A method, device and storage medium for online distribution of active power of power grid generator sets

Abstract

The present invention discloses a method, device, and storage medium for online distribution of active power of power grid generator sets. The method comprises: obtaining upper and lower thresholds of predicted power grid output to determine power grid generator set output constraints; determining prediction errors based on the predicted power grid output upper and lower thresholds to calculate the total power grid output; obtaining the number of power grid flow constraints and constructing power grid generator set flow constraints based on the total power grid output; obtaining the total number of power grid generator sets and the number of power grid startup constraints to construct power grid generator set startup constraints; and randomly distributing active power for all generator sets in a desired online data sample of each power grid operation mode based on the phased active power distribution of the power grid generator sets, taking into account the power grid generator set output constraints, power grid generator set flow constraints, and power grid generator set startup constraints. The active power distribution values of the power grid generator sets generated by the present invention can meet randomness requirements and have high generation efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of power system automation control, and in particular to a method, a device and a storage medium for online distribution of active power of power grid generator sets. Background Art

[0002] In recent years, with the gradual advancement of artificial intelligence (AI) research in power grid simulation and analysis, the demand for power grid simulation sample data has increased significantly. On the one hand, the training of most AI models requires massive sample sizes, typically in deep learning. On the other hand, to avoid model overfitting, certain sample distribution requirements are also in place. Currently, existing power grid operation mode samples primarily come from online and offline data. Online and offline data samples each have their own advantages and disadvantages: Online data samples are automatically generated from actual operating modes, resulting in a large sample size but uneven distribution, many similar samples, and low representativeness. Offline data samples are manually adjusted from operating modes, resulting in high representativeness but a small sample size that fails to cover all power grid operating conditions. The quantity and quality of online and offline data samples are insufficient to support the requirements of AI algorithms. To better conduct data analysis and power grid characterization research, effective sample supplementation and screening should be conducted based on the distribution characteristics of the power grid data itself, increasing sample diversity and balance. This will facilitate the effective extraction of power grid operation patterns and provide sufficient sample data support for AI technologies.

[0003] Generating operational model samples suitable for AI applications should at least include determining the total grid load distribution, allocating active power to generators, distributing loads at access points, and allocating reactive power to generators according to specific load levels, and automatically adjusting convergence. Generating completely random active power allocation schemes under certain constraints is a fundamental and crucial issue.

[0004] Currently, the most direct method for randomly allocating active power to power grid units is to determine the total load on the grid, then have each generator unit independently generate active power distribution according to a completely random process. The process then determines whether the constraints are met and discards any that do not. This method generates a large number of invalid generator unit output arrangements, which are unlikely to meet the total output or power flow constraints of the generator units. As the scale of the power grid continues to expand, the constraints on the output distribution of the grid generator units continue to increase, making it increasingly unsuitable for each generator unit to independently distribute active power according to a completely random process. Summary of the Invention

[0005] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method, device and storage medium for online distribution of active power of power grid generator sets to solve the technical problem of low efficiency of the existing method for distributing active power of power grid generator sets. A method for generating online data samples of power grid operation mode is also provided to solve the technical problems of low typicality of online data samples and slow generation speed of offline data samples.

[0006] In order to solve the above technical problems, the present invention is implemented by adopting the following technical solutions:

[0007] In a first aspect, the present invention provides a method for online distribution of active power of power generation units in a power grid, comprising:

[0008] Obtain the upper and lower thresholds of the power grid's predicted output to determine the power grid generator set output constraints;

[0009] Determining the prediction error and calculating the total power grid output according to the upper and lower thresholds of the power grid prediction output;

[0010] Obtaining the number of power flow constraints of the power grid, and constructing the power flow constraints of the power grid generator units based on the total power grid output, the distribution of the generator units in the power grid power flow constraints, and the upper and lower output thresholds of the power grid power flow constraints;

[0011] Obtain the total number of power grid generator sets and the number of power grid startup constraints, and construct power grid generator set startup constraints;

[0012] According to the phased active power distribution of the power grid generator sets, the active power is randomly distributed for all the generator sets in the online data samples of each power grid operation mode to be generated, taking into account the output constraints, power flow constraints and startup constraints of the power grid generator sets.

[0013] Furthermore, the method for determining the prediction error and calculating the total power grid output according to the upper and lower thresholds of the power grid predicted output includes:

[0014] According to the upper and lower thresholds of the power grid predicted output, the maximum range of the power grid predicted output is determined according to the maximum prediction error;

[0015] Formula (1) is used to classify the maximum range of the predicted power output of the power grid:

[0016]

[0017] Where: roundup(·) represents the rounding function; δ% represents the maximum prediction error; λ represents the preset bin interval; P lmax Indicates the upper threshold of the power grid forecast output, P lmin Indicates the lower threshold of the predicted power output of the power grid;

[0018] According to formula (2), the total power output P of the power grid is calculatedz :

[0019] P z =(1-δ%)P min +z×λ (2)

[0020] Where: P z Indicates the total power output of the power grid, z=1,2…Z; P min is the minimum active output constraint matrix of the power grid generator set: P min =(P 1min , P 2min ...P imin ...P Nmin ), N is the total number of power grid generators, P imin Represents the lower threshold of active output of generator set i, i=1,2…N.

[0021] Furthermore, the constructing of the power grid generator set startup constraints includes constructing a minimum startup number constraint matrix, constructing a maximum startup number constraint matrix, and constructing a power grid generator set startup constraint matrix.

[0022] Furthermore, the minimum boot number constraint matrix St min And the maximum number of boot constraints matrix St max They are:

[0023] St min =(St 1min ...St omin ...St Omin ) (3)

[0024] St max =(St 1max ...St omax ...St Omax ) (4)

[0025] Where: St i Indicates the start / stop status of generator set i; when St i =1 indicates that generator set i is put into operation. St omax are the minimum and maximum startup quantities of the power grid startup constraint o, respectively.

[0026] Furthermore, the power grid generator startup constraint matrix OSt is:

[0027]

[0028] Where, OSt NO Indicates whether the generator set N is within the grid startup constraint o. If it is within the grid startup constraint o, then OSt NOTake 1; otherwise, OSt NO Take 0.

[0029] Furthermore, the constructing of power grid generator set flow constraints includes constructing a power grid flow maximum constraint matrix, constructing a power grid flow minimum constraint matrix, and constructing a power grid flow constraint matrix.

[0030] Furthermore, the grid flow maximum constraint matrix F max and the minimum constraint matrix F of the power flow min They are:

[0031] F max =(P z ,P R1max ,P R2max ···P Rfmax ···P RFmax ) (6)

[0032] F min =(P z ,P R1min ,P R2min ···P Rfmin ···P RFmin ) (7)

[0033] Where: R Rfmax and R Rfmin They are the upper and lower thresholds of the power flow constraint f, and they satisfy P i is the active power output of generator set i, and the number of grid flow constraints is F, R f represents the set of generators constrained by the fth power flow, f = 1, 2…F.

[0034] Furthermore, the power grid flow constraint matrix FSt is:

[0035]

[0036] Where: FSt NF Indicates whether the generator set N is within the grid flow constraint f. If it is within the grid flow constraint f, then FSt NF Take 1; otherwise, FSt NF Take 0; the first column of FSt represents the total output constraint of the power grid generator sets.

[0037] Furthermore, after constructing the power-on constraints of the power grid generator sets, K power-on state matrices of the power grid generator sets that simultaneously meet the power-on constraints of the power grid generator sets and the power flow constraints are generated according to the number K of online data samples of the power grid operation mode that need to be generated, and the K power-on state matrices of the power grid generator sets are merged into the power-on and shutdown matrix of the power grid generator sets.

[0038] Furthermore, a method for generating a startup state matrix of K generator sets that simultaneously satisfies the startup constraints of the power grid generator sets and the power grid flow constraints includes the following steps:

[0039] According to the operation probability of the generator set in the online data sample of the current grid operation mode, the startup status of the generator set in the online data sample of each grid operation mode is determined by using random numbers, and the startup status matrix of the generator set is preliminarily generated:

[0040] ST k =(ST k1 ,ST k2 ..ST ki ..ST kN ) (9)

[0041] Where: ST ki is the startup status of generator set i in sample k, k = 1, 2…K;

[0042] Combine the initially generated generator set startup state matrix and the grid generator set startup constraints to construct judgment matrix A and judgment matrix B. Combine the initially generated generator set startup state matrix and the grid power flow constraints to construct judgment matrix C and judgment matrix D:

[0043] A=ST k OSt-St min (10)

[0044] B=St max -ST k OSt (11)

[0045] C=ST k FSt-F min (12)

[0046] D=F max -ST k FSt (13)

[0047] Where:

[0048] If all elements in matrices A, B, C, and D are greater than or equal to zero, the corresponding generator set startup state matrix is retained; otherwise, the corresponding generator set startup state matrix is discarded.

[0049] Repeat the above steps until K generator set startup state matrices are generated.

[0050] Furthermore, the startup state ST of generator set i in sample k ki Use formula (14) to calculate:

[0051]

[0052] np k =npf(0,1)

[0053] ST pki =rand(1,1) (14)

[0054] Where: np k is the probability of the unit in sample k being put into operation; npf(0,1) generates a random number evenly distributed between [0,1]; rand(1,1) generates a random number evenly distributed between [0,1] with one row and one column.

[0055] Furthermore, active power is randomly allocated to all generator sets in the online data samples of each grid operation mode to be generated, including:

[0056] Construct the phased allocation matrix TP of generator set i i ;

[0057] Construct the unassigned matrix TS of generator set i i ;

[0058] Based on matrix TP i and matrix TS i Calculate the maximum allocation matrix E of generator set i cmax and the minimum allocation matrix E cmin :

[0059]

[0060] Where, E ar represents the amount of power that has been allocated in different power flow constraints f; E umax and E umin They represent the maximum power schedulable value and the minimum power schedulable value of the unassigned units under different power flow constraints f; repmat(P max , K,1) means that P max Copy it into a matrix of K rows and 1 column; P max is the maximum active output constraint matrix of the power grid generator set: P max =(P 1max ,P 2max ...P imax …P Nmax ); where P imax Represents the upper threshold of active output of generator set i; repmat(P min , K,1) means that P min Copy it into a matrix of K rows and 1 column;

[0061] Based on the maximum allocation matrix E of generator set i cmax and the minimum allocation matrix Ecmin , calculate and determine the power range of the generator set i output in the online data sample k of the grid operation mode [P ai ,P bi ]:

[0062]

[0063] Where, FSt(i,:) means taking out the i-th row of FSt as a new matrix, Rmin() means taking out the minimum value of each row to form a new matrix, Rmax() means taking out the maximum value of each row to form a new matrix, min() means taking out the minimum value of the elements in the two matrices to form a new matrix, and max() means taking out the maximum value of the elements in the two matrices to form a new matrix; Rai and Rbi are the minimum and maximum power outputs of generator set i in sample k, respectively.

[0064] Calculate the phased power distribution matrix TP of generator set i+1 i+1 for:

[0065] TP i+1 =(TS i -TS i+1 )(npF(K,1,0,1)*(Pai-Pbi)+Pbi)+TP i (17)

[0066] Where: npf(K,1,0,1) generates a matrix with K rows and 1 column randomly distributed in [0,1];

[0067] According to the calculation of the stage distribution power matrix TP of generator set i+1 i+1 The active power distribution value of the generator set in K samples is calculated one by one using the method.

[0068] In a second aspect, the present invention provides an online active power distribution device for power grid generator sets, comprising:

[0069] An acquisition module is used to obtain upper and lower thresholds of the power grid forecast output to determine the power grid generator set output constraints;

[0070] A calculation module, configured to determine a prediction error and calculate a total power grid output according to upper and lower thresholds of the power grid predicted output;

[0071] A power flow constraint module is used to obtain the number of power grid power flow constraints, and to construct the power grid generator power flow constraints based on the total power grid output, the distribution of the generator sets in the power grid power flow constraints, and the upper and lower output thresholds of the power grid power flow constraints;

[0072] The startup constraint module is used to obtain the total number of power grid generator sets and the number of power grid startup constraints, and to construct the startup constraints of power grid generator sets;

[0073] The power allocation module is used to randomly allocate active power to all generator sets in the online data samples of each grid operation mode to be generated based on the phased active power distribution of the grid generator sets, taking into account the output constraints of the grid generator sets, the grid power generation flow constraints, and the grid generator set startup constraints.

[0074] Furthermore, the online active power distribution device for the power grid generator set further comprises:

[0075] The generator set startup and shutdown module is used to generate K generator set startup state matrices that simultaneously meet the grid generator set startup constraints and grid power flow constraints according to the number K of online data samples of grid operation mode that need to be generated, and merge the K generator set startup state matrices into the generator set startup and shutdown matrix.

[0076] Furthermore, the generator set startup and shutdown module includes:

[0077] The generator set operation unit is used to determine the generator set's startup status in each grid operation mode online data sample based on the generator set's startup probability in the current grid operation mode online data sample using random numbers, and preliminarily generate the generator set startup status matrix:

[0078] ST k =(ST k1 ,ST k2 ..ST ki ..ST kN ) (18)

[0079] Where: ST ki is the startup status of generator set i in sample k, k = 1, 2…K;

[0080] And, for calculating the startup state ST of generator set i in sample k using formula (19): ki :

[0081]

[0082] np k =npf(0,1)

[0083] ST pki =rand(1,1) (19)

[0084] Where: np k is the probability of the unit in sample k being put into operation; npf(0,1) generates a random number evenly distributed between [0,1]; rand(1,1) generates a random number evenly distributed between [0,1] with 1 row and 1 column;

[0085] The judgment unit is used to construct the judgment matrix A and the judgment matrix B by combining the initially generated generator set startup state matrix and the grid generator set startup constraints, and to construct the judgment matrix C and the judgment matrix D by combining the initially generated generator set startup state matrix and the grid power flow constraints:

[0086] A=ST k OSt-St min (20)

[0087] B=St max -ST k OSt (21)

[0088] C=ST k FSt-F min (twenty two)

[0089] D=F max -ST k FSt (23)

[0090] Where: If all elements in matrices A, B, C, and D are greater than or equal to zero, the corresponding generator set startup state matrix is retained; otherwise, the corresponding generator set startup state matrix is discarded;

[0091] Repeat the judgment steps until K generator set startup state matrices are generated.

[0092] Furthermore, the power distribution module includes:

[0093] The first construction unit is used to construct the phased allocation matrix TP of the generator set i i ;

[0094] The second construction unit is used to construct the unassigned matrix TS of the generator set i i ;

[0095] The first calculation unit is used to calculate the i and matrix TS i Calculate the maximum allocation matrix E of generator set i cmax and the minimum allocation matrix E cmin :

[0096]

[0097] Where, E ar represents the amount of power that has been allocated in different power flow constraints f; E umax and E umin They represent the maximum power schedulable value and the minimum power schedulable value of the unassigned units under different power flow constraints f; repmat(P max , K,1) means that Pmax Copy it into a matrix of K rows and 1 column; P max is the maximum active output constraint matrix of the power grid generator set: P max =(P 1max ,P 2max …P imax ...P Nmax ); where P imax Represents the upper threshold of active output of generator set i; repmat(P min , K,1) means that P min Copy it into a matrix of K rows and 1 column;

[0098] The second calculation unit is used to calculate the maximum allocation matrix E of the generator set i. cmax and the minimum allocation matrix E cmin , calculate and determine the power range of the generator set i output in the online data sample k of the grid operation mode [P ai ,P bi ]:

[0099]

[0100] Where, FSt(i,:) means taking out the i-th row of FSt as a new matrix, Rmin() means taking out the minimum value of each row to form a new matrix, Rmax() means taking out the maximum value of each row to form a new matrix, min() means taking out the minimum value of the elements in the two matrices to form a new matrix, and max() means taking out the maximum value of the elements in the two matrices to form a new matrix; Rai and Rbi are the minimum and maximum power outputs of generator set i in sample k, respectively.

[0101] The third calculation unit is used to calculate the phased distribution power matrix TP of the generator set i+1 i+1 for:

[0102] TP i+1 =(TS i -TS i+1 )(npF(K,1,0,1)*(Pai-Pbi)+Pbi)+TP i (26)

[0103] Where: npf(K,1,0,1) generates a matrix with K rows and 1 column randomly distributed in [0,1];

[0104] According to the calculation of the stage distribution power matrix TP of generator set i+1 i+1 The active power distribution value of the generator set in K samples is calculated one by one using the method.

[0105] In a third aspect, the present invention provides an online distribution device for active power of power generation sets in a power grid, comprising a processor and a storage medium;

[0106] The storage medium is used to store instructions;

[0107] The processor is configured to operate according to the instructions to execute the steps of the method for online distribution of active power of power generation sets in a power grid as described in the first aspect.

[0108] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for online distribution of active power of power generation sets in a power grid as described in the first aspect.

[0109] In a fifth aspect, the present invention provides a method for generating online data samples of power grid operation modes, comprising the steps of the method for online distribution of active power of power generation sets in the power grid as described in any one of the first aspects.

[0110] Compared with the prior art, the present invention has the following beneficial effects:

[0111] The present invention constructs power flow constraints and power-on constraints for power grid generator sets based on multiple constraints on the active output of power grid units, and randomly distributes active power to all generator sets in the online data samples of each power grid operation mode to be generated by using a matrix iterative algorithm of a random function; it has the advantages of higher distribution efficiency, stronger randomness, and the ability to meet multiple types of constraints at the same time while improving efficiency; in addition, the present invention effectively supplements and screens samples, increases sample diversity and balance, does not cause deviation in the stability characteristic analysis results, and provides conditions for more effectively extracting power grid operation laws. BRIEF DESCRIPTION OF THE DRAWINGS

[0112] Figure 1 The figure shows a structural block diagram of an online active power distribution device for power grid generator sets provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0113] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. Unless there is a conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0114] The term "and / or" in this document simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone. Additionally, the character " / " in this document generally indicates an "or" relationship between the related objects.

