CN112559963B - Dynamic parameter identification method and device for power distribution network - Google Patents
Dynamic parameter identification method and device for power distribution network Download PDFInfo
- Publication number
- CN112559963B CN112559963B CN202011309392.1A CN202011309392A CN112559963B CN 112559963 B CN112559963 B CN 112559963B CN 202011309392 A CN202011309392 A CN 202011309392A CN 112559963 B CN112559963 B CN 112559963B
- Authority
- CN
- China
- Prior art keywords
- probability
- variable
- data
- power distribution
- distribution network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000009826 distribution Methods 0.000 title claims abstract description 173
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 27
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 230000007704 transition Effects 0.000 claims description 19
- 230000001364 causal effect Effects 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 8
- 238000007476 Maximum Likelihood Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 4
- 230000003068 static effect Effects 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 4
- 238000005381 potential energy Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 18
- 238000010586 diagram Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 7
- 238000004590 computer program Methods 0.000 description 7
- 208000025174 PANDAS Diseases 0.000 description 3
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 description 3
- 240000004718 Panda Species 0.000 description 3
- 235000016496 Panda oleosa Nutrition 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Analysis (AREA)
- Operations Research (AREA)
- Computational Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Mathematical Physics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Mathematical Optimization (AREA)
- Health & Medical Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Pure & Applied Mathematics (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Health & Medical Sciences (AREA)
- Water Supply & Treatment (AREA)
- Evolutionary Biology (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Probability & Statistics with Applications (AREA)
- Educational Administration (AREA)
- Algebra (AREA)
- Development Economics (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a method and a device for identifying dynamic parameters of a power distribution network. The method comprises the following steps: preprocessing the collected operation data of the power distribution network and the external environment data to generate a dynamic parameter identification sample of the power distribution network; discretizing a dynamic parameter identification sample of the power distribution network to obtain a discretized sample; based on a pre-established probability map model, acquiring parameters of the probability map model according to the discretized sample; the probability map model is a two-time slice probability map model; and acquiring dynamic parameters of the power distribution network based on a confidence propagation algorithm according to the observation variables and the probability map model after acquiring the parameters. According to the method, the dynamic parameters of the power distribution network can be deduced by utilizing the probability graph model under the condition that part of measured data or external environment data are missing, and the method is beneficial to improving the parameter identification precision under the condition that part of power distribution area operation modes are suddenly changed or external environment is suddenly changed.
Description
Technical Field
The invention belongs to the field of data-driven power distribution network parameter identification, and particularly relates to a power distribution network dynamic parameter identification method and device.
Background
The emphasis of energy conservation and loss reduction is on a power distribution network, and the physical parameters of a power distribution line are the basis of power grid loss calculation. Because the power distribution network is large in scale, the power distribution automation level of each area is different, and accurate physical parameters of a power distribution line are difficult to obtain in partial power supply areas. In actual operation, the physical parameters of the distribution network are closely related to the factors such as the ambient temperature, the current-carrying capacity of the distribution network, and the like, and dynamic characteristics can be presented along with the change of the operation environment, so that the physical parameters of the static distribution network stored in a Production Management System (PMS) are inaccurate. The operation state of the distribution network cannot be accurately mastered by the distribution network dispatching part in a part of areas, so that the possibility of power failure accidents caused by improper manual operation is increased. The mode of manually checking the dynamic parameters is low in efficiency and high in cost, and how to utilize advanced measurement system data to perform data-driven power distribution network dynamic parameter identification becomes a key for realizing intelligent power grid situation awareness.
Disclosure of Invention
The invention aims to: aiming at the problems in the prior art, the invention provides a method and a device for identifying the dynamic parameters of the power distribution network based on a probability graph model, which are used for analyzing the influence of the running state of the power distribution network and external environment factors on the dynamic characteristics of the parameters of the power distribution network, deducing the probability distribution of the dynamic parameters of the power distribution network by using the probability graph model and realizing the accurate identification of the dynamic parameters of the power distribution network.
The technical scheme is as follows: in order to achieve the above objective, in one aspect, the present invention provides a method for identifying dynamic parameters of a power distribution network, including the following steps:
preprocessing the collected operation data of the power distribution network and the external environment data to generate a dynamic parameter identification sample of the power distribution network;
discretizing a dynamic parameter identification sample of the power distribution network to obtain a discretized sample;
based on a pre-established probability map model, acquiring parameters of the probability map model according to a discretization sample, wherein the probability map model is a two-time-slice probability map model;
and acquiring dynamic parameters of the power distribution network based on a confidence propagation algorithm according to the observation variables and the probability map model after acquiring the parameters.
Further, the step of establishing the probability map model includes:
selecting the temperature, the humidity, the voltage drop of the feeder line section and the transmission power of the feeder line section at one moment as observation variables in the probability map model, and selecting the impedance of the line at the moment as hidden variables of the probability map model;
according to the causal relationship between each observed variable and hidden variable in a single time slice, each variable is added into a probability map model one by one, and a static Bayesian network is constructed;
setting an initial time slice, and designating prior probability distribution of variables under the time slice;
and designating the causal relationship of each state between adjacent time slices, and constructing a transition model.
Further, the generating a power distribution network dynamic parameter identification sample includes:
performing secondary spline interpolation on the external environment data to ensure that the frequencies of the external environment data with different data sources are identical;
merging the data of different data sources, and eliminating redundant fields in the data;
and eliminating repeated data in the operation data of the power distribution network and carrying out data emptying.
Further, the discretizing the dynamic parameter identification sample of the power distribution network has the following calculation formula:
wherein Z is hidden variable, m is the number of divided discrete intervals, N count (z=s) is the number of samples of the hidden variable in the data at state s; n (N) amount (Z) is the total number of samples.
Further, the acquiring parameters of the probability map model includes:
based on the probability quality function, obtaining an initial probability distribution table of each variable in the discretized sample;
calculating a conditional probability distribution table among variables according to the discretization sample based on a maximum expected algorithm;
counting a transition probability distribution table of each variable from the time t to the time t+1 from continuous data samples on a time axis;
the correctness of the conditional probability distribution table is determined by checking whether the sum of the probability distributions of each variable is 1, checking whether the conditional probability distribution is consistent with causal relationships in the bayesian network.
Further, the probability mass function is expressed as:
wherein the method comprises the steps ofThe probability of being initially in state s for the hidden variable; n (N) count (z=s) is the number of samples of the hidden variable in the data at state s; n (N) amount (Z) is the total number of samples.
Further, the calculating a conditional probability distribution table among variables according to the maximum expectation algorithm in the discretized sample comprises the following steps:
the posterior probability of the hidden variable is calculated according to the initial value of the conditional probability or the conditional probability obtained by the previous iteration, and is used as the current expected value of the hidden variable, and the expression is as follows:
P posterior (Z)＝P(Z|X；θ cpt )
wherein Z is a hidden variable, P posterior (Z) posterior probability of hidden variable, θ cpt The method is characterized in that the method is a conditional probability distribution table in a probability graph model, and X is an observation variable;
updating a conditional probability distribution table with a maximum likelihood function as a target, wherein the conditional probability distribution table has the expression:
wherein m is the number of hidden variable states, P (X, Z; theta) cpt ) Is the expectation of the hidden variable obtained from the sample;
and when the probability of the training data sample is maximum according to the conditional probability distribution table in the probability graph model, the iteration of the maximum expected algorithm is ended.
Further, the transition probability distribution table has the expression:
wherein the method comprises the steps ofThe probability of transition of the hidden variable Z from the state 1 to the state 2 from t-1 to t time is represented; n (N) count (s 1 ,s 2 ) Representing the times of transition from the state 1 to the state 2 of the hidden variable Z from t-1 to t in the acquired historical data; n (N) amount (s 1 ) Representing the number of samples of the acquired historical data in which the hidden variable Z is in state 1.
Further, the obtaining the dynamic parameters of the power distribution network based on the belief propagation algorithm includes:
initializing probability distribution of each variable according to the sample;
a certain state variable Y in the network is randomly selected, and the confidence of the node is replaced by b (Y t )：
Wherein phi (Y) t ,X t ) For the likelihood function between the corresponding state variable Y and the observed variable X at the moment t, the joint compatibility of the node Y at the moment t is represented, G is the first-order neighborhood of the node Y, m xY (Y t ) A message passed to node Y for node x;
updating information between variables:
wherein ψ (Y) t ,Y t-1 ) The potential energy between the nodes from the t-1 moment to the t moment at the node Y is obtained;
until the convergence condition is satisfied:
b (n) (Y t )-b (n-1) (Y t )＜10 -5
confidence b (Y) t ) As a result of the inference of probability distribution of hidden variables in each state interval.
In another aspect, the present invention provides a dynamic parameter identification device for a power distribution network, including:
the data preprocessing module is used for preprocessing the collected operation data of the power distribution network and the external environment data to generate a dynamic parameter identification sample of the power distribution network;
the discretization processing module is used for discretizing the dynamic parameter identification sample of the power distribution network to obtain a discretized sample;
the model parameter determining module is used for acquiring parameters of the probability map model according to the discretized sample based on a pre-established probability map model; the probability map model is a two-time slice probability map model;
the power distribution network dynamic parameter generation module is used for acquiring power distribution network dynamic parameters based on a confidence coefficient propagation algorithm according to the observation variables and the probability map model after the parameters are acquired.
The beneficial effects are that:
1. according to the method, the influence and the factor of the dynamic parameters of the power distribution network are analyzed according to the data statistics and the priori knowledge, the problem that the dynamic parameters cannot be acquired in a part of power distribution areas is solved, the parameter identification precision under the condition of abrupt change of the operation mode of the part of power distribution areas or abrupt change of the external environment is improved, the operation state of the power distribution network is mastered and analyzed by power distribution network dispatching personnel, and the final identification result can provide a good basis for upper-layer application of a power distribution automation system;
2. the invention carries out scientific analysis on the collected operation data and meteorological data of the power distribution network, provides a basis for situation awareness of the power distribution network, is beneficial to realizing a panoramic visible and controllable power distribution network, and is beneficial to providing more reliable, safe and economic electric energy for power grid companies.
Drawings
FIG. 1 is a flowchart of a method for identifying dynamic parameters of a power distribution network according to an embodiment of the present invention;
FIG. 2 is a diagram of a dynamic Bayesian model structure for line physical parameter identification in accordance with an embodiment of the present invention;
FIG. 3 is a flow chart of raw data preprocessing in accordance with an embodiment of the present invention;
FIG. 4 is a schematic diagram of reasoning about dynamic parameters of a power distribution network based on a belief propagation algorithm according to an embodiment of the present invention;
fig. 5 is a simplified topology model of a medium voltage distribution network used in simulation experiments in accordance with an embodiment of the present invention.
Detailed Description
The invention is further described below in connection with specific embodiments. The following examples are only for more clearly illustrating the technical aspects of the present invention, and are not intended to limit the scope of the present invention.
Fig. 1 is a flowchart of a dynamic parameter identification method for a power distribution network according to an embodiment of the present invention, as shown in fig. 1, the method includes the following steps:
step 1, preprocessing the collected operation data and external environment data of the power distribution network to generate a dynamic parameter identification sample of the power distribution network.
According to one embodiment, raw data may be preprocessed in the form of a DataFrame using Pandas.
The data required by the data-driven line physical parameter identification comprises power distribution network operation data and external environment data: the operation data of the power distribution network come from the intelligent ammeter acquisition data, including node voltage, current, active power and reactive power acquired once every 15 min; external environmental data was from 58238 weather station data (supplied by meteomanz. Com), including regional temperature and humidity collected once every 3 hours. The original data mainly has three problems, namely, the original data come from different data sources, and the data in the different data sources need to be combined and integrated in one data frame; secondly, the dimension of the data needs to be reduced, and the original data has too many attributes, which is not beneficial to data modeling; thirdly, the data has missing values and outliers, and data cleaning is needed.
To solve the above problem, in one embodiment, the following operations may be taken with respect to the original data:
performing secondary spline interpolation on the external environment data to ensure that the frequencies of the external environment data with different data sources are identical;
merging the data of different data sources, and eliminating redundant fields in the data;
and eliminating repeated data in the operation data of the power distribution network and carrying out data emptying.
In a specific example, as shown in fig. 3, further use may be made of:
taking into account the mismatch of the acquisition frequency of two types of data from different data sources, inserting data point complement data into every two external environment data by using a spline interpolation method, wherein the spline interpolation method is characterized in that interpolation is carried out on the time sequence of temperature and humidity by using an inter 1d function in a python scipy module package, and the weather history data are complemented to match the frequencies of the two data sources;
merging data from different data sources through a merge function in a pandas module, and eliminating redundant fields in original data by using a drop function;
repeatedly processing the data blank and the data; eliminating repeated data in the operation data by utilizing a drop_redundant function; and detecting the missing proportion of the variable by using pandas.isnull.sum (), and under the condition that the missing rate is lower (less than 95%) and the importance is lower, carrying out data nulling by using a dropna function to finally obtain a cleaned dynamic parameter identification data sample of the power distribution network. The structure of the samples is shown in table 1 below.
Table 1 distribution line parameter identification data samples
And 2, discretizing the dynamic parameter identification sample of the power distribution network to obtain a discretized sample.
In one embodiment, the data samples may be discretized using a maximum entropy algorithm, the calculation formula of which is as follows:
wherein Z is hiddenVariable, m is the number of divided discrete intervals, N count (z=s) is the number of samples of the variable in state s in the data; n (N) amount (Z) is the total number of samples. Each variable in the network is distributed to the respective state space according to the condition of maximum mutual information entropy, and the mutual information entropy between the states is maximum when the number of samples contained in each discrete interval is the same without priori knowledge. The granularity of discretization was chosen to be 10% of the number of samples, and the result of discretization for each variable is shown in table 2 below.
TABLE 2 variable discretization results at 10% discrete particle size
And step 3, acquiring parameters of the probability map model according to the discretized sample based on a pre-established probability map model.
The probability map model is a two-time slice probability map model.
The probability map model is pre-established, and according to one embodiment, the step of establishing the probability map model may include:
selecting the temperature T at T t Humidity H t Voltage drop DeltaV of feeder line segment t Feeder segment transmission power S t As an observation variable in the probability map model, the impedance Z of the line at that moment is selected t As hidden variables of the probability map model;
according to the causal relationship between each observed variable and hidden variable in a single time slice, each variable is added into a probability map model one by one, and a static Bayesian network is constructed;
setting an initial time slice, and designating prior probability distribution of variables under the time slice;
and designating the causal relationship of each state between adjacent time slices, and constructing a transition model.
The model building process is further described below.
In particular, consideration of dynamic parameters of a power distribution network is related to external environment and power distribution network operation stateAccording to the relation between these variables and line impedance parameters, the temperature T at T time is selected t Humidity H t Voltage drop DeltaV of feeder line segment t Feeder segment transmission power S t The expression of the observation variable X as the probability map model, namely the probability map model at the time t, is as follows:
X t ＝{T t ,H t ,ΔV t ,S t }
considering that the power distribution network line is relatively short, the application ignores the capacitance to ground of the line when constructing the probability map model. Thus hidden variable Y of probability map model at time t t For impedance of the line at that moment, Z is used t And (3) representing.
After the random variables of the model are determined, the order of the variables is selected according to causal relationships. T representing external environment t And H t Are all influencing factors of the line impedance and are therefore the line impedance Z t And a parent node. The expression of the line node voltage drop is as follows:
wherein V is 1 ，V 2 Representing the voltage amplitude difference between node 1 and node 2, P, Q are the active power and reactive power flowing between node 1 and node 2, respectively, R, X are the resistance and reactance of the lines connecting node 1 and node 2, respectively, I R ，I X Representing the corresponding active and reactive currents, respectively. In addition S t Characterizing apparent power; z is Z t Is the line impedance. From the above, it can be seen that the apparent power and line impedance are the influencing factors of the voltage drop, and therefore they are the line voltage drop DeltaV t Is a parent node of (c). And finally, starting from an empty graph, adding the variables into the probability graph model one by one according to the dependency relationship among the variables to form a directed acyclic graph.
On the basis, an initial time slice is selected, prior probability distribution of variables under the time slice is designated, causal relations of states between adjacent time slices are designated, a transfer model is built, and the construction of a dynamic parameter identification model of the power distribution network based on the two-time slice probability map model is completed. And obtaining the two-time-slice probability map model for identifying the dynamic parameters of the power distribution network in fig. 2.
According to one embodiment, the parameters of the probability map model may be obtained by:
based on the probability quality function, obtaining an initial probability distribution table of each variable in the discretized sample;
calculating a conditional probability distribution table among variables according to the discretization sample based on a maximum expected algorithm;
counting a transition probability distribution table of each variable from the time t to the time t+1 from continuous data samples on a time axis;
the correctness of the conditional probability distribution table is determined by checking whether the sum of the probability distributions of each variable is 1, checking whether the conditional probability distribution is consistent with causal relationships in the bayesian network.
In one embodiment, the parameters of the probability map model may be obtained in the form of an initial probability distribution table, a conditional probability distribution table, and a transition probability distribution table:
obtaining an initial probability distribution table of each variable from the discretized sample by calculating a probability mass function, wherein the structure of the initial probability distribution table is shown in the following table 3;
TABLE 3 initial probability distribution Table
The sum of all elements in the initial probability vector is 1, each element P i Obtained by calculating a probability mass function (Probability Mass Function, PMF):
wherein the method comprises the steps ofIs hidden and changedThe probability that the quantity is initially in state s; n (N) count (z=s) is the number of samples of the hidden variable in the data at state s; n (N) amount (Z) is the total number of samples.
The conditional probability distribution table between variables is calculated from the discretized samples by the maximum expectation algorithm, and in the case where the state number of hidden variables is m and the state numbers of observed variables are n, k, v and h, respectively, the conditional probability distribution table is a matrix of m× (n×k×v×h), and the structure thereof is shown in table 4 below.
TABLE 4 conditional probability distribution Table
The conditional probability distribution may be obtained by EM algorithm. The EM algorithm first initializes the probability distribution and then iterates in two steps until convergence. The two-step iterative process is as follows:
1) Step E calculation (estimation Step): calculating posterior probability of the hidden variable according to the initial value of the conditional probability or the conditional probability obtained by the previous iteration, and taking the posterior probability as the current expected value of the hidden variable:
P posterior (Z)＝P(Z|X；θ cpt )
wherein P is posterior (Z) posterior probability of hidden variable, θ cpt Is a parameter of a conditional probability distribution table in a DBN (deep belief network ).
2) M steps calculation (Maximization Step): updating the conditional probability distribution table with the maximization of likelihood function as the target:
wherein m is the number of hidden variable states, P (X, Z; theta) cpt ) Expectations for hidden variables obtained from samples. And when the probability of the training data sample is maximum according to the conditional probability distribution table in the probability graph model, the iteration of the maximum expected algorithm is ended. The maximum expected algorithm can carry out maximum likelihood estimation on parameters from the incomplete data set, and is suitable for probability map model conditional probability distribution calculation under the condition of power distribution network acquisition data missing.
And counting a transition probability distribution table of each variable from the time t to the time t+1 from continuous data samples on a time axis. The transition probability distribution is a parameter representing the timing transition of the variable in the DBN and can be calculated by the following formula:
wherein the method comprises the steps ofThe probability of transition of the hidden variable Z from the state 1 to the state 2 from t-1 to t time is represented; n (N) count (s 1 ,s 2 ) Representing the times of transition from the state 1 to the state 2 of the hidden variable Z from t-1 to t in the acquired historical data; n (N) amount (s 1 ) Representing the number of samples of the acquired historical data in which the hidden variable Z is in state 1.
Finally, the correctness of the derived parameters can be checked by checking if the sum of the probability distributions of each variable is 1, and if the conditional probability distribution is consistent with causal relationships in the bayesian network.
And 4, acquiring dynamic parameters of the power distribution network based on a confidence coefficient propagation algorithm according to the observation variables and the probability map model after acquiring the parameters.
According to one embodiment, a belief propagation algorithm is used to infer the power distribution network dynamic parameters where the observed variables are known, as shown in fig. 4, which includes:
1) Initializing probability distribution of each variable according to the sample;
2) A certain state variable Y in the network is randomly selected, and the confidence level of the node can be expressed as b (Y t ) And the confidence level is matched with the adjacent node and all the nodes passing through the adjacent nodeInformation m that edge passes to the node xY (Y t ) In direct proportion, the confidence of a node can be replaced with a probability:
wherein phi (Y) t ,X t ) And G is a first-order neighborhood of the node, namely a set of all nodes adjacent to the node. m is m xY (Y t ) The message passed to node Y for node x indicates the effect that node x has on node Y at time t.
3) Updating information between variables:
wherein ψ (Y) t ,Y t-1 ) The potential energy between the nodes from the t-1 moment to the t moment at the node Y is used for reflecting the compatibility between hidden variables.
4) Continuously repeating the steps 2) and 3) to perform continuous iteration of message propagation and confidence coefficient updating until the convergence condition is met:
b (n) (Y t )-b (n-1) (Y t )＜10 -5
5) And taking the confidence level of the final hidden variable as the inferred result of probability distribution of the hidden variable in each state interval. The final reasoning result of the DBN is a probability distribution, and the DBN model can provide all the situations which can occur at the moment and the occurrence probability thereof compared with single-point parameter identification. For comparison with the conventional single-point line parameter identification model, the final single-point identification result of the line impedance parameter is obtained from the sample { Z ] in the historical data in the state interval 1 ,Z 2 ,…,Z N And (3) calculating, wherein root mean square can be used as a point identification result of the line parameters.
Wherein N is the number of samples in the same state as the identification result in the historical data, Z i For the impedance value of the sample,the root mean square of the impedance values of these samples. The parameter identification results of the 14 lines using two 10kV feeders connected by a tie switch in fig. 5 as the parameter identification objects are shown in table 5 below.
TABLE 5 dynamic parameter identification results for power distribution networks
Because the lengths of the lines are different, the average error rate of the parameter identification of each line is taken as an evaluation standard, the average error rate of the impedance parameter identification of the proposed probability map model is 3.80%, and the average error rate of the reactance parameter identification is 9.05%.
In another embodiment, the present invention provides a power distribution network dynamic parameter identification device, including:
the data preprocessing module is used for preprocessing the collected operation data of the power distribution network and the external environment data to generate a dynamic parameter identification sample of the power distribution network;
the discretization processing module is used for discretizing the dynamic parameter identification sample of the power distribution network to obtain a discretized sample;
the model parameter determining module is used for acquiring parameters of the probability map model according to the discretized sample based on a pre-established probability map model; the probability map model is a two-time slice probability map model;
the power distribution network dynamic parameter generation module is used for acquiring power distribution network dynamic parameters based on a confidence coefficient propagation algorithm according to the observation variables and the probability map model after the parameters are acquired.
In summary, the invention constructs a novel power distribution network dynamic parameter identification probability graph model to solve the problem of power distribution network parameter identification errors caused by operation condition changes and data errors. The problem of uncertainty existing in the identification process of the dynamic parameters of the power distribution network is solved by utilizing the knowledge in the probability theory field, and accurate line physical parameters are provided for situation awareness and line loss calculation of the power distribution network. The model provided by the invention can improve the accuracy and robustness of the distribution line impedance parameter identification, improves the intelligent degree of distribution network analysis and management, and provides a parameter basis for distribution network dispatcher to master, analyze and control the distribution network operation mode.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The present invention has been disclosed in the preferred embodiments, but the invention is not limited thereto, and the technical solutions obtained by adopting equivalent substitution or equivalent transformation fall within the protection scope of the present invention.
Claims (8)
1. The dynamic parameter identification method for the power distribution network is characterized by comprising the following steps of:
preprocessing the collected operation data of the power distribution network and the external environment data to generate a dynamic parameter identification sample of the power distribution network;
discretizing a dynamic parameter identification sample of the power distribution network to obtain a discretized sample;
based on a pre-established probability map model, acquiring parameters of the probability map model according to a discretization sample, wherein the probability map model is a two-time-slice probability map model;
acquiring dynamic parameters of the power distribution network based on a confidence coefficient propagation algorithm according to the observation variables and the probability graph model after the parameters are acquired;
the acquiring the parameters of the probability map model comprises the following steps:
based on the probability quality function, obtaining an initial probability distribution table of each variable in the discretized sample;
calculating a conditional probability distribution table among variables according to the discretization sample based on a maximum expected algorithm;
counting a transition probability distribution table of each variable from the time t to the time t+1 from continuous data samples on a time axis;
checking whether the probability distribution of each variable is 1 or not, checking whether the conditional probability distribution is consistent with the causal relationship in the Bayesian network, and determining the correctness of the conditional probability distribution table;
the maximum expectation-based algorithm calculates a conditional probability distribution table among variables according to a discretization sample, and comprises the following steps:
the posterior probability of the hidden variable is calculated according to the initial value of the conditional probability or the conditional probability obtained by the previous iteration, and is used as the current expected value of the hidden variable, and the expression is as follows:
P posterior (Z)＝P(Z|X；θ cpt )
wherein Z is a hidden variable, P posterior (Z) posterior probability of hidden variable, θ cpt The method is characterized in that the method is a conditional probability distribution table in a probability graph model, and X is an observation variable;
updating a conditional probability distribution table with a maximum likelihood function as a target, wherein the conditional probability distribution table has the expression:
wherein m is the number of hidden variable states, P (X, Z; theta) cpt ) Is the expectation of the hidden variable obtained from the sample;
and when the probability of the training data sample is maximum according to the conditional probability distribution table in the probability graph model, the iteration of the maximum expected algorithm is ended.
2. The method of claim 1, wherein the step of establishing the probability map model comprises:
selecting the temperature, the humidity, the voltage drop of the feeder line section and the transmission power of the feeder line section at one moment as observation variables in the probability map model, and selecting the impedance of the line at the moment as hidden variables of the probability map model;
according to the causal relationship between each observed variable and hidden variable in a single time slice, each variable is added into a probability map model one by one, and a static Bayesian network is constructed;
setting an initial time slice, and designating prior probability distribution of variables under the time slice;
and designating the causal relationship of each state between adjacent time slices, and constructing a transition model.
3. The method of claim 1, wherein generating the power distribution network dynamic parameter identification sample comprises:
performing secondary spline interpolation on the external environment data to ensure that the frequencies of the external environment data with different data sources are identical;
merging the data of different data sources, and eliminating redundant fields in the data;
and eliminating repeated data in the operation data of the power distribution network and carrying out data emptying.
4. The method according to claim 1, wherein the discretizing the distribution network dynamic parameter identification sample is calculated as follows:
wherein Z is hidden variable, m is the number of divided discrete intervals, N count (z=s) is the number of samples of the hidden variable in the data at state s; n (N) amount (Z) is the total number of samples.
5. The method of claim 1, wherein the probability mass function is expressed as:
wherein the method comprises the steps ofThe probability of being initially in state s for the hidden variable; n (N) count (z=s) is the number of samples of the hidden variable in the data at state s; n (N) amount (Z) is the total number of samples.
6. The method of claim 1, wherein the transition probability distribution table is expressed as:
wherein the method comprises the steps ofThe probability of transition of the hidden variable Z from the state 1 to the state 2 from t-1 to t time is represented; n (N) count (s 1 ,s 2 ) Representing the times of transition from the state 1 to the state 2 of the hidden variable Z from t-1 to t in the acquired historical data; n (N) amount (s 1 ) Representing the number of samples of the acquired historical data in which the hidden variable Z is in state 1.
7. The method of claim 1, wherein the obtaining the dynamic parameters of the power distribution network based on the belief propagation algorithm comprises:
initializing probability distribution of each variable according to the sample;
a certain state variable Y in the network is randomly selected, and the confidence of the node is replaced by b (Y t )：
Wherein phi (Y) t ,X t ) For the likelihood function between the corresponding state variable Y and the observed variable X at the moment t, the joint compatibility of the node Y at the moment t is represented, G is the first-order neighborhood of the node Y, m xY (Y t ) A message passed to node Y for node x;
updating information between variables:
wherein ψ (Y) t ,Y t-1 ) The potential energy between the nodes from the t-1 moment to the t moment at the node Y is obtained;
until the convergence condition is satisfied:
b (n) (Y t )-b (n-1) (Y t )＜10 -5
confidence b (Y) t ) As a result of the inference of probability distribution of hidden variables in each state interval.
8. The utility model provides a distribution network dynamic parameter discernment device which characterized in that includes:
the data preprocessing module is used for preprocessing the collected operation data of the power distribution network and the external environment data to generate a dynamic parameter identification sample of the power distribution network;
the discretization processing module is used for discretizing the dynamic parameter identification sample of the power distribution network to obtain a discretized sample;
the model parameter determining module is used for acquiring parameters of the probability map model according to the discretized sample based on a pre-established probability map model; the probability map model is a two-time slice probability map model;
the power distribution network dynamic parameter generation module is used for acquiring power distribution network dynamic parameters based on a confidence coefficient propagation algorithm according to the observation variables and the probability map model after the parameters are acquired;
the acquiring the parameters of the probability map model comprises the following steps:
based on the probability quality function, obtaining an initial probability distribution table of each variable in the discretized sample;
calculating a conditional probability distribution table among variables according to the discretization sample based on a maximum expected algorithm;
counting a transition probability distribution table of each variable from the time t to the time t+1 from continuous data samples on a time axis;
checking whether the probability distribution of each variable is 1 or not, checking whether the conditional probability distribution is consistent with the causal relationship in the Bayesian network, and determining the correctness of the conditional probability distribution table;
the maximum expectation-based algorithm calculates a conditional probability distribution table among variables according to a discretization sample, and comprises the following steps:
the posterior probability of the hidden variable is calculated according to the initial value of the conditional probability or the conditional probability obtained by the previous iteration, and is used as the current expected value of the hidden variable, and the expression is as follows:
P posterior (Z)＝P(Z|X；θ cpt )
wherein Z is a hidden variable, P posterior (Z) posterior probability of hidden variable, θ cpt The method is characterized in that the method is a conditional probability distribution table in a probability graph model, and X is an observation variable;
updating a conditional probability distribution table with a maximum likelihood function as a target, wherein the conditional probability distribution table has the expression:
wherein m is the number of hidden variable states, P (X, Z; theta) cpt ) Is the expectation of the hidden variable obtained from the sample;
and when the probability of the training data sample is maximum according to the conditional probability distribution table in the probability graph model, the iteration of the maximum expected algorithm is ended.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011309392.1A CN112559963B (en) | 2020-11-20 | 2020-11-20 | Dynamic parameter identification method and device for power distribution network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011309392.1A CN112559963B (en) | 2020-11-20 | 2020-11-20 | Dynamic parameter identification method and device for power distribution network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112559963A CN112559963A (en) | 2021-03-26 |
| CN112559963B true CN112559963B (en) | 2024-02-06 |
Family
ID=75044678
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011309392.1A Active CN112559963B (en) | 2020-11-20 | 2020-11-20 | Dynamic parameter identification method and device for power distribution network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112559963B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113158309B (en) * | 2021-04-09 | 2022-08-19 | 天津大学 | Heating and ventilation equipment operation strategy identification method |
| CN114696316A (en) * | 2022-03-03 | 2022-07-01 | 国网江苏省电力有限公司电力科学研究院 | Distribution network parameter identification method and device considering probability distribution |
| CN115464237A (en) * | 2022-08-19 | 2022-12-13 | 北京理工大学 | Real-time control method, system and device for electric arc additive manufacturing equipment |
| CN117520869B (en) * | 2024-01-04 | 2024-03-29 | 国网江苏省电力有限公司苏州供电分公司 | Method and system for identifying parameter interval of medium-voltage distribution network based on dynamic Bayesian network |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109933953A (en) * | 2019-04-11 | 2019-06-25 | 东南大学 | Composition of Switching State of Distribution Network discrimination method based on probability graph model |
| CN111262243A (en) * | 2020-03-04 | 2020-06-09 | 国网浙江省电力有限公司 | Intelligent identification and optimization method for operation mode of park power distribution system |
| CN111381498A (en) * | 2020-03-09 | 2020-07-07 | 常熟理工学院 | Expectation maximization identification method of multi-sensor based on multi-rate variable time-lag state space model |
-
2020
- 2020-11-20 CN CN202011309392.1A patent/CN112559963B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109933953A (en) * | 2019-04-11 | 2019-06-25 | 东南大学 | Composition of Switching State of Distribution Network discrimination method based on probability graph model |
| CN111262243A (en) * | 2020-03-04 | 2020-06-09 | 国网浙江省电力有限公司 | Intelligent identification and optimization method for operation mode of park power distribution system |
| CN111381498A (en) * | 2020-03-09 | 2020-07-07 | 常熟理工学院 | Expectation maximization identification method of multi-sensor based on multi-rate variable time-lag state space model |
Non-Patent Citations (1)
| Title |
|---|
| Distribution line parameter estimation considering dynamic operating states with a probabilistic graphical model;Wei Jiang, et al.;Electrical Power and Energy Systems;第2-8页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112559963A (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112559963B (en) | Dynamic parameter identification method and device for power distribution network | |
| López et al. | Parsimonious short-term load forecasting for optimal operation planning of electrical distribution systems | |
| Deng et al. | A missing power data filling method based on improved random forest algorithm | |
| Fioriti et al. | A novel stochastic method to dispatch microgrids using Monte Carlo scenarios | |
| CN109002933B (en) | Distribution line route variable relation model optimization method based on Relieff and t-SNE | |
| CN107104442B (en) | Method for calculating probability load flow of power system including wind power plant by considering parameter ambiguity | |
| CN109933953B (en) | Power distribution network switch state identification method based on probability map model | |
| Cheng et al. | Enhanced state estimation and bad data identification in active power distribution networks using photovoltaic power forecasting | |
| Raptis et al. | Total power quality index for electrical networks using neural networks | |
| CN106022614A (en) | Data mining method of neural network based on nearest neighbor clustering | |
| CN110348114B (en) | Non-precise fault identification method for power grid completeness state information reconstruction | |
| CN104573876A (en) | Wind power plant short-period wind speed prediction method based on time sequence long memory model | |
| CN112734128A (en) | 7-day power load peak value prediction method based on optimized RBF | |
| CN109993665B (en) | Online safety and stability assessment method, device and system for power system | |
| CN115508770B (en) | KL-NB algorithm-based electric energy meter operation state online evaluation method | |
| LU500551B1 (en) | Virtual load dominant parameter identification method based on incremental learning | |
| CN112379325A (en) | Fault diagnosis method and system for intelligent electric meter | |
| CN115062864A (en) | Method and system for decomposing-integrating prediction of medium-term load of power distribution network | |
| CN110826237A (en) | Bayesian belief network-based wind power equipment reliability analysis method and device | |
| CN113937764A (en) | Low-voltage distribution network high-frequency measurement data processing and topology identification method | |
| CN111654392A (en) | Low-voltage distribution network topology identification method and system based on mutual information | |
| Wang et al. | Transmission network dynamic planning based on a double deep-Q network with deep ResNet | |
| CN114065634A (en) | Data-driven power quality monitoring and stationing optimization method and device | |
| Guo et al. | On-line prediction of transient stability using decision tree method—Sensitivity of accuracy of prediction to different uncertainties | |
| CN111864728B (en) | Important equipment identification method and system for reconfigurable power distribution network |
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 |