DC channel outgoing power and the load respectively. P zinj is the provincial injection power after considering new energy power generation.

3. The method according to claim 2, characterized in that, Fitting the area transfer factor includes, 1) Establish a relationship between provincial injection power and inter-provincial tie-line power flow P fl = ZP zinj (2) Among them is the power flow of the inter-provincial tie line, L is the number of inter-provincial tie lines, is the provincial injection power, K is the number of provinces, is the regional transfer factor matrix, 2) Substitute the sample data at N times into the relationship formula (2), use the least squares method to solve the area transfer factor matrix, and construct a loss function for a single inter-provincial tie-line Among them is the theoretical tidal current calculated using the fitting region transfer factor, is the actual tidal current of the sample, N is the number of samples, and the least squares method aims to obtain a set of parameter values that minimize the loss function value. 3) Use the matrix method of the least squares method to solve the multiple linear regression. Since the sum of the provincial injection powers of each province is zero, set Province 1 as the balancing province, and only need to solve k - 1 parameters, J(b) = 1 / 2(P zinj b - P fl ) T (P zinj b - P fl ) b = ((P zinj ) T P zinj ) -1 (P zinj ) T P fl (3) where \(P\) zinj \(\in\mathbb{R}\) N×K-1 is the injected power for the regions of \(N\) samples, \(b\in\mathbb{R}\) K-1 is the parameter array, corresponding to \(k - 1\) provinces is the estimated power flow array for \(N\) samples, \(P\) fl \(\in\mathbb{R}\) N is the actual power flow array for \(N\) samples, \(J(b)\) is defined as the loss function is the partial derivative of the loss function with respect to the array \(b\), and after arrangement, the solution of the parameter is \(b = ((P\) zinj ) T P zinj ) -1 (P\) zinj ) T P fl , 4) Calculate the parameter array b of L tie-lines and combine them by row. Finally, add a zero column to the first column to obtain the area transfer factor matrix.

4. The method according to claim 3, characterized in that, According to the goodness of fit R 2 Selecting the regional transfer factor for the time period with the highest corresponding fitting accuracy includes: 1) Goodness-of-fit R 2 Used as an indicator to measure the goodness of fit in regression analysis where n is the number of samples, and y i is the true value of the sample, is the fitted value of the sample, is the arithmetic mean of the true values of the samples; represents the sum of squared errors after precise fitting, represents the sum of squared errors after rough fitting using the sample mean, and R 2 ranges from 0 to 1. The closer it is to 1, the closer the sum of squared errors after fitting is to 0, that is, the higher the fitting accuracy. 2) Divide the samples into multiple sample groups with different time spans according to the time scale, 3) Perform least squares fitting on multiple sample groups respectively to obtain the corresponding goodness of fit R 2 value 4) Compare the goodness-of-fit R values corresponding to multiple sample groups, and select the group of samples with the highest R value to obtain the regional transfer factor at the corresponding time scale, and substitute it into the subsequent model analysis and power flow prediction. 2 value, and select the group of samples with the highest R 2 value to obtain the regional transfer factor at the corresponding time scale, and substitute it into the subsequent model analysis and power flow prediction.

5. The method according to claim 4, characterized in that, Establishing the simulated power grid dispatching optimization model P1 includes: 1) Determine the objective function as Among them, P k，t , are respectively the output of thermal power units, wind energy curtailment, photovoltaic energy curtailment, and load shedding in the k-th province at time t. f() is the power generation cost of thermal power units. B, D, and C are respectively the penalties for wind energy curtailment, photovoltaic energy curtailment, and load shedding. T is the total number of time periods, and K is the total number of regions. 2) Introduce the tie-line power flow constraint P flow = ZP inj (6) where, P flow ∈ R T×L is the tie-line power flow matrix, L is the number of tie-lines, Z is the area transfer factor obtained in Step 4, P inj ∈ R T×K is the provincial injection power matrix, and this constraint reflects the relationship between tie-line power flow and provincial injection power 3) Introduce the provincial power balance constraint Among them, P k，w，t is the wind power output of the k-th province at time t, and P k，v，t is the photovoltaic power output of the k-th province at time t, and P l，k，t is the amount of electricity transmitted to the k-th province by the l-th tie line at time t, and P k，d，t is the load of the k-th province at time t. Here, the power transmitted by the DC channel is regarded as a part of the load, is the set of all time periods, is the set of all provinces. This constraint indicates that the load of any province at any time is equal to the sum of the thermal power generation output, photovoltaic power generation output, wind power generation output of the province, the electricity transmitted by all tie lines to the province, and the load shedding of the province. 4) Introduce the tie-line capacity constraint where, P l，max is the capacity of the l-th tie line, and P l，t is the power flow on the l-th tie line at time t. is the set of all tie lines. This constraint indicates that the positive and negative power flows on any tie line at any time cannot exceed the capacity of the tie line. 5) Introduce the thermal power unit output constraint Among them, is the lower limit of the output of the thermal power units in the k-th province at time t, is the upper limit of the output of the thermal power units in the k-th province at time t. This constraint indicates that the output value of the thermal power units in any province at any time is between its upper and lower limits. 6) Introduce the wind turbine unit output constraint Among them, is the lower limit of the output of wind turbines in the k-th province at time t, is the upper limit of the output of wind turbines in the k-th province at time t. This constraint indicates that the output value of wind turbines in any province at any time is between its upper and lower limits. 7) Introduce the photovoltaic unit output constraint Among them, is the lower limit of the output of the photovoltaic and wind turbine units in the k-th province at time t, is the upper limit of the output of the photovoltaic and wind turbine units in the k-th province at time t. This constraint indicates that the output value of the photovoltaic and wind turbine units in any province at any time is between its upper and lower limits. 8) Introduce the wind curtailment constraint This constraint indicates that the wind curtailment energy of any province at any time is equal to the maximum output of the wind turbines in that province minus the actual output of the wind turbines in that province, 9) Introduce the photovoltaic curtailment constraint This constraint indicates that the photovoltaic curtailment energy of any province at any time is equal to the maximum output of the photovoltaic units in that province minus the actual output of the photovoltaic units in that province, Establish a simulated power grid dispatching model P1 for the Northwest Power Grid, 6. The method according to claim 1, characterized in that, Constructing the Lagrangian function to calculate the marginal value of the tie-line capacity includes: 1) Relax the tie-line capacity constraint Construct the Lagrangian function with the objectives of minimizing the generation cost, maximizing the new energy consumption, and minimizing the load shedding respectively L 1 = f 1 + λ 1 (-P l，max - P l，t ) + μ 1 (P l，t - P l，max ) L 2 = f 2 + λ 2 (-P l，max - P l，t ) + μ 2 (P l，t - P l，max ) L 3 = f 3 + λ 3 (-P l，max - P l，t ) + μ 3 (P l，t - P l，max ) Among them, L 1 used to calculate the marginal value of tie-line capacity based on reducing generation costs, L 2 used to calculate the marginal value of tie-line capacity based on improving new energy consumption, L 3 used to calculate the marginal value of tie-line capacity based on reducing load shedding 2) Solve the dual problem corresponding to the Lagrangian function to obtain the marginal value of the tie-line capacity, Among them, λ 1 + μ 1 is the marginal value of the tie-line capacity based on reducing the generation cost, λ 2 + μ 2 is the marginal value of the tie-line capacity based on improving the accommodation of new energy, and λ 3 + μ 3 is the marginal value of the tie-line capacity based on reducing the load shedding.

Images