CN111062610A - Power system state estimation method and system based on information matrix sparse solution - Google Patents
Power system state estimation method and system based on information matrix sparse solution Download PDFInfo
- Publication number
- CN111062610A CN111062610A CN201911294697.7A CN201911294697A CN111062610A CN 111062610 A CN111062610 A CN 111062610A CN 201911294697 A CN201911294697 A CN 201911294697A CN 111062610 A CN111062610 A CN 111062610A
- Authority
- CN
- China
- Prior art keywords
- information matrix
- matrix
- node
- active
- reactive
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 239000011159 matrix material Substances 0.000 title claims abstract description 266
- 238000000034 method Methods 0.000 title claims abstract description 78
- 238000009826 distribution Methods 0.000 claims abstract description 58
- 230000008569 process Effects 0.000 claims abstract description 35
- 238000004364 calculation method Methods 0.000 claims abstract description 33
- 230000008030 elimination Effects 0.000 claims abstract description 25
- 238000003379 elimination reaction Methods 0.000 claims abstract description 25
- 230000011664 signaling Effects 0.000 claims abstract description 12
- 238000005457 optimization Methods 0.000 claims abstract description 11
- 238000000354 decomposition reaction Methods 0.000 claims description 24
- 238000003860 storage Methods 0.000 claims description 21
- 238000005259 measurement Methods 0.000 claims description 17
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 238000012163 sequencing technique Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 9
- 238000002347 injection Methods 0.000 claims description 8
- 239000007924 injection Substances 0.000 claims description 8
- 238000003780 insertion Methods 0.000 claims description 5
- 230000037431 insertion Effects 0.000 claims description 5
- 238000012217 deletion Methods 0.000 claims description 2
- 230000037430 deletion Effects 0.000 claims description 2
- 230000015654 memory Effects 0.000 abstract description 25
- 239000013598 vector Substances 0.000 description 26
- 238000010586 diagram Methods 0.000 description 12
- 238000004590 computer program Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 230000005540 biological transmission Effects 0.000 description 4
- 239000000243 solution Substances 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000006835 compression Effects 0.000 description 3
- 238000007906 compression Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- VZSRBBMJRBPUNF-UHFFFAOYSA-N 2-(2,3-dihydro-1H-inden-2-ylamino)-N-[3-oxo-3-(2,4,6,7-tetrahydrotriazolo[4,5-c]pyridin-5-yl)propyl]pyrimidine-5-carboxamide Chemical compound C1C(CC2=CC=CC=C12)NC1=NC=C(C=N1)C(=O)NCCC(N1CC2=C(CC1)NN=N2)=O VZSRBBMJRBPUNF-UHFFFAOYSA-N 0.000 description 1
- 102000002274 Matrix Metalloproteinases Human genes 0.000 description 1
- 108010000684 Matrix Metalloproteinases Proteins 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Mathematical Physics (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Computational Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Operations Research (AREA)
- Pure & Applied Mathematics (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Algebra (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Complex Calculations (AREA)
Abstract
The invention discloses a method and a system for estimating the state of a power system, wherein the method comprises the following steps: acquiring telemetering and remote signaling data of primary equipment of the power system in real time; constructing a node branch model of the power system; determining non-zero element distribution of an active Jacobian matrix according to the node branch model, and determining non-zero element distribution of an active information matrix; performing node optimization numbering based on active information matrix non-zero element distribution; calculating an active Jacobian matrix and a reactive Jacobian matrix according to the new node sequence to obtain a new active information matrix and a new reactive information matrix, and performing factorization by adopting a Gaussian elimination method to obtain an active information matrix factor table and a reactive information matrix factor table; and estimating the state of the power system based on the active information matrix factor table and the reactive information matrix factor table. By utilizing the method and the device for solving the information matrix and estimating the state of the power system, the memory operation in the solving process can be reduced, the calculation load is reduced, and the calculation efficiency of the power grid state estimation is improved.
Description
Technical Field
The invention relates to the technical field of power system dispatching automation, in particular to a power system state estimation method and system for optimizing the solving process of a state estimation information matrix.
Background
Most of State Estimation programs of actual production running systems in current scheduling control systems are Fast Decomposed State Estimation (FDSE) based on the least square principle, and iterative calculation formulas of the State Estimation programs are shown as the following formula.
In the formula: z is a measurement vector (m-dimension),
is a state quantity (n dimension), R is a measurement error variance matrix, h is a nonlinear measurement vector function,
is a Jacobian matrix, G ═ HTR-1H is an information matrix, and H is an information matrix,
z generally includes branch power, node injection power, and node voltage amplitude measurements,
the state quantity correction amount is obtained.
Calculating the information matrix G ═ HTR-1H can be described as a Sparse matrix multiplication problem (SPMM); the equation (1-a) is a linear equation system solving problem, and is generally solved mathematically based on a trigonometric decomposition method, and since G is a symmetric real matrix, it is generally decomposed into LDU, where L ═ UT is a unit lower triangular matrix, and D is a diagonal matrix.
With the construction of a data platform during cloud and real time regulation, the state estimation calculation scale of the power system is increased day by day, for a state and provincial real-time data platform, the number of state estimation calculation nodes reaches 1-2 ten thousand nodes, and the number of measurement nodes reaches 5-7 ten thousand nodes, so that the existing state calculation program is difficult to meet the requirements of real-time calculation and analysis in efficiency. In FDSE, the calculation of the information matrix and the LU factorization of the coefficient matrix of the linear equation set of the iterative formula occupy most of the time of state estimation, and the improvement of the calculation speed of the state estimation information matrix and the LU factorization plays a significant role in accelerating the state of the whole power system.
Although sparse storage is adopted for solving the information matrix in the traditional state estimation program, operation is still carried out according to dense vectors in the matrix multiplication and factor decomposition processes, so that the calculation efficiency is reduced. In state estimation, the computation of the information matrix can be represented by the formula C ═ AB, where A, B is a real sparse matrix. Let the dimensions of the Sparse matrix A, B be m1 × n1 and m2 × n2, respectively, and n1 ═ m2, A, B are stored in csr (compressed Sparse row) format, ai _ nz, bi _ nz are the number of non-zero elements in the ith row A, B, respectively. Let C (i) be the ith row vector of C, C (i, j) be the nonzero element value of column (i) of C (i), solve C ═ AB by cumulative summation of product row by row, have
The matrix C is also typically a sparse matrix. Since the size of the space occupied by the sparse vector C (i)/matrix C in the CSR format cannot be known in advance, the conventional implementation method generally uses a dense vector C (i), the dimension of which is defined as n2, and C (i) is initialized or virtually calculated in advance before C (i, j) is calculated to obtain the storage space required by C (i), which consumes a long time and affects the calculation efficiency. For example, dense vector C (i) is adopted, all elements in C (i) need to be initialized to zero before progressive product accumulation summation is adopted due to non-zero element distribution of C (i) which is unknown in advance, and the CSR format storage of C is realized by sparse compression storage after C (i) operation is finished, so that the main problem of operation is that the original zero elements which do not need to be stored in sparse matrix C need to be initialized to zero in dense vector C (i).
In the state estimation problem solving process, when the number n of the computing nodes is large, a large amount of memory initialization (or memory allocation release) work needs to be carried out, so that the computing efficiency is reduced, when a single-precision floating point float type and a non-zero index are adopted to adopt a 4-byte integer type in the information matrix calculation of a 3 ten-thousand-node-scale power grid, the memory to be initialized in the solution C (i) reaches 0.228MB, the memory to be initialized in the solution C reaches 68.66MB, when the node scale is larger or the memory to be initialized in the double-precision mode is larger, the data transmission quantity between a cache and the memory is greatly increased, the cache hit rate is lower, and the computing efficiency of the information matrix is seriously influenced.
Non-zero injection elements are generated in the sparse matrix factorization process, the distribution of the non-zero elements of the factor table cannot be determined before the elimination operation according to the rows/columns, the conventional method for decomposing according to the rows/columns is to spread sparse vectors of rows of the coefficient matrix into dense vectors, the elimination operation is carried out based on the dense vectors, and after the operation of each row/column is finished, sparse compression storage is adopted, so that CSR/CSC format storage of the factor table is realized. Similar to the SPMM problem, the main problem with this implementation is that the zero elements that are not needed to be stored originally when the sparse vector is expanded into the dense vector are stored in the dense vector, and when the number n of the computation nodes is large, even if the program can improve the hit rate by using the spatial locality, a large amount of data still needs to be transmitted between the cache and the main memory, which affects the efficiency of triangle decomposition.
Disclosure of Invention
The invention aims to provide a power system state estimation method and a power system state estimation system, which can reduce memory operation in the solving process, reduce the calculation load and improve the calculation efficiency of power grid state estimation.
The technical scheme adopted by the invention is as follows:
in one aspect, the present invention provides a method for estimating a state of an electrical power system, including:
acquiring telemetering and remote signaling data of primary equipment of the power system in real time;
constructing a node branch model of the power system based on the acquired data;
determining non-zero element distribution of the active Jacobian matrix according to the node branch model;
determining the non-zero element distribution of the active information matrix according to the non-zero element distribution of the active Jacobian matrix;
carrying out node optimized numbering based on the active information matrix non-zero element distribution structure;
calculating an active Jacobian matrix and a reactive Jacobian matrix according to the new node sequence, and further obtaining a new active information matrix and a new reactive information matrix;
performing factorization on the new active information matrix and the new reactive information matrix respectively to obtain an active information matrix factor table and a reactive information matrix factor table;
and estimating the state of the power system based on the active information matrix factor table and the reactive information matrix factor table.
According to the invention, node optimized numbering is carried out based on active information matrix non-zero element distribution, and node optimized numbering is not carried out based on node outgoing degree and node measurement number, so that the number of non-zero elements generated in the subsequent information matrix factorization process can be reduced; on the basis, the new active information matrix and the new reactive information matrix are respectively subjected to factorization, so that the work of expanding sparse vectors into dense vectors and the low efficiency caused by frequent initialization and operation of array memories in the decomposition process are avoided, the calculation efficiency of high-dimensional sparse matrix triangular decomposition can be effectively improved, the memory operation in the solving process is reduced, the calculation load is reduced, and the calculation efficiency of power grid state estimation is improved.
Optionally, the active jacobian matrix non-zero element distribution and the active information matrix non-zero element distribution, and the active information matrix and the reactive information matrix after node number optimization are respectively sparsely stored by using a C + + standard template library association container definition structure; and storing the active information matrix factor table and the reactive information matrix factor table into a data structure defined by a C + + standard template library correlation container, or converting the active information matrix factor table and the reactive information matrix factor table into row-by-row/column-by-column sparse storage. The storage and operation of the sparse matrix are realized based on the STL association container, unnecessary memory operations in three processes of SPMM, node optimized numbering, triangle decomposition and the like are reduced, the data transmission quantity between the memory and the cache can be effectively reduced, and the calculation efficiency of state estimation is remarkably improved.
Optionally, the primary devices of the power system telemeter remote signaling data and concurrently obtain the remote signaling data from a real-time library list table set in the dispatching automation system respectively for the types of the primary devices.
