CN114243690A - Power grid active safety correction method and device, electronic equipment and storage medium - Google Patents
Power grid active safety correction method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN114243690A CN114243690A CN202111534432.7A CN202111534432A CN114243690A CN 114243690 A CN114243690 A CN 114243690A CN 202111534432 A CN202111534432 A CN 202111534432A CN 114243690 A CN114243690 A CN 114243690A
- Authority
- CN
- China
- Prior art keywords
- power grid
- target
- sample
- overload
- generator set
- 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.)
- Pending
Links
- 238000012937 correction Methods 0.000 title claims abstract description 153
- 238000000034 method Methods 0.000 title claims abstract description 85
- 238000003860 storage Methods 0.000 title claims abstract description 17
- 230000035945 sensitivity Effects 0.000 claims abstract description 52
- 238000009826 distribution Methods 0.000 claims abstract description 21
- 230000000694 effects Effects 0.000 claims abstract description 16
- 230000008030 elimination Effects 0.000 claims abstract description 16
- 238000003379 elimination reaction Methods 0.000 claims abstract description 16
- 238000012549 training Methods 0.000 claims description 34
- 230000001105 regulatory effect Effects 0.000 claims description 12
- 238000011176 pooling Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000005728 strengthening Methods 0.000 abstract description 14
- 230000006870 function Effects 0.000 description 30
- 238000005457 optimization Methods 0.000 description 16
- 238000004422 calculation algorithm Methods 0.000 description 11
- 239000013598 vector Substances 0.000 description 11
- 239000011159 matrix material Substances 0.000 description 10
- 230000008569 process Effects 0.000 description 10
- 230000009471 action Effects 0.000 description 8
- 230000002787 reinforcement Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 150000001875 compounds Chemical class 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000010248 power generation Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- QBPFLULOKWLNNW-UHFFFAOYSA-N chrysazin Chemical compound O=C1C2=CC=CC(O)=C2C(=O)C2=C1C=CC=C2O QBPFLULOKWLNNW-UHFFFAOYSA-N 0.000 description 1
- 230000003750 conditioning effect Effects 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000013144 data compression Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000010252 digital analysis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000013401 experimental design Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000010801 machine learning Methods 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
- 239000002243 precursor Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000012795 verification 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
-
- 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
- H02J3/002—Flicker reduction, e.g. compensation of flicker introduced by non-linear load
-
- 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
- H02J3/24—Arrangements for preventing or reducing oscillations of 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
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
- H02J3/466—Scheduling the operation of the generators, e.g. connecting or disconnecting generators to meet a given demand
-
- 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
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
- H02J3/48—Controlling the sharing of the in-phase component
-
- 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
- H02J2203/10—Power transmission or distribution systems management focussing at grid-level, e.g. load flow analysis, node profile computation, meshed network optimisation, active network management or spinning reserve management
-
- 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]
Abstract
The application provides a power grid active safety correction method and device, electronic equipment and a storage medium. The power grid active safety correction method comprises the following steps: when the target line is overloaded, inputting the state information in the power grid into a trained power grid active safety correction strengthening model, and determining a target generator set needing output regulation in the power grid; and calculating the target output adjustment quantity of the target generator set based on the target sensitivity of the overloaded target line to the target generator set. When the multi-line overload is processed, the output of the generator set can be adjusted according to the calculated target output adjustment quantity of the generator set on the basis of the current running state information of the power grid and the target sensitivity of the overload line to the generator set, the line overload is eliminated efficiently, and the elimination effect of the current power flow distribution condition of the power grid and the output adjustment quantity on the line overload is comprehensively considered.
Description
Technical Field
The present application relates to the field of power grid optimization technologies, and in particular, to a method and an apparatus for correcting active power safety of a power grid, an electronic device, and a storage medium.
Background
In the operation of an electric power system, the control of the power flow of a line within a reasonable threshold value is a basic requirement for ensuring the safe and stable operation of the system, under the general condition, the frequency deviation caused by the fluctuation of the system load and the power fluctuation of a connecting line between systems can be adjusted by automatic power generation control, and the overload of the line caused by the change of the system operation mode caused by factors such as maintenance plans, line faults and the like needs to be rearranged by the output of a unit to eliminate the overload condition of the line so as to ensure the safe operation of the system, and the active safety correction is mainly realized by rearranging the output of a generator to redistribute the power flow so as to achieve the purposes of reducing the frequency deviation of the system and eliminating the out-of-limit line power (namely the overload phenomenon of the line).
The method for correcting the active safety of the power grid in the prior art can be divided into two categories, namely an optimization method and a traditional sensitivity algorithm, wherein the optimization method is to construct an optimization model by establishing an objective function and determining safety constraints, and the active safety correction of the power grid is realized by solving mathematical programming, but when the optimization method is used for correcting the active safety, adjustment units involved in adjustment in power grid adjustment are too many, and the traditional sensitivity algorithm is to transfer the load of an overload circuit to other circuits so as to realize the purpose of eliminating the overload.
Disclosure of Invention
In view of the above, an object of the present application is to provide a method, an apparatus, an electronic device, and a storage medium for correcting grid active power safety of a power grid, where when handling multi-line overload, the present application may adjust output of a generator set based on current operating state information of the power grid and target sensitivity of an overloaded line to the generator set, and a target output adjustment amount of the generator set obtained by calculation is used as a basis, so that line overload is efficiently eliminated, and a current power flow distribution condition of the power grid and an effect of the output adjustment amount on line overload are comprehensively considered.
The embodiment of the application provides a power grid active safety correction method and device of a power grid, electronic equipment and a storage medium, wherein the power grid active safety correction method comprises the following steps:
when detecting that at least one target overload circuit exists in the power grid, inputting the current operation state information of the power grid into a trained target power grid active safety correction model, and determining a target generator set which needs output regulation in the power grid; the target power grid active safety correction model is obtained by training an initial power grid active safety correction model based on initial operation state information of a power grid with a sample overload line, sample output adjustment quantity of a sample generator set for output adjustment in the power grid and an overload elimination effect of the adjusted sample overload line;
calculating a target output adjustment quantity of the target generator set based on the target sensitivity of the target overload line to the target generator set;
and adjusting the output of the target generator set according to the target output adjustment quantity.
