CN113346504B - Active power distribution network voltage control method based on data knowledge driving - Google Patents
Active power distribution network voltage control method based on data knowledge driving Download PDFInfo
- Publication number
- CN113346504B CN113346504B CN202110672346.6A CN202110672346A CN113346504B CN 113346504 B CN113346504 B CN 113346504B CN 202110672346 A CN202110672346 A CN 202110672346A CN 113346504 B CN113346504 B CN 113346504B
- Authority
- CN
- China
- Prior art keywords
- data
- distribution network
- power distribution
- voltage
- strategy
- 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
- 238000009826 distribution Methods 0.000 title claims abstract description 93
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000011217 control strategy Methods 0.000 claims abstract description 30
- 238000005457 optimization Methods 0.000 claims abstract description 22
- 238000004364 calculation method Methods 0.000 claims abstract description 17
- 238000004088 simulation Methods 0.000 claims abstract description 10
- 238000012795 verification Methods 0.000 claims abstract description 9
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 8
- 238000013499 data model Methods 0.000 claims abstract description 8
- 230000009471 action Effects 0.000 claims description 28
- 230000001105 regulatory effect Effects 0.000 claims description 27
- 239000003990 capacitor Substances 0.000 claims description 17
- 230000003993 interaction Effects 0.000 claims description 10
- 239000004576 sand Substances 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 239000000126 substance Substances 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000009825 accumulation Methods 0.000 claims description 2
- 230000003044 adaptive effect Effects 0.000 claims description 2
- 230000001174 ascending effect Effects 0.000 claims description 2
- 238000005520 cutting process Methods 0.000 claims description 2
- 238000012216 screening Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000013500 data storage Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 230000002787 reinforcement Effects 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 231100000136 action limit Toxicity 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000035699 permeability Effects 0.000 description 1
- 230000001737 promoting effect Effects 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
- H02J3/12—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network 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
- H02J3/04—Circuit arrangements for ac mains or ac distribution networks for connecting networks of the same frequency but supplied from different sources
- H02J3/06—Controlling transfer of power between connected networks; Controlling sharing of load between connected 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
- H02J3/12—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load
- H02J3/16—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load by adjustment of reactive power
-
- 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]
-
- 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
Abstract
The invention discloses an active power distribution network voltage control method based on data knowledge driving. Firstly, establishing an active power distribution network reactive power optimization model based on second-order cone optimal power flow; secondly, performing big data simulation based on the established reactive power optimization model to form a data expansion strategy, screening and combining historical data and solution expansion strategy constraints, establishing strategy constraints through the data model to form accumulative data knowledge, and establishing a reactive voltage control strategy knowledge map taking power distribution network data and a control strategy as a core; and finally, matching the control strategy and parameter results by matching similar states in the knowledge map according to the constructed reactive voltage control knowledge map of the power distribution network according to the current network state and adopting an improved time series similar data retrieval algorithm through similarity calculation, sectional verification and slope verification, performing safety verification and optimal solution, generating a control strategy and updating the state strategy in the knowledge map.
Description
Technical Field
The invention belongs to the power industry and relates to voltage control of a power distribution network, in particular to an active power distribution network voltage control method based on data knowledge driving.
Background
At present, the permeability of distributed power sources in a power distribution network is continuously increased, and in 2050, the consumption of non-fossil energy accounts for more than 50% of the consumption of primary energy in China, so that new challenges are provided for the voltage stability and the regulation and control operation complexity of a system. Meanwhile, ensuring higher standard power quality is an inevitable trend in power development and is an objective demand for promoting social stability development. The over-voltage and the low-voltage not only reduce the energy conversion efficiency, but also easily cause the equipment inrush current and harmonic waves to damage the user equipment, and even cause safety accidents. At present, a reactive power adjusting method is mainly adopted to control voltage within a designated range, and specifically, the reactive power adjusting method comprises means of changing the voltage of an on-load tap adjusting bus of a main transformer of a transformer substation, switching a capacitor bank to compensate reactive power and the like through cooperative control of reactive power adjusting equipment, so that the aims of improving the quality of electric energy, operating economy and the like are fulfilled. At present, the mathematical solution is mostly solved by an optimization-based or reinforcement learning method. In the optimization-based method, due to the fact that the transformer tap is arranged, the shunt capacitors are discrete and the load flow equation is nonlinear, the problem of mixed integer nonlinear programming is generally solved, the second-order cone relaxation has high accuracy in solving the problem of reactive voltage control, but along with the increase of the distribution network scale and controllable equipment, the solving complexity is exponentially increased, the calculation time is long, and the real-time regulation and control requirements are difficult to meet. At the same time, there is a convergence problem, possibly falling into local optimality. The reactive power control method based on the reinforcement learning method can meet the real-time control requirement in speed, has certain inspiration and robustness, is insufficient in interpretability and easy to fall into dimension disaster, is easy to cause extreme strategy conditions in free action exploration during learning, and cannot guarantee safety.
Disclosure of Invention
The invention aims to provide a data knowledge-driven active power distribution network voltage control method, a knowledge graph-based reactive voltage control model can well represent data and strategies, and strategy constraints are established through the data model to form accumulative data knowledge. By improving the time sequence retrieval algorithm, the data retrieval requirement of the power distribution network can be better met.
In order to achieve the purpose, the active power distribution network voltage control method based on data knowledge driving comprises the following steps:
step 1: considering that discrete reactive voltage regulation equipment comprises an on-load tap changing transformer, a grouping switching capacitor and a voltage regulator, and establishing an active power distribution network reactive power optimization model based on second-order cone optimal power flow;
step 2: based on the active power distribution network reactive power optimization model established in the step 1, carrying out big data simulation by utilizing the existing load and photovoltaic data to form a data expansion strategy, combining the historical voltage regulation strategy of the power distribution network with the result of the data expansion strategy, establishing strategy constraints through the data model to form accumulative data knowledge, and establishing a power distribution network reactive voltage control strategy knowledge map taking power distribution network data and control strategies as the core;
And 3, step 3: and (3) based on the power distribution network reactive voltage control knowledge graph constructed in the step (2), matching similar states in the knowledge graph by using the current network state, adopting an improved time series similar data retrieval algorithm, performing security check and optimization solution through similarity calculation, segmentation check, slope check, matching control strategy and parameter results, generating a control strategy and updating the state strategy in the power distribution network reactive voltage control knowledge graph.
The beneficial effects of the invention are as follows: the reactive voltage control model based on the knowledge graph can well represent data and strategies, and strategy constraints are established through the data model to form data knowledge capable of being accumulated. By improving the time sequence retrieval algorithm, the data retrieval requirements of the power distribution network can be better met, and the generated strategy is excellent in reactive power optimization effect; compared with an optimization algorithm, the time consumption of strategy generation is greatly reduced, and the strategy interpretability is improved. The man-machine interaction generation strategy improves the regulation and control level of power personnel under different reactive voltage control scenes, is suitable for the rapid regulation and control of a power grid, and performs valuable exploration in knowledge data reasoning and interaction in the field of power systems.
Drawings
FIG. 1 is a power distribution network reactive voltage control strategy knowledge graph data model diagram of the present invention;
FIG. 2 is a flow diagram of knowledge-graph based policy generation;
fig. 3 is an interaction diagram of a reactive voltage control test system of a power distribution network.
Detailed Description
The invention is described in further detail below with reference to the following figures and specific examples:
a voltage control method of an active power distribution network based on data knowledge driving comprises the following steps:
step 1: considering that discrete reactive power voltage regulation equipment comprises an on-load voltage regulating transformer, a group switching capacitor and a voltage regulator, and establishing an active power distribution network reactive power optimization model based on second-order cone optimal power flow;
and 2, step: based on the active power distribution network reactive power optimization model established in the step 1, carrying out big data simulation by utilizing the existing load and photovoltaic data to form a data expansion strategy, combining the historical voltage regulation strategy of the power distribution network with the result of the data expansion strategy, establishing strategy constraint through the data model to form accumulative data knowledge, and establishing a power distribution network reactive voltage control strategy knowledge map taking the power distribution network data and the control strategy as the core, wherein the data model is shown in figure 1;
and 3, step 3: based on the power distribution network reactive voltage control knowledge graph constructed in the step 2, matching the similar state in the knowledge graph with the current network state, adopting an improved time series similar data retrieval algorithm, performing security check and optimization solution through similarity calculation, segmentation check, slope check, matching control strategy and parameter results, generating a control strategy and updating the state strategy in the power distribution network reactive voltage control knowledge graph, as shown in fig. 2.
In the above technical solution, the active power distribution network voltage control method based on data knowledge driving further includes the following steps:
and 4, step 4: human-computer interaction online decision is added, time scale, equipment action and key point voltage are controlled, and a power distribution network voltage control test system is developed based on the method so as to improve the convenience and interactivity of reactive voltage control operation.
In the above technical solution, the active power distribution network reactive power optimization model based on the second-order cone optimal power flow is as follows: the method comprises the steps that the goal of minimizing the total operation cost is achieved on the premise that voltage constraints are met, the voltage constraints comprise power flow balance constraints, ohm law corresponding constraints, node voltage constraints and apparent power equality constraints, and the total operation cost comprises line loss, on-load tap changer switching loss, voltage regulator switching loss and grouping switching capacitor switching loss;
the expression for minimizing the total operation cost is as follows:
wherein the content of the first and second substances,
representing the active loss of the line, C, operating during time tpIs composed of
Switching loss factor of loss, COLTCFor the switching loss factor of on-load tap changers, CCBSwitching loss factor, C, for switched capacitors in groupsVRIs the switching loss coefficient of the voltage regulator, wherein N is the proportional switching loss of the reactive voltage regulating equipment of the power distribution network OLTCNumber of on-load tap changers, NCBNumber of capacitors switched for groups, NVRIn order to count the number of the voltage regulators,
for the nth on-load tap changer gear at time tth,
the capacitor gear is switched for the nth group at time tth,
the gear of the nth voltage regulator is the time T, and the T is the operation time range of reactive voltage control of the power distribution network;
the power flow balance constraint expression is as follows:
the ohm law corresponding constraint expression is as follows:
wherein PGtRepresenting the active power of the line, QG, at time ttIndicating the reactive power of the line at time t, QDtRepresenting the reactive demand of the line, u, at time ttRepresenting the square of the magnitude of the voltage at all nodes at time t
The vector of the composition is then calculated,
representing the square of the magnitude of the voltage at node i at time t, ltRepresenting the square of the magnitude of all line currents at time t
The vector of the composition is then calculated,
representing the square of the amplitude of the current on the line linking node i and node j at time t, TptVector P representing gear data of all reactive voltage regulating equipment of power distribution networktRepresenting the active power flow of the line at time t, QtRepresenting the reactive power flow of the line at time t, PijAnd QijRepresenting the active and reactive power flows on the line connecting node i and node j respectively,
represents an arbitrary time t;
The node voltage constraint expression is:
wherein the content of the first and second substances,urepresents the upper limit of the square of the voltage amplitude,
representing the lower limit of the square of the voltage amplitude, N is the set of nodes,
represents an arbitrary node;
the quadratic equation of the apparent power equality constraint model can be relaxed into inequality constraint, and the expression is as follows:
wherein the content of the first and second substances,
representing the active power on the line linking node i and node j at time t,
representing the reactive power, l, on the line linking node i and node j at time tijRepresenting the square of the magnitude of the current, u, on a line linking node i and node jiRepresents the square of the voltage magnitude at node i;
the service life of the on-load tap changer and the grouping switching capacitor is influenced by frequent actions, so the on-load tap changer and the grouping switching capacitor are subjected to the action in the dispatching period TDThe number of switching times and the gear position in the gear are limited. Solving the reactive voltage regulation model becomes a mixed integer nonlinear problem, and big data simulation can be carried out by utilizing the existing load and photovoltaic data, so that a stage control strategy is obtained.
In the above technical solution, the specific implementation method of step 2 is:
based on the active power distribution network reactive power optimization model established in the step 1, the historical voltage regulation strategy of the power distribution network and the data expansion strategy constraint are screened out and combined, and a reactive voltage control strategy knowledge graph taking power distribution network data and a power distribution network reactive voltage control strategy as a core is established. The traditional reactive voltage control model needs complex solving calculation, has high requirements on hardware computing power and long time consumption for strategy generation, and is usually repeated calculation on similar data of the same model, so that the resource utilization rate is low and the historical control strategy is not effectively summarized and applied.
Constructing a power distribution network reactive voltage control strategy knowledge map based on power distribution network system data, wherein the power distribution network system data comprises active data, reactive data, a distributed power supply, load prediction time sequence data and optimal gears of a voltage regulating device in power distribution network historical data and simulation data; the knowledge graph comprises entities, entity relations and attributes, the entities comprise the reactive power distribution network voltage regulating equipment state, power distribution network system data, action strategies and strategy results, and the entity relations are that the voltage regulating equipment state, the system data space, the control strategies and the power grid parameter results are sequentially connected, as shown in figure 1.
State s of reactive voltage regulation equipment of power distribution networktThe method comprises the following steps that S belongs to S, wherein S is a state combination set of all reactive power voltage regulating devices of the power distribution network and contains gear conditions of all the reactive power voltage regulating devices; the expression mode of the data space of the power distribution network system is
Wherein n isdIs a system data sequence, d is a controllable voltage regulating equipment ordinal number, NsThe number of the pressure regulating devices can be controlled;
from stThe result of the state of the voltage regulating equipment is s based on the matching of the data space path to the next dispatching cyclet+1Namely, the action state strategy pi is generated according to the gear state change situation of the equipment before and after the action iThe state condition of each voltage regulating device is combined into a finite state set.
Because the path from the system state to the action state strategy is obtained by solving based on the second-order cone optimal power flow algorithm, and the result already restrains active, reactive and voltage regulating equipment and the like, the action strategy linked with the initial state of the knowledge graph is a result with restraint. Meanwhile, in order to completely represent the action strategy effect, the network parameters of the power distribution network after action are constructed into a strategy result entity, so that the strategy result is evaluated, and strategy selection constraint is further realized through interaction.
In the above technical solution, the concrete implementation method of the similarity calculation in step 3 is:
matching the initial state of the power distribution network system according to the current timestamp t to obtain the entity initial state s meeting the conditions in the knowledge graphtGiven a new power distribution network system data sequence nsAnd retrieving the most similar system data sequence n in the data spacedObtaining the state and action strategy s of the voltage regulating equipment in the next dispatching cyclet+1；
The data space comprises an active sequence, a reactive sequence, a distributed power output sequence and a load output sequence, and the data sequence data n of the most similar systemdWith a given system data sequence n sThe sampling period and the number of the sampling periods are the same.
In the above technical solution, the retrieving of the data sequence n of the most similar system data in the data spacedThe method comprises the following steps of measuring the similarity by adopting an Euclidean distance function, wherein the expression is as follows:
wherein n isdFor the most similar system data sequence data in data space, nsFor distribution network system data sequences, wiFor data importance weight, ε is similarity threshold, and sequences n satisfying the formula are retaineddThe adjacent control strategy entity is a reactive voltage control strategy obtained under the condition of a similarity threshold epsilon, and the expression is as follows:
dist(nd,ns)≤ε
in the above technical solution, the similarity calculation and matching in step 3 further adopts a method of segment check and slope check, and the specific implementation process is as follows:
the specific implementation method of the segmented verification comprises the following steps: defining a sliding window
Window sw combines data sequence ndCutting into q segments eachThe segment length is fixed to be p, the Euclidean distance function is adopted to carry out segment matching calculation, a data sequence meeting the distance condition is reserved, and if the similarity of the segment data sequence is greater than a similarity threshold value, the calculation is stopped; and accumulating and counting the Euclidean distance of the finally reserved data sequence.
The specific implementation method of the slope verification comprises the following steps: adding slope filtering on the basis of linear accumulation of calculation results of Euclidean distance functions, and sliding matching n of window sw sAnd ndData sequence, characteristic point sequence section obtained by sliding matching
And
n in the window with unit sliding window lengthsAnd ndWhen the difference product of the data sequence segments is negative and less than kappa, wherein kappa is a negative slope trend threshold, n is indicatedsAnd ndThe slope trend of the data sequence segment is opposite and the difference value is larger, and the data sequence segment is removed from the alternative similar data sequence, so that the retrieval precision can be improved, and the expression is as follows:
obtaining m most similar system data sequence data through segmented verification and slope filtering, wherein the corresponding number of s is not more than mt+1Action strategy, if m is 1, then there is only one kind of st+1Action policy, then st+1Is the corresponding matching item; if m>1, according to the reactive voltage regulation operation cost pair st+1Sorting in ascending order and selecting the action strategy with the lowest operation cost; and if m is 0, performing real-time optimization simulation based on the active power distribution network reactive power optimization model established in the step 1, and storing corresponding data into a knowledge graph.
In the above technical solution, the security check in step 3 is:
and verifying that the power flow of the action strategy is converged and the voltage is not out of limit for the action strategy result obtained under the condition that the number m of the most similar system data sequence data is not 0.
In the above technical solution, the step 3 retrieves the corresponding result tail entity data in the data space as the reactive power voltage regulation equipment action strategy, and based on the constructed reactive power voltage regulation strategy knowledge graph, the data space is limited by the knowledge graph matching action strategy, so as to accurately restrict part of equipment action limits and times, reduce the data retrieval range, and implement scene-adaptive and interactive strategy generation.
In the above technical solution, the specific implementation method of step 4 is:
changing retrieval data scheduling period TDGenerating three time scale control schemes of daily level, hour level and minute level according to the operation time range T of reactive voltage control of the power distribution network, and carrying out adaptive adjustment according to output and load states and real-time requirements;
the adaptability adjustment is as follows: the daily control is adopted for a power grid with unobvious long-time output and load change, the small-scale control is adopted under the condition of normal fluctuation, and the minute-scale control is adopted in an emergency state, so that the daily control and the minute-scale control are not generally adopted due to higher cost and frequent action.
In the reactive voltage regulation process, the voltage fluctuation condition of an important load node is the key for influencing the quality of electric energy. Important load points also change according to different scenes, so that the corresponding state results of the strategy are constrained in the data space retrieval process, and the key node selection can be adjusted through real-time interaction in the strategy generation process, and the voltage fluctuation range of the important nodes is accurately limited.
In the technical scheme, the power distribution network voltage control test system realizes data storage, retrieval and interaction, realizes data storage and management based on the most popular graph database Neo4, and adopts an artificial intelligence markup language technology to realize human-computer interaction, as shown in fig. 3. The reactive voltage control knowledge map of the power distribution network is stored in a map database Neo4j, a regulation and control person inputs the data in Chinese characters, the system analyzes input meanings by using AIML, generates instructions according to analysis results to complete operations such as map query search, updating and management, and completes interaction according to question and answer logic, so that scene random fluctuation and control target complex change in the reactive voltage control process of the power distribution network are realized, multiple schemes are provided for selection, and time scales, equipment action conditions and special scene voltage regulation requirements are considered, so that interactivity and applicability are improved. In the inquiry and recommendation process, interactive results are presented in various forms such as character visualization, graph visualization, data tables and the like, so that the recommendation information can be intuitively mastered at multiple angles.
Those not described in detail in this specification are well within the skill of the art.
Claims (9)
1. A voltage control method of an active power distribution network based on data knowledge driving is characterized by comprising the following steps: the method comprises the following steps:
Step 1: considering that discrete reactive power voltage regulation equipment comprises an on-load voltage regulating transformer, a group switching capacitor and a voltage regulator, and establishing an active power distribution network reactive power optimization model based on second-order cone optimal power flow;
the active power distribution network reactive power optimization model based on the second-order cone optimal power flow is as follows: the method comprises the following steps of achieving the goal of minimizing total operation cost on the premise of meeting voltage constraints, wherein the voltage constraints comprise a power flow balance constraint, an ohm law corresponding constraint, a node voltage constraint and an apparent power equality constraint;
and 2, step: based on the active power distribution network reactive power optimization model established in the step 1, carrying out big data simulation by utilizing the existing load and photovoltaic data to form a data expansion strategy, combining the historical voltage regulation strategy of the power distribution network with the result of the data expansion strategy, establishing strategy constraints through the data model to form accumulative data knowledge, and establishing a power distribution network reactive voltage control strategy knowledge map taking power distribution network data and control strategies as the core;
constructing a power distribution network reactive voltage control strategy knowledge map based on power distribution network system data, wherein the power distribution network system data comprises active data, reactive data, a distributed power supply, load prediction time sequence data and optimal gears of a voltage regulating device in power distribution network historical data and simulation data; the knowledge graph comprises entities, entity relations and attributes, the entities comprise the reactive power distribution network voltage regulating equipment state, power distribution network system data, action strategies and strategy results, and the entity relations are that the voltage regulating equipment state, the system data space, the control strategies and the power grid parameter results are sequentially connected;
And step 3: and (3) based on the power distribution network reactive voltage control knowledge graph constructed in the step (2), matching the similar state in the knowledge graph with the current network state, adopting an improved time series similar data retrieval algorithm, performing safety check and optimization solution through similarity calculation, segmentation check, slope check, matching control strategy and parameter results, generating a control strategy and updating the state strategy in the power distribution network reactive voltage control knowledge graph.
2. The active power distribution network voltage control method based on data knowledge driving of claim 1, characterized in that: it also includes the following steps:
and 4, step 4: and man-machine interaction online decision is added, and time scale, equipment action and key point voltage are controlled.
3. The active power distribution network voltage control method based on data knowledge driving of claim 1, characterized in that: the total operation cost comprises line loss, on-load tap changer switching loss, voltage regulator switching loss and grouping switching capacitor switching loss;
the expression for minimizing the total operation cost is as follows:
wherein the content of the first and second substances,
representing the active loss of the line, C, operating during time tpIs composed of
Switching loss factor of loss, C OLTCSwitching loss factor, C, for on-load tap changersCBSwitching loss factor, C, for switched capacitors in groupsVRIs the switching loss coefficient of the voltage regulator, wherein N is the proportional switching loss of the reactive voltage regulating equipment of the power distribution networkOLTCNumber of on-load tap changers, NCBNumber of capacitors switched for groups, NVRIn order to count the number of the voltage regulators,
for the nth on-load tap changer gear at time tth,
the capacitor gear is switched for the nth group at time tth,
the gear of the nth voltage regulator is the time T, and the T is the operation time range of reactive voltage control of the power distribution network;
the power flow balance constraint expression is as follows:
the ohm law corresponding constraint expression is as follows:
wherein PGtRepresenting the active power of the line, QG, at time ttIndicating the reactive power of the line at time t, QDtRepresenting the reactive demand of the line, u, at time ttRepresenting the square of the magnitude of the voltage at all nodes at time t
The vector of the composition is then calculated,
representing the square of the magnitude of the voltage at node i at time t, ltRepresenting the square of the magnitude of all line currents at time t
The vector of the composition is then calculated,
representing the square of the amplitude of the current on the line linking node i and node j at time t, TptVector P representing gear data of all reactive voltage regulating equipment of power distribution network tRepresenting the active power flow of the line at time t, QtRepresenting the reactive power flow of the line at time t, PijAnd QijRepresenting the active and reactive power flows on the line connecting node i and node j respectively,
represents an arbitrary time t;
the node voltage constraint expression is:
wherein the content of the first and second substances,urepresents the upper limit of the square of the voltage amplitude,
representing the lower limit of the square of the voltage amplitude, N is the set of nodes,
represents an arbitrary node;
the quadratic equality relaxation of the apparent power equality constraint model is inequality constraint, and the expression is as follows:
wherein the content of the first and second substances,
representing the active power on the line linking node i and node j at time t,
representing the reactive power, l, on the line linking node i and node j at time tijRepresenting the square of the magnitude of the current, u, on a line linking node i and node jiRepresents the square of the voltage magnitude at node i;
the service life of the on-load tap changer and the grouping switching capacitor is influenced by frequent actions, so the on-load tap changer and the grouping switching capacitor are subjected to the action in the dispatching period TDThe number of switching times and gears in the system are limited, the solution of the reactive voltage regulation model is changed into a mixed integer nonlinear problem, and the existing load and photovoltaic data are utilized to carry out big data simulation, so that a stage control strategy is obtained.
4. The active power distribution network voltage control method based on data knowledge driving of claim 1 is characterized in that: state s of reactive voltage regulation equipment of power distribution network tThe method comprises the following steps that S belongs to S, wherein S is a state combination set of all reactive power voltage regulating devices of the power distribution network and contains gear conditions of all the reactive power voltage regulating devices; the expression mode of the data space of the power distribution network system is
Wherein n isdIs a system data sequence, d is a controllable voltage regulating equipment ordinal number, NsThe number of the pressure regulating devices can be controlled;
from stThe result of the state of the voltage regulating equipment is s based on the matching of the data space path to the next dispatching cyclet+1I.e. generating an action state strategy from the gear state change of the equipment before and after actionπiAnd the state conditions of the voltage regulating devices are combined into a finite state set.
5. The active power distribution network voltage control method based on data knowledge driving of claim 1, characterized in that: the concrete implementation method of similarity calculation in the step 3 is as follows:
matching the initial state of the power distribution network system according to the current timestamp t to obtain the entity initial state s meeting the conditions in the knowledge graphtGiven a new distribution network system data sequence nsAnd retrieving the most similar system data sequence n in the data spacedObtaining the state and action strategy s of the voltage regulating equipment in the next dispatching cyclet+1；
The data space comprises an active power sequence, a reactive power sequence, a distributed power output sequence and a load output sequence, and the most similar system data sequence n dWith a given system data sequence nsThe sampling period and the number of the sampling periods are the same.
6. The active power distribution network voltage control method based on data knowledge driving of claim 5 is characterized in that: the retrieval of the most similar system data sequence data n in the data spacedThe method comprises the following steps of measuring the similarity by adopting an Euclidean distance function, wherein the expression is as follows:
wherein n isdFor the most similar system data sequence data in data space, nsFor distribution network system data sequences, wiFor data importance weight, ε is similarity threshold, and sequences n satisfying the formula are retaineddThe adjacent control strategy entity is a reactive voltage control strategy obtained under the condition of a similarity threshold epsilon, and the expression is as follows:
dist(nd,ns)≤ε。
7. the active power distribution network voltage control method based on data knowledge driving of claim 1, characterized in that: the similarity calculation and matching in the step 3 also adopts a method of segment check and slope check, and the specific implementation process is as follows:
the specific implementation method of the segmented verification comprises the following steps: defining a sliding window
Window sw combines data sequence ndCutting the data into q sections, fixing the length of each section as p, performing section matching calculation by adopting an Euclidean distance function, reserving a data sequence meeting a distance condition, and stopping calculation if the similarity of the section data sequence is greater than a similarity threshold value; accumulating and counting the Euclidean distance of the finally reserved data sequence;
The specific implementation method of the slope verification comprises the following steps: adding slope filtering on the basis of linear accumulation of calculation results of Euclidean distance functions, and sliding matching n of window swsAnd ndData sequence, characteristic point sequence section obtained by sliding matching
And
n in the window with unit sliding window lengthsAnd ndWhen the difference product of the data sequence segments is negative and less than kappa, wherein kappa is a negative slope trend threshold, n is indicatedsAnd ndAnd eliminating the data sequence section from the alternative similar data sequence with opposite slope trend and larger difference value so as to improve the retrieval precision, wherein the expression is as follows:
obtaining m most similar system data sequence data through segmented verification and slope filtering, wherein the corresponding number of s is not more than mt+1Action strategy, if m is 1, then there is only one kind of st+1Action policy, then st+1Is the corresponding matching item; if m>1, according to the reactive voltage regulation operation cost pair st+1Sorting in ascending order and selecting the action strategy with the lowest operation cost; and if m is 0, performing real-time optimization simulation based on the active power distribution network reactive power optimization model established in the step 1, and storing corresponding data into a knowledge graph.
8. The active power distribution network voltage control method based on data knowledge driving of claim 1, characterized in that: the safety check in the step 3 is as follows:
And verifying that the power flow of the action strategy is converged and the voltage is not out of limit for the action strategy result obtained under the condition that the number m of the most similar system data sequence data is not 0.
9. The active power distribution network voltage control method based on data knowledge driving of claim 2 is characterized in that: the specific implementation method of the step 4 comprises the following steps:
changing retrieval data scheduling period TDGenerating three time scale control schemes of a daily level, an hour level and a minute level according to the operation time range T of the reactive voltage control of the power distribution network, and carrying out adaptive adjustment according to the output and load states and real-time requirements;
the adaptability adjustment is as follows: the day-level control is adopted for the power grid with unobvious long-time output and load change, the hour-level control is adopted under the condition of normal fluctuation, and the minute-level control is adopted in the emergency state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110672346.6A CN113346504B (en) | 2021-06-17 | 2021-06-17 | Active power distribution network voltage control method based on data knowledge driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110672346.6A CN113346504B (en) | 2021-06-17 | 2021-06-17 | Active power distribution network voltage control method based on data knowledge driving |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113346504A CN113346504A (en) | 2021-09-03 |
| CN113346504B true CN113346504B (en) | 2022-06-28 |
Family
ID=77476229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110672346.6A Active CN113346504B (en) | 2021-06-17 | 2021-06-17 | Active power distribution network voltage control method based on data knowledge driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113346504B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113850481B (en) * | 2021-09-07 | 2022-12-16 | 华南理工大学 | Power system scheduling service assistant decision method, system, device and storage medium |
| CN116345470B (en) * | 2023-05-31 | 2023-08-08 | 国网冀北电力有限公司 | Knowledge-graph-fused power system power flow optimization method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106786631A (en) * | 2017-03-06 | 2017-05-31 | 天津大学 | Distributed power source voltage ＆ var control strategy setting method on the spot based on cone planning |
| CN106921164A (en) * | 2017-04-05 | 2017-07-04 | 广东电网有限责任公司东莞供电局 | The MIXED INTEGER Second-order cone programming method and system of distribution voltage power-less collaboration optimization |
| WO2018049737A1 (en) * | 2016-09-18 | 2018-03-22 | 国电南瑞科技股份有限公司 | Safe correction calculation method based on partition load control |
| CN111064201A (en) * | 2019-12-31 | 2020-04-24 | 东南大学 | Power distribution network voltage optimization and regulation method based on network topology optimization control |
-
2021
- 2021-06-17 CN CN202110672346.6A patent/CN113346504B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018049737A1 (en) * | 2016-09-18 | 2018-03-22 | 国电南瑞科技股份有限公司 | Safe correction calculation method based on partition load control |
| CN106786631A (en) * | 2017-03-06 | 2017-05-31 | 天津大学 | Distributed power source voltage ＆ var control strategy setting method on the spot based on cone planning |
| CN106921164A (en) * | 2017-04-05 | 2017-07-04 | 广东电网有限责任公司东莞供电局 | The MIXED INTEGER Second-order cone programming method and system of distribution voltage power-less collaboration optimization |
| CN111064201A (en) * | 2019-12-31 | 2020-04-24 | 东南大学 | Power distribution network voltage optimization and regulation method based on network topology optimization control |
Non-Patent Citations (2)
| Title |
|---|
| 邢海军 等.基于多种主动管理策略的配电网综合无功优化.《电网技术》.2015,第39卷(第06期), * |
| 郭清元 等.基于混合整数二阶锥规划的新能源配电网电压无功协同优化模型.《中国电机工程学报》.2017,(第05期), * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113346504A (en) | 2021-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Two-stage dynamic reactive power dispatch strategy in distribution network considering the reactive power regulation of distributed generations | |
| CN113346504B (en) | Active power distribution network voltage control method based on data knowledge driving | |
| CN107358332A (en) | A kind of dispatching of power netwoks runs lean evaluation method | |
| CN109687469B (en) | Intelligent soft switching voltage control method for active power distribution network based on opportunity constraint planning | |
| CN109586298B (en) | Multi-direct-current receiving-end power grid comprehensive load optimization control method and system | |
| CN110635518A (en) | Source network load and storage optimization method based on photovoltaic high permeability | |
| Li et al. | Deep reinforcement learning-based adaptive voltage control of active distribution networks with multi-terminal soft open point | |
| CN107292489A (en) | A kind of dispatching of power netwoks runs lean evaluation system | |
| Yu et al. | Design of machine learning algorithm for tourism demand prediction | |
| CN107171339A (en) | A kind of distribution network voltage idle work optimization method containing microgrid | |
| He et al. | Hierarchical optimal energy management strategy of hybrid energy storage considering uncertainty for a 100% clean energy town | |
| CN116826847A (en) | Dynamic network reconstruction and reactive voltage adjustment collaborative optimization method, device and equipment | |
| CN115313403A (en) | Real-time voltage regulation and control method based on deep reinforcement learning algorithm | |
| Li et al. | Distributed deep reinforcement learning for integrated generation‐control and power‐dispatch of interconnected power grid with various renewable units | |
| Aydin et al. | Comparative analysis of multi-criteria decision making methods for the assessment of optimal SVC location | |
| CN111193295A (en) | Distribution network flexibility improvement robust optimization scheduling method considering dynamic reconfiguration | |
| Bagheri Tolabi et al. | Novel hybrid fuzzy-intelligent water drops approach for optimal feeder multi objective reconfiguration by considering multiple-distributed generation | |
| Liu et al. | An AGC dynamic optimization method based on proximal policy optimization | |
| CN114285090A (en) | New energy limit consumption capability evaluation method based on single station-partition-whole network | |
| Zhao et al. | Distribution network reconfiguration digital twin model based on bi-level dynamical time division | |
| CN113673065A (en) | Loss reduction method for automatic reconstruction of power distribution network | |
| Wenlong et al. | Reactive power optimization of distribution network based on case-based reasoning | |
| CN112117772A (en) | Energy storage fast optimization configuration method for considering output resistor plug under new energy access | |
| Liu | Research on the Application of Artificial Intelligence Technology in Power System Intelligent Dispatching Automation | |
| CN113949068B (en) | Voltage control method based on power grid loss reduction grading partition |
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 |