Optionally, if m is defined as the active measurement number, n is the number of nodes calculated by the power grid, the active jacobian matrix HpForming an active information matrix by simulating multiplication operation for an m-x (n-1) dimension sparse matrix, wherein the active information matrix is as follows:
wherein
An active power measurement error variance matrix is obtained;
sparse matrix Hp、GpThe non-zero element distribution is stored by adopting an associated container map and set definition structure provided by C + + STL, and comprises the following steps: defining a structure: set < int > SparseRowNzSet is used for storing a certain row of non-zero element column number set, and the column number is represented by an index value of int type (integer variable) in the set; defining a structure typedefstd, map < int, SparseRowNzSet > SparseMatrixNzMap for storing sparse matrix Hp、GpThe distribution of non-zero elements, the row number, is represented by an int-type key index.
In an analog multiplication operation
In this case, only G is carried out without actually carrying out numerical calculationpDetermination of non-zero-element structure, for Hp、GpThe medium zero element is not stored.
Due to HpAll elements in the system are directly and completely independent of each other, so that a successful Jacobian matrix H can be formed in a parallel modepA non-zero element distribution.
Optionally, defining an active information matrix GpThe non-zero element distribution of (a) is g _ nz _ map, which is stored in a sparedmatrix xnzmap structure, and the node optimization numbering based on the active information matrix non-zero element distribution structure comprises:
obtaining node combinations of all outgoing lines of the initial coefficient matrix based on g _ nz _ map, defining the node combinations as d _ bus _ map, and storing the node combinations in a typedef std structure which is predefined based on a C + + STL associated container, wherein map is less than int, std set is less than int > DegreeBusMap structure;
using simulated gaussesActive information matrix G by elimination method and semi-dynamic minimum algorithmpThe nodes are sorted, and node sets of all outgoing degrees in the sorting process are respectively stored in a DegreeBusMap structure;
each element in the DegreeBusMap structure takes the outgoing degree value as a key value, and the actual value is a node set with the outgoing degree equal to the key value. And in the node optimization numbering process, only the processing of network structure change is carried out, and actual numerical operation is not carried out.
Optionally, for the active information matrix GpIn the process of sequencing the nodes, the insert and erase operations provided by C + + STL are used to realize the insertion of the newly added injection elements in the set and the deletion of the eliminated nodes. The method can avoid self-design of complex data structures and algorithm operations such as sequencing, adding and deleting, and the like, and improve the development and programming efficiency.
Optionally, the active information matrix G is subjected to a simulated Gaussian elimination method and a semi-dynamic minimum algorithmpThe node ordering includes:
s1, determining the node elements in the current g _ nz _ map and d _ bus _ map;
s2, selecting the node with the minimum node outgoing degree from the current d _ bus _ map as a current elimination node, and marking the node as k;
s3, correcting a set of associated nodes of the node k and the associated nodes thereof according to the g _ nz _ map, and recording the set as Link _ set;
s4, deleting the node k from the g _ nz _ map according to the key value of the node k, and numbering the currently deleted node k;
s5, correcting the d _ bus _ map according to the deleted g _ nz _ map of the node to serve as the updated current d _ bus _ map;
s6, repeating the steps S1 to S6 until all the nodes in the initial d _ bus _ map are sorted, and finishing the sorting operation.
The core idea of the semi-dynamic minimum degree sorting algorithm is that the nodes with the minimum non-zero elements are numbered preferentially in the decomposition process, the number of the non-zero elements of the row where the rest nodes are located is modified when the nodes are eliminated, and the numbered nodes and the edges starting from the numbered nodes do not participate in the subsequent analog elimination operation any more.
OptionalAfter the nodes are numbered optimally, the successful Jacobian matrix H is calculated in parallel according to the new node sequencepAnd a reactive Jacobian matrix HqThen active information matrix
And a reactive information matrix
Wherein,
is a reactive measurement error variance matrix, GqIs an n multiplied by n dimensional symmetric sparse matrix;
h is to bep、Hq、GpAnd GqAnd storing a sparse matrix by adopting an associated container definition structure provided by C + + STL, wherein zero elements in the matrix are not stored.
Optionally, the nodes are numbered optimally and then stored by adopting a sparseRowMap structurepAnd GqAnd performing factorization by adopting a Gaussian elimination method, calling insert operation of an association container by a newly generated non-zero injection element, or performing non-zero insertion by taking map as an association array. After the elimination operation based on spareRowMap, the non-zero column index of the factor table will be automatically sorted.
The present invention also provides a power system state estimation system, including:
the data acquisition module is used for acquiring telemetering and remote signaling data of primary equipment of the power system in real time;
the node branch model building module is used for building a node branch model of the power system based on the obtained data;
the active Jacobian matrix determining module is used for determining the non-zero element distribution of the active Jacobian matrix according to the node branch model;
the active information matrix determining module is used for determining the non-zero element distribution of the active information matrix according to the non-zero element distribution of the active Jacobian matrix;
the node sequencing module is used for carrying out node optimized numbering based on the active information matrix non-zero element distribution structure;
the information matrix determining module is used for calculating an active Jacobian matrix and a reactive Jacobian matrix according to the new node sequence so as to obtain a new active information matrix and a new reactive information matrix;
the factor decomposition module is used for respectively performing factor decomposition on the new active information matrix and the new reactive information matrix to obtain an active information matrix factor table and a reactive information matrix factor table;
and the subsequent processing module is used for estimating the state of the power system based on the active information matrix factor table and the reactive information matrix factor table.
