CN107134790B - Power distribution network reactive power optimization control sequence determination method based on big data - Google Patents
Power distribution network reactive power optimization control sequence determination method based on big data Download PDFInfo
- Publication number
- CN107134790B CN107134790B CN201610109681.4A CN201610109681A CN107134790B CN 107134790 B CN107134790 B CN 107134790B CN 201610109681 A CN201610109681 A CN 201610109681A CN 107134790 B CN107134790 B CN 107134790B
- Authority
- CN
- China
- Prior art keywords
- matrix
- data
- reactive power
- load
- distribution network
- 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.)
- Active
Links
- 238000005457 optimization Methods 0.000 title claims abstract description 113
- 238000009826 distribution Methods 0.000 title claims abstract description 93
- 238000000034 method Methods 0.000 title claims abstract description 52
- 239000011159 matrix material Substances 0.000 claims abstract description 133
- 238000001228 spectrum Methods 0.000 claims abstract description 21
- 239000013598 vector Substances 0.000 claims description 18
- 230000001186 cumulative effect Effects 0.000 claims description 10
- 238000012544 monitoring process Methods 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 10
- 229940050561 matrix product Drugs 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 230000003595 spectral effect Effects 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 238000004519 manufacturing process Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 230000003416 augmentation Effects 0.000 claims description 3
- 238000007726 management method Methods 0.000 description 14
- 238000005516 engineering process Methods 0.000 description 11
- 238000004422 calculation algorithm Methods 0.000 description 7
- 238000013473 artificial intelligence Methods 0.000 description 6
- 238000013459 approach Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000013178 mathematical model Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 3
- 230000003190 augmentative effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 239000003990 capacitor Substances 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 125000002950 monocyclic group Chemical group 0.000 description 1
- 230000035699 permeability Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 238000002922 simulated annealing Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/18—Arrangements for adjusting, eliminating or compensating reactive power in networks
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/30—Reactive power compensation
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention relates to a method for determining a reactive power optimization control sequence of a power distribution network based on big data, which comprises the steps of utilizing the existing historical load data and historical optimization data of a power system as data sources, calculating the average spectrum radius of a load random matrix by constructing a reactive power optimization random matrix of the power distribution network in combination with a single-loop law, calculating a correlation coefficient of historical loads and current loads according to the average spectrum radius, obtaining load days with larger correlation by comparing the magnitude of the correlation coefficient, and using the reactive power optimization control sequence of the day as the current optimization sequence. The technical scheme of the invention solves the problem of reactive power operation optimization of the power distribution network.
Description
Technical Field
The invention relates to the technical field of power distribution networks, in particular to a power distribution network reactive power optimization control sequence determination method based on big data.
Background
The reactive power optimization of the power distribution network aims at improving the voltage quality and reducing the network loss, takes reactive power compensation or regulation equipment as a control means, and belongs to the problem of nonlinear programming. In the face of power distribution network data with high dimension, strong coupling and high randomness under a big data background, the traditional reactive power optimization method based on linear programming or nonlinear programming mainly has the problem that the global optimal solution is difficult to obtain, and the reactive power optimization method based on artificial intelligence algorithms such as genetic algorithm, simulated annealing algorithm and tabu search algorithm mainly has the problem that the convergence speed is slow or the convergence is difficult. Meanwhile, the increasing scale of the power distribution network data enables information in massive historical data to be richer and valuable information capable of being mined to be increased, so that the big data technology indicates a new exploration direction for the power distribution network reactive power optimization technology.
Along with the fact that big data are deep into the aspects of life of people, big data technology has become a new trend of scientific and technological development. The information contained in the big data is from multiple, heterogeneous and independent data sources, and complex association exists between the data. The development of data acquisition and data storage technology enables the original data to be remarkably increased in both dimension and scale, and in addition, the multivariate and heterogeneous characteristics and the complex incidence relation of the big data increase the difficulty of data processing and effective information mining. Big data technology brings great challenges and great opportunities to people. The large data figure is full in the industrial production field, the traffic field and the internet field, the wave of the large data has impacted various aspects of people's life, and people have entered the large data era.