[0115] Example 1:

[0116] This embodiment introduces a method for online distribution of active power of power generation units in a power grid, which specifically includes the following steps:

[0117] Step 1: Obtain the upper and lower thresholds of the power grid forecast output to determine the power grid generator output constraints;

[0118] Step 2: Determine the prediction error and calculate the total power grid output according to the upper and lower thresholds of the power grid prediction output;

[0119] Step 3: Obtain the number of power flow constraints, and construct the power flow constraints of the power grid generator units based on the total power grid output, the distribution of the generator units in the power flow constraints, and the upper and lower output thresholds of the power flow constraints;

[0120] Step 4: Obtain the total number of power grid generator sets and the number of power grid startup constraints, and construct power grid generator set startup constraints;

[0121] Step 5: According to the phased active power distribution of the power grid generator sets, considering the output constraints, power flow constraints, and startup constraints of the power grid generator sets, the active power is randomly distributed for all the generator sets in the online data samples of each power grid operation mode to be generated.

[0122] In the online active power distribution method for power grid generator sets provided by the embodiment of the present invention, the active power output range of the power grid generator set i satisfies: P imin ≤P i ≤P imax ; Where: the total number of generator sets is N, i=1,2…N; P imin and P imax are the lower threshold value and the upper threshold value of the active output of the generator set i respectively; based on the active output range of the power grid generator set i, the power grid generator set output constraint in the embodiment of the present invention includes the minimum active output constraint matrix of the power grid generator set: P min =(P 1min , P 2min ...P imin ...P Nmin ), and the maximum active output constraint matrix of the power grid generator set: P max =(P 1max ,P 2max ...Pimax …P Nmax ).

[0123] As an embodiment of the present invention, a method for calculating the total power grid output by determining a prediction error based on upper and lower thresholds of the power grid predicted output includes:

[0124] According to the upper and lower thresholds of the power grid predicted output, the maximum range of the power grid predicted output is determined according to the maximum prediction error;

[0125] Formula (1) is used to classify the maximum range of the predicted power output of the power grid:

[0126]

[0127] Where: roundup(·) represents the rounding function; δ% represents the maximum prediction error; λ represents the preset bin interval; P lmax Indicates the upper threshold of the power grid forecast output, P lmin Indicates the lower threshold of the predicted power output of the power grid;

[0128] According to formula (2), the total power output P of the power grid is calculated z :

[0129] P z =(1-δ%)P min +z×λ (2)

[0130] Where: P z Indicates the total power grid output, z=1,2…Z;

[0131] Specifically, the grid load prediction error range is conventionally set at 0.5%-2%. In the embodiment of the present invention, the maximum range of the grid predicted output is determined according to the maximum load prediction error of 2%. In addition, the preset grading interval in the embodiment of the present invention is grading according to 1000MW. The grading interval of 1000MW is small enough for the grid. Of course, the grading interval can also be preset to other values such as 1200MW or 1300MW according to the specific grid operation conditions. Therefore, the total grid output P in the embodiment of the present invention is z The calculation formula is as follows:

[0132] P z =(1-2%)P min +z×1000 (2')

[0133] In the online active power distribution method for power grid generator sets provided by an embodiment of the present invention, constructing power grid generator set startup constraints includes constructing a minimum startup number constraint matrix, constructing a maximum startup number constraint matrix, and constructing a power grid generator set startup constraint matrix.

[0134] Specifically, the minimum power-on number constraint matrix St in the embodiment of the present invention ismin And the maximum number of boot constraints matrix St max They are:

[0135] St min =(St 1min ...St omin ...St Omin ) (3)

[0136] St max =(St 1max ...St omax ...St Omax ) (4)

[0137] Where: St i Indicates the start / stop status of generator set i; when St i =1 indicates that generator set i is put into operation. St omax are the minimum and maximum startup quantities of the power grid startup constraint o, respectively.

[0138] In the embodiment of the present invention, the power generation unit startup constraint matrix OSt is:

[0139]

[0140] Where, OSt NO Indicates whether the generator set N is within the grid startup constraint o. If it is within the grid startup constraint o, then OSt NO Take 1; otherwise, OSt NO Take 0

[0141] In the online active power distribution method for power grid generator sets provided by an embodiment of the present invention, constructing power grid generator set flow constraints includes constructing a power grid flow maximum constraint matrix, constructing a power grid flow minimum constraint matrix, and constructing a power grid flow constraint matrix.

[0142] Specifically, in the embodiment of the present invention, the maximum constraint matrix F of the power grid flow is max and the minimum constraint matrix F of the power flow min They are:

[0143] F max =(P z ,P R1max ,P R2max ···P Rfmax ···P RFmax ) (6)

[0144] F min =(P z ,P R1min ,P R2min ···PRfmin ···P RFmin ) (7)

[0145] Where: R Rfmax and R Rfmin They are the upper and lower thresholds of the power flow constraint f, and they satisfy P i is the active power output of generator set i, and the number of grid flow constraints is F, R f represents the set of generators constrained by the fth power grid flow, f = 1, 2…F.

[0146] The power grid flow constraint matrix FSt in the embodiment of the present invention is:

[0147]

[0148] Where: FSt NF Indicates whether the generator set N is within the grid flow constraint f. If it is within the grid flow constraint f, then FSt NF Take 1; otherwise, FSt NF Take 0; the first column of FSt represents the total output constraint of the power grid generator sets.

[0149] As an embodiment of the present invention, after constructing the power-grid generator set startup constraints, K generator set startup state matrices that simultaneously meet the power-grid generator set startup constraints and the power-grid flow constraints are generated according to the number K of online data samples of the power-grid operation mode that need to be generated, and the K generator set startup state matrices are merged into a generator set startup and shutdown matrix.

[0150] In an embodiment of the present invention, a method for generating a startup state matrix of K generator sets that simultaneously satisfies both the startup constraints of the power grid generator sets and the power grid flow constraints comprises the following steps:

[0151] According to the operation probability of the generator set in the online data sample of the current grid operation mode, the startup status of the generator set in the online data sample of each grid operation mode is determined by using random numbers, and the startup status matrix of the generator set is preliminarily generated:

[0152] ST k =(ST k1 ,ST k2 ..ST ki ..ST kN ) (9)

[0153] Where: ST ki is the startup status of generator set i in sample k, k = 1, 2…K;

[0154] Combine the initially generated generator set startup state matrix and the grid generator set startup constraints to construct judgment matrix A and judgment matrix B. Combine the initially generated generator set startup state matrix and the grid power flow constraints to construct judgment matrix C and judgment matrix D:

[0155] A=ST k OSt-St min (10)

[0156] B=St max -ST k OSt (11)

[0157] C=ST k FSt-F min (12)

[0158] D=F max -ST k FSt (13)

[0159] Where:

[0160] If all elements in matrices A, B, C, and D are greater than or equal to zero, the corresponding generator set startup state matrix is retained; otherwise, the corresponding generator set startup state matrix is discarded.

[0161] Repeat the above steps until K generator set startup state matrices are generated.

[0162] Among them, the starting state ST of generator set i in sample k involved in formula (9) is ki Use formula (14) to calculate:

[0163]

[0164] np k =npf(0,1)

[0165] ST pki =rand(1,1) (14)

[0166] Where: np k is the probability of the unit in sample k being put into operation; rand(1,1) generates a random number with 1 row and 1 column uniformly distributed between [0,1]; npf(0,1) generates a random number evenly distributed between [0,1];

[0167] It should be further explained that npf(0,1) can select uniform distribution, normal distribution, etc. according to data requirements. If the current power grid operation status requires generating all operation modes, uniform distribution is selected. If the current power grid operation status requires generating an operation mode near a certain operation point, normal distribution is selected.

[0168] As an embodiment of the present invention, performing random active power distribution for all generator sets in the generated online data samples of each grid operation mode specifically includes the following steps:

[0169] Step a: Construct the phased allocation matrix TP of generator set i i :

[0170]

[0171] Among them, P k1 ,P k2 …P ki-1 represents the output allocation value of the generator set in sample k, and the generator set with unallocated output is set to 0;

[0172] Step b: Construct the unassigned matrix TS of generator set i i :

[0173]

[0174] Among them, TS k1 …TS ki Indicates that in sample k, the generator set that has been allocated or is being allocated output is assigned a value of 0; the generator set that has not been allocated power is assigned a value of 1; TS1 is a matrix of all 1s, TS N is an all-zero matrix, and the others are generated by analogy at once.

[0175] Step c: Based on matrix TP i and matrix TS i Calculate the maximum allocation matrix E of generator set i cmax and the minimum allocation matrix E cmin :

[0176]

[0177] Where, E ar represents the amount of power that has been allocated in different power flow constraints f; E umax and E umin They represent the maximum power schedulable value and the minimum power schedulable value of the unassigned units under different power flow constraints f; repmat(P max , K,1) means that P max , copied into a matrix of K rows and 1 column; P max is the maximum active output constraint matrix of the power grid generator set: P max =(P 1max ,P 2max …P imax …P Nmax ); where P imaxRepresents the upper threshold of active output of generator set i; repmat(P min , K,1) means that P min Copy it into a matrix of K rows and 1 column;

[0178] Step d: Based on the maximum allocation matrix E of generator set i cmax and the minimum allocation matrix E cmin , calculate and determine the power range of the generator set i output in the online data sample k of the grid operation mode [P ai ,P bi ]:

[0179]

[0180] Where, FSt(i,:) means taking out the i-th row of FSt as a new matrix, Rmin() means taking out the minimum value of each row to form a new matrix, Rmax() means taking out the maximum value of each row to form a new matrix, min() means taking out the minimum value of the elements in the two matrices to form a new matrix, and max() means taking out the maximum value of the elements in the two matrices to form a new matrix; Rai and Rbi are the minimum and maximum power outputs of generator set i in sample k, respectively.

[0181] Step e: Calculate the phased power distribution matrix TP of generator set i+1 i+1 for:

[0182] TP i+1 =(TS i -TS i+1 )(npF(K,1,0,1)*(Pai-Pbi)+Pbi)+TP i (17)

[0183] Where: npf(K,1,0,1) generates a matrix with K rows and 1 column randomly distributed in [0,1];

[0184] Step f: Calculate the phased power matrix TP of generator set i+1 i+1 The method is used to calculate the active power distribution value of the generator set in K samples one by one; finally, the stage distribution power matrix TP is calculated. N+1 , thus completing the active power distribution of all generator sets in sample K.

[0185] The embodiment of the present invention provides an online active power distribution method for power grid generator sets. The method determines the output constraints of the power grid generator sets and constructs the power flow constraints and startup constraints of the power grid generator sets based on the various constraints on the active power output of the power grid generator sets. The method calculates the power distribution of the generator set i in K samples in succession by using a matrix iteration algorithm of a random function to complete the active power distribution of the power grid generator sets. The method has stronger randomness and higher distribution efficiency. While improving efficiency, it can meet multiple types of constraints and will not cause deviations in the stability characteristic analysis results, thus providing conditions for more effectively extracting the operation laws of the power grid.