Advantageous effects
Compared with the prior art, the invention has the following progress:
(1) node optimized numbering is carried out based on active information matrix non-zero element distribution, and the active information matrix and the reactive information matrix after node sequencing are subjected to factorization by adopting a Gaussian elimination method, so that the number of non-zero elements generated in the factorization process of the information matrix can be reduced, the work of expanding sparse vectors into dense vectors is avoided, the low efficiency caused by frequently initializing array memories and calculating in the factorization process is avoided, and the calculation efficiency of high-dimensional sparse matrix triangular decomposition is improved;
(2) by realizing the storage and operation of the sparse matrix based on the STL association container, unnecessary memory operations in three processes of SPMM, node optimized numbering, triangle decomposition and the like are reduced, the data transmission quantity between the memory and the cache can be effectively reduced, and the calculation efficiency of state estimation is obviously improved;
(3) the STL-associated-container-based programming can effectively utilize an STL advanced and efficient algorithm to realize the management of memory data, avoid self-writing of complex data structures and algorithms, reduce the programming complexity and effectively improve the programming efficiency and the program quality;
(4) the method can be effectively compatible with the existing software and hardware platform in the current dispatching control system, does not change the deployment architecture and the overall flow of the state estimation application, and can reuse the existing program function.
Drawings
FIG. 1 is a schematic diagram illustrating a process of solving an information matrix in the state estimation method according to the present invention;
FIG. 2 is a schematic diagram of a non-zero distribution of a Jacobian matrix;
FIG. 3 is a schematic diagram of information matrix solution;
fig. 4 is a schematic diagram illustrating a fast node sorting flow based on STL association containers.
Detailed Description
The following further description is made in conjunction with the accompanying drawings and the specific embodiments.
Although sparse storage is adopted for solving the information matrix in the traditional state estimation program, operation is still carried out according to dense vectors in the matrix multiplication and factor decomposition processes, so that the calculation efficiency is reduced. The invention conception of the invention is as follows: sparse algorithms are adopted in the processes of solving the information matrix, matrix multiplication and factorization, so that the calculation efficiency of state estimation is improved.
Example 1
The present embodiment is a method for estimating a state of an electric power system, and referring to fig. 1, the method includes:
acquiring telemetering and remote signaling data of primary equipment of the power system in real time;
constructing a node branch model of the power system based on the acquired data;
determining non-zero element distribution of the active Jacobian matrix according to the node branch model;
determining the non-zero element distribution of the active information matrix according to the non-zero element distribution of the active Jacobian matrix;
carrying out node optimized numbering based on the active information matrix non-zero element distribution structure;
calculating an active Jacobian matrix and a reactive Jacobian matrix according to the new node sequence, and further obtaining a new active information matrix and a new reactive information matrix;
performing factorization on the new active information matrix and the new reactive information matrix respectively to obtain an active information matrix factor table and a reactive information matrix factor table;
and estimating the state of the power system based on the active information matrix factor table and the reactive information matrix factor table.
Node optimized numbering is carried out based on active information matrix non-zero element distribution, and node optimized numbering is not carried out based on node outgoing degree and node measurement number, so that the number of non-zero elements generated in the subsequent information matrix factorization process can be reduced; on the basis, the new active information matrix and the new reactive information matrix are respectively subjected to factorization, a Gaussian elimination method can be adopted, the work of expanding sparse vectors into dense vectors and the low efficiency caused by the need of frequently initializing array memories and calculating in the decomposition process can be avoided, the calculation efficiency of high-dimensional sparse matrix triangular decomposition can be effectively improved, the memory operation in the solving process is reduced, the calculation load is reduced, and the calculation efficiency of power grid state estimation is improved.
On the basis, in this embodiment, for active jacobian matrix non-zero element distribution and active information matrix non-zero element distribution, and the active information matrix and reactive information matrix after node optimized numbering, C + + standard template library associated container definition structures are respectively adopted for sparse storage; and storing the active information matrix factor table and the reactive information matrix factor table into a data structure defined by a C + + standard template library correlation container, or converting the active information matrix factor table and the reactive information matrix factor table into row-by-row/column-by-column sparse storage. The storage and operation of the sparse matrix are realized based on the STL association container, unnecessary memory operations in three processes of SPMM, node optimized numbering, triangle decomposition and the like are reduced, the data transmission quantity between the memory and the cache can be effectively reduced, and the calculation efficiency of state estimation is remarkably improved.
Examples 1 to 1
Based on the basic flow of embodiment 1, the present embodiment specifically describes the flow of the power system state estimation method.
First, acquisition of remote signaling and remote measuring data of primary equipment of power system
And the primary equipment of each power system telemeters remote signaling data and reads the data from a real-time library list table which is set in the dispatching automation system respectively aiming at the types of the primary equipment in parallel. The primary equipment of the power system comprises a generator, a transformer, a line, a load, a capacitance reactor, a circuit breaker, a disconnecting switch and the like.
Second, node branch model construction
According to a relational power grid model in a dispatching automation system, data including connection relations of primary equipment, equipment parameters and the like are subjected to topology analysis, and then a node branch model is formed, which can refer to the prior art.
Three, Jacobian matrix formation
Forming an active Jacobian matrix H in parallel according to the node branch modelpNon-zero element distribution, as shown in FIG. 2, the active information matrix is formed by analog multiplication
Non-zero bin distribution, as shown in fig. 3.