The big data technology emphasizes information technology and communication technology at the beginning of birth, and is then applied to other fields including large-scale power grid data analysis and the like. With the popularization of smart grid construction, especially the gradual improvement and perfection of a grid data acquisition system and an online operation monitoring system, grid operation data (especially distribution grid operation data) show explosive growth. In addition, the permeability of distributed power supplies, electric vehicles and flexible loads is gradually increased, the scale of the power distribution network is continuously enlarged, and the structure is gradually complicated, so that the data of the power distribution network is high in dimension, strong in coupling performance and high in randomness, and huge pressure is directly brought to load flow calculation and reactive power optimization.
Therefore, the reactive power optimization potential of the large power data is mined, and the optimization of the reactive power resource operation of the power distribution network is the necessary requirement for realizing reactive power resource coordination control technology research and corresponding component development. At present, reactive power optimization of a power distribution network mainly depends on reactive power compensation equipment such as a capacitor bank in a transformer substation and at the head end of a feeder line, and an on-load adjustable transformer for adjusting reactive power and voltage of the whole network.
The reactive operation optimization is dynamic optimization which takes the minimum electric energy loss (or cost) of a system, the highest voltage qualified rate of each node and the minimum switching times of a transformer tap, a capacitor and a reactor as objective functions on the basis of the fixed and unchangeable grid structure of a power grid, and is essentially a large-scale nonlinear mixed integer optimization problem. At present, various optimization algorithms have achieved a lot of achievements, and the optimization algorithms are mainly divided into a traditional reactive power optimization method and an artificial intelligence optimization method.
The traditional reactive power optimization method has many problems: (1) the method depends on an accurate mathematical model, but the accurate mathematical model is complex and difficult to adapt to the requirement of real-time control, and the rough mathematical model has larger error. (2) The requirement on the initial point is strict, the true optimum can be achieved only when the initial point is more forceful from the global optimum solution, otherwise, only suboptimal solution or infeasible solution can be generated. (3) The reactive power optimization method based on the derivative information has certain limits on objective functions and constraint conditions, such as continuity, differentiability and the like, and also simplifies the approximation processing if necessary. In addition, the artificial intelligence optimization method has some defects: (1) because the artificial intelligence algorithm does not depend on the derivation of the model, loop iteration is needed, the convergence speed is low, and the real-time performance is difficult to achieve. (2) The iteration times of the algorithm are determined, the convergence time is difficult to predict, and the convergence time is unstable. Furthermore, the method is simple. No matter the traditional optimization method or the artificial intelligence method, the problem of dimension disaster is faced when the reactive power optimization problem of the large-scale power distribution network is solved.
Therefore, the invention aims to overcome the defects of the existing reactive power optimization technology and provide a method for determining a reactive power optimization control sequence of a power distribution network based on big data.
Disclosure of Invention
The invention aims to provide a method for determining a reactive power optimization control sequence of a power distribution network based on big data, and the method solves the problem of reactive power operation optimization of the power distribution network.
In order to achieve the purpose, the invention adopts the following technical scheme: a method for determining a reactive power optimization control sequence of a power distribution network based on big data comprises the following steps:
step 1: acquiring a reactive power optimization data source of the power distribution network;
step 2: constructing a power distribution network reactive power optimization random matrix by combining a single-loop law according to the power distribution network reactive power optimization data source;
and step 3: the power distribution network reactive power optimization random matrix is subjected to equal singular value equivalence transformation, and the average spectrum radius of the reactive power optimization random matrix is obtained by using an average spectrum radius obtaining method; calculating a correlation coefficient between the historical load and the current load according to the average spectrum radius;
and 4, step 4: and selecting the date with stronger correlation coefficient according to the correlation coefficient of the historical load and the current load, and taking the reactive power optimization sequence on the date as an optimization result.