[0186] Example 2:

[0187] like Figure 1 As shown, the embodiment of the present invention provides an online distribution device for active power of power generation sets in a power grid, which can be used to implement the method described in the first embodiment, specifically including:

[0188] An acquisition module is used to obtain upper and lower thresholds of the power grid forecast output to determine the power grid generator set output constraints;

[0189] A calculation module, configured to determine a prediction error and calculate a total power grid output according to upper and lower thresholds of the power grid predicted output;

[0190] A power flow constraint module is used to obtain the number of power grid power flow constraints, and to construct the power grid generator power flow constraints based on the total power grid output, the distribution of the generator sets in the power grid power flow constraints, and the upper and lower output thresholds of the power grid power flow constraints;

[0191] The startup constraint module is used to obtain the total number of power grid generator sets and the number of power grid startup constraints, and to construct the startup constraints of power grid generator sets;

[0192] The power allocation module is used to randomly allocate active power to all generator sets in the online data samples of each grid operation mode to be generated based on the phased active power distribution of the grid generator sets, taking into account the output constraints of the grid generator sets, the grid power generation flow constraints, and the grid generator set startup constraints.

[0193] As an embodiment of the present invention, the online active power distribution device of the power grid generator set further includes:

[0194] The generator set startup and shutdown module is used to generate K generator set startup state matrices that simultaneously meet the grid generator set startup constraints and grid power flow constraints according to the number K of online data samples of grid operation mode that need to be generated, and merge the K generator set startup state matrices into the generator set startup and shutdown matrix.

[0195] Specifically, the generator set startup and shutdown module in the embodiment of the present invention includes:

[0196] The generator set operation unit is used to determine the generator set's startup status in each grid operation mode online data sample based on the generator set's startup probability in the current grid operation mode online data sample using random numbers, and preliminarily generate the generator set startup status matrix:

[0197] ST k =(ST k1 ,ST k2 ..ST ki ..ST kN ) (18)

[0198] Where: ST ki is the startup status of generator set i in sample k, k = 1, 2…K;

[0199] And, for calculating the startup state ST of generator set i in sample k using formula (19): ki :

[0200]

[0201] np k =npf(0,1)

[0202] ST pki =rand(1,1) (19)

[0203] Where: np k is the probability of the unit in sample k being put into operation; npf(0,1) generates a random number evenly distributed between [0,1]; rand(1,1) generates a random number evenly distributed between [0,1] with 1 row and 1 column;

[0204] The judgment unit is used to construct the judgment matrix A and the judgment matrix B by combining the initially generated generator set startup state matrix and the grid generator set startup constraints, and to construct the judgment matrix C and the judgment matrix D by combining the initially generated generator set startup state matrix and the grid power flow constraints:

[0205] A=ST k OSt-St min (20)

[0206] B=St max -ST k OSt (21)

[0207] C=ST k FSt-F min (twenty two)

[0208] D=F max -ST k FSt (23)

[0209] Where: If all elements in matrices A, B, C, and D are greater than or equal to zero, the corresponding generator set startup state matrix is retained; otherwise, the corresponding generator set startup state matrix is discarded;

[0210] Repeat the judgment steps until K generator set startup state matrices are generated.

[0211] As an embodiment of the present invention, the rate distribution module in the online active power distribution device of the power grid generator set includes:

[0212] The first construction unit is used to construct the phased allocation matrix TP of the generator set i i ;

[0213] The second construction unit is used to construct the unassigned matrix TS of the generator set i i ;

[0214] The first calculation unit is used to calculate the i and matrix TS i Calculate the maximum allocation matrix E of generator set i cmax and the minimum allocation matrix E cmin :

[0215]

[0216] Where, E ar represents the amount of power that has been allocated in different power flow constraints f; E umax and E umin They represent the maximum power schedulable value and the minimum power schedulable value of the unassigned units under different power flow constraints f; repmat(P max , K,1) means that P max Copy it into a matrix of K rows and 1 column; P max is the maximum active output constraint matrix of the power grid generator set: P max =(P 1max ,P 2max …P imax ...P Nmax ); where P imax Represents the upper threshold of active output of generator set i; repmat(P min , K,1) means that P min Copy it into a matrix of K rows and 1 column;

[0217] The second calculation unit is used to calculate the maximum allocation matrix E of the generator set i. cmax and the minimum allocation matrix E cmin , calculate and determine the power range of the generator set i output in the online data sample k of the grid operation mode [P ai ,Pbi ]:

[0218]

[0219] Where, FSt(i,:) means taking out the i-th row of FSt as a new matrix, Rmin() means taking out the minimum value of each row to form a new matrix, Rmax() means taking out the maximum value of each row to form a new matrix, min() means taking out the minimum value of the elements in the two matrices to form a new matrix, and max() means taking out the maximum value of the elements in the two matrices to form a new matrix; Rai and Rbi are the minimum and maximum power outputs of generator set i in sample k, respectively.

[0220] The third calculation unit is used to calculate the phased distribution power matrix TP of the generator set i+1 i+1 for:

[0221] TP i+1 =(TS i -TS i+1 )(npF(K,1,0,1)*(Pai-Pbi)+Pbi)+TP i (26)

[0222] Where: npf(K,1,0,1) generates a matrix with K rows and 1 column randomly distributed in [0,1];

[0223] According to the calculation of the stage distribution power matrix TP of generator set i+1 i+1 The active power distribution value of the generator set in K samples is calculated one by one by using the method, thereby completing the active power distribution of all generator sets in sample K.

[0224] The online active power distribution device for the power grid generator set provided in the embodiment of the present invention and the online active power distribution method for the power grid generator set provided in Example 1 are based on the same technical concept and can produce the beneficial effects as described in Example 1. For the contents not described in detail in this embodiment, please refer to Example 1.

[0225] Example 3:

[0226] An embodiment of the present invention provides an online active power distribution device for power grid generator sets, comprising a processor and a storage medium;

[0227] The storage medium is used to store instructions;

[0228] The processor is configured to operate according to the instructions to execute the steps of any one of the methods in the first embodiment.

[0229] Example 4:

[0230] An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the methods in the first embodiment are implemented.

[0231] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0232] Embodiment 5:

[0233] The present invention provides a method for generating online data samples of power grid operation modes, comprising the steps of the method for online distribution of active power of power grid generator sets as described in the first embodiment.

[0234] The method for generating online data samples of power grid operation modes provided in an embodiment of the present invention effectively supplements and screens samples, overcomes the problems of low typicality of online data samples and slow generation speed of offline data samples, increases sample diversity and balance, provides conditions for more effective extraction of power grid operation laws, and provides sufficient sample data support for artificial intelligence technology.

[0235] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0236] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0237] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0238] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for online distribution of active power of power generation units in a power grid, characterized in that: include: Obtain the upper and lower thresholds of the power grid's predicted output to determine the power grid generator set output constraints; Determining the prediction error and calculating the total power grid output according to the upper and lower thresholds of the power grid prediction output; Obtaining the number of power flow constraints of the power grid, and constructing the power flow constraints of the power grid generator units based on the total power grid output, the distribution of the generator units in the power grid power flow constraints, and the upper and lower output thresholds of the power grid power flow constraints; Obtain the total number of power grid generator sets and the number of power grid startup constraints, and construct power grid generator set startup constraints; According to the phased active power distribution of the power grid generator sets, the active power is randomly distributed to all the generator sets in the online data samples of each grid operation mode to be generated, taking into account the output constraints, power flow constraints and startup constraints of the power grid generator sets; The method of random distribution of active power includes: Building a generator set The phased allocation matrix and the unassigned matrix ; Matrix-based and Calculate generator sets The maximum and minimum allocation matrices of; Generator-based The maximum and minimum allocation matrix calculation determines the grid operation mode of the generator sets in the online data sample Output power range [P ai , P bi ]; The phased power distribution matrix of the generator set is calculated based on the following formula: ； Where: is a matrix with K rows and 1 column randomly distributed between [0,1]; K is the number of online data samples of power grid operation mode; According to the calculation of the generator set The phased power allocation matrix The method is to calculate The active power distribution value of the generator sets in the sample.

2. The method for online distribution of active power of power generation units in a power grid according to claim 1, characterized in that: The method for determining the prediction error and calculating the total power grid output according to the upper and lower thresholds of the power grid predicted output includes: According to the upper and lower thresholds of the power grid predicted output, the maximum range of the power grid predicted output is determined according to the maximum prediction error; Formula (1) is used to classify the maximum range of the predicted power output of the power grid: （1） Where: represents the ceiling function; represents the maximum prediction error; Indicates the preset bin interval; Indicates the upper threshold of the power grid forecast output, Indicates the lower threshold of the predicted power output of the power grid; Calculate the total power grid output according to formula (2): : （2） Where: Indicates the total power output of the grid. ; is the minimum active output constraint matrix of the power grid generator set: , N is the total number of power grid generators, Indicates a generator set The lower threshold of active output, .

3. The method for online distribution of active power of power generation units in a power grid according to claim 2, characterized in that: The constructing of power grid generator set startup constraints includes constructing a minimum startup number constraint matrix, constructing a maximum startup number constraint matrix, and constructing a power grid generator set startup constraint matrix.

4. The method for online distribution of active power of power generation units in a power grid according to claim 3, characterized in that: The minimum startup number constraint matrix and the maximum number of boot constraints matrix They are: （3） （4） Where: Indicates a generator set The stop state; when When the generator set Commissioning, when When the generator set Stop operation and meet , the total number of grid startup constraints is , For the A set of generator sets with grid startup constraints, ; and Grid startup constraints The minimum and maximum number of startups.

5. The method for online distribution of active power of power generation units in a power grid according to claim 4, characterized in that: The grid generator set startup constraint matrix for: （5） Where, Indicates whether the generator set N is within the grid startup constraint o. If it is within the grid startup constraint o, then Take 1; otherwise, Take 0.

6. The method for online distribution of active power of power generation units in a power grid according to any one of claims 1 to 5, characterized in that: The constructing of power grid generator set flow constraints includes constructing a power grid flow maximum constraint matrix, constructing a power grid flow minimum constraint matrix and constructing a power grid flow constraint matrix.

7. The method for online distribution of active power of power generation units in a power grid according to claim 6, characterized in that: The maximum constraint matrix of the power grid flow and the minimum constraint matrix of power flow They are: （6） （7） Where: and Grid flow constraints The output upper threshold and output lower threshold, and meet , For generator sets The active power output of the grid is , Indicates the A set of generators constrained by grid flow, .

8. The method for online distribution of active power of power generation units in a power grid according to claim 7, characterized in that: The power grid flow constraint matrix for: （8） Where: Indicates a generator set Is there a power grid flow constraint? In the case of power flow constraints In Take 1; otherwise, Take 0; The first column represents the total output constraint of the power grid generators.

9. The method for online distribution of active power of power generation units in a power grid according to claim 1, characterized in that: After the power generation unit startup constraints are established, the number of online data samples of the power grid operation mode to be generated is , generate a grid generator set startup constraint and grid power flow constraint. A generator set startup status matrix, and The generator set startup status matrices are merged into the generator set startup and shutdown matrix.

10. The method for online distribution of active power of power generation units in a power grid according to claim 9, characterized in that: Generate a grid that satisfies both the grid generator startup constraints and the grid power flow constraints. The method for generating a generator set startup state matrix comprises the following steps: According to the operation probability of the generator set in the online data sample of the current grid operation mode, the startup status of the generator set in the online data sample of each grid operation mode is determined by using random numbers, and the startup status matrix of the generator set is preliminarily generated: （9） Where: For generator set i in the sample In the power-on state, ; Combine the initially generated generator set startup state matrix and the grid generator set startup constraints to construct judgment matrix A and judgment matrix B. Combine the initially generated generator set startup state matrix and the grid power flow constraints to construct judgment matrix C and judgment matrix D: （10） （11） （12） （13） If all elements in matrices A, B, C, and D are greater than or equal to zero, the corresponding generator set startup state matrix is retained; otherwise, the corresponding generator set startup state matrix is discarded. Repeat the above steps until the The generator set startup status matrix.

11. The method for online distribution of active power of power generation units in a power grid according to claim 10, characterized in that: Generator set i in sample Power-on status Use formula (14) to calculate: （14） Where: For samples The probability of the unit being put into operation in ; To generate a Random numbers evenly distributed between To generate a 1 row and 1 column A random number uniformly distributed between .

12. The method for online distribution of active power of power generation units in a power grid according to claim 11, characterized in that: Matrix-based and Calculate generator sets Maximum allocation matrix and the minimum allocation matrix The calculation process is: （15） Where, Indicates the constraints of different power grid flows The amount of power already allocated in and Respectively represent different power grid flow constraints The maximum power schedulable value and the minimum power schedulable value of the unallocated units; Indicates that Copy A matrix with 1 row and 1 column; is the maximum active output constraint matrix of the power grid generator set: ;in, Indicates a generator set The upper threshold of active output; Indicates that Copy it into a matrix of K rows and 1 column; Generator-based The maximum and minimum allocation matrix calculation determines the grid operation mode of the generator sets in the online data sample Output power range [P ai , P bi The calculation process of ] is: （16） Where, Indicates removal of The rows are used as new matrices, It means taking out the minimum value of each row into a new matrix. It means taking out the maximum value of each row into a new matrix. It means taking out the minimum value of the elements in the two matrices to form a new matrix. It means taking out the maximum value of the elements in two matrices to synthesize a new matrix; The generator sets in the calculated sample k are The minimum and maximum output power.

13. An online active power distribution device for power grid generator sets, characterized in that: The device comprises: An acquisition module is used to obtain upper and lower thresholds of the power grid forecast output to determine the power grid generator set output constraints; A calculation module, configured to determine a prediction error and calculate a total power grid output according to upper and lower thresholds of the power grid predicted output; A power flow constraint module is used to obtain the number of power grid power flow constraints, and to construct the power grid generator power flow constraints based on the total power grid output, the distribution of the generator sets in the power grid power flow constraints, and the upper and lower output thresholds of the power grid power flow constraints; The startup constraint module is used to obtain the total number of power grid generator sets and the number of power grid startup constraints, and to construct the startup constraints of power grid generator sets; The power allocation module is used to randomly allocate active power to all generator sets in the online data samples of each grid operation mode to be generated, based on the phased active power distribution of the grid generator sets, taking into account the output constraints of the grid generator sets, the grid power flow constraints, and the grid generator set startup constraints; Wherein, the power distribution module includes: The first construction unit is used to construct a generator set The phased allocation matrix ; The second construction unit is used to construct a generator set The unassigned matrix ; The first calculation unit is used to calculate the matrix based on and matrix Calculate generator sets Maximum allocation matrix and the minimum allocation matrix ; The second calculation unit is used for Maximum allocation matrix and the minimum allocation matrix , calculate and determine the online data sample of the power grid operation mode Medium generator set Output power range ; The third calculation unit is used to calculate the generator set The phased power allocation matrix for: Where: is a matrix with K rows and 1 column randomly distributed between [0,1]; K is the number of online data samples of power grid operation mode; According to the calculation of the generator set The phased power allocation matrix The method is to calculate The active power distribution value of the generator sets in the sample.

14. The online active power distribution device for power grid generator sets according to claim 13, characterized in that: Also includes: The generator set startup and shutdown module is used to generate the number of online data samples of the grid operation mode required , generate a grid generator set startup constraint and grid power flow constraint. A generator set startup status matrix, and The generator set startup status matrices are merged into the generator set startup and shutdown matrix.

15. The online active power distribution device for power grid generator sets according to claim 14, characterized in that: The generator set startup and shutdown module includes: The generator set operation unit is used to determine the generator set's startup status in each grid operation mode online data sample based on the generator set's startup probability in the current grid operation mode online data sample using random numbers, and preliminarily generate the generator set startup status matrix: （18） Where: For generator set i in the sample In the power-on state, ; And, used to calculate the generator set i in the sample using formula (19) Power-on status : （19） Where: For samples The probability of the unit being put into operation in ; To generate a Random numbers evenly distributed between To generate a 1 row and 1 column A random number uniformly distributed between The judgment unit is used to construct the judgment matrix A and the judgment matrix B by combining the initially generated generator set startup state matrix and the grid generator set startup constraints, and to construct the judgment matrix C and the judgment matrix D by combining the initially generated generator set startup state matrix and the grid power flow constraints: （20） （21） （22） （23） If all elements in matrices A, B, C, and D are greater than or equal to zero, the corresponding generator set startup state matrix is retained; otherwise, the corresponding generator set startup state matrix is discarded. Repeat the judgment steps until a The generator set startup status matrix.

16. The online active power distribution device for power grid generator sets according to claim 15, characterized in that: The first calculation unit is based on the matrix and matrix Calculate generator sets Maximum allocation matrix and the minimum allocation matrix The calculation process includes: （24） Where, Indicates the constraints of different power grid flows The amount of power already allocated in and Respectively represent different power grid flow constraints The maximum power schedulable value and the minimum power schedulable value of the unallocated units; Indicates that Copy A matrix with 1 row and 1 column; is the maximum active output constraint matrix of the power grid generator set: ;in, Indicates a generator set The upper threshold of active output; Indicates that Copy it into a matrix of K rows and 1 column; The second calculation unit is based on the generator set Maximum allocation matrix and the minimum allocation matrix , calculate and determine the online data sample of the power grid operation mode Medium generator set Output power range The calculation process includes: （25） Where, Indicates removal of The rows are used as new matrices, It means taking out the minimum value of each row into a new matrix. It means taking out the maximum value of each row into a new matrix. It means taking out the minimum value of the elements in the two matrices to form a new matrix. It means taking out the maximum value of the elements in two matrices to synthesize a new matrix; The generator sets in the calculated sample k are The minimum and maximum output power.

17. An online active power distribution device for power grid generator sets, characterized in that: including processor and storage medium; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method for online distribution of active power of power generation sets in a power grid according to any one of claims 1 to 12.

18. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for online distribution of active power of power generation sets in a power grid as claimed in any one of claims 1 to 12 are implemented.

19. A method for generating online data samples of power grid operation mode, characterized in that: The method comprises the steps of the method for online distribution of active power of power grid generator sets as described in any one of claims 1 to 12.

Images