Further, a trained target power grid active power safety correction model is obtained through the following method:
inputting initial operation state information of a power grid with a sample overload line into an initial power grid active safety correction model, and determining a sample generator set needing output regulation in the power grid;
calculating the sample output adjustment quantity of the sample generator set according to the sensitivity of the sample overload line to the sample generator set;
determining adjusted running state information in the power grid with the sample overload line according to the sample output adjustment quantity;
updating network parameters of the initial power grid active safety correction model according to the adjusted running state information, and judging whether the overload in the adjusted sample overload line in the power grid is completely eliminated or not;
and if all the active power safety correction models are eliminated, finishing the training of the initial power grid active power safety correction model to obtain the target power grid active power safety correction model.
Further, after the network parameters of the initial power grid active safety correction model are updated according to the adjusted running state information and whether the overload in the adjusted sample overload line in the power grid is completely eliminated is judged, the power grid installation correction method includes:
and if not, continuing to train the initial power grid active safety correction model until the overload in the adjusted sample overload line in the power grid is completely eliminated.
Further, the target power grid active safety correction model includes a convolution layer, a pooling layer, and a full connection layer, and the method includes the steps of inputting the current operation state information of the power grid into the trained target power grid active safety correction model, and determining a target generator set in the power grid, which needs to perform output regulation, including:
inputting the running state information of the power grid into the convolution layer, and determining the initial running state characteristic corresponding to the current running state information;
inputting the initial running state features into the pooling layer, and determining important running state features corresponding to the current running state information;
and inputting the important operating state characteristics into the full-connection layer for weighting, and determining a target generator set needing output regulation.
Further, the adjusting output of the target generator set according to the target output adjustment amount includes:
based on the target output adjustment quantity, output adjustment is carried out on the target generator set, and target power flow distribution information of the power grid is determined;
judging whether the overload in the adjusted overload line in the power grid is completely eliminated or not according to the target power flow distribution information;
and if all the output power is eliminated, determining that the output power adjustment is effective.
Further, after the determining, according to the target power flow distribution information, whether the overload in the adjusted overload line in the power grid is completely eliminated, the power grid active safety correction method further includes:
and if not, determining that the output regulation is invalid, and regulating the load in the power grid of the target overload circuit.
The embodiment of the present application further provides an active safety correction device of electric wire netting, active safety correction device includes:
the determining module is used for inputting the current operation state information of the power grid into a trained target power grid active safety correction model when at least one target overload circuit is detected to exist in the power grid, and determining a target generator set which needs to perform output regulation in the power grid; the target power grid active safety correction model is obtained by training an initial power grid active safety correction model based on initial operation state information of a power grid with a sample overload line, sample output adjustment quantity of a sample generator set for output adjustment in the power grid and an overload elimination effect of the adjusted sample overload line;
the calculation module is used for calculating a target output adjustment quantity of the target generator set based on the target sensitivity of the target overload circuit to the target generator set;
and the adjusting module is used for adjusting the output of the target generator set according to the target output adjustment quantity.
Further, a trained target power grid active power safety correction model is obtained through the following method:
inputting initial operation state information of a power grid with a sample overload line into an initial power grid active safety correction model, and determining a sample generator set needing output regulation in the power grid;
calculating the sample output adjustment quantity of the sample generator set according to the sensitivity of the sample overload line to the sample generator set;
determining adjusted running state information in the power grid with the sample overload line according to the sample output adjustment quantity;
updating network parameters of the initial power grid active safety correction model according to the adjusted running state information, and judging whether the overload in the adjusted sample overload line in the power grid is completely eliminated or not;
and if all the active power safety correction models are eliminated, finishing the training of the initial power grid active power safety correction model to obtain the target power grid active power safety correction model.
An embodiment of the present application further provides an electronic device, including: the system comprises a processor, a memory and a bus, wherein the memory stores machine readable instructions executable by the processor, when an electronic device runs, the processor and the memory are communicated through the bus, and the machine readable instructions are executed by the processor to execute the steps of the power grid active safety correction method.
The embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the power grid active safety correction method as described above are performed.
Compared with the prior art, the method, the device, the electronic equipment and the storage medium for correcting the active power safety of the power grid can be used for adjusting the output of the generator set based on the current running state information of the power grid and the target sensitivity of an overload circuit to the generator set when the multi-circuit overload is processed, the circuit overload is efficiently eliminated, the effect of eliminating the circuit overload by the current power flow distribution condition of the current power grid and the output adjustment quantity is comprehensively considered, and the phenomenon that other branch circuits are overloaded when the overload circuit is eliminated is avoided.
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 shows a flowchart of a method for correcting grid active safety of a power grid according to an embodiment of the present application;
fig. 2 shows a flowchart of another grid active safety correction method for a power grid according to an embodiment of the present application;
fig. 3 shows a flowchart of training of a trained power grid active safety correction strengthening model in another power grid active safety correction method for a power grid provided in an embodiment of the present application;
fig. 4 shows a schematic structural diagram of a node system for comparing a grid active safety correction method in a grid active safety correction method of a power grid provided in an embodiment of the present application;
fig. 5 is a schematic structural diagram illustrating an active safety correction device of a power grid according to an embodiment of the present application;
fig. 6 shows a schematic structural diagram of an electronic device provided in an embodiment of the present application.
In the figure, 500-active safety correction device; 510-a determination module; 520-a calculation module; 530-a conditioning module; 600-an electronic device; 610-a processor; 620-memory; 630-bus.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. The components of the embodiments of the present application, as generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations and, thus, the following detailed description of the embodiments of the present application, which is provided in the figures, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. Every other embodiment that can be obtained by a person skilled in the art without making creative efforts based on the embodiments of the present application falls within the protection scope of the present application.
Firstly, researches show that the power grid active safety correction method in the prior art can be divided into two categories, namely an optimization method and a traditional sensitivity algorithm, wherein the optimization method is to construct an optimization model by establishing an objective function and determining safety constraints, and utilize the solution of mathematical programming to realize the active safety correction of the power grid, but when the optimization method is used for performing the active safety correction, and when the optimization method is used for performing the active safety correction, too many adjusting units related to adjustment in power grid adjustment are caused, discomfort is caused, and a plurality of circuits with lower load rates still exist in the power grid after the adjustment is applied in an actual system, and the traditional sensitivity algorithm is to transfer the load of an overload circuit to other circuits, so that the purpose of eliminating the overload is realized.
Based on this, the embodiment of the application provides a method and a device for correcting the active power safety of a power grid of the power grid, an electronic device and a storage medium, when multi-line overload is processed, the output of a generator set can be adjusted according to the calculated target output adjustment quantity of the generator set based on the current running state information of the power grid and the target sensitivity of an overload line to the generator set, the line overload is eliminated efficiently, and the elimination effect of the current power flow distribution condition of the power grid and the output adjustment quantity on the line overload is comprehensively considered.
Referring to fig. 1, fig. 1 is a flowchart of a method for correcting the active power safety of a power grid according to an embodiment of the present application. As shown in fig. 1, the method for correcting the active safety of the power grid provided in the embodiment of the present application includes:
s101, when at least one target overload circuit is detected in the power grid, inputting the current operation state information of the power grid into a trained target power grid active safety correction model, and determining a target generator set which needs to perform output regulation in the power grid; the target power grid active safety correction model is obtained by training an initial power grid active safety correction model based on initial operation state information of a power grid with a sample overload line, sample output adjustment quantity of a sample generator set for output adjustment in the power grid and an overload elimination effect of the adjusted sample overload line.
In the step, in the process of power grid operation, when the power grid operation mode changes due to factors such as maintenance plans and line faults, and further a target overload line appears, the power grid including the overloaded target overload line and the operation state information of the current operation state of the power grid represented by other lines which are not overloaded are input into a trained power grid active safety correction strengthening model, so that the trained power grid active safety correction strengthening model outputs a target generator set which can determine output regulation required in the power grid, the operation state information comprises line load information, power grid flow distribution information, load information of each node and output regulation capacity of the generator set, therefore, the operation state characteristics of the trained power grid active safety correction strengthening model not only take the line overload information in the target overload line in the current power grid into account, and factors such as power flow distribution information of the power grid, output adjusting capacity of the generator set, load information of other lines in the current power grid and the like are also considered.
The trained power grid active safety correction strengthening model is obtained by training an initial power grid active safety correction model, after state information in a power grid is input into the trained power grid active safety correction strengthening model, firstly, the operation state features of the operation state information are extracted through a convolution network in the trained power grid active safety correction strengthening model, then, the operation state features are subjected to strengthening learning based on a strengthening learning network in the trained power grid active safety correction strengthening model, and a target generator set needing output regulation is determined.
Furthermore, the trained power grid active safety correction strengthening model can set different objective functions according to specific application environments, and the embodiment provided by the application takes the minimum power adjustment amount of a target generator set as an objective function; wherein the formula of the objective function is:
wherein, Δ PkAnd representing the power adjustment amount of the kth unit.
Here, the objective function must satisfy the following constraint:
firstly, under the condition of not counting network loss, the output of the sample generator set and the node load of the power grid where the sample overload line is located should satisfy the following formula:
∑i∈NPi＝∑j∈LLj；
in the formula, PiIs the active power of the sample generator set i, LjIs the active load of node j.
Secondly, the sample output adjustment quantity of the sample generator set needing output adjustment should meet the following inequality constraints:
in the formula, Pi maxThe output upper limit of the i node of the sample generator set is obtained; pi minThe output upper limit of the i node of the sample generator set is obtained; piThe active output of the i node of the sample generator set is obtained.
Thirdly, the flow of the sample line in the adjusted sample line should satisfy the following inequality constraint:
in the formula (I), the compound is shown in the specification,
is the maximum power flow limit value in the sample line L;
is the minimum tidal current limit in the sample line L; pLIs the actual active power in the sample line L.
In the first, second and third constraint conditions, the equality constraint means that at any time, the power generated by the sample generator set is equal to the power consumed by the load; the inequality constraint indicates that the relevant equipment within the grid is within safe operating limits under qualified power quality conditions.
The initial power grid active safety correction model adopted in the embodiment provided by the application is a DQN algorithm model, and the application adopts a neural network represented by a convolutional network in the DQN algorithm to approximate a value function in a reinforcement learning model, wherein the value function is expressed as: qπ(s, a), the approximate equation of the convolution network approximating to the value function in the initial power grid active safety correction model is Q (s, a; theta) approximately equal to Qπ(s,a)。
Wherein α is a learning rate; s is a state matrix represented by the operating state feature, and in the embodiment provided by the present application, the operating state matrix s represented by the operating state feature is the following matrix:
in the formula: when i ═ j, sijThe load is a load carried by the node i, wherein the output power to the power grid is a positive value, and the power requested to the power grid is a negative value; when i ≠ j, sijCompared with the traditional sensitivity algorithm which only considers the influence of the generator set on the overload line, the operation state matrix provided by the embodiment not only considers the influence of the generator set on the overload line, but also considers the influence of factors such as line load, power flow distribution information, a topological structure of a power grid, node load rate and the like on the sample output regulating quantity of the sample generator set, so that on the basis of eliminating a target overload line in the power grid, the stability condition of adjusting the whole power grid is realized, and the problems of repeated adjustment and low efficiency existing in the traditional sensitivity algorithm are solved.
The Q learning algorithm is used for discretizing the actual state and the action of the regulating unit in the determined power grid when reinforcement learning in the initial power grid active safety correction model is applied to determining the actual problem of the regulating unit in the power grid, and the higher discretization granularity weakens the generalization capability of the Q learning algorithm, while the lower granularity brings about a dimension disaster, wherein the dimension disaster is a phenomenon that in the problem related to calculation of vectors or matrixes, the calculated amount is exponentially multiplied along with the increase of the dimension. Dimensional disaster relates to the fields of digital analysis, sampling, combination, machine learning, data mining, and databases.
The initial power grid active safety correction model utilizes strong fitting capacity of a convolutional network to construct a corresponding relation between the state of a regulating unit and the action of the regulating unit in the power grid in an actual application scene, directly represents a value function of the state and the action, and expands the application range of reinforcement learning to the continuous and high-dimensional actual problem.
The iterative formula of the state and action value function in the Q learning algorithm is as follows:
Qt+1(st,at)＝Qt(st,at)+αδt；
wherein Q ist+1A value function for t +1 iterations; qtAs a function of the value of the t-th iteration; alpha is the learning rate; stFor the state of the t-th iteration, atAn act of a t-th iteration; deltatIs the time difference error of the t-th cycle.
Here, δtThe expression for the time difference error represented is:
δt＝rt+1+γQt(st+1,a′)-Qt(st,at)；
where a is the state st+1An action to be performed; stThe state of the t iteration; st+1The state of the t +1 th iteration; r ist+1The reward value is an optimized parameter for optimizing the parameters of the power grid active safety correction and enhancement model according to the sample output adjustment quantity, the adjustment cost of each sample generator and the line load rate; a istAn act of a t-th iteration; gamma is the attenuation factor.
Further, a trained target power grid active power safety correction model is obtained through the following method:
inputting the initial operation state information of the power grid with the sample overload line into an initial power grid active safety correction model, and determining a sample generator set which needs to perform output regulation in the power grid.
Here, the target power grid active safety correction model includes a convolution layer, a pooling layer, and a full connection layer, and the determining of the target generator set that needs to perform output regulation in the power grid by inputting the current operating state information of the power grid into the trained target power grid active safety correction model includes:
and inputting the running state information of the power grid into the convolution layer, and determining the initial running state characteristic corresponding to the current running state information.
The operator can manually set the initial operation states of various power grids with at least one target overload line through the matpower.
Inputting the initial running state features into the pooling layer, determining important running state features corresponding to the current running state information, wherein the pooling layer is used for performing data compression on the initial running state features in the convolutional layer and extracting the important running state features in the initial running state features.
And inputting the important operation state features into the full-connection layer for weighting, and determining a target generator set needing output adjustment, wherein the full-connection layer weights the initial operation state features and the important operation state features in the convolution layer and the pooling layer, and maps the feature space transformation to a sample mark space.
The convolutional layer is the core of a target power grid active safety correction model and has the functions of realizing the characteristic extraction of initial running state information of input running state information, the pooling layer is used for compressing and extracting important running state characteristics of data and parameters, the fully-connected layer weights the convolutional layer and the pooling layer characteristics, the characteristic space is transformed and mapped to a sample mark space, and a target generator set needing output regulation is determined.
The output variation of the sample generator set which needs to perform output adjustment in the power grid is different, and then the influence on the safety and the stability of the power grid is different, so that the sample generator set which needs to perform output adjustment in the power grid can ensure that the adjustment amount of the sample generator set is minimum under the condition of determining to eliminate the overload elimination precursor in the sample line, the sample generator set which needs to perform output adjustment in the initial power grid active safety correction model provided in the embodiment is a, and the sample generator set which needs to perform output adjustment in order to eliminate the overload line of the power grid sample is a, which can be specifically expressed as:
a＝{a1,a2,…,ai,…,an}；
here, ai1 is expressed as a sample generator set for increasing output; a isi-1 is denoted as sample genset as reduced output sample genset; a isiAnd 0 represents that the sample generator set does not participate in regulating the power grid and the sample overload line.
And calculating the sample output adjustment quantity of the sample generator set according to the sensitivity of the sample overload line to the sample generator set.
After a sample generator set needing output adjustment is determined, a sample output adjustment quantity of the sample generator set is determined based on the sample sensitivity of an overload line to the sample generator set, wherein the sensitivity refers to the sensitivity of the active output of the sample generator set to the active power flow of each branch, and specifically, when the active power flow of the branch slightly changes, the sample output adjustment quantity and the sample generator set needing output adjustment output from a trained target power grid active safety correction model are slightly changed.
Here, the sensitivity index proposed by the present embodiment can be derived from a power flow equation existing in the power grid.
Here, the grid may be represented by n nonlinear network equations under steady state conditions:
f(u,x)＝0；
thus, u represents a column vector of a sample generator set needing output regulation in the power grid; x represents a column vector of a sample output adjustment quantity of the generator set; the equation form varies with the selection of the control variables and the state variables and the change of the coordinate form, and the equation becomes:
f(u0,x0)＝0；
after the operation state of the power grid changes, x and u generate deviation amounts delta x and delta u respectively, and at the moment, the balance equation of the power grid becomes:
f(Δu+u0,Δx+x0)＝0；
and (3) adopting Taylor series expansion at the operating point for the above formula, and omitting high-order terms above the second order:
combining the above formula with the control action set of the sample generator set needing output regulation, the following formula is changed:
here, the above equation is a basic form of the sensitivity equation, and the linear relationship between the controlled variable Δ x and the controlled variable Δ u is determined according to the sensitivity equation:
in the formula (I), the compound is shown in the specification,
s is the matrix vector of sensitivity.
According to the matrix vector of the sensitivity, the sample output adjustment quantity is calculated as follows:
wherein, the delta u is a sample output regulating quantity vector; Δ x is the overload vector of the sample overload line; and a is a sample generator set needing output regulation.
And determining the adjusted running state information in the power grid with the sample overload line according to the sample output adjustment quantity.
And updating the current operation state information in the power grid according to the sample output adjustment amount, and determining the adjusted operation state information in the power grid with the sample overload line, wherein the adjusted operation state information comprises updated network parameters, a target generator set needing output adjustment, an iterated target generator set needing output adjustment and the sample output adjustment amount, and the network parameters are updating functions of the reward values.
And updating the network parameters of the initial power grid active safety correction model according to the adjusted running state information, and judging whether the overload in the adjusted sample overload line in the power grid is completely eliminated.
And judging whether the overload in the sample overload line adjusted in the power grid is completely eliminated or not by using matpower according to the network parameters.
Here, the updating function is a parameter updating function after overload occurs on each sample line in the power grid, which is determined according to the sample output adjustment amount of the sample generator set for output adjustment in the power grid, the economic adjustment cost of each sample generator and the line load rate.
Here, the parameter updating function after each sample line in the power grid is overloaded is a reward function, and an expression of the reward value function is as follows:
wherein, FijAdjusting cost (ten thousand yuan/MW & h) of a sample generator set i needing output adjustment on the jth capacity section; delta PGijAdjusting the active output (MW) of a sample generator set i needing output adjustment on the jth capacity segment; k is a unit set; m is the specified capacity segment number of the economic adjustment cost of the sample generator set needing output adjustment; mu.siThe load rate of the ith sample overload line is; n is the number of system lines; t is the iteration times in each round of learning; k is a radical of1、k2For sample generatorsA power generation processing cost coefficient and an overall load factor.
Therefore, the power generation processing cost coefficient and the whole load rate coefficient of the sample generator can be set in a user-defined mode according to requirements.
After updating the network parameters, carrying out playback updating on the initial power grid active safety correction model through the adjusted running state information generated in the training process, and further updating the reward value function; the experience information is adjusted running state information, and is composed of updated network parameters, a target generator set needing output adjustment, an iterated target generator set needing output adjustment and sample output adjustment.
In the model training, the used minimum loss function is trained to obtain a target power grid active safety correction model, and the expression of the minimum loss function is as follows:
Lt(θt)＝E(s，a，r，s′)(y-Q(s，a；θt))2；
in the formula, Lt(θt) Is the loss function at t; thetatThe parameter is a parameter of an initial power grid active safety correction model at t; s is a sample generator set which needs to perform output regulation; a is a sample output adjustment amount corresponding to a sample generator set needing output adjustment; e(s，a，r，s′)Is a mathematical expectation; y is an initial power grid active safety correction model network, and an expression of the optimal training target value y is as follows:
in the formula:
parameters for training an initial power grid active safety correction model; gamma is an attenuation factor; s' is the state of the next iteration; a is a sample output adjustment amount corresponding to a sample generator set needing output adjustment in the next iteration; r represents the active safety correction model of the target initial power gridAnd under the condition of the type parameter, executing the accumulated discount reward obtained when the sample output adjustment quantity is executed, wherein the accumulated discount reward is a parameter updating function after each sample line in the power grid is overloaded.
The grid active safety correction strengthening model is based on gradient updating weight, and the weight can be obtained by solving partial derivatives of a minimization loss function:
in the formula:
representing the gradient calculation.
The active safety correction model of the target power grid and the active safety correction model of the initial power grid have the same neural network structure, when the value function in the reinforcement learning model is approximated by a nonlinear function in the training process of the active safety correction model of the initial power grid, the value of the reinforcement learning model is easy to update and is easy to vibrate, and unstable learning behaviors are presented, so the active safety correction model of the target power grid is introduced, the active safety correction model of the target power grid is identical to the active safety correction model of the reinforcement initial power grid in structure, but in the active safety correction model of the target initial power grid, parameters can be independently updated after a certain number of steps, so the Q value of the active safety correction model of the reinforcement power grid is temporarily fixed in the training process, and the learning process is more stable.
And if all the active power safety correction models are eliminated, finishing the training of the initial power grid active power safety correction model to obtain the target power grid active power safety correction model.
If all the training data are eliminated, the training of the current round is finished, and the next round of training is started. And when the relevant parameters of the power grid active safety correction model tend to be stable, finishing training to obtain the target power grid active safety correction model.
In the process of inputting the initial operation state information of the power grid with the sample overload line into the initial power grid active safety correction model for training, one round of training cannot eliminate all overload, the mark for ending one round of training is to further update the network parameters of the initial power grid active safety correction model, and multiple rounds of training are needed to eliminate all loads.
And if not, continuing to train the initial power grid active safety correction model until the overload in the adjusted sample overload line in the power grid is completely eliminated.
And S102, calculating a target output adjustment quantity of the target generator set based on the target sensitivity of the target overload circuit to the target generator set.
In the step, after a target generator set needing output regulation is determined, a target output regulation quantity of the target generator set is determined based on a target sensitivity of an overloaded target overload line to the target generator set, wherein the target sensitivity refers to a sensitivity of node active output to branch active power flow, and refers to a slight change between the target output regulation quantity and the output target generator set needing output regulation when the target generator set is slightly changed.
Here, the target sensitivity index proposed in this embodiment may be derived from a power flow equation, and the power grid under a steady-state condition may be represented by a nonlinear network equation:
f(u，x)＝0；
thus, u represents the column vector of the target generator set in the power grid for which output regulation is required; x represents a column vector of a target output adjustment of the target generator set; the equation form varies with the choice of the control variables and the state variables, and the change of the coordinate form, and under stable operation, the equation becomes:
f(u0，x0)＝0；
when the operation state of the power grid changes, x and u generate deviation amounts Δ x and Δ u, respectively, and at this time, the balance equation of the system becomes:
f(Δu+u0，Δx+x0)＝0；
and (3) adopting Taylor series expansion at the operating point for the above formula, and omitting high-order terms above the second order:
combining the above formula with the control action of the target generator set needing output regulation, the following formula is changed:
here, the above equation is a basic form of a target sensitivity equation, and a linear relationship between a target generator set Δ u requiring output adjustment in the power grid and a target output adjustment amount Δ x of the target generator set is determined according to the target sensitivity equation as follows:
in the formula (I), the compound is shown in the specification,
and S is a target sensitivity matrix.
And according to the sensitivity matrix, calculating a target output adjustment quantity as follows:
wherein, the delta u is a target output regulating quantity vector; Δ x is a target overload line quantity vector; a is the action policy set.
S103, performing output adjustment on the target line in the power grid according to the target output adjustment quantity.
In the step, new power flow distribution information in the power grid is calculated and determined according to the target output adjustment quantity, whether overload in a target overload line in the power grid is completely eliminated or not is judged according to the new power flow distribution information, and whether other safety correction modes are needed to eliminate the overload or not is determined according to the result and elimination degree of the overload elimination in the target overload line.
Thus, the other safety correction means include, but are not limited to, using a means of cutting line load to eliminate overload in the target overload line.
Fig. 3 is a flowchart of training of a trained power grid active power safety correction strengthening model provided in the embodiment of the present application, and the specific flow is as follows:
firstly, an initial power grid active safety correction model is initialized.
Obtaining initial operation state information of the power grid with the sample overload line.
An initial operating condition characteristic is determined.
And determining a sample generator set which needs to be subjected to output regulation in the power grid based on the initial operation state characteristics.
A sample output adjustment is determined.
Determining the adjusted operating state information.
And updating the network parameters of the initial power grid active safety correction model.
It is determined whether the overload in the sample overload line is completely removed.
After all the elimination is determined, whether the training times reach the preset training times is judged.
And if not, judging whether the sample generator set for output regulation in the power grid reaches the upper cycle limit.
And if the cycle upper limit is not reached, re-determining the initial operation state characteristics.
And if the upper limit of the circulation is determined to be reached, judging whether the training times reach the preset training times again.
And determining to acquire the initial operation state information of the power grid with the sample overload line again.
And after the preset training times are determined, updating the network parameters of the initial power grid active power safety correction model.
Compared with the prior art, the method for correcting the active power safety of the power grid can be used for adjusting the output of the generator set based on the current running state information of the power grid and the target sensitivity of an overload circuit to the generator set when the multi-circuit overload is processed, effectively eliminating the circuit overload, comprehensively considering the current power flow distribution condition of the power grid and the effect of the output adjustment quantity on eliminating the circuit overload, and reducing the phenomenon that the overload circuit is eliminated, but the load overload of other branches is aggravated.
Referring to fig. 2, fig. 2 is a flowchart of a method for correcting active safety of a power grid according to another embodiment of the present application. As shown in fig. 2, the method for correcting the active safety of the power grid provided in the embodiment of the present application includes:
s201, when at least one target overload circuit is detected in the power grid, inputting the current operation state information of the power grid into a trained target power grid active safety correction model, and determining a target generator set which needs to perform output regulation in the power grid; the target power grid active safety correction model is obtained by training an initial power grid active safety correction model based on initial operation state information of a power grid with a sample overload line, sample output adjustment quantity of a sample generator set for output adjustment in the power grid and an overload elimination effect of the adjusted sample overload line.
S202, calculating a target output adjustment quantity of the target generator set based on the target sensitivity of the target overload circuit to the target generator set.
And S203, performing output adjustment on the target generator set based on the target output adjustment amount, and determining target tide distribution information of the power grid.
In this step, new target load flow distribution information in the power grid is recalculated based on the target output adjustment amount.
Here, the power grid refers to a power grid after regional division, and specific division rules may be customized according to an application environment, for example: the provincial level, the city level, the county level and the like can be divided according to the region.
And S204, judging whether the overload in the adjusted overload circuit in the power grid is completely eliminated or not according to the target load flow distribution information.
And S205, if all the output power is eliminated, determining that the output power adjustment is effective.
Further, if not all the output power is eliminated, determining that the output power adjustment is invalid, and adjusting the load in the power grid where the target overload circuit is located.
If not, determining that the output regulation is invalid, and performing a regulation mode combining multiple modes on the load in the power grid where the load target overload line is located, or directly using other regulation modes.
Other ways of adjustment include, but are not limited to, adjustment using the load in the target overload line.
The descriptions of S201 to S202 may refer to the descriptions of S101 to S102, and the same technical effects can be achieved, which are not described in detail.
Compared with the prior art, the method for correcting the active power safety of the power grid can be used for adjusting the output of the generator set based on the current running state information of the power grid and the target sensitivity of an overload circuit to the generator set when the multi-circuit overload is processed, effectively eliminating the circuit overload, comprehensively considering the current power flow distribution condition of the power grid and the effect of the output adjustment quantity on eliminating the circuit overload, and reducing the phenomenon that the overload circuit is eliminated, but the load overload of other branches is aggravated.
Taking an IEEE39 node system as an example, the verification result of the power grid active safety correction method provided by the embodiment of the present application is compared with the conventional adjustment strategy, and the validity of the method provided herein is verified.
As shown in fig. 4, in the experimental design, a Pycharm platform is adopted to establish a target power grid active power safety correction model in a tensflo environment, and a power grid target load flow is calculated by using Matpower.
Firstly, initializing an initial power grid active safety correction model, determining initialization parameters, and setting the following table 1 as an initialization parameter table of the initial power grid active safety correction model:
TABLE 1
Here, after the lines 23-24 are disconnected, the lines 21-22, 16-21 are overloaded, with the two lines being overloaded at 135.8367MW and 59.2052MW, respectively. Exceeding the maximum transmission power of 15.09% and 9.87% of the line.
When the method for correcting the active safety of the power grid is adopted for adjustment, the line is corrected in sequence according to the severity of line overload, a state matrix is formed firstly, and the model is input to obtain the number 34 and the number 36 of units participating in adjustment, wherein the number 34 of units is an added output unit, and the number 36 of units is a subtracted output unit. The sensitivity of the lines 21-22 to the two sets of units can be obtained, the variation of the unit output is 146.1346MW, and the overload of the lines 21-22 is eliminated through the adjustment. And updating the current state matrix by the model to obtain the next cycle of the units 32 and 39 participating in the adjustment, wherein the output of the unit 32 is increased by 82.4613MW, the output of the unit 39 is reduced correspondingly, and the overload condition of the line 26-29 is eliminated through the adjustment of the current cycle.
If a traditional sensitivity method is adopted, firstly, the lines 21-22 are adjusted, the number 36 and number 38 machine sets are selected through calculation in the first adjustment, the output 38 is increased by 36 according to a corresponding adjustment strategy to reduce the output, the output variation of the machine set is limited to 60.2732MW due to the limitation of the maximum output of the machine set, the load rate of the lines 21-22 is reduced through the adjustment in the current adjustment, the load rate is reduced from 1.1509 to 1.1027, and the out-of-limit condition is not eliminated; the second round selects the sets 30 and 37, the output adjustment of the sets is 47.0846MW and-47.0846 MW respectively, and the overload of the lines 21-22 is eliminated through the adjustment of the second round. Then, the overload of the line 16-21 is eliminated, firstly, the 32 # machine set and the 39 # machine set are selected according to the corresponding strategies, 76.5642MW is increased and the corresponding output force is reduced respectively, at this time, although the out-of-limit phenomenon of the line 26-29 is eliminated, the line overload out-of-limit of the line 22-35 occurs, the output force of the 30 # machine set is increased by 50.6247MW according to the adjustment strategy, the corresponding output force of the 39 # machine set is reduced, the overload phenomenon of the line 16-21 can be eliminated, and new line overload does not occur. If the optimization method is adopted, the minimum adjustment amount is taken as a target function, and a planning method is adopted to obtain a corresponding adjustment result. The results of the three strategies for adjusting the output and the adjustment cost are shown in tables 2 and 3:
TABLE 2
As can be seen from table 2: the optimization method adjustment strategy relates to 8 generators, the number of the adjusted generators is too large, and the actual scheduling execution difficulty is increased. Although the sensitivity method can reduce the number of the units participating in the adjustment, a new overload circuit appears in the adjustment process, and the overload range is expanded. The method achieves the aim of eliminating overload lines with minimum units.
TABLE 3
As can be seen from table 3: the method provided by the invention brings the adjustment cost into the selection basis of the unit, and compared with the optimization method and the sensitivity method, the adjustment cost is reduced by about 45% and 47% respectively.
In order to further embody the advantages of the method compared with the optimization method and the sensitivity method, the initial load rates of the lines are ranked from high to low and then numbered, and the lines are adjusted by three strategies, in the initial state of the initial power grid active safety correction model, the lines 21-22 and 16-21 are respectively overloaded by 15.09% and 9.87%, the load level of most lines is between 20% and 80%, but the load rates of 13 lines are less than 10%, and the adjustment result can be obtained as follows: the traditional sensitivity method achieves the purpose of eliminating overload by transferring the load of an overload line to the other three heavy load lines in the adjusting process, but after the adjustment is finished, 5 lines with high load rate appear in the system, and when the high load rate lines of the system are more, new overload lines are easy to generate. The adjustment result of the optimization method still has more lines with lower load rate, and the overall condition of the system is not greatly improved.
It can be seen from comparison of the three adjustment strategies that the traditional sensitivity method does not consider the unit adjustment capacity, the traditional sensitivity method selects the unit with the highest sensitivity as the adjustment unit, which may result in insufficient adjustment capacity of the adjustment unit, and may cause new line overload during adjustment, so that the line in an overload state is enlarged, and the units participating in adjustment are excessive in an optimization method, and meanwhile, neither of the two methods takes the adjustment cost into consideration for unit selection, but when the application handles multi-line overload, the target output adjustment quantity of the generator set can be calculated based on the current operation state information of the power grid and the target sensitivity of the overload line to the generator set, and the output adjustment of the generator set is performed based on the target output adjustment quantity, so that the line overload is efficiently eliminated, and the current power flow distribution condition of the power grid and the elimination effect of the output adjustment quantity on the line overload are comprehensively considered, the phenomenon that the power of other branches is overloaded when the overload circuit is eliminated is avoided.
Please refer to fig. 5 and 5, which are schematic structural diagrams of an active safety calibration device of a power grid according to an embodiment of the present disclosure. As shown in fig. 5, the active safety correction device 500 of the power grid includes:
a determining module 510, configured to, when it is detected that at least one target overload line exists in the power grid, input current operation state information of the power grid into a trained target power grid active safety correction model, and determine a target generator set that needs to perform output regulation in the power grid; the target power grid active safety correction model is obtained by training an initial power grid active safety correction model based on initial operation state information of a power grid with a sample overload line, sample output adjustment quantity of a sample generator set for output adjustment in the power grid and an overload elimination effect of the adjusted sample overload line;
a calculating module 520, configured to calculate a target output adjustment of the target generator set based on a target sensitivity of the target overload line to the target generator set;
and an adjusting module 530, configured to adjust the output of the target generator set according to the target output adjustment amount.
Compared with the prior art, the active safety correction device of the power grid provided by the embodiment of the application can be based on the current running state information of the power grid and the target sensitivity of an overload circuit to a generator set when the multi-circuit overload is processed, the calculated target output regulating quantity of the generator set is used as the basis for output regulation of the generator set, the circuit overload is eliminated efficiently, the elimination effect of the current power flow distribution condition of the current power grid and the output regulating quantity on the circuit overload is comprehensively considered, the elimination of the overload circuit is reduced, and the phenomenon of overload of other branch circuits is aggravated.
Referring to fig. 6, fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present disclosure. As shown in fig. 6, the electronic device 600 includes a processor 610, a memory 620, and a bus 630.
The memory 620 stores machine-readable instructions executable by the processor 610, when the electronic device 600 runs, the processor 610 communicates with the memory 620 through the bus 630, and when the machine-readable instructions are executed by the processor 610, the steps of the power grid active power safety correction method in the method embodiments shown in fig. 1 to fig. 5 may be executed.
An embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method for correcting the active power safety of the power grid in the method embodiments shown in fig. 1 to fig. 2 may be executed.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and there may be other divisions when actually implemented, and for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or units through some communication interfaces, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer-readable storage medium executable by a processor. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present application, and are used for illustrating the technical solutions of the present application, but not limiting the same, and the scope of the present application is not limited thereto, and although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope disclosed in the present application; such modifications, changes or substitutions do not depart from the spirit and scope of the exemplary embodiments of the present application, and are intended to be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A power grid active safety correction method is characterized by comprising the following steps:
when detecting that at least one target overload circuit exists in the power grid, inputting the current operation state information of the power grid into a trained target power grid active safety correction model, and determining a target generator set which needs output regulation in the power grid; the target power grid active safety correction model is obtained by training an initial power grid active safety correction model based on initial operation state information of a power grid with a sample overload line, sample output adjustment quantity of a sample generator set for output adjustment in the power grid and an overload elimination effect of the adjusted sample overload line;
calculating a target output adjustment quantity of the target generator set based on the target sensitivity of the target overload line to the target generator set;
and adjusting the output of the target generator set according to the target output adjustment quantity.
2. The power grid active safety correction method according to claim 1, characterized in that the trained target power grid active safety correction model is obtained by:
inputting initial operation state information of a power grid with a sample overload line into an initial power grid active safety correction model, and determining a sample generator set needing output regulation in the power grid;
calculating the sample output adjustment quantity of the sample generator set according to the sensitivity of the sample overload line to the sample generator set;
determining adjusted running state information in the power grid with the sample overload line according to the sample output adjustment quantity;
updating network parameters of the initial power grid active safety correction model according to the adjusted running state information, and judging whether the overload in the adjusted sample overload line in the power grid is completely eliminated or not;
and if all the active power safety correction models are eliminated, finishing the training of the initial power grid active power safety correction model to obtain the target power grid active power safety correction model.
3. The grid active safety correction method according to claim 2, wherein after the updating of the network parameters of the initial grid active safety correction model according to the adjusted operating state information and the determination of whether all the overloads in the adjusted sample overload line in the grid are eliminated, the grid active safety correction method comprises:
and if not, continuing to train the initial power grid active safety correction model until the overload in the adjusted sample overload line in the power grid is completely eliminated.
4. The grid active safety correction method according to claim 1, wherein the target grid active safety correction model includes a convolutional layer, a pooling layer, and a full-link layer, and the step of inputting the current operation state information of the grid into the trained target grid active safety correction model to determine a target generator set in the grid that needs to be regulated in terms of output includes:
inputting the running state information of the power grid into the convolution layer, and determining the initial running state characteristic corresponding to the current running state information;
inputting the initial running state features into the pooling layer, and determining important running state features corresponding to the current running state information;
and inputting the important operation state characteristics and the initial operation state characteristics into the full connection layer for weighting, and determining a target generator set needing output regulation.
5. The grid active safety correction method according to claim 1, wherein the adjusting the target generator set according to the target output adjustment amount comprises:
based on the target output adjustment quantity, output adjustment is carried out on the target generator set, and target power flow distribution information of the power grid is determined;
judging whether the overload in the adjusted overload line in the power grid is completely eliminated or not according to the target power flow distribution information;
and if all the output power is eliminated, determining that the output power adjustment is effective.
6. A grid active safety correction device, the grid active safety correction device comprising:
the determining module is used for inputting the current operation state information of the power grid into a trained target power grid active safety correction model when at least one target overload circuit is detected to exist in the power grid, and determining a target generator set which needs to perform output regulation in the power grid; the target power grid active safety correction model is obtained by training an initial power grid active safety correction model based on initial operation state information of a power grid with a sample overload line, sample output adjustment quantity of a sample generator set for output adjustment in the power grid and an overload elimination effect of the adjusted sample overload line;
the calculation module is used for calculating a target output adjustment quantity of the target generator set based on the target sensitivity of the target overload circuit to the target generator set;
and the adjusting module is used for adjusting the output of the target generator set according to the target output adjustment quantity.
7. The grid active safety correction device according to claim 6, characterized in that the trained target grid active safety correction model is obtained by:
inputting initial operation state information of a power grid with a sample overload line into an initial power grid active safety correction model, and determining a sample generator set needing output regulation in the power grid;
calculating the sample output adjustment quantity of the sample generator set according to the sensitivity of the sample overload line to the sample generator set;
determining adjusted running state information in the power grid with the sample overload line according to the sample output adjustment quantity;
updating network parameters of the initial power grid active safety correction model according to the adjusted running state information, and judging whether the overload in the adjusted sample overload line in the power grid is completely eliminated or not;
and if all the active power safety correction models are eliminated, finishing the training of the initial power grid active power safety correction model to obtain the target power grid active power safety correction model.
8. The grid active safety correction device according to claim 7, wherein after the updating of the network parameters of the initial grid active safety correction model according to the adjusted operating state information and the determination of whether all the overloads in the adjusted sample overload line in the grid are eliminated, the grid active safety correction method comprises:
and if not, continuing to train the initial power grid active safety correction model until the overload in the adjusted sample overload line in the power grid is completely eliminated.
9. An electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, the processor and the memory communicate with each other through the bus when an electronic device is operated, and the machine-readable instructions are executed by the processor to perform the steps of the power grid active safety correction method according to any one of claims 1 to 5.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has a computer program stored thereon, and the computer program is executed by a processor to perform the steps of the grid active safety correction method according to any one of the claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111534432.7A CN114243690A (en) | 2021-12-15 | 2021-12-15 | Power grid active safety correction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111534432.7A CN114243690A (en) | 2021-12-15 | 2021-12-15 | Power grid active safety correction method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114243690A true CN114243690A (en) | 2022-03-25 |
Family
ID=80756349
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111534432.7A Pending CN114243690A (en) | 2021-12-15 | 2021-12-15 | Power grid active safety correction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114243690A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115833101A (en) * | 2022-12-06 | 2023-03-21 | 北京百度网讯科技有限公司 | Power scheduling method and device, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080281474A1 (en) * | 2002-09-03 | 2008-11-13 | Patel Sureshchandra B | System of Super Super Decoupled Loadflow Computation for Electrical Power System |
| CN104821605A (en) * | 2015-04-13 | 2015-08-05 | 国家电网公司 | Active safety correction method based on improved particle swarm optimization algorithm |
| CN107069708A (en) * | 2017-03-14 | 2017-08-18 | 国网湖北省电力公司 | A kind of power grids circuits strategy for security correction method based on extreme learning machine |
| CN108110754A (en) * | 2017-12-29 | 2018-06-01 | 山东大学 | A kind of strategy for security correction method for considering to coordinate between poly zoarium |
| CN109193809A (en) * | 2018-08-14 | 2019-01-11 | 河海大学 | A kind of electric system strategy for security correction optimization method based on sensitivity matrix |
| CN111262275A (en) * | 2020-03-26 | 2020-06-09 | 西安电子科技大学 | Global cooperative control method for simulated space solar power station |
-
2021
- 2021-12-15 CN CN202111534432.7A patent/CN114243690A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080281474A1 (en) * | 2002-09-03 | 2008-11-13 | Patel Sureshchandra B | System of Super Super Decoupled Loadflow Computation for Electrical Power System |
| CN104821605A (en) * | 2015-04-13 | 2015-08-05 | 国家电网公司 | Active safety correction method based on improved particle swarm optimization algorithm |
| CN107069708A (en) * | 2017-03-14 | 2017-08-18 | 国网湖北省电力公司 | A kind of power grids circuits strategy for security correction method based on extreme learning machine |
| CN108110754A (en) * | 2017-12-29 | 2018-06-01 | 山东大学 | A kind of strategy for security correction method for considering to coordinate between poly zoarium |
| CN109193809A (en) * | 2018-08-14 | 2019-01-11 | 河海大学 | A kind of electric system strategy for security correction optimization method based on sensitivity matrix |
| CN111262275A (en) * | 2020-03-26 | 2020-06-09 | 西安电子科技大学 | Global cooperative control method for simulated space solar power station |
Non-Patent Citations (1)
| Title |
|---|
| 王鹏翔: "基于机器自学习的电网安全校正算法研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》, 15 September 2011 (2011-09-15) * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115833101A (en) * | 2022-12-06 | 2023-03-21 | 北京百度网讯科技有限公司 | Power scheduling method and device, electronic equipment and storage medium |
| CN115833101B (en) * | 2022-12-06 | 2023-11-14 | 北京百度网讯科技有限公司 | Power scheduling method, device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220210028A1 (en) | Cloud software service resource allocation method based on qos model self-correctio | |
| CN108108532A (en) | With the method for particle cluster algorithm optimization power electronic circuit | |
| CN110120670B (en) | DPV-containing power distribution network reactive voltage optimization method, terminal equipment and storage medium | |
| CN110504716B (en) | Photovoltaic inverter reactive mode optimization selection method, terminal equipment and storage medium | |
| CN109066710A (en) | A kind of multi-objective reactive optimization method, apparatus, computer equipment and storage medium | |
| CN114243690A (en) | Power grid active safety correction method and device, electronic equipment and storage medium | |
| JP6902487B2 (en) | Machine learning system | |
| CN112054519B (en) | Power distribution network low voltage optimization treatment method, system and equipment | |
| JP6067289B2 (en) | Reduced model creation device, creation method and creation program for power system | |
| CN108539799B (en) | method and device for scheduling wind power in power grid | |
| CN110649633A (en) | Power distribution network reactive power optimization method and system | |
| CN111160808A (en) | Distributed event-triggered power system economic dispatching method with uncertain parameters | |
| CN115860170A (en) | Power quality optimization method of power distribution system considering power electronic load | |
| CN115713057A (en) | Analog integrated circuit design parameter automatic optimization method based on deep neural network | |
| CN115269177A (en) | Dynamic calibration optimization method for engine model parameters | |
| CN112103949B (en) | Power grid disturbance stability control method and device, electronic equipment and storage medium | |
| CN108448569B (en) | Method for identifying and improving electromechanical transient simulation numerical convergence of large power grid | |
| CN112733863A (en) | Image feature extraction method, device, equipment and storage medium | |
| CN110580523B (en) | Error calibration method and device for analog neural network processor | |
| CN116154745A (en) | Distributed energy storage planning method and device for direct-current power distribution network | |
| CN113972659B (en) | Energy storage configuration method and system considering random power flow | |
| CN111694595B (en) | Software behavior adjusting method based on error tolerance | |
| CN111030119B (en) | Power grid reactive power flow adjusting method and device based on improved direct current flow algorithm | |
| CN116542341A (en) | Training of machine learning model, data processing method, device and medium | |
| JP2022036809A (en) | Operation support system, operation support method, and program |
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 |