The method comprises the following steps that 1, a power distribution network reactive power optimization data source is obtained according to power distribution management system data; and integrating, storing and processing the data of the power distribution management system through a distributed data source to form a power distribution network reactive power optimization data source.
The process of constructing the distribution network reactive power optimization random matrix by combining the single-loop law according to the distribution network reactive power optimization data source in the step 2 comprises the following steps:
setting omega phi as a power distribution network reactive power optimization data source set phin,tRepresenting the nth bus, the t-th sampling data in a single day by a matrix phiiI-1, 2,3, …, L denotes the reactive power optimization data source matrix on day i, { ΦiExpressing a reactive power optimization data source matrix sequence; the number of the power grid buses is N, the sampling frequency of data in a single day is T, and the reactive power optimization data source matrix is expressed as follows:
thus, an active load data source matrix P is obtainediAnd a reactive power data source matrix QiThen the load data source matrix is represented as:
defining a cumulative load function is (t):
wherein s (τ) represents the load at τ, is (T) represents the cumulative load at T hours, and when the single-day sampling frequency of the load data is T, the approximate discretization of the cumulative load function is represented as:
wherein T is 1,2,3, …, T, sn,kRepresenting the nth bus and the kth sampling load; for load data source matrix Si＝{snt}2N×TCalculating the totalLoad data source matrix:
let the sequence for the history cumulative load matrix be { ShiAnd (6) defining a reactive power optimization load random matrix sequence (A) when the accumulative load matrix Sd is predicted currentlyiIn which A isiIs a 2N × 2T dimension augmentation matrix:
Ai＝[Shi,Sd],i＝1,2,3,…,L
then the matrix A is augmentediNamely the reactive power optimization random matrix of the power distribution network.
The process of solving the average spectrum radius of the reactive power optimization random matrix by using the average spectrum radius solving method in the step 3 comprises the following steps:
setting the reactive power optimization random matrix as X as A as Xij},xijIs complex, i.e. matrix X ∈ CN×MIs an N × M-dimensional complex matrix
For an M-dimensional row vector, X is then represented as
Wherein the row vector
The matrix X is standardized and converted into a standard non-Hermitian matrix by rows:
in the formula (I), the compound is shown in the specification,
is the mean of the row vectors of the random matrix,
is the standard deviation of the row vectors of the random matrixMean of transformed row vectors
Standard deviation thereof
Obtaining a standard non-Hermitian matrix after row-by-row conversion
Will be provided with
Conversion into singular value equivalent square matrix Xu∈CN×NIs an N × N-dimensional complex matrix composed of
Is converted into an equivalent square matrix Xu∈CN×NSatisfies the following conditions:
wherein, U is a unitary matrix,
represents XuSatisfy the requirement of
From the single-loop law, for L singular values the equivalent random matrix XuThe matrix product Z ═ ZijExpressed as:
the matrix product Z is processed by row standardization and converted into a standard matrix product
Wherein, σ (z)i) Is the standard deviation of the ith row vector of the matrix Z;
is provided with
Representation matrix
Characteristic root of (1), then average spectral radius κMSRComprises the following steps:
the correlation coefficient of the historical load and the current load in the step 3 is defined as R, which is determined by the following formula:
the correlation coefficient Ri∈ (0,1), wherein R is close to 1, indicating that the historical load data has strong correlation with the current predicted load data, and R is close to 0, indicating that the historical load data has weak correlation with the current predicted load data.
The power distribution management system data comprises power distribution automation system data, scheduling automation system data, power grid meteorological information system data, power quality monitoring management system data, production management system data, geographic information system data, power utilization information acquisition system data, distribution transformer load monitoring system data, load control system data, marketing service management system data, ERP system data, 95598 customer service system data and economic and social data.