Active jacobian matrix HpIs a sparse matrix of m x (n-1), m is the active power measurement number, n is the number of nodes calculated by the power grid, and an active information matrix
In (1),
an active power measurement error variance matrix is obtained;
sparse matrix Hp、GpThe non-zero element distribution is stored by adopting an associated container map and set definition structure provided by C + + STL, and comprises the following steps: defining a structure: set < int > SparseRowNzSet is used for storing a certain row of non-zero element column number set, and the column number is represented by an index value of int type in the set; defining a structure: typedefstd, map < int, SparseRowNzSet > SparseMatrixNzMap for storing sparse matrix Hp、GpA distribution of non-zero elements, the row number being represented by an int-type key index;
in an analog multiplication operation
In this case, only G is carried out without actually carrying out numerical calculationpDetermination of non-zero-element structure, for Hp、GpThe medium zero element is not stored.
Due to HpAll elements in the system are directly and completely independent of each other, so that a successful Jacobian matrix H can be formed in a parallel modepA non-zero element distribution.
Three, quick node minimum node optimization numbering
In the fast decomposition state estimation, the node optimization numbering is performed based on the node outgoing degree (based on the node admittance matrix Y) and the node measurement number (Jacobian matrix H) in the traditional method, and although the numbering method is simple, the Y, H non-zero element distribution is inconsistent with the non-zero element distribution of the coefficient G of the linear equation set in the FDSE iterative equation. In order to reduce the number of non-zero elements generated in the factorization process of the information matrix, the invention carries out node sequencing based on the non-zero element structure of the active information matrix.
In the sorting process, a structure typedef std is defined based on a C + + STL association container, map is less than int, std set is less than int > DegreeBusMap and used for recording node sets of all outgoing degrees in the sorting process, key values are outgoing degree values, and actual values are node set sets of which the outgoing degrees are equal to the key values. The core idea of the semi-dynamic minimum degree sorting algorithm is that the nodes with the minimum non-zero elements are numbered preferentially in the decomposition process, the number of the non-zero elements of the row where the rest nodes are located is modified when the nodes are eliminated, and the numbered nodes and the edges starting from the numbered nodes do not participate in the subsequent analog elimination operation any more. And in the node sequencing process, insert and delete of the newly added injection elements in the set are realized by using insert and erase operations provided by a C + + standard template library, so that the self-designed complex data structure and algorithm operations such as sequencing, adding and deleting of the complex data structure are avoided. Only the processing of network structure change is carried out in the node optimization numbering process, and actual numerical operation is not carried out.
Specifically, the node optimization sequencing process is shown in fig. 4, and an active information matrix G is definedpThe non-zero element distribution of (a) is g _ nz _ map, which is stored in a sparedmatrix xnzmap structure, and the node optimization numbering based on the active information matrix non-zero element distribution structure comprises:
obtaining node combinations of all outgoing lines of the initial coefficient matrix based on g _ nz _ map, defining the node combinations as d _ bus _ map, and storing the node combinations in a typedef std structure which is predefined based on a C + + STL associated container, wherein map is less than int, std set is less than int > DegreeBusMap structure;
active information matrix G by adopting simulated Gaussian elimination method and semi-dynamic minimum algorithmpThe nodes are sorted, and node sets of all outgoing degrees in the sorting process are respectively stored in a DegreeBusMap structure;
each element in the DegreeBusMap structure takes the outgoing degree value as a key value, and the actual value is a node set with the outgoing degree equal to the key value.
Wherein, the active information matrix G is processed by adopting a simulated Gaussian elimination method and a semi-dynamic minimum algorithmpThe node ordering includes:
s1, determining the node elements in the current g _ nz _ map and d _ bus _ map;
s2, selecting the node with the minimum node outgoing degree from the current d _ bus _ map as a current elimination node, and marking the node as k;
s3, correcting a set of associated nodes of the node k and the associated nodes thereof according to the g _ nz _ map, and recording the set as Link _ set;
s4, deleting the node k from the g _ nz _ map according to the key value of the node k, and numbering the currently deleted node k;
s5, correcting the d _ bus _ map according to the deleted g _ nz _ map of the node to serve as the updated current d _ bus _ map;
s6, repeating the steps S1 to S6 until all the nodes in the initial d _ bus _ map are sorted, and finishing the sorting operation.
Four, active and reactive jacobian matrix and information matrix formation
After the nodes are numbered optimally, the successful Jacobian matrix H is calculated in parallel according to the new node sequencepAnd a reactive Jacobian matrix HqThen active information matrix
And a reactive information matrix
Wherein,
is a reactive measurement error variance matrix, GqIs an n multiplied by n dimensional symmetric sparse matrix;
h is to bep、Hq、GpAnd GqAnd storing a sparse matrix by adopting an associated container definition structure provided by C + + STL, wherein zero elements in the matrix are not stored. GpAnd GqAnd sparse storage is performed by adopting an associated container map and set definition structure provided by C + + STL. Defining the structure: map < int, float > SparseRowMap is used for storing sparse matrix row nonzero elements, column numbers are represented by int-type key indexes, and sparse matrix nonzero elements are represented by float-type associated values; the structure typedefstd is defined, map < int, SparseRowMap > SparseMatrixMap for storing sparse matrices, the row number being represented by a key index of int type.
Active jacobian matrix H in fast decoupled state estimationpAnd a reactive Jacobian matrix HqIndependent of each other, and adopts parallel computing method to realize HpAnd HqParallel computation of (2).
Information matrix factorization
Due to the active information matrix GpAnd GqThe factor tables are independent of each other, so that the nodes are stored by adopting a SparseRowMap structure after being numbered optimallypAnd GqAnd carrying out factorization by adopting a Gaussian elimination method in parallel to obtain a corresponding factor table.
Because the distribution of the non-zero elements of the factor table cannot be determined before the elimination operation, the traditional method is to spread the sparse vectors of the rows of the coefficient matrix into dense vectors according to the rows, carry out the elimination operation based on the dense vectors, and adopt sparse compression storage after the operation of each row is finished so as to realize the sparse format storage of the factor table.
