Method, system and equipment for quickly calculating three-phase unbalanced power flow of power distribution network and medium

Abstract

The invention provides a power distribution network three-phase unbalanced power flow rapid calculation method, system and device and a medium, and the method comprises the steps: building a voltage correction equation set according to the voltage data of each node in the branch data of a power distribution network; solving the voltage correction equation set by using a parallel computing method to obtain a correction value of the voltage data; carrying out load flow calculation on the power distribution network based on the convergence condition of the corrected value of the voltage data; according to the method, the voltage correction equation set established according to the voltage data of each node in the branch data of the power distribution network is solved through the parallel computing technology to realize correction of the voltage data, and the convergence of the correction value of the voltage data is judged, so that the calculation accuracy of unbalanced load flow calculation of the power distribution network is improved; and rapid solving of the complex power distribution network power flow is realized.

Description

Technical Field

[0001] The patent application of this invention belongs to the technical field of distribution network, and specifically relates to a method, system, equipment and medium for quickly calculating three-phase unbalanced power flow in a distribution network. Background Art

[0002] At present, distribution network flow calculation is a basic electrical calculation to study the steady-state operation of distribution systems. It determines the electrical state of the entire network based on given network structure and operating conditions, and provides a quantitative analysis basis for judging the rationality, safety, reliability and economy of system operation mode and planning and design schemes.

[0003] With the access of distributed resources and the construction and development of new power systems, many scholars have begun to pay attention to and study the fast power flow algorithms of complex distribution networks. Traditional power flow calculation methods are generally proposed for high-voltage transmission networks. Distribution networks are different from high-voltage transmission networks. Distribution networks have the following typical characteristics: distribution network points are numerous and widespread, and the branch parameter R / X ratio is large; and the three-phase parameter asymmetry and three-phase load imbalance problems of the distribution network are prominent. The loads are connected in three-phase star and triangle modes and single-phase lines to the ground and between phases; multi-phase unbalanced grounding or ungrounded operation, therefore, the traditional power flow calculation method is not suitable for distribution networks; and when performing distribution network power flow calculations, there are two modes: three-phase system balanced operation and three-phase system unbalanced operation. When the three-phase system is in balanced operation, the middle When the neutral point voltage is 0, the influence of the grounding method on the power flow results can be ignored. However, when the three-phase system is unbalanced, the influence of the neutral point grounding method on the solution model cannot be ignored during power flow calculation. For high-voltage distribution networks, a three-phase four-wire neutral point direct grounding method is generally adopted, a medium-voltage distribution network adopts a three-phase three-wire neutral point ungrounded method, and a low-voltage distribution network adopts a three-phase four-wire neutral point repeated grounding method. Therefore, when performing power flow calculations for multi-voltage distribution networks, it is necessary to consider the grounding methods of various equipment and transformers, which leads to low accuracy of power flow calculation results and a long calculation process. Summary of the invention

[0004] In order to overcome the above-mentioned deficiencies of the prior art, the present invention patent application proposes a method for quickly calculating the three-phase unbalanced power flow of a distribution network, including:

[0005] According to the voltage data of each node in the distribution network branch data, a voltage correction equation group is established;

[0006] Solving the voltage correction equation group by a parallel computing method to obtain a correction value of the voltage data;

[0007] Based on the convergence of the correction value of the voltage data, power flow calculation is performed on the distribution network.

[0008] Preferably, the voltage correction equations are established according to the voltage data of each node in the distribution network branch data, including:

[0009] According to the voltage data of each node in the distribution network branch data, the Jacobian matrix is ​​established;

[0010] Establishing a voltage correction equation group based on the Jacobian matrix, the power balance equation and the preset initial value of the voltage data;

[0011] The voltage data includes one or more of the following: voltage amplitude and voltage phase angle.

[0012] Preferably, establishing a Jacobian matrix according to the voltage data of each node in the distribution network branch data includes:

[0013] According to the voltage data of each node in the distribution network branch data, an admittance matrix is ​​established;

[0014] According to the admittance matrix, obtaining position information of non-zero elements in the admittance matrix;

[0015] Based on the position information of the non-zero elements in the admittance matrix, the admittance matrix is ​​sparsely processed to obtain a Jacobian matrix.

[0016] Preferably, the method of solving the voltage correction equation group by a parallel computing method to obtain the correction value of the voltage data includes:

[0017] Based on the initial value of the voltage data, using the power balance equation, calculating the power balance value of each node in the distribution network;

[0018] The power balance value of each node in the distribution network is transmitted to the GPU end, and based on the power balance value and the called cuSPARSE library, the voltage correction equation group is solved by using the Newton-Raphson method to obtain the correction value of the voltage data;

[0019] The power balance value includes: an active power balance difference and a reactive power balance difference.

[0020] Preferably, the power balance equation corresponds to the following:

[0021]

[0022] Among them, ΔP i α Represents the active power balance difference of phase α at node i in the distribution network; α=a,b,c; represents the active power injected into phase α of node i in the distribution network; V i α represents the α-phase voltage at node i in the distribution network; represents the β-phase voltage of node j in the distribution network; j = 1…n; n represents the total number of nodes in the distribution network; β = a, b, c; represents the conductance between phase α and phase β between nodes i and j in the distribution network; It represents the susceptance between phase α and phase β between nodes i and j in the distribution network; It represents the phase angle between phase α and phase β between nodes i and j in the distribution network; represents the reactive power balance difference of phase α at node i in the distribution network; It represents the reactive power injected into phase α of node i in the distribution network.

[0023] Preferably, the flow calculation for the distribution network based on the convergence of the correction value of the voltage data includes:

[0024] Step S1: Compare the correction value and the convergence precision to obtain the convergence judgment result of the correction value; when the absolute value of the correction value is not less than the convergence precision, the correction value does not converge, and step S2 is executed; when the absolute value of the correction value is less than the convergence precision, the correction value converges, and step S3 is executed;

[0025] Step S2: Based on the obtained initial value of the new voltage data of each node in the distribution network, a new voltage correction equation group is established, and the new voltage correction equation group is solved on the GPU side by using the Newton-Raphson method to obtain correction data as the correction value of the new voltage data, and the correction value is sent to the CPU side, and step S1 is executed;

[0026] Step S3: performing power flow calculation on the distribution network to obtain corresponding power flow calculation results.

[0027] Preferably, the power flow calculation result includes one or more of the following: balanced node power, PQ node voltage, PQ node injection power, PV node voltage and PV node injection power.

[0028] Preferably, the new voltage correction equations are established based on the initial values ​​of the new voltage data of each node in the distribution network, including:

[0029] Calculating a new power balance value of each node in the distribution network based on the acquired initial value of the new voltage data;

[0030] A new voltage correction equation group is established according to the new power balance value and the initial value of the new voltage data.

[0031] Preferably, the initial value of the new voltage data includes the following acquisition process:

[0032] The node voltage correction calculation formula is used to correct the initial value of the voltage data to obtain a new initial value of the voltage data.

[0033] Preferably, the node voltage correction calculation formula is as follows:

[0034]

[0035] in, Indicates the corrected voltage amplitude; Indicates the voltage amplitude before correction; ΔV i α Indicates the correction value of voltage amplitude; Indicates the corrected voltage phase angle; Indicates the voltage phase angle before correction; Indicates the correction value of the voltage phase angle.

[0036] Based on the same inventive concept, the patent application of the present invention also provides a system for quickly calculating three-phase unbalanced power flow in a distribution network, comprising:

[0037] Equation group establishment module: used to establish a voltage correction equation group according to the voltage data of each node in the distribution network branch data;

[0038] An equation group solving module: used for solving the voltage correction equation group by a parallel computing method to obtain a correction value of the voltage data;

[0039] The power flow calculation module is used to perform power flow calculation on the distribution network based on the convergence of the correction value of the voltage data.

[0040] Preferably, the equation system establishment module is specifically used for:

[0041] According to the voltage data of each node in the distribution network branch data, the Jacobian matrix is ​​established;

[0042] Establishing a voltage correction equation group based on the Jacobian matrix, the power balance equation and the preset initial value of the voltage data;

[0043] The voltage data includes one or more of the following: voltage amplitude and voltage phase angle.

[0044] Preferably, the equation group establishment module establishes a Jacobian matrix according to the voltage data of each node in the distribution network branch data, including:

[0045] According to the voltage data of each node in the distribution network branch data, an admittance matrix is ​​established;

[0046] According to the admittance matrix, obtaining position information of non-zero elements in the admittance matrix;

[0047] Based on the position information of the non-zero elements in the admittance matrix, the admittance matrix is ​​sparsely processed to obtain a Jacobian matrix.

[0048] Preferably, the equation solving module is specifically used for:

[0049] Based on the initial value of the voltage data, using the power balance equation, calculating the power balance value of each node in the distribution network;

[0050] The power balance value of each node in the distribution network is transmitted to the GPU end, and based on the power balance value and the called cuSPARSE library, the voltage correction equation group is solved by using the Newton-Raphson method to obtain the correction value of the voltage data;

[0051] The power balance value includes: an active power balance difference and a reactive power balance difference.

[0052] Preferably, the power balance equation in the equation solving module corresponds to the following:

[0053]

[0054] Among them, ΔP i α Represents the active power balance difference of phase α at node i in the distribution network; α=a,b,c; represents the active power injected into phase α of node i in the distribution network; V i α represents the α-phase voltage at node i in the distribution network; represents the β-phase voltage of node j in the distribution network; j = 1…n; n represents the total number of nodes in the distribution network; β = a, b, c; represents the conductance between phase α and phase β between nodes i and j in the distribution network; It represents the susceptance between phase α and phase β between nodes i and j in the distribution network; It represents the phase angle between phase α and phase β between nodes i and j in the distribution network; represents the reactive power balance difference of phase α at node i in the distribution network; It represents the reactive power injected into phase α of node i in the distribution network.

[0055] Preferably, the power flow calculation module specifically includes:

[0056] Step S1: Compare the correction value and the convergence precision to obtain the convergence judgment result of the correction value; when the absolute value of the correction value is not less than the convergence precision, the correction value does not converge, and step S2 is executed; when the absolute value of the correction value is less than the convergence precision, the correction value converges, and step S3 is executed;

[0057] Step S2: Based on the obtained initial value of the new voltage data of each node in the distribution network, a new voltage correction equation group is established, and the new voltage correction equation group is solved on the GPU side by using the Newton-Raphson method to obtain correction data as the correction value of the new voltage data, and the correction value is sent to the CPU side, and step S1 is executed;

[0058] Step S3: performing power flow calculation on the distribution network to obtain corresponding power flow calculation results.

[0059] Preferably, the power flow calculation results in the power flow calculation module include one or more of the following: balanced node power, PQ node voltage, PQ node injection power, PV node voltage and PV node injection power.

[0060] Preferably, the power flow calculation module establishes a new voltage correction equation group based on the initial value of the new voltage data of each node in the distribution network, including:

[0061] Calculating a new power balance value of each node in the distribution network based on the acquired initial value of the new voltage data;

[0062] A new voltage correction equation group is established according to the new power balance value and the initial value of the new voltage data.

[0063] Preferably, the initial value of the new voltage data in the power flow calculation module includes the following acquisition process:

[0064] The node voltage correction calculation formula is used to correct the initial value of the voltage data to obtain a new initial value of the voltage data.

[0065] Preferably, the node voltage correction calculation formula in the power flow calculation module is as follows:

[0066]

[0067] in, Indicates the corrected voltage amplitude; Indicates the voltage amplitude before correction; ΔV i α Indicates the correction value of voltage amplitude; Indicates the corrected voltage phase angle; Indicates the voltage phase angle before correction; Indicates the correction value of the voltage phase angle.

[0068] Based on the same inventive concept, the present patent application also provides a computer device, including: one or more processors;

[0069] A memory for storing one or more programs;

[0070] When the one or more programs are executed by the one or more processors, the method for fast calculation of three-phase unbalanced power flow in the distribution network as mentioned above is implemented.

[0071] Based on the same inventive concept, the patent application of the present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed, the method for quickly calculating the three-phase unbalanced power flow of the distribution network as described above is implemented.

[0072] Compared with the closest prior art, the patent application of this invention has the following beneficial effects:

[0073] The patent application of the present invention provides a method, system, device and medium for quickly calculating the three-phase unbalanced power flow of a distribution network, including: establishing a voltage correction equation group according to the voltage data of each node in the distribution network branch data; solving the voltage correction equation group by a parallel computing method to obtain the correction value of the voltage data; performing power flow calculation on the distribution network based on the convergence of the correction value of the voltage data; the patent application of the present invention solves the voltage correction equation group established according to the voltage data of each node in the distribution network branch data by parallel computing technology to realize the correction of voltage data, and judges the convergence of the corrected voltage data, which is conducive to improving the calculation accuracy of the distribution network unbalanced power flow calculation, and can realize the rapid solution of complex distribution network power flow. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 A schematic diagram of a flow chart of a method for rapidly calculating a three-phase unbalanced power flow in a distribution network provided in the patent application of the present invention;

[0075] Figure 2 A schematic diagram of the execution flow of a CPU+GPU hybrid programming model when a parallel computing method is used in a method for fast calculation of three-phase unbalanced power flow in a distribution network provided in the patent application of the present invention;

[0076] Figure 3 A schematic diagram of a convergence judgment flow of a correction value of voltage data in a method for quickly calculating a three-phase unbalanced power flow in a distribution network provided in the patent application of the present invention;

[0077] Figure 4 A schematic diagram of the overall framework of a method for fast calculation of three-phase unbalanced power flow in a distribution network provided in the patent application of the present invention;

[0078] Figure 5 A schematic diagram of the structure of a distribution network three-phase unbalanced power flow rapid calculation system provided in the patent application of the present invention. DETAILED DESCRIPTION

[0079] The specific implementation methods of the patent application of the present invention are further described in detail below in conjunction with the accompanying drawings.

[0080] Embodiment 1:

[0081] The present invention patent application provides a method for quickly calculating the three-phase unbalanced power flow in a distribution network, and the flow chart is as follows: Figure 1 As shown, including:

[0082] Step 1: Establish a voltage correction equation group based on the voltage data of each node in the distribution network branch data;

[0083] Step 2: Solve the voltage correction equations by using a parallel computing method to obtain a correction value of the voltage data;

[0084] Step 3: Based on the convergence of the corrected value of the voltage data, perform power flow calculation on the distribution network.

[0085] Specifically, step 1 includes:

[0086] According to the voltage data of each node in the distribution network branch data, the Jacobian matrix is ​​established;

[0087] Establishing a voltage correction equation group based on the Jacobian matrix, the power balance equation and the preset initial value of the voltage data;

[0088] The voltage data includes one or more of the following: voltage amplitude and voltage phase angle.

[0089] The Jacobian matrix is ​​established according to the voltage data of each node in the distribution network branch data, including:

[0090] According to the voltage data of each node in the distribution network branch data, an admittance matrix is ​​established;

[0091] According to the admittance matrix, obtaining position information of non-zero elements in the admittance matrix;

[0092] Based on the position information of the non-zero elements in the admittance matrix, sparse processing is performed on the admittance matrix to obtain a Jacobian matrix;

[0093] The process of establishing the admittance matrix specifically includes:

[0094] According to the three-phase admittance matrix data of each component in the distribution network branch data, the overall node admittance matrix Y of the three-phase network can be established:

[0095]

[0096] Where n represents the number of network nodes in the distribution network; the matrix elements in Y are sub-matrices between different nodes, such as Y 1nrepresents the complex admittance submatrix between the first node and the nth node; Y 2n represents the complex admittance submatrix between the second node and the nth node, and the other elements are similar; the form of the three-phase network node admittance matrix is ​​consistent with the symmetrical system, with Y ij represents the complex admittance submatrix between nodes i and j in the distribution network; i, j = 1…n, which is a 3×3 order complex matrix, that is,

[0097]

[0098] in, represents the self-admittance of phase a between nodes i and j; represents the mutual admittance between phase a and phase b between nodes i and j; represents the mutual admittance between phase c and phase b between nodes i and j, and the other elements are similar. Therefore, the order of the admittance matrix is ​​not n×n, but 3n×3n, where n is the number of network nodes.

[0099] After the three-phase admittance matrix is ​​formed, the corresponding network equation is:

[0100]

[0101] in, represents the node injected current; represents the node voltage; Y represents the overall node admittance matrix of the three-phase network; where, represents the injected current of phase a, phase b and phase c of the first network node, ..., represents the injected current of phase a, phase b and phase c of the nth network node; represents the injected voltage of phase a, phase b and phase c of the first network node, ..., represents the injected voltage of phase a, phase b and phase c of the nth network node; n represents the number of network nodes;

[0102] Expanding the above formula, we can get the α (α = a, b, c) phase injection current of the i-th network node

[0103]

[0104] Where, β = a, b, c; n represents the number of network nodes; j = 1…n; represents the mutual admittance between phase α and phase β between nodes i and j in the distribution network; represents the β-phase node voltage at node j;

[0105] Take the conjugate of both sides and multiply them by the α-phase node voltage of node i The α-phase injection power of node i can be obtained

[0106]

[0107] in, represents the α-phase node voltage of node i; represents the β-phase node voltage of node j; the node voltage is expressed in polar coordinates, that is, And let y is the imaginary part; then the α-phase injected power of node i is The polar coordinate form of is as follows:

[0108]

[0109] The active power P of phase α of the corresponding node i i α and reactive power They are:

[0110]

[0111]

[0112] Among them, V i α represents the α-phase voltage at node i in the distribution network; represents the β-phase voltage of node j in the distribution network; j = 1…n; n represents the total number of nodes in the distribution network; y is the imaginary part; β = a, b, c; represents the conductance between phase α and phase β between nodes i and j in the distribution network; It represents the susceptance between phase α and phase β between nodes i and j in the distribution network; It represents the phase angle between phase α and phase β between nodes i and j in the distribution network; That is, the voltage phase angle between phase α and phase β between nodes i and j in the distribution network is the difference between the phase angle of the voltage of phase α at node i and the phase angle of the voltage of phase β at node j;

[0113] The traditional admittance matrix generally stores all Y array elements in the computer, and the subscripts of the elements in the array correspond to the Y array one by one. Although the array form of the traditional method is intuitive and easy to understand, and it is also convenient to process the data of the Y array, it will store a large number of zero elements in the array, which not only wastes a lot of storage space, but also makes such a Y array more time-consuming to access. The more elements in the array, the longer it takes to read. The average outgoing line degree of each node in the distribution network is 3 to 5. Therefore, when the scale of the system is large, the admittance matrix is ​​quite sparse, and the Jacobian matrix in block form is as sparse as the admittance matrix. In the process of calculating the power flow of the distribution network, the patent application of the present invention performs sparse processing on the admittance matrix according to the position information of the non-zero elements in the admittance matrix, that is, the admittance matrix and the Jacobian matrix established according to the admittance matrix are stored by sparse matrix storage technology to improve the memory responsiveness of the power flow calculation and reduce the memory demand for the computer. The specific process is as follows:

[0114] The formation and storage of the node admittance matrix Y matrix is ​​completed by using an array of a specified format. YD, YS, YSt and Th1 are four one-dimensional arrays used to store the admittance matrix, which store respectively:

[0115] YD: stores the diagonal elements of the admittance matrix, i.e., the node self-admittance, all of which are initially set to 0.

[0116] YS: Store the upper triangular non-zero elements of the admittance matrix by row, all initially set to 0.

[0117] YSt: Stores the serial numbers of the non-zero elements in the upper triangle of the admittance matrix by row, all of which are initially set to 0.

[0118] Yhl: The number of rows and columns storing the admittance matrix. Initially, the admittance matrix arrays YD and Ys are empty. The admittance matrix is ​​formed by reading the branch data one by one. In order to make the reading and writing of the admittance matrix more suitable for parallel power flow calculation, the node self-admittance array YD and the mutual admittance array YS are calculated separately. The mutual admittance array YS without considering the branch is calculated first. When reading the branch data one by one, the branch number Yst should be calculated first. Since only the upper triangular matrix is ​​stored in this sparse model, the upper triangular matrix property is considered, that is, the row number is less than the column number, and the calculation formula for calculating the mutual admittance number is given as follows:

[0119]

[0120] Among them, L k1 represents the head end node of branch k; L k2 It represents the terminal node of branch k; Yhl represents the number of rows and columns of the admittance matrix, which is also the total number of nodes in the distribution network; p represents the pth element of the sequence array; YSt[p] represents the value of the pth element in the YSt array.

[0121] After solving the serial number array of non-zero elements in the admittance matrix, the admittance values ​​of non-zero elements can be written into the upper triangular mutual admittance array. At this time, it should be noted that the admittance value is in complex form, so each element in each array should store the real part and the imaginary part, that is, the two real values ​​of conductance and admittance. Since the two values ​​are generally used continuously, the real part and the imaginary part of each admittance element should be stored continuously. At this time, the relationship between the arrays YSt and YS is shown in Table 1 below:

[0122] Table 1. Storage format of sparse array in n-node system

[0123]

[0124] With this storage method, the memory space savings can be obtained as:

[0125] NS=2n 2 -(2n+4l+1)

[0126] Among them, NS represents the amount of data saved by using sparse array storage; 2n 2 represents the data volume of the original admittance matrix n×n; 2n represents the data volume of the self-admittance array; 4l represents the sum of the data volumes of the mutual admittance array and the mutual admittance sequence number array; l represents the number of branches of the system.

[0127] In the process of using Newton's method for power flow calculation, the most important part is the formation of the Jacobian matrix and the calculation of the elements in the sub-matrix. The Jacobian matrix has the following characteristics: the Jacobian matrix is ​​an asymmetric sparse matrix, in which the elements are not symmetric, but the sparsity of the elements is symmetric. The sub-matrix H of the Jacobian matrix ij 、N ij 、M ij , L ij It has the same sparsity structure as the node admittance matrix; the elements of the Jacobian matrix are all functions related to the node voltage phasor (amplitude and phase angle). During iterative operations, as the voltage phasor is continuously corrected by new values, the values ​​of each element also need to be recalculated; the non-diagonal elements in the Jacobian matrix are all related to the real part of each element in the admittance matrix G ij and the imaginary part B ij Closely related, when Y ij =G ij +yB ij When H in the Jacobian matrix ij 、N ij 、M ij , L ij are all zero, indicating that the Jacobian matrix is ​​also relatively sparse.

[0128] When Newton's method using polar coordinates forms the Jacobian matrix (abbreviated as J matrix), the array formed by the upper and lower triangular storage method similar to the admittance matrix can be used to directly calculate the non-zero elements in the J matrix. Unlike the admittance matrix, the admittance value array and the sequence number array no longer only record the upper triangular elements, but all the elements of the admittance matrix. ij 、N ij 、M ij , L ij The structures of are different. When forming the Jacobian matrix of the polar coordinate Newton method, it is necessary to consider the sparse structural characteristics of each of the four sub-matrices and their corresponding relationship with the admittance matrix: For the upper left sub-matrix, except for the diagonal elements, only the non-zero elements H of each odd row need to be calculated. ij 、N ij , then by H ij =L ij 、N ij =-M ij To form the corresponding even-numbered rows of non-zero elements, but for the diagonal elements, the above relationship does not exist (i.e. H ij ≠L ij 、N ij ≠-M ij ), so it needs to be calculated separately; for the remaining 3 sub-matrices, the upper right corner sub-matrix calculates the non-zero elements H by row ij 、M ij 、Calculate the non-zero elements H in the lower left corner submatrix by row ij 、N ij And the lower right corner submatrix calculates the non-zero elements H in sequence by row ij The basic process of fast calculation of Jacobian matrix for polar coordinate power flow calculation based on symmetric sparse matrix technology is as follows:

[0129] i) Quick determination of non-zero elements

[0130] Judgment of the non-zero elements of the upper left corner submatrix (i≠j): For H in each row and odd column ij The element is non-zero. If H ij ≠0, then we can get N ij 、M ij , L ij are not zero, and from the symmetry of the non-zero positions of the elements, we can get H ji 、N ji 、M ji , L ji All are not zero, that is, judging one element can get the non-zeroness of 8 elements; for the non-zero element judgment of the upper right corner sub-matrix (lower left corner sub-matrix) (i≠j): for each column of H ij Perform non-zero judgment, if H ij ≠0, then we can get M ij is not zero. From the symmetry of the non-zero position of the element, we can get Hji 、N ji All are not zero, that is, judging one element can get the non-zeroness of four elements; judging the non-zero elements of the lower right corner submatrix (i≠j): for each column of H ij Perform non-zero judgment, if H ij ≠0, then we can get H from the symmetry of the non-zero position of the element ji It is not zero, that is, judging 1 element can obtain the non-zeroness of 2 elements.

[0131] ii) Rapid determination of calculation elements:

[0132] When eliminating the elements of the J matrix by column, the positions of the non-zero elements obtained in the previous step can be used to directly determine the valid calculation elements according to the four-corner rule, thus reducing a large number of invalid calculations.

[0133] iii) Record of non-zero elements in the upper triangle:

[0134] Since the structure of the J array is similar to that of the Y array when it is just formed, the position of the non-zero elements also corresponds to the Y array. The non-zeroness of the array is immediately judged, and the coordinates of all the non-zero elements H in the upper triangle are recorded and used in the subsequent elimination calculation and back substitution to solve the voltage phase angle Δθ of node i. i and the voltage amplitude of the node ΔV i process to fully utilize symmetric sparse matrix technology.

[0135] For the PQ node in the distribution network, when the active power is injected into the α phase of the node i and reactive power After determination, the corresponding power balance equation can be established, that is, the calculation formula corresponding to the power balance equation is as follows:

[0136]

[0137] For a PV node in the system, when the active power is injected into the α phase of node i and node voltage After determination, the corresponding active power balance equation can be established, and the expression is as follows:

[0138]

[0139] Among them, ΔP i α Represents the active power balance difference of phase α at node i in the distribution network; α=a,b,c; represents the active power injected into phase α of node i in the distribution network; V i α represents the α-phase voltage at node i in the distribution network; represents the β-phase voltage of node j in the distribution network; j = 1…n; n represents the total number of nodes in the distribution network; β = a, b, c; represents the conductance between phase α and phase β between nodes i and j in the distribution network; It represents the susceptance between phase α and phase β between nodes i and j in the distribution network; It represents the phase angle between phase α and phase β between nodes i and j in the distribution network; represents the reactive power balance difference of phase α at node i in the distribution network; represents the reactive power injected into phase α of node i in the distribution network; represents the α-phase node voltage of node i;

[0140] At the same time, reactive power constraints must be added:

[0141]

[0142] in, represents the injected reactive power of node i; Q imin represents the minimum three-phase reactive power that node i can inject; Q imax represents the maximum three-phase reactive power that node i can inject;

[0143] According to the Jacobian matrix, the power balance equation and the initial value of the voltage data, the established voltage correction equation group is as follows:

[0144]

[0145] Among them, ΔP represents the active power balance difference of the distribution network; ΔQ represents the reactive power balance difference of the distribution network; ΔP1 a represents the active power balance difference of phase a at the first node in the distribution network, ..., ΔP n c represents the c-phase active power balance difference of the n-th node; ΔQ1 α represents the reactive power balance difference of phase a at the first node in the distribution network, ..., represents the reactive power balance difference of phase c of the nth node; a, b, c represent phases respectively; n represents the total number of nodes in the distribution network; Δθ represents the correction value of the voltage phase angle; ΔV represents the correction value of the voltage amplitude; Δθ1 α represents the correction value of the voltage phase angle of phase a at the first node of the distribution network, ..., It represents the correction value of the phase angle of the phase c voltage at the nth node of the distribution network; ΔV1 α represents the correction value of the voltage amplitude of phase a at the first node of the distribution network, ..., represents the correction value of the voltage amplitude of phase c at the nth node of the distribution network; represents the Jacobian matrix composed of sub-matrices H, N, J, and L.

[0146] Step 2 includes:

[0147] Based on the initial value of the voltage data, using the power balance equation, calculating the power balance value of each node in the distribution network;

[0148] The power balance value of each node in the distribution network is transmitted to the GPU end, and based on the power balance value and the called cuSPARSE library, the voltage correction equation group is solved by using the Newton-Raphson method to obtain the correction value of the voltage data;

[0149] The power balance value includes: an active power balance difference and a reactive power balance difference.

[0150] In the process of solving the voltage correction equation group, it is necessary to first calculate the element value of each sub-matrix in the Jacobian matrix. The element calculation formula of each sub-matrix in the Jacobian matrix for i≠j and α≠β is as follows:

[0151]

[0152]

[0153]

[0154]

[0155] The calculation formula for the elements of each sub-matrix in the Jacobian matrix for i=j and α=β is as follows:

[0156]

[0157]

[0158]

[0159]

[0160] Among them, P i α represents the active power of phase α at node i; represents the β-phase voltage phase angle of node j; represents the β-phase voltage phase angle of node i; represents the phase angle of the α-phase voltage at node i; represents the β-phase voltage phase angle of node i; V i αrepresents the α-phase voltage at node i in the distribution network; represents the β-phase voltage of node j in the distribution network; V i β represents the β-phase voltage of node i in the distribution network; i,j=1…n; n represents the total number of nodes in the distribution network; represents the conductance between phase α and phase β between nodes i and j in the distribution network; represents the conductance between phase α and phase β of node i in the distribution network; represents the α-phase self-conductance of node i in the distribution network; It represents the susceptance between phase α and phase β between nodes i and j in the distribution network; It represents the susceptance between phase α and phase β of node i in the distribution network; represents the self-susceptance of phase α of node i in the distribution network; It represents the voltage phase angle between phase α and phase β between nodes i and j in the distribution network; It represents the voltage phase angle between phase α and phase β of node i in the distribution network; represents the reactive power of phase α at node i in the distribution network; α, β = a, b, c; represents the β-phase voltage phase angle of node i;

[0161] The initial value of the α-phase voltage amplitude at a given node i and the initial value of the phase angle of the α phase voltage at node i Under the condition of , the active power balance difference ΔP of the α phase corresponding to node i can be calculated by the power balance equation i α and reactive power balance difference At the same time, the voltage correction equations can be solved by calculating the elements in the Jacobian matrix, and then the correction value ΔV of the α-phase voltage amplitude of node i can be calculated. i α And the correction value of voltage phase angle

[0162] In order to increase the speed of solving the voltage correction equations, the patent application of the present invention adopts parallel computing technology. Parallel computing is the calculation done on parallel computers, that is, the calculation of multiple instruction streams and multiple data streams is completed on multiple processor units within a unit time. General computing based on GPU is a branch of parallel computing. Due to the different design goals of CPU and GPU, there are great differences between the two in architecture, parallel level and performance. The purpose of CPU is to enable the execution unit to obtain data and instructions as quickly as possible, so it adopts complex control logic and branch prediction, as well as a large number of caches to improve the efficiency of a few execution pipelines, and is extremely sensitive to delays; while GPU improves the overall throughput of computing and data by operating a large number of simple stream processors together.

[0163] GPU general computing generally adopts a hybrid programming model: it is mainly responsible for some large-scale data-intensive parallel computing tasks, giving full play to its inherent strong parallel computing capabilities. Therefore, general computing is to complement the high bandwidth of the CPU and the powerful parallel processing capabilities of the GPU with strong branch processing capabilities, which has great advantages in terms of cost, cost performance and development prospects.

[0164] The CUDA architecture is a brand-new architecture designed specifically for GPU computing. It can use NVIDIA's GPU for parallel computing, allowing the GPU to give full play to its inherent advantages in handling complex computing problems. The CUDA architecture fully mobilizes the powerful parallel computing capabilities of the GPU. CUDA-enabled GPUs can solve many complex computing tasks more efficiently than CPUs. At the same time, CUDA also supports multiple language programming environments, such as C, C++, Java, etc., making the GPU move from a dedicated platform to a general platform and has been widely used in many fields.

[0165] CUDA is a CPU+GPU hybrid programming model. Figure 2As shown in the figure, the CPU and GPU are used as the host and device in the model respectively, forming a parallel computing environment combining a host and multiple devices. The CPU is mainly responsible for serial computing and logical event processing, while the GPU undertakes large-scale highly parallel computing tasks. During the execution of the entire program, CUDA assigns the parallel tasks in the program to the GPU for processing. The CUDA parallel computing function called in the host code and running on the device is called a kernel function, which must be defined by the __global__ function type qualifier. A kernel function is just a small step in the CUDA program of the CPU and GPU mixed programming mode, and cannot be executed as a complete program. A complete CUDA program consists of multiple steps, including serial processing steps executed in the CPU and parallel steps executed by the GPU. CUDA has two key features: threads can have two levels of parallelism in the GPU; and communication is achieved at a lower level through fence synchronization and shared memory.

[0166] In the whole process of calculating the power flow problem, solving the modified equations is the most time-consuming part. The solution of the linear equation system Ax=b can take advantage of the powerful computing power of GPU parallel computing. The key to parallel power flow calculation based on CUDA is to regard the computing task as a parallel computing of data streams. In the stream model, the program is regarded as a series of management of data streams. The elements in the stream are processed by the instructions in the kernel. The kernel runs each stream element and writes the result to an output stream.

[0167] The numerical solutions to algebraic linear equations are generally divided into two categories: direct method and iterative method. The direct method is a method that can obtain the exact solution of the equation through a finite number of operations, while the iterative method is a process that uses a certain limit process to gradually approach the exact solution of the linear equation system. The matrix form of the iterative method of the algebraic linear equation system Ax=b can be written as follows:

[0168] X (k+1) =BX k +F

[0169] Among them, B and F represent two iteration matrices with different elements; X k represents the vector at the kth iteration; X (k+1) represents the vector at the k+1th iteration; for any initial vector X (0)and constant vector F, the necessary and sufficient condition for the iterative method to converge is that the spectral radius ρ(B) of the iteration matrix B is less than 1. Direct methods for solving equations usually include Gauss elimination, LU decomposition, Cholesky decomposition, etc. Sparse direct methods need to consider the injection of non-zero elements. However, sparse iterative methods do not need to consider the injection of non-zero elements during the solution process, and can simply implement parallel computing in GPUs. In addition, iterative methods are superior to direct methods in terms of scalability and the memory required for computing.

[0170] cuSPARSE is a CUDA-based library for sparse linear algebra and graphics computing. It provides a flexible high-level interface for processing sparse matrices and solving sparse linear equations, which can easily realize the transmission of sparse matrices between the host and the device. The cuSPARSE library provides some sparse mathematical operations, such as matrix-vector multiplication, matrix transposition, etc., and also provides some iterative solvers for solving linear equations. In addition, it supports multiple sparse matrix formats, such as compressed row sparse matrix format and mixed coefficient matrix format, making it easy to transmit sparse matrix data between the host and the device.

[0171] In summary, the patent application of the present invention adopts a parallel computing method when solving the voltage correction equation group. The method is an inner and outer double-layer iteration framework, which consists of two parts: CPU and GPU. Among them, the CPU is responsible for controlling the basic calculation process of the Newton-Raphson method, including the setting of the initial value of the iteration, the construction of the preprocessor, the correction of the iteration variables and the judgment of the convergence, which belongs to the outer iteration part of the algorithm; the GPU is responsible for the calculation of the elements of the admittance matrix, the formation of the correction equation group and the repeated solution of the correction equation group, which belongs to the inner iteration part of the algorithm. The data correlation between the calculation steps in the outer iteration of the algorithm is strong, and serial calculation is mainly performed, while the inner iteration mainly involves the operation between sparse matrices and vectors, and the data correlation is weak, which is suitable for parallel calculation. At the same time, the outer iteration part and the inner iteration part of the algorithm maintain good communication, and the two cooperate with each other to jointly realize the calculation of the three-phase unbalanced power flow of the distribution network.

[0172] Step 3, such as Figure 3 As shown, including:

[0173] Step S1: Compare the correction value and the convergence precision to obtain the convergence judgment result of the correction value; when the absolute value of the correction value is not less than the convergence precision, the correction value does not converge, and step S2 is executed; when the absolute value of the correction value is less than the convergence precision, the correction value converges, and step S3 is executed;

[0174] Step S2: Based on the obtained initial value of the new voltage data of each node in the distribution network, a new voltage correction equation group is established, and the new voltage correction equation group is solved on the GPU side by using the Newton-Raphson method to obtain correction data as the correction value of the new voltage data, and the correction value is sent to the CPU side, and step S1 is executed;

[0175] Step S3: performing power flow calculation on the distribution network to obtain corresponding power flow calculation results.

[0176] When the correction value of the α phase voltage angle of node i is less than the set convergence accuracy, it satisfies And the correction value of the α-phase voltage amplitude of node i is less than the convergence accuracy, that is, When , the iteration converges and the calculation ends. Otherwise, the node voltage correction calculation formula is used to correct the initial value of each node voltage, and then and As the initial value of the voltage amplitude and phase angle of phase α at node i, repeat the calculation in the previous step until the iteration converges.

[0177] The power flow calculation results include one or more of the following: balanced node power, PQ node voltage, PQ node injection power, PV node voltage and PV node injection power;

[0178] For PV nodes, in each iteration process, only the voltage phase angle is modified, while the voltage amplitude remains unchanged; in addition, the injected reactive power of each phase is calculated and tested. If the constraints are met, the calculation can continue. Otherwise, the given voltage amplitude must be modified, or the PV node must be changed to a PQ node, and other nodes must be selected as PV nodes.

[0179] The balance node is also a voltage reference point, and its voltage amplitude and phase angle remain unchanged during the iteration process, so there is no need to correct them. However, after the iteration, the three-phase power of the balance node needs to be calculated.

[0180] The method of establishing a new voltage correction equation group based on the initial value of the new voltage data of each node in the distribution network obtained includes:

[0181] Calculating a new power balance value of each node in the distribution network based on the acquired initial value of the new voltage data;

[0182] A new voltage correction equation group is established according to the new power balance value and the initial value of the new voltage data.

[0183] The initial value of the new voltage data includes the following acquisition process:

[0184] The node voltage correction calculation formula is used to correct the initial value of the voltage data to obtain a new initial value of the voltage data.

[0185] The node voltage correction calculation formula is as follows:

[0186]

[0187] in, Indicates the corrected voltage amplitude; Indicates the voltage amplitude before correction; ΔV i α Indicates the correction value of voltage amplitude; Indicates the corrected voltage phase angle; Indicates the voltage phase angle before correction; Δθ i α Indicates the correction value of the voltage phase angle.

[0188] The invention patent application provides a method for fast calculation of three-phase unbalanced power flow in a distribution network, which is used to solve the problems of slow power flow calculation, large memory space occupation, inaccurate calculation results, etc. caused by the three-phase unbalance problem during the operation of a complex distribution network; Figure 4 As shown in the figure, first, the three-phase node admittance matrix and three-phase power balance equation of the distribution network are established; secondly, the distribution law of non-zero elements of the admittance matrix and Jacobian matrix in the three-phase power flow calculation of the Newton method is studied, and the sparse matrix technique is used to store large-scale matrices to reduce the memory requirements of the computer for large-scale power flow calculation; finally, the values ​​of each element of the Jacobian matrix are calculated synchronously through parallel computing technology and the correction equation group is solved in parallel, giving full play to the characteristics of the heterogeneous platform and greatly improving the efficiency of power flow calculation. In this method, the three-phase power flow model of the distribution equipment, especially the distribution transformer, is effectively established, and the memory space of the admittance matrix and the Jacobian matrix is ​​reduced through the sparse matrix technology. Then, the computational efficiency of the unbalanced power flow calculation of the distribution network is improved through parallel computing technology, and the rapid solution of the power flow of complex distribution networks is realized.

[0189] Embodiment 2:

[0190] Based on the same inventive concept, the present invention patent application also provides a fast calculation system for three-phase unbalanced power flow in a distribution network.

[0191] The structural composition diagram is as follows Figure 5 As shown, including:

[0192] Equation group establishment module: used to establish a voltage correction equation group according to the voltage data of each node in the distribution network branch data;

[0193] An equation group solving module: used for solving the voltage correction equation group by a parallel computing method to obtain a correction value of the voltage data;

[0194] The power flow calculation module is used to perform power flow calculation on the distribution network based on the convergence of the correction value of the voltage data.

[0195] The equation group establishment module is specifically used for:

[0196] According to the voltage data of each node in the distribution network branch data, the Jacobian matrix is ​​established;

[0197] Establishing a voltage correction equation group based on the Jacobian matrix, the power balance equation and the preset initial value of the voltage data;

[0198] The voltage data includes one or more of the following: voltage amplitude and voltage phase angle.

[0199] The equation group establishment module establishes a Jacobian matrix according to the voltage data of each node in the distribution network branch data, including:

[0200] According to the voltage data of each node in the distribution network branch data, an admittance matrix is ​​established;

[0201] According to the admittance matrix, obtaining position information of non-zero elements in the admittance matrix;

[0202] Based on the position information of the non-zero elements in the admittance matrix, the admittance matrix is ​​sparsely processed to obtain a Jacobian matrix.

[0203] The equation solving module is specifically used for:

[0204] Based on the initial value of the voltage data, using the power balance equation, calculating the power balance value of each node in the distribution network;

[0205] The power balance value of each node in the distribution network is transmitted to the GPU end, and based on the power balance value and the called cuSPARSE library, the voltage correction equation group is solved by using the Newton-Raphson method to obtain the correction value of the voltage data;

[0206] The power balance value includes: an active power balance difference and a reactive power balance difference.

[0207] Preferably, the power balance equation in the equation solving module corresponds to the following:

[0208]

[0209] Among them, ΔP i α Represents the active power balance difference of phase α at node i in the distribution network; α=a,b,c; represents the active power injected into phase α of node i in the distribution network; Vi α represents the α-phase voltage at node i in the distribution network; represents the β-phase voltage of node j in the distribution network; j = 1…n; n represents the total number of nodes in the distribution network; β = a, b, c; represents the conductance between phase α and phase β between nodes i and j in the distribution network; It represents the susceptance between phase α and phase β between nodes i and j in the distribution network; It represents the phase angle between phase α and phase β between nodes i and j in the distribution network; represents the reactive power balance difference of phase α at node i in the distribution network; It represents the reactive power injected into phase α of node i in the distribution network.

[0210] The power flow calculation module specifically includes:

[0211] Step S1: Compare the correction value and the convergence precision to obtain the convergence judgment result of the correction value; when the absolute value of the correction value is not less than the convergence precision, the correction value does not converge, and step S2 is executed; when the absolute value of the correction value is less than the convergence precision, the correction value converges, and step S3 is executed;

[0212] Step S2: Based on the obtained initial value of the new voltage data of each node in the distribution network, a new voltage correction equation group is established, and the new voltage correction equation group is solved on the GPU side by using the Newton-Raphson method to obtain correction data as the correction value of the new voltage data, and the correction value is sent to the CPU side, and step S1 is executed;

[0213] Step S3: performing power flow calculation on the distribution network to obtain corresponding power flow calculation results.

[0214] The power flow calculation results in the power flow calculation module include one or more of the following: balanced node power, PQ node voltage, PQ node injection power, PV node voltage and PV node injection power.

[0215] The power flow calculation module establishes a new voltage correction equation group based on the initial value of the new voltage data of each node in the distribution network, including:

[0216] Calculating a new power balance value of each node in the distribution network based on the acquired initial value of the new voltage data;

[0217] A new voltage correction equation group is established according to the new power balance value and the initial value of the new voltage data.

[0218] The initial value of the new voltage data in the power flow calculation module includes the following acquisition process:

[0219] The node voltage correction calculation formula is used to correct the initial value of the voltage data to obtain a new initial value of the voltage data.

[0220] The node voltage correction calculation formula in the power flow calculation module is as follows:

[0221]

[0222] in, Indicates the corrected voltage amplitude; Indicates the voltage amplitude before correction; Indicates the correction value of voltage amplitude; Indicates the corrected voltage phase angle; Indicates the voltage phase angle before correction; Indicates the correction value of the voltage phase angle.

[0223] Embodiment 3:

[0224] Based on the same inventive concept, the patent application of the present invention also provides a computer device, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function, so as to implement the steps of a method for fast calculation of three-phase unbalanced power flow in a distribution network in the above embodiment.

[0225] Embodiment 4:

[0226] Based on the same inventive concept, the patent application of the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It can be understood that the computer-readable storage medium here can include both built-in storage media in a computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the steps of a method for quickly calculating a three-phase unbalanced power flow in a distribution network in the above embodiment.

[0227] It should be understood by those skilled in the art that the embodiments of the present invention patent application can be provided as methods, systems, or computer program products. Therefore, the present invention patent application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention patent application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0228] The present invention patent 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 invention patent application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks 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 generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. 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.

[0229] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate 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 A function specified in one or more boxes.

[0230] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0231] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the patent application of the present invention rather than to limit its protection scope. Although the patent application of the present invention is described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that after reading the patent application of the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the application, but these changes, modifications or equivalent substitutions are all within the protection scope of the claims to be approved.

Claims

1. A method for quickly calculating three-phase unbalanced power flow in a distribution network, characterized in that: include: According to the voltage data of each node in the distribution network branch data, a voltage correction equation group is established; Solving the voltage correction equation group by a parallel computing method to obtain a correction value of the voltage data; Based on the convergence of the correction value of the voltage data, power flow calculation is performed on the distribution network.

2. The method according to claim 1, characterized in that The voltage correction equations are established according to the voltage data of each node in the distribution network branch data, including: According to the voltage data of each node in the distribution network branch data, the Jacobian matrix is ​​established; Establishing a voltage correction equation group based on the Jacobian matrix, the power balance equation and the preset initial value of the voltage data; The voltage data includes one or more of the following: voltage amplitude and voltage phase angle.

3. The method according to claim 2, characterized in that The Jacobian matrix is ​​established according to the voltage data of each node in the distribution network branch data, including: According to the voltage data of each node in the distribution network branch data, an admittance matrix is ​​established; According to the admittance matrix, obtaining position information of non-zero elements in the admittance matrix; Based on the position information of the non-zero elements in the admittance matrix, the admittance matrix is ​​sparsely processed to obtain a Jacobian matrix.

4. The method according to claim 2, characterized in that The method of solving the voltage correction equation group by a parallel computing method to obtain the correction value of the voltage data includes: Based on the initial value of the voltage data, using the power balance equation, calculating the power balance value of each node in the distribution network; The power balance value of each node in the distribution network is transmitted to the GPU end, and based on the power balance value and the called cuSPARSE library, the voltage correction equation group is solved by using the Newton-Raphson method to obtain the correction value of the voltage data; The power balance value includes: an active power balance difference and a reactive power balance difference.

5. The method according to claim 4, characterized in that The power balance equation corresponds to the following: Among them, ΔP i α Represents the active power balance difference of phase α at node i in the distribution network; α=a,b,c; represents the active power injected into phase α of node i in the distribution network; V i α represents the α-phase voltage at node i in the distribution network; represents the β-phase voltage of node j in the distribution network; j = 1…n; n represents the total number of nodes in the distribution network; β = a, b, c; represents the conductance between phase α and phase β between nodes i and j in the distribution network; It represents the susceptance between phase α and phase β between nodes i and j in the distribution network; It represents the phase angle between phase α and phase β between nodes i and j in the distribution network; represents the reactive power balance difference of phase α at node i in the distribution network; It represents the reactive power injected into phase α of node i in the distribution network.

6. The method according to claim 2, characterized in that The step of performing power flow calculation on the distribution network based on the convergence of the correction value of the voltage data includes: Step S1: Compare the correction value and the convergence precision to obtain the convergence judgment result of the correction value; when the absolute value of the correction value is not less than the convergence precision, the correction value does not converge, and step S2 is executed; when the absolute value of the correction value is less than the convergence precision, the correction value converges, and step S3 is executed; Step S2: Based on the obtained initial value of the new voltage data of each node in the distribution network, a new voltage correction equation group is established, and the new voltage correction equation group is solved on the GPU side by using the Newton-Raphson method to obtain correction data as the correction value of the new voltage data, and the correction value is sent to the CPU side, and step S1 is executed; Step S3: performing power flow calculation on the distribution network to obtain corresponding power flow calculation results.

7. The method according to claim 6, characterized in that The power flow calculation result includes one or more of the following: balanced node power, PQ node voltage, PQ node injection power, PV node voltage and PV node injection power.

8. The method according to claim 6, characterized in that The method of establishing a new voltage correction equation group based on the initial value of the new voltage data of each node in the distribution network obtained includes: Calculating a new power balance value of each node in the distribution network based on the acquired initial value of the new voltage data; A new voltage correction equation group is established according to the new power balance value and the initial value of the new voltage data.

9. The method according to claim 8, characterized in that The initial value of the new voltage data includes the following acquisition process: The node voltage correction calculation formula is used to correct the initial value of the voltage data to obtain a new initial value of the voltage data.

10. The method according to claim 9, characterized in that The node voltage correction calculation formula is as follows: in, Indicates the corrected voltage amplitude; Indicates the voltage amplitude before correction; ΔV i α Indicates the correction value of voltage amplitude; Indicates the corrected voltage phase angle; Indicates the voltage phase angle before correction; Indicates the correction value of the voltage phase angle.

11. A fast calculation system for three-phase unbalanced power flow in a distribution network, characterized in that: include: Equation group establishment module: used to establish a voltage correction equation group according to the voltage data of each node in the distribution network branch data; An equation group solving module: used for solving the voltage correction equation group by a parallel computing method to obtain a correction value of the voltage data; The power flow calculation module is used to perform power flow calculation on the distribution network based on the convergence of the correction value of the voltage data.

12. The system according to claim 11, characterized in that The equation group establishment module is specifically used for: According to the voltage data of each node in the distribution network branch data, the Jacobian matrix is ​​established; Establishing a voltage correction equation group based on the Jacobian matrix, the power balance equation and the preset initial value of the voltage data; The voltage data includes one or more of the following: voltage amplitude and voltage phase angle.

13. The system of claim 12, wherein: The equation group establishment module establishes a Jacobian matrix according to the voltage data of each node in the distribution network branch data, including: According to the voltage data of each node in the distribution network branch data, an admittance matrix is ​​established; According to the admittance matrix, obtaining position information of non-zero elements in the admittance matrix; Based on the position information of the non-zero elements in the admittance matrix, the admittance matrix is ​​sparsely processed to obtain a Jacobian matrix.

14. The system of claim 12, wherein: The power flow calculation module is specifically used for: Based on the initial value of the voltage data, using the power balance equation, calculating the power balance value of each node in the distribution network; The power balance value of each node in the distribution network is transmitted to the GPU end, and based on the power balance value and the called cuSPARSE library, the voltage correction equation group is solved by using the Newton-Raphson method to obtain the correction value of the voltage data; The power balance value includes: an active power balance difference and a reactive power balance difference.

15. A computer device, characterized in that: include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the method for quickly calculating the three-phase unbalanced power flow in the distribution network according to any one of claims 1 to 10 is implemented.

16. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed, the method for quickly calculating the three-phase unbalanced power flow of the distribution network as described in any one of claims 1 to 10 is implemented.

Images