Compared with the closest prior art, the technical scheme provided by the invention has the following excellent effects:
the method for determining the reactive power optimization control sequence of the power distribution network based on the big data is used for solving the problem of operation optimization of reactive power of the power distribution network under the condition that the grid structure of the power distribution network is not changed; the method relies on historical data of the power system, provides an optimization control scheme for the current reactive power optimization according to the historical data, can adapt to the development of the intelligent power grid, makes up the defects of the traditional reactive power optimization method and the artificial intelligence method, does not depend on an accurate mathematical model, has high convergence speed, and can relieve the dimension disaster; meanwhile, load correlation is obtained by utilizing the average spectrum radius of the load random matrix, a reactive power optimization sequence is determined according to the correlation, modeling on a system is avoided, a relatively reliable optimization result can be obtained, and the method has the advantages of high convergence rate, high calculation efficiency and the like. Whether the reactive power flow distribution of the power system is reasonable or not directly influences the safety and stability of the power system and is closely related to economic benefits. The voltage of the system is reduced due to insufficient reactive power, the electric equipment cannot be reasonably used, and even a series of accidents such as voltage collapse can be caused; the excess of reactive power also causes the quality of the system voltage to deteriorate, endangers the safety of the system and the equipment, and the excessive reactive power standby wastes unnecessary investment. Reasonable reactive power supply configuration and switching-in and switching-out can effectively reduce the network loss, ensure the voltage quality, prevent the occurrence of accidents or prevent the expansion of the accidents, thereby improving the economical efficiency, the safety and the stability of the operation of the power system.
Drawings
FIG. 1 is a flow chart of a method of an embodiment of the present invention;
FIG. 2 is a diagram of a big data architecture according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to examples.
Example 1:
the invention aims to provide a method for determining a reactive power optimization control sequence of a power distribution network based on big data. The method comprises the steps of utilizing the existing historical load data and the historical optimization data of the power system as data sources, calculating the average spectrum radius of the load random matrix by constructing a power distribution network reactive power optimization random matrix and combining a single-loop law, calculating the correlation coefficient of the historical load and the current load according to the average spectrum radius, obtaining the load day with larger correlation by comparing the correlation coefficient, and using the day reactive power optimization control sequence as the current optimization sequence.
The flow of the method for determining the reactive power optimization control sequence of the power distribution network based on the big data is shown in figure 1, and the method comprises the following steps:
step 1: and acquiring a reactive power optimization data source of the power distribution network by using the existing power distribution management system data.
The intelligent power distribution network has rich data sources, most local cities at present have a plurality of power distribution management systems, including data sources such as a power distribution automation system, a scheduling automation system, a power grid meteorological information system, an electric energy quality monitoring management system, a production management system, a geographic information system, an electricity consumption information acquisition system, a distribution transformer load monitoring system, a load control system, a marketing service management system, an ERP system, a 95598 customer service system and economic and social data, and the overall conditions of the data sources are shown in table 1. The data sources already cover a plurality of management services such as scheduling, operation and inspection, marketing and the like, and cover most of power grid monitoring and information acquisition of 110kV and below multi-voltage levels. The intelligent power distribution and utilization big data application has rich data source types, and covers power distribution and utilization automation, information data, user data, social and economic data and the like of power distribution and transformation, power distribution substations, power distribution switchgears, electric meters, electric energy quality and the like from the data source types.
TABLE 1 typical Power distribution System data Source
| Numbering | Data source system | Data format |
| 1 | Distribution automation | Structured data |
| 2 | Production management system | Structured data |
| 3 | Geographic information system | Semi-structured/unstructured data |
| 4 | Dispatching automation system | Structured data |
| 5 | Power consumption information collection | Structured data |
| 6 | Load control system | Structured data |
| 7 | Load monitoring system | Structured/semi-structured data |
| 8 | Marketing business application system | Structured data |
| 9 | Electric energy quality monitoring system | Structured/semi-structured data |
| 10 | Power grid meteorological information system | Unstructured data |
| 11 | 95598 customer service system | Unstructured data |
| 12 | ERP system | Structured data |
| 13 | Regional socioeconomic data | Structured/unstructured data |
After the data of the power distribution management system is obtained, the existing distributed data source integration, storage and processing technology is adopted to integrate and form the power distribution network reactive power optimization data source according to the big data architecture shown in fig. 2.
Step 2: according to the obtained reactive power optimization data source of the power distribution network, a reactive power optimization random matrix of the power distribution network is constructed by combining a single-loop law, and the method comprises the following steps:
firstly, the random matrix for reactive power optimization control of the power grid is determined to be a random matrix which takes random data of a power system obeying certain distribution as elements. The reactive power optimization element random matrix has the statistical characteristic of the random matrix, obeys the correlation law in the random matrix theory, reflects the structural characteristic of the power grid and contains the operation information of the power grid. The method mainly comprises the steps of selecting a reactive power optimization load random matrix and applying the reactive power optimization load random matrix to reactive power optimization of a power distribution network.
Setting omega phi as a power distribution network reactive power optimization data source set phin,tRepresents the nth bus and the tth bus in a single daySampling data by matrix phiiI-1, 2,3, …, L denotes the reactive power optimization data source matrix on day i, { ΦiAnd expressing a reactive power optimization data source matrix sequence. The number of the power grid buses is N, the sampling frequency of data in a single day is T, and then the reactive power optimization data source matrix can be expressed as follows:
thus, an active load data source matrix P can be obtainediAnd a reactive power data source matrix QiThen the load data source matrix can be represented as
In order to enable the matrix to reflect the spatial distribution characteristics of the load data and the time sequence distribution characteristics of the load data, a cumulative load function is (t) is defined
Wherein s (tau) represents the load at the time of tau, is (T) represents the accumulated load at T hours, and when the single-day sampling frequency of the load data is T, the accumulated load function can be approximately discretized and represented as
Where T is 1,2,3, …, T, sn,kThe nth bus, the kth sample load is indicated. For load data source matrix Si＝{snt}2N×TMatrix capable of calculating accumulative load data source
The sequence for the historical cumulative load matrix is ShiFourthly, a reactive power optimization load random matrix sequence { A } can be defined by predicting the accumulative load matrix Sd currentlyiIn which A isiFor 2N × 2T dimension augmentation matrix
Ai＝[Shi,Sd],i＝1,2,3,…,L (6)
Then the matrix A is augmentediNamely the reactive power optimization random matrix of the power distribution network.
And step 3: and performing equal singular value equivalent transformation on the obtained reactive power optimization random matrix of the power distribution network, and solving the average spectrum radius of the reactive power optimization random matrix by using an average spectrum radius solving method. And solving the correlation coefficient of the historical load and the current load according to the average spectrum radius.
(1) Method for calculating average spectrum radius
Setting the reactive power optimization random matrix as X as A as Xij},xijIs complex, i.e. matrix X ∈ CN×MFor N × M-dimensional complex matrix
For an M-dimensional row vector, X can be represented as
Wherein the row vector
Firstly, the matrix X is standardized and converted into a standard non-Hermitian matrix according to rows
In the formula
Is the mean of the row vectors of the random matrix,
is the standard deviation of the random matrix row vector and the mean value of the transformed row vector
Standard deviation of
Obtaining a standard non-Hermitian matrix after row-by-row conversion
To obtain the eigenvalues of the random matrix, the method needs to be implemented
Conversion into singular value equivalent square matrix Xu∈CN×NFrom
Is converted into an equivalent square matrix Xu∈CN×NSatisfy the requirement of
Wherein U is a unitary matrix, and U is a unitary matrix,
represents XuSatisfy the requirement of
From the single-loop law, for L singular values the equivalent random matrix XuThe matrix product Z ═ ZijCan be expressed as
Taking K as 1, standardizing the matrix product Z according to rows, and converting the matrix product Z into a standard matrix product
The transformation is as follows
Where σ (z)i) Is the ith row vector of the matrix ZStandard deviation of (2).
Representation matrix
The mean spectral radius can then be expressed as
(2) Method for calculating correlation coefficient
For the solved reactive power optimization load random matrix sequence { AiInstruction of
Xi＝Ai,i＝1,2,3,…,L (12)
From a random matrix sequence { XiSolving a singular value equivalent matrix sequence { X ] of the random matrixui}，XuiAnd
having the same covariance matrix, matrix XuiThe distribution of the characteristic values of (A) can reflect the matrix AiEigenvalue distribution of the covariance matrix.
The obtained average spectral radius sequence { kappa }MSRiThe characteristic root distribution characteristics of the covariance matrix are reflected. According to the monocyclic law, characteristic roots
Distributed in a unit circle of the complex plane, then have
Then there is an average spectral radius kMSRi∈(0,1]. When k isMSRiWhen the correlation value approaches 1, the eigenvalue of the covariance matrix is uniformly distributed along the direction of each characteristic root, so that the covariance matrix is uniformly distributed in spectrum, and the correlation of the original augmented matrix is weaker; when k isMSRiWhen the variance matrix approaches 0, the eigenvalues of the covariance matrix are distributed and concentrated along a certain characteristic root direction, the covariance matrix spectrum distribution is concentrated, and the original augmented matrix isThe correlation is strong. Defining a correlation coefficient R, wherein R is defined as
The correlation coefficient Ri∈ (0, 1). when R approaches 1, it indicates that the historical load data has a strong correlation with the current predicted load data, and when R approaches 0, it indicates that the historical load data has a weak correlation with the current predicted load data.
And 4, step 4: and 3, obtaining a correlation coefficient of the historical load and the current load according to the step 3, comparing the magnitude of the correlation coefficient, selecting the date with the stronger correlation coefficient, and taking the reactive power optimization sequence of the current day as the optimization result of the reactive power optimization big data method of the power distribution network.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and those skilled in the art should understand that although the above embodiments are referred to: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be accorded the scope of the claims appended hereto.
Claims (6)
1. A method for determining a reactive power optimization control sequence of a power distribution network based on big data is characterized by comprising the following steps:
step 1: acquiring a reactive power optimization data source of the power distribution network;
step 2: constructing a power distribution network reactive power optimization random matrix by combining a single-loop law according to the power distribution network reactive power optimization data source;
and step 3: the power distribution network reactive power optimization random matrix is subjected to equal singular value equivalence transformation, and the average spectrum radius of the reactive power optimization random matrix is obtained by using an average spectrum radius obtaining method; calculating a correlation coefficient between the historical load and the current load according to the average spectrum radius;
and 4, step 4: and selecting the date with stronger correlation coefficient according to the correlation coefficient of the historical load and the current load, and taking the reactive power optimization sequence on the date as an optimization result.
2. The method for determining the reactive power optimization control sequence of the power distribution network based on the big data as claimed in claim 1, wherein: the method comprises the following steps that 1, a power distribution network reactive power optimization data source is obtained according to power distribution management system data; and integrating, storing and processing the data of the power distribution management system through a distributed data source to form a power distribution network reactive power optimization data source.
3. The big data-based power distribution network reactive power optimization control sequence determination method according to claim 1 or 2, wherein: the process of constructing the distribution network reactive power optimization random matrix by combining the single-loop law according to the distribution network reactive power optimization data source in the step 2 comprises the following steps:
setting omega phi as a power distribution network reactive power optimization data source set phin,tRepresenting the nth bus, the t-th sampling data in a single day by a matrix phiiI 1,2,3, L denotes the reactive power optimization data source matrix on day i, { Φ ·iExpressing a reactive power optimization data source matrix sequence; the number of the power grid buses is N, the sampling frequency of data in a single day is T, and the reactive power optimization data source matrix is expressed as follows:
thus, an active load data source matrix P is obtainediAnd a reactive power data source matrix QiThen the load data source matrix is represented as:
defining a cumulative load function is (t):
wherein s (τ) represents the load at τ, is (T) represents the cumulative load at T hours, and when the single-day sampling frequency of the load data is T, the approximate discretization of the cumulative load function is represented as:
wherein T is 1,2,3, T, sn,kRepresenting the nth bus and the kth sampling load; for load data source matrix Si＝{snt}2N×TCalculating an accumulated load data source matrix:
let the sequence for the history cumulative load matrix be { ShiAnd (6) defining a reactive power optimization load random matrix sequence (A) when the accumulative load matrix Sd is predicted currentlyiIn which A isiIs a 2N × 2T dimension augmentation matrix:
Ai＝[Shi,Sd],i＝1,2,3,···,L
then the matrix A is augmentediNamely the reactive power optimization random matrix of the power distribution network.
4. The method for determining the reactive power optimization control sequence of the power distribution network based on the big data as claimed in claim 1, wherein: the process of solving the average spectrum radius of the reactive power optimization random matrix by using the average spectrum radius solving method in the step 3 comprises the following steps:
setting the reactive power optimization random matrix as X as A as Xij},xijIs complex, i.e. matrix X ∈ CN×MIs an N × M-dimensional complex matrix, and is set as an M-dimensional row vector, then X is expressed as the row vector
The matrix X is standardized and converted into a standard non-Hermitian matrix by rows:
in the formula, the mean value of the row vectors of the random matrix is the standard deviation of the row vectors of the random matrix, and the standard deviation of the mean value of the converted row vectors is converted according to rows to obtain the standard non-Hermitian matrix
Will be provided with
Conversion into singular value equivalent square matrix Xu∈CN×NIs an N × N-dimensional complex matrix composed of
Is converted into an equivalent square matrix Xu∈CN×NSatisfies the following conditions:
wherein, U is a unitary matrix,
represents XuSatisfy the requirement of
From the single-loop law, for L singular values the equivalent random matrix XuThe matrix product Z ═ ZijExpressed as:
the matrix product Z is processed by row standardization and converted into a standard matrix product
Wherein, σ (z)i) Is the standard deviation of the ith row vector of the matrix Z;
is provided with
Representation matrix
Characteristic root of (1), then average spectral radius κMSRComprises the following steps:
5. the method for determining the reactive power optimization control sequence of the power distribution network based on the big data as claimed in claim 4, wherein: the correlation coefficient of the historical load and the current load in the step 3 is defined as R, which is determined by the following formula:
the correlation coefficient Ri∈ (0,1), wherein R is close to 1, indicating that the historical load data has strong correlation with the current predicted load data, and R is close to 0, indicating that the historical load data has weak correlation with the current predicted load data.
6. The method for determining the reactive power optimization control sequence of the power distribution network based on the big data as claimed in claim 2, wherein: the power distribution management system data comprises power distribution automation system data, scheduling automation system data, power grid meteorological information system data, power quality monitoring management system data, production management system data, geographic information system data, power utilization information acquisition system data, distribution transformer load monitoring system data, load control system data, marketing service management system data, ERP system data, 95598 customer service system data and economic and social data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610109681.4A CN107134790B (en) | 2016-02-26 | 2016-02-26 | Power distribution network reactive power optimization control sequence determination method based on big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610109681.4A CN107134790B (en) | 2016-02-26 | 2016-02-26 | Power distribution network reactive power optimization control sequence determination method based on big data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107134790A CN107134790A (en) | 2017-09-05 |
| CN107134790B true CN107134790B (en) | 2020-07-03 |
Family
ID=59721908
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610109681.4A Active CN107134790B (en) | 2016-02-26 | 2016-02-26 | Power distribution network reactive power optimization control sequence determination method based on big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107134790B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109546659B (en) * | 2018-10-19 | 2022-04-19 | 北京交通大学 | Power distribution network reactive power optimization method based on random matrix and intelligent scene matching |
| CN109586309B (en) * | 2018-10-19 | 2022-04-26 | 北京交通大学 | Power distribution network reactive power optimization method based on big data free entropy theory and scene matching |
| CN111860617A (en) * | 2020-07-01 | 2020-10-30 | 国家电网有限公司 | Comprehensive optimization operation method for power distribution network |
| CN114677019B (en) * | 2022-03-30 | 2024-06-28 | 山东大学 | Power utilization characteristic identification method and system based on random source matrix and analytic hierarchy process |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101930213A (en) * | 2009-10-22 | 2010-12-29 | 上海交通大学 | Aggregation predicted control system based on model predicted control and control method thereof |
| CN103401248A (en) * | 2013-07-17 | 2013-11-20 | 华南理工大学 | Random reactive optimization method for power distribution network including wind power plant |
-
2016
- 2016-02-26 CN CN201610109681.4A patent/CN107134790B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101930213A (en) * | 2009-10-22 | 2010-12-29 | 上海交通大学 | Aggregation predicted control system based on model predicted control and control method thereof |
| CN103401248A (en) * | 2013-07-17 | 2013-11-20 | 华南理工大学 | Random reactive optimization method for power distribution network including wind power plant |
Non-Patent Citations (2)
| Title |
|---|
| Xing He;Qian Ai;Robert Caiming Qiu;Wentao Huang;Longjian.A Big Data Architecture Design for Smart Grids Based on Random Matrix Theory.《 IEEE Transactions on Smart Grid》.2015, * |
| Xinyi Xu;Xing He;Qian Ai;Robert Caiming Qiu.A Correlation Analysis Method for Power Systems Based on Random Matrix Theory.《 IEEE Transactions on Smart Grid》.2015, * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107134790A (en) | 2017-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Research and application of a hybrid forecasting framework based on multi-objective optimization for electrical power system | |
| Bessa et al. | Spatial-temporal solar power forecasting for smart grids | |
| CN107134790B (en) | Power distribution network reactive power optimization control sequence determination method based on big data | |
| CN101888087A (en) | Method for realizing distributed super-short-term area load forecasting in distribution network terminal | |
| CN106503851A (en) | A kind of improved Short-Term Load Forecasting Method based on wavelet analysises | |
| Xiao et al. | Short-term power load interval forecasting based on nonparametric Bootstrap errors sampling | |
| Cai et al. | Gray wolf optimization-based wind power load mid-long term forecasting algorithm | |
| Stephen et al. | Non-Gaussian residual based short term load forecast adjustment for distribution feeders | |
| CN116151464A (en) | Photovoltaic power generation power prediction method, system and storable medium | |
| Peng et al. | Multi-objective robust optimization allocation for energy storage using a novel confidence gap decision method | |
| Deng et al. | Probabilistic forecasting and simulation of electricity markets via online dictionary learning | |
| Bo et al. | The forecasting model of reactive power based on SVM | |
| Wang et al. | Short-term wind power prediction based on DBSCAN clustering and support vector machine regression | |
| Kader et al. | Active power control strategy for wind farms based on power prediction errors distribution considering regional data | |
| Lodetti et al. | MAE and RMSE analysis of K-means predictive algorithm for photovoltaic generation | |
| Rouwhorst et al. | Improving clustering-based forecasting of aggregated distribution transformer loadings with gradient boosting and feature selection | |
| Ananthu et al. | A study of 100kWp PV Plant Output Power Forecasting: A case study | |
| CN116307111A (en) | Reactive load prediction method based on K-means clustering and random forest algorithm | |
| CN112348235B (en) | Wind-solar bus load self-adaptive prediction method, device and computer equipment | |
| Zhou et al. | The analysis and prediction of power consumption in industry based on machine learning | |
| CN108009668B (en) | Large-scale load adjustment prediction method applying machine learning | |
| Liu et al. | SFCL: electricity consumption forecasting of CNN-LSTM based on similar filter | |
| He et al. | A hybrid prediction interval model for short-term electric load forecast using Holt-Winters and Gate Recurrent Unit | |
| Zhang et al. | Deep learning network for energy storage scheduling in power market environment short-term load forecasting model | |
| Liu et al. | Application of Optimization Techniques in the Design and Operation of Microgrid |
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 |