This embodiment is directed to the information matrix G adopting the SparseRowMap structurepAnd GqDirectly decomposing the data by rows by adopting a Gaussian elimination method, calling insert operation of an associated container by a newly generated non-zero injection element or realizing efficient insertion of the non-zero element by taking a map as an associated array, and carrying out elimination operation based on a SparseRowMapThe column index of the non-zero elements of the factor table is automatically sorted. Therefore, the work of expanding the sparse vector into the dense vector and the low efficiency caused by the need of frequently initializing the array memory and calculating the array memory according to the row decomposition are avoided, and the calculation efficiency of the high-dimensional sparse matrix triangular decomposition is effectively improved.
Sixthly, storing the information matrix factor table
In this embodiment, the active information matrix factor table and the reactive information matrix factor table are stored in a data structure defined by a C + + STL association container or are converted into row-by-row/column-by-column sparse storage for use by a subsequent program.
Seven, subsequent state estimation calculation
The subsequent state estimation calculation is performed based on the information matrix factor table stored in the step six, including iterative calculation of node voltage and the like, and the prior art can be adopted, which is not considered as the research content of the invention and is not repeated.
Example 2
Based on the same inventive concept as embodiment 1, the present embodiment is a power system state estimation system, including:
the data acquisition module is used for acquiring telemetering and remote signaling data of primary equipment of the power system in real time;
the node branch model building module is used for building a node branch model of the power system based on the obtained data;
the active Jacobian matrix determining module is used for determining the non-zero element distribution of the active Jacobian matrix according to the node branch model;
the active information matrix determining module is used for determining the non-zero element distribution of the active information matrix according to the non-zero element distribution of the active Jacobian matrix;
the node sequencing module is used for carrying out node optimized numbering based on the active information matrix non-zero element distribution structure;
the information matrix determining module is used for calculating an active Jacobian matrix and a reactive Jacobian matrix according to the new node sequence so as to obtain a new active information matrix and a new reactive information matrix;
the factor decomposition module is used for performing factor decomposition on the new active information matrix and the new reactive information matrix by adopting a Gaussian elimination method respectively, and storing an active information matrix factor table and a reactive information matrix factor table obtained through decomposition;
and the subsequent processing module is used for estimating the state of the power system based on the active information matrix factor table and the reactive information matrix factor table.
The specific functions of the modules are implemented as the specific implementation manner in reference to example 1-1.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a system for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction system which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While the present invention has been described with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, which are illustrative and not restrictive, and it will be apparent to those skilled in the art that various changes and modifications can be made therein without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (10)
1. A method for estimating a state of an electric power system, comprising:
acquiring telemetering and remote signaling data of primary equipment of the power system in real time;
constructing a node branch model of the power system based on the acquired data;
determining non-zero element distribution of the active Jacobian matrix according to the node branch model;
determining the non-zero element distribution of the active information matrix according to the non-zero element distribution of the active Jacobian matrix;
carrying out node optimized numbering based on the active information matrix non-zero element distribution structure;
calculating an active Jacobian matrix and a reactive Jacobian matrix according to the new node sequence, and further obtaining a new active information matrix and a new reactive information matrix;
performing factorization on the new active information matrix and the new reactive information matrix respectively to obtain an active information matrix factor table and a reactive information matrix factor table;
and estimating the state of the power system based on the active information matrix factor table and the reactive information matrix factor table.
2. The method according to claim 1, wherein the active jacobian matrix non-zero element distribution and the active information matrix non-zero element distribution, and the active information matrix and the reactive information matrix after node number optimization are respectively sparsely stored by using a C + + standard template library association container definition structure; and storing the active information matrix factor table and the reactive information matrix factor table into a data structure defined by a C + + standard template library correlation container, or converting the active information matrix factor table and the reactive information matrix factor table into row-by-row/column-by-column sparse storage.
3. A method according to claim 1 or 2, characterized in that each power system primary unit telemeters remote signalling data and retrieves it from a real-time library list set in the scheduling automation system for each primary unit type.
4. A method according to claim 1 or 2, characterized in that if m is defined as the number of active measurements and n is the number of nodes of the grid calculation, the active jacobian matrix H is the number of active jacobian nodespForming an active information matrix by simulating multiplication operation for an m-x (n-1) dimension sparse matrix, wherein the active information matrix is as follows:
wherein
An active power measurement error variance matrix is obtained;
sparse matrix Hp、GpThe non-zero element distribution is stored by adopting an associated container map and set definition structure provided by C + + STL, and comprises the following steps: defining a structure: set < int > SparseRowNzSet for storing a certain row of non-zero element column number set, wherein the column number is represented by an index value of int type in the set; defining a structure: typedefstd, map < int, SparseRowNzSet > SparseMatrixNzMap, for storing sparse matrix Hp、GpThe distribution of non-zero elements, the row number, is represented by an int-type key index.
5. Method according to claim 4, characterized in that the active information matrix G is definedpThe non-zero element distribution of (a) is g _ nz _ map, which is stored in a sparedmatrix xnzmap structure, and the node optimization numbering based on the active information matrix non-zero element distribution structure comprises:
obtaining node combinations of all outgoing lines of the initial coefficient matrix based on g _ nz _ map, defining the node combinations as d _ bus _ map, and storing the node combinations in a typedef std structure which is predefined based on a C + + STL associated container, wherein map is less than int, std set is less than int > DegreeBusMap structure;
active information matrix G by adopting simulated Gaussian elimination method and semi-dynamic minimum algorithmpThe nodes are sorted, and node sets of all outgoing degrees in the sorting process are respectively stored in a DegreeBusMap structure;
each element in the DegreeBusMap structure takes the outgoing degree value as a key value, and the actual value is a node set with the outgoing degree equal to the key value.
6. The method of claim 5, wherein the active information matrix G is alignedpIn the process of sequencing the nodes, the insert and erase operations provided by C + + STL are used to realize the insertion of the newly added injection elements in the set and the deletion of the eliminated nodes.
7. Method according to claim 5 or 6, characterized in that the active information matrix G is subjected to simulated Gaussian elimination and semi-dynamic minimizationpThe node ordering includes:
s1, determining the node elements in the current g _ nz _ map and d _ bus _ map;
s2, selecting the node with the minimum node outgoing degree from the current d _ bus _ map as a current elimination node, and marking the node as k;
s3, correcting a set of associated nodes of the node k and the associated nodes thereof according to the g _ nz _ map, and recording the set as Link _ set;
s4, deleting the node k from the g _ nz _ map according to the key value of the node k, and numbering the currently deleted node k;
s5, correcting the d _ bus _ map according to the deleted g _ nz _ map of the node to serve as the updated current d _ bus _ map;
s6, repeating the steps S1 to S6 until all the nodes in the initial d _ bus _ map are sorted, and finishing the sorting operation.
8. The method of claim 1, wherein after the nodes are numbered optimally, the successful jacobian matrix H is computed in parallel from the new node orderpAnd a reactive Jacobian matrix HqThen active information matrix
And a reactive information matrix
Wherein,
is a reactive measurement error variance matrix, GqIs an n multiplied by n dimensional symmetric sparse matrix;
h is to bep、Hq、GpAnd GqAnd storing a sparse matrix by adopting an associated container definition structure provided by C + + STL, wherein zero elements in the matrix are not stored.
9. The method as claimed in claim 8, wherein the numbering is optimized for nodes and then the G stored using SparseRowMap structurepAnd GqAnd performing factorization by adopting a Gaussian elimination method, calling insert operation of an association container by a newly generated non-zero injection element, or performing non-zero insertion by taking map as an association array.
10. A power system state estimation system, comprising:
the data acquisition module is used for acquiring telemetering and remote signaling data of primary equipment of the power system in real time;
the node branch model building module is used for building a node branch model of the power system based on the obtained data;
the active Jacobian matrix determining module is used for determining the non-zero element distribution of the active Jacobian matrix according to the node branch model;
the active information matrix determining module is used for determining the non-zero element distribution of the active information matrix according to the non-zero element distribution of the active Jacobian matrix;
the node sequencing module is used for carrying out node optimized numbering based on the active information matrix non-zero element distribution structure;
the information matrix determining module is used for calculating an active Jacobian matrix and a reactive Jacobian matrix according to the new node sequence so as to obtain a new active information matrix and a new reactive information matrix;
the factor decomposition module is used for respectively performing factor decomposition on the new active information matrix and the new reactive information matrix to obtain an active information matrix factor table and a reactive information matrix factor table;
and the subsequent processing module is used for estimating the state of the power system based on the active information matrix factor table and the reactive information matrix factor table.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911294697.7A CN111062610B (en) | 2019-12-16 | 2019-12-16 | Power system state estimation method and system based on information matrix sparse solution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911294697.7A CN111062610B (en) | 2019-12-16 | 2019-12-16 | Power system state estimation method and system based on information matrix sparse solution |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111062610A true CN111062610A (en) | 2020-04-24 |
| CN111062610B CN111062610B (en) | 2022-07-29 |
Family
ID=70301898
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911294697.7A Active CN111062610B (en) | 2019-12-16 | 2019-12-16 | Power system state estimation method and system based on information matrix sparse solution |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111062610B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112084198A (en) * | 2020-09-16 | 2020-12-15 | 云南电网有限责任公司 | Direct column writing method and device for node admittance matrix in compressed storage form |
| CN113095668A (en) * | 2021-04-07 | 2021-07-09 | 国网山东省电力公司电力科学研究院 | Power system state estimation edge server configuration method and system based on plant station |
| CN113139718A (en) * | 2021-04-07 | 2021-07-20 | 国网山东省电力公司电力科学研究院 | Parallel estimation method and system for power system state |
| CN113326474A (en) * | 2021-05-31 | 2021-08-31 | 山东大学 | Optimization calculation method and system for regulation sensitivity of elements of Jacobian matrix of interconnected large power grid |
| CN113886765A (en) * | 2021-09-30 | 2022-01-04 | 广东电网有限责任公司 | Method and device for detecting error data injection attack |
| CN117290653A (en) * | 2023-11-24 | 2023-12-26 | 巨霖科技(上海)有限公司 | Matrix solving method and system based on EDA system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102836A (en) * | 2014-07-14 | 2014-10-15 | 国家电网公司 | Method for quickly estimating robust state of power system |
| CN104184144A (en) * | 2014-09-05 | 2014-12-03 | 国家电网公司 | Robust state estimation method used for multi-voltage-class power grid model |
| US20170221530A1 (en) * | 2016-01-29 | 2017-08-03 | System Operator Of The United Power System | Computerized device for the marginal state assessment of power systems |
| CN107579516A (en) * | 2017-09-13 | 2018-01-12 | 国电南瑞科技股份有限公司 | A kind of method for improving Power system state estimation calculating speed |
| CN109255541A (en) * | 2018-09-20 | 2019-01-22 | 华北电力大学 | A kind of power distribution network robust state estimation method the sum of multiplied based on least square and one |
| CN109840651A (en) * | 2017-11-28 | 2019-06-04 | 国家电网公司 | A kind of bilinearity WLAV method for estimating state considering zero injection-constraint |
-
2019
- 2019-12-16 CN CN201911294697.7A patent/CN111062610B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102836A (en) * | 2014-07-14 | 2014-10-15 | 国家电网公司 | Method for quickly estimating robust state of power system |
| CN104184144A (en) * | 2014-09-05 | 2014-12-03 | 国家电网公司 | Robust state estimation method used for multi-voltage-class power grid model |
| US20170221530A1 (en) * | 2016-01-29 | 2017-08-03 | System Operator Of The United Power System | Computerized device for the marginal state assessment of power systems |
| CN107579516A (en) * | 2017-09-13 | 2018-01-12 | 国电南瑞科技股份有限公司 | A kind of method for improving Power system state estimation calculating speed |
| CN109840651A (en) * | 2017-11-28 | 2019-06-04 | 国家电网公司 | A kind of bilinearity WLAV method for estimating state considering zero injection-constraint |
| CN109255541A (en) * | 2018-09-20 | 2019-01-22 | 华北电力大学 | A kind of power distribution network robust state estimation method the sum of multiplied based on least square and one |
Non-Patent Citations (1)
| Title |
|---|
| 罗玉春等: "KLU稀疏直接求解器在状态估计中的应用", 《中国电力》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112084198A (en) * | 2020-09-16 | 2020-12-15 | 云南电网有限责任公司 | Direct column writing method and device for node admittance matrix in compressed storage form |
| CN112084198B (en) * | 2020-09-16 | 2022-09-16 | 云南电网有限责任公司 | Direct column writing method and device for node admittance matrix in compressed storage form |
| CN113095668A (en) * | 2021-04-07 | 2021-07-09 | 国网山东省电力公司电力科学研究院 | Power system state estimation edge server configuration method and system based on plant station |
| CN113139718A (en) * | 2021-04-07 | 2021-07-20 | 国网山东省电力公司电力科学研究院 | Parallel estimation method and system for power system state |
| CN113095668B (en) * | 2021-04-07 | 2023-05-05 | 国网山东省电力公司电力科学研究院 | Power system state estimation edge server configuration method and system based on plant station |
| CN113326474A (en) * | 2021-05-31 | 2021-08-31 | 山东大学 | Optimization calculation method and system for regulation sensitivity of elements of Jacobian matrix of interconnected large power grid |
| CN113326474B (en) * | 2021-05-31 | 2022-05-31 | 山东大学 | Optimization calculation method and system for regulation sensitivity of elements of Jacobian matrix of interconnected large power grid |
| CN113886765A (en) * | 2021-09-30 | 2022-01-04 | 广东电网有限责任公司 | Method and device for detecting error data injection attack |
| CN113886765B (en) * | 2021-09-30 | 2023-09-29 | 广东电网有限责任公司 | Method and device for detecting error data injection attack |
| CN117290653A (en) * | 2023-11-24 | 2023-12-26 | 巨霖科技(上海)有限公司 | Matrix solving method and system based on EDA system |
| CN117290653B (en) * | 2023-11-24 | 2024-02-20 | 巨霖科技(上海)有限公司 | Matrix solving method and system based on EDA system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111062610B (en) | 2022-07-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111062610B (en) | Power system state estimation method and system based on information matrix sparse solution | |
| CN105528191A (en) | Data accumulation apparatus and method, and digital signal processing device | |
| CN111682530B (en) | Method, device, equipment and medium for determining out-of-limit probability of voltage of power distribution network | |
| CN114492814A (en) | Method, device and medium for simulating energy of target system based on quantum computation | |
| CN112054519A (en) | Power distribution network low voltage optimization treatment method, system and equipment | |
| CN107579516B (en) | Method for improving state estimation calculation speed of power system | |
| CN113326343B (en) | Road network data storage method and system based on multi-level grids and file indexes | |
| CN117808101A (en) | Neural network reasoning method, system and storage medium based on FPGA | |
| CN113947237A (en) | AMS-TCN-based wind power error correction method | |
| CN110489798B (en) | Fine-grained efficient design method of resolver for active power distribution network real-time simulation | |
| CN113094899A (en) | Random power flow calculation method and device, electronic equipment and storage medium | |
| CN101561797A (en) | Method and device for singular value and feature value composition of matrix on processing system | |
| CN108595148A (en) | Division function implementation method, circuit, chip and system | |
| CN117290653A (en) | Matrix solving method and system based on EDA system | |
| CN104897176B (en) | A kind of photogrammetric block adjustment method of multi-core parallel concurrent | |
| CN114970165A (en) | Digital entity automatic modeling method, system and equipment for cable in bridge | |
| CN115481364A (en) | Parallel computing method for large-scale elliptic curve multi-scalar multiplication based on GPU (graphics processing Unit) acceleration | |
| CN115173421A (en) | Probability optimal power flow calculation method based on progressive Latin hypercube sampling | |
| CN111797564A (en) | Method and system for obtaining correlation sample of high-dimensional distributed photovoltaic output | |
| CN112217215A (en) | PSD-BPA-based large-scale power system random load flow calculation method | |
| Pan et al. | Compact multigrid | |
| CN118569686A (en) | Power grid user load data complement method, terminal and medium based on low-rank matrix complement and variation modal decomposition | |
| CN114336793B (en) | Method for determining flexibility of alternating current/direct current hybrid power distribution network | |
| CN112952843B (en) | Distributed photovoltaic probability power flow calculation method for power distribution network based on Copula theory | |
| CN113836481B (en) | Matrix computing circuit, method, electronic device, and computer-readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |