A method and system for loss allocation of a power distribution network containing distributed power sources

Abstract

The application discloses a kind of power distribution network containing distributed power supply loss allocation method and system. Including obtaining data information link table according to the node topological graph power flow result under the total grid connection of distributed power supply;According to the data information link table, the power contribution coefficient matrix of each pure power supply and / or each pure load to each node is calculated / or power consumption coefficient matrix;According to the proportional sharing principle, the occupation coefficient matrix of each distributed power supply and / or load to the power of each line is obtained;According to the active power loss of line in data information link table and the occupation coefficient matrix of each distributed power supply and / or load to each branch, the total amount of power distribution network distributed power supply and / or load loss allocation is obtained.The application can be applied to complex structure, transaction subject variety, line connection variety Large power grid, can significantly improve its network loss allocation calculation accuracy.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power grid loss allocation, and particularly relates to a power grid loss allocation method and system for a distribution network containing distributed power supply. BACKGROUND

[0002] After the distributed power supply (DG) is connected to the distribution network, the power flow direction and the network loss of the distribution network are changed. The influence degree of the distribution network loss is affected by multiple factors such as the grid-connected capacity, grid-connected position, grid-connected penetration rate and grid-connected operation mode of the distributed power supply. At present, the distributed power supply does not bear the corresponding network loss proportion in the power flow and transmission of the distribution network, so that the power supply party and the user bear the excess network loss, which also leads to that the distributed power supply cannot share the economic benefits brought by the grid connection and cannot bring a good reference direction for the economic planning and operation of the distribution network. Therefore, it is of great significance to study the network loss allocation caused by the grid connection of the distributed power supply. However, the influence of the distributed power supply on the network loss still lacks accurate measuring means to evaluate the contribution degree of the distributed power supply to the network loss of the distribution network.

[0003] In recent years, there have been many research results on the network loss allocation problem of the distribution network containing distributed power supply. The marginal network loss coefficient method (MLC) considers the network power flow, calculates the network loss allocation of each node according to the network loss change caused by the power change of each load or generator node. This allocation method can reflect the marginal network loss cost caused by each node to the network and can provide certain economic signals. However, this method is only suitable for networks with small impedance ratio (R / X), and the allocation results may be unbalanced, so a correction coefficient needs to be introduced to correct the network loss allocation of each node.

[0004] The proportional allocation method (Pro-rata method) allocates the network loss to each load or generator according to the power level of the load or generator in a certain proportion. The calculation is simple, but since this method does not consider the network structure, line distance and node position, the allocation results cannot provide correct economic signals. The branch current decomposition method (BCDM) based on circuit theory calculates the network loss allocation of each node according to the branch current connected to the root node, which establishes a connection between the branch network loss and the node current, and can accurately calculate the network loss allocation of the traditional distribution network. However, for the distribution network containing distributed power supply, since this method allocates the network loss change to both the load and the DG, it may cause cross-subsidies in space and time, and further lead to unfair allocation results.

[0005] Z-BUS method is a kind of allocation method based on power flow analysis and admittance matrix, which can well reflect the influence of network topology on network loss, but this allocation method cannot flexibly select loss allocation nodes, and the network loss allocation of pure generator nodes is often much higher than that of other nodes, and the allocation result is unfair;The edge method obtains the incremental network loss coefficient from the power flow to allocate the network loss to the network users and market participants.

[0006] Due to the high nonlinearity of the DG-containing distribution network loss allocation problem, the above allocation methods cannot provide a closed solution, and the traditional power flow tracking method uses the proportional allocation principle and does not fundamentally explain the path information of generator and load power flow from the circuit theory and power flow principle.In addition, the network loss rate is one of the important evaluation indexes of power grid enterprises, and the input of DG will increase the network loss management burden of power grid enterprises, reduce the equipment utilization rate of power grid network, and cause the economic interests of power supply and use enterprises to be damaged.Therefore, it has important research value for the economic good operation of power supply and use enterprises to study a fair and reasonable DG-containing distribution network loss allocation method. SUMMARY

[0007] In view of the unreasonable problem of DG allocation network loss in the existing DG-containing distribution network, the application provides a knowledge graph and network topology-based active distribution network loss allocation method and system, which considers the structural characteristics and calculation parameters of DG integration into the distribution network, can improve the calculation accuracy of the distribution network loss allocation and ensure the objectivity of the DG participation in the network loss allocation result, and the research method is beneficial to the economic planning and operation of the DG-containing distribution network.

[0008] A knowledge graph and network topology-based active distribution network loss allocation method for one of the purposes of the application, comprising the following steps:

[0009] S1, obtaining a data information linked list according to the node topology graph under the condition that all distributed power sources are connected to the grid, wherein the data information linked list includes distributed power source power, power consumption load power, branch power, branch active power loss, node and branch number;

[0010] S2, searching for pure power source nodes and / or pure load nodes in the data information linked list, calculating the power contribution coefficient matrix of each pure power source node to each node and / or the power extraction coefficient matrix of each pure load node to each node in the node topology graph according to the data information linked list;The pure power source node is a node with only power;The pure load node is a node with only load;According to the power contribution coefficient matrix of each pure power source node to each node and / or the power extraction coefficient matrix of each pure load node to each node, the occupation coefficient matrix of each pure power source node and / or each pure load node to each line power is obtained;

[0011] S3, multiplying the line active loss in the data information chain table with the occupation coefficient matrix of each pure power node and / or each pure load node to each line power to obtain the total amount of network loss allocation borne by each pure power node and / or each pure load node in the distribution network.

[0012] Further technical solutions include that the step S1 further includes the following steps:

[0013] S101, selecting a distribution network loss abnormality recognition ontology, entity and attribute based on the electrical parameters in the data information chain table;

[0014] S102, extracting a feature representing network loss abnormality recognition from the selected distribution network loss abnormality recognition ontology, entity and attribute;

[0015] S103, establishing a network loss abnormality judgment rule according to the feature of network loss abnormality recognition;

[0016] S104, constructing a distribution network loss abnormality recognition knowledge graph according to the network loss abnormality judgment rule;

[0017] S105, identifying a network loss abnormal line according to the distribution network loss abnormality recognition knowledge graph; and performing a checking process on the identified network loss abnormal line.

[0018] Further technical solutions include that in the step S101, the distribution network loss abnormality recognition ontology, entity and attribute represent different attributes of the same type of entity in each row, and the data form representing the entity type is saved in a source database, constituting unstructured data of network loss abnormality features; and the knowledge and its corresponding relationship are extracted in the form of triples, constituting a distribution network loss abnormality knowledge graph prototype;

[0019] In the construction method of the above distribution network loss abnormality knowledge graph prototype, the entity involved represents a certain type of specific physical device, such as a line, and the corresponding associated attribute is, for example, current, and the form of each entity is represented as (entity | entity corresponding identifier, associated attribute set of the identified entity); in the step S101, the method for constituting a primary knowledge graph of the distribution network according to the triple form is to extract knowledge and its corresponding relationship in the form of “entity-attribute-attribute value” from the source database to form a knowledge graph triple, and the source database field saving method can be used to constitute a distribution network knowledge graph prototype.

[0020] Further technical solutions include that in the step S102, the method for extracting the feature representing network loss abnormality recognition includes:

[0021] According to the word segmentation algorithm, the unstructured data representing the network loss abnormal characteristics are processed, the TF-IDF algorithm is used to analyze the weight of the entities and the abnormal forms corresponding to the network loss contained in the word segmentation results, and the entities and the corresponding abnormal characteristics are selected according to the weight size to represent the network loss abnormal characteristics.

[0022] The word segmentation algorithm is used to process the unstructured data representing the distribution network loss abnormal characteristics, and the TF-IDF algorithm is used to extract the entities and the network loss abnormal characteristics with a proportion greater than a set value in the distribution network loss abnormal reason.

[0023] The further technical solution includes: in the step S103, the network loss abnormal judgment rule is established according to the features recognized by the network loss abnormality and the industry common sense.

[0024] The further technical solution includes: in the step S104, the method for constructing the distribution network loss abnormal recognition knowledge graph according to the network loss abnormal judgment rule includes: establishing a triple relationship between the entities and the abnormal characteristics with a proportion of line loss abnormal characteristics greater than a set value according to the obtained network loss abnormal characteristics, combining the distribution network loss abnormal knowledge graph prototype to form a distribution network line loss abnormal knowledge graph; setting the distribution network loss abnormal judgment rule with an abnormal characteristic proportion greater than a set value, arranging it into a triple form according to the sequence of the judgment rule, and then forming the network loss abnormal recognition knowledge graph.

[0025] The further technical solution includes: in the step S105, the method for identifying the network loss abnormal line according to the distribution network loss abnormal recognition knowledge graph includes: sorting the line network loss abnormality possibility of the distribution network loss abnormal knowledge graph according to the K-means clustering algorithm, searching for the line with abnormal distribution network loss value according to the size order of the network loss abnormality possibility, and performing manual checking processing.

[0026] The further technical solution includes: in the step S2, the calculation method of the power contribution coefficient in the power contribution coefficient matrix includes:

[0027]

[0028] In the formula:

[0029] Pj,g represents the power contribution coefficient of the power supply g on the node j, and the node j is the end node of the line k; Pj,g represents the power contribution coefficient of the power supply g on the node j, and the node j is the end node of the line k;

[0030] Lk represents the active power flowing through the line k;

[0031] Nj represents the total active power flowing into the node j, and the node j is the end node of the line k;

[0032] k∈-j represents the branch associated with node j.

[0033] The further technical solution includes: in the step S2, for the searched pure power supply node and / or pure load node, the subsequent data information chain table search is no longer participated.

[0034] The further technical solution includes: the node topology graph under the condition that all the distributed power supplies are connected to the grid includes a distributed power supply four-node topology graph, and for the distributed power supply four-node topology graph, the calculation method of the power contribution coefficient of each pure power supply node to each node includes:

[0035]

[0036] In the formula:

[0037] B represents the power contribution coefficient matrix of each pure power supply node to each node;

[0038] Pa, Pb, Pc, Pd, Pe represent the active power flowing through all lines a, b, c, d and e in the distributed power supply four-node topology graph;

[0039] PG1, PG2 represent the active power flowing into all power supplies G1 and G2.

[0040] The further technical solution includes: the node topology graph under the condition that all the distributed power supplies are connected to the grid includes a distributed power supply four-node topology graph, and in the distributed power supply four-node topology graph, the calculation method of the occupation coefficient matrix of each pure power supply node to each line power includes:

[0041]

[0042] In the formula:

[0043] β represents the occupation coefficient matrix of each pure power supply node to each line power;

[0044] Pa, Pb, Pc, Pd, Pe represent the active power flowing through all lines a, b, c, d and e in the distributed power supply four-node topology graph; PG1, PG2 represent the active power flowing into all power supplies G1 and G2.

[0045] The further technical solution includes: the node topology graph under the condition that all the distributed power supplies are connected to the grid includes a distributed power supply four-node topology graph, and in the distributed power supply four-node topology graph, the calculation method of the total network loss allocation amount of each pure power supply node in the distribution network includes:

[0046]

[0047] In the formula:

[0048] P a : represents the active power flowing into line a;

[0049] P G1,loss , P G2,loss : represents the total amount of network loss allocation undertaken by the pure power node where all power sources G1 and G2 are located;

[0050] P a,loss , P b,loss , P c,loss , P d,loss , P e,loss : represents the loss of all lines a, b, c, d, e;

[0051] P G1 , P G2 : represents the active power flowing into all power sources G1 and G2.

[0052] Further technical solutions include: the node topology graph under the condition that all distributed power sources are grid-connected includes a distributed power source four-node topology graph, and in the distributed power source four-node topology graph, the calculation method of the power draw coefficient matrix of each pure load node to each node includes:

[0053]

[0054] In the formula:

[0055] δ: represents the power draw coefficient matrix of each pure load node to each node in the distributed power source four-node topology graph;

[0056] P a , P b , P c , P d , P e : represents the active power flowing through all lines a, b, c, d, e in the distributed power source four-node topology graph;

[0057] P L3 , P L4 : represents the active power flowing through all loads L3 and L4 in the distributed power source four-node topology graph.

[0058] Further technical solutions include: in the distributed power source four-node topology graph, the calculation method of the total amount of network loss allocation undertaken by each pure load node in the distribution network includes:

[0059]

[0060] In the formula:

[0061] P a,loss , P b,loss , P c,lossP d,loss P e,loss : represents the loss of all lines a, b, c, d, e;

[0062] P d P e : represents the active power flowing through all lines d, e in the four-node topology diagram of the distributed power supply;

[0063] P L3,loss P L4,loss : represents the total amount of network loss allocation borne by the pure load nodes where the loads L3 and L4 are located in the four-node topology diagram of the distributed power supply;

[0064] P L3 P L4 : represents the active power flowing through all loads L3 and L4 in the four-node topology diagram of the distributed power supply.

[0065] A network loss allocation system for a distribution network containing a distributed power supply to achieve the second purpose of the application, comprising a data acquisition module, a power contribution coefficient calculation module, an occupation coefficient calculation module, and a total network loss calculation module;

[0066] The data acquisition module is used to obtain a data information linked list according to the power flow results of the node topology diagram under the condition that all the distributed power supplies are connected to the grid, and the data information linked list includes distributed power supply power, power of electrical load, branch power, branch active loss, node and branch number;

[0067] The power contribution coefficient calculation module is used to search for each pure power supply node and / or each pure load node in the data information linked list, and calculate the power contribution coefficient matrix of each pure power supply node to each node and / or the power consumption coefficient matrix of each pure load node to each node according to the data information linked list;

[0068] The occupation coefficient calculation module is used to obtain the occupation coefficient matrix of each pure power supply node and / or each pure load node to the power of each line according to the power contribution coefficient matrix of each pure power supply node to each node and / or the power consumption coefficient matrix of each pure load node to each node;

[0069] The total network loss calculation module is used to obtain the total amount of network loss allocation borne by each pure power supply node and / or each pure load node in the distribution network according to the active loss of the line in the data information linked list and the occupation coefficient matrix of each pure power supply node and / or each pure load node to the power of each line.

[0070] A non-transitory computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement any step of the network loss allocation method for a distribution network containing a distributed power supply to achieve the third purpose of the application.

[0071] Advantages:

[0072] The power distribution network loss allocation method can quickly and accurately analyze the power distribution relationship between each power supply and load according to the topological position and electrical connection relationship of the power supply and load and other conditions, determine the proportion of the power supply and load in bearing the system loss and the use of the power supply and load on the transmission line, understand the influence degree of each power supply and load on the power grid, and reasonably collect fees and formulate power generation plans according to the actual use of the power supply and load on the power grid; the method not only provides a scheme for accurately allocating the network loss among the power supply and load, but also provides a useful economic signal for system operation; meanwhile, the method can be applied to power distribution networks with complex structures, various transaction subjects and various line connections, and can significantly improve the calculation accuracy of network loss allocation. BRIEF DESCRIPTION OF DRAWINGS

[0073] Figure 1 A flowchart of a DG-containing power distribution network loss allocation method based on a knowledge graph and network topology is shown.

[0074] Figure 2 An equivalent topological graph of a distributed power supply connected to a grid is shown.

[0075] Figure 3 A flowchart of a knowledge graph-based network loss anomaly identification method is shown.

[0076] Figure 4 An equivalent DG-containing power distribution network of IEEE 28 nodes is shown.

[0077] Figure 5 A partial example of an IEEE 28-node equivalent DG-containing power distribution network loss anomaly judgment knowledge graph is shown.

[0078] Figure 6 A cluster center of the line loss rate of the 28 lines of the IEEE 28-node equivalent DG-containing power distribution network is shown.

[0079] Figure 7 The K-means clustering result of the 28 lines of the IEEE 28-node equivalent DG-containing power distribution network is shown.

[0080] Figure 8 A process diagram of forward flow tracing of a four-node topological graph of a distributed power supply based on network topology is shown.

[0081] Figure 9 A process diagram of reverse flow tracing of a four-node topological graph of a distributed power supply based on network topology is shown. DETAILED DESCRIPTION

[0082] The following specific embodiments are used to explain the technical solutions of the claims of the present application, so that those skilled in the art can understand the claims of the present application. The protection scope of the present application is not limited to the following specific embodiments. The technical solutions of the claims of the present application which are different from the following specific embodiments and contain the technical solutions of the claims of the present application are also within the protection scope of the present application.

[0083] The specific embodiments of the present application are further specifically described below in combination with the drawings.

[0084] S1, obtaining a data information chain table according to the power flow result of the node topological graph under the condition that all distributed power sources are connected to the grid, wherein the data information chain table comprises distributed power source power, power consumption load power, branch power, branch active power loss, node and branch number;

[0085] Taking the IEEE 28-node topological graph as an example, the data information chain table shown in Table 1 and Table 2 is obtained by performing power flow calculation on the IEEE 28-node topological graph before and after the DG is connected to the grid, and the unit is kW;

[0086]

[0087] Table 1 Change in power of distribution network lines before and after introduction of DG

[0088]

[0089]

[0090] Table 2 Change in power loss of distribution network lines before and after introduction of DG

[0091] According to the power flow relationship of the equivalent topological graph of the distributed power source connected to the grid shown in FIG. 1, Figure 2 the power relationship in the formula can be expressed as Figure 2

[0092]

[0093] In the formula,

[0094] S G and S A are the powers flowing through the first and last ends of the line;

[0095] S loss is the power loss of the line;

[0096] S DG and S L are the powers of the distributed power source and the load.

[0097] When S G > 0, the power grid supplies power to the power consumption load adjacent to the distributed power source, and S G ​The positive and negative decision determines whether the distributed power source participates in the power flow of the power grid.

[0098] According to the judgment S A The sign of S L And the size relationship of S DG The operation scene of the DG can be divided into "no power generation", "local consumption" and "surplus on-grid" according to the size relationship of S

[0099]

[0100] The application scene suitable for the verification of the network loss allocation method described in the embodiment is "surplus on-grid"; because when the DG is in the "local consumption" scene, there is a certain interest game relationship between the power supplier and the power load, and there is no fair and clear division principle at present. Therefore, the DG "surplus on-grid" is adopted as the simulation verification research scene of the network loss allocation method based on network topology active distribution network described in the application in the embodiment.

[0101] Preferably, it further comprises constructing a distribution network loss abnormality identification knowledge graph, designing a network loss abnormal line rapid search method, identifying a network loss abnormal line and performing checking processing, Figure 3 As shown in the figure, it is a flow chart for constructing a distribution network loss abnormality identification knowledge graph, and the method specifically comprises the following steps:

[0102] S101: Based on the data information chain table, the ontology, entity and attribute involved in the distribution network loss abnormality identification are selected, the entity capable of judging the network loss abnormality characteristics and the associated attributes thereof are selected, a triple form is formed, and a distribution network knowledge graph prototype is constructed;

[0103] The ontology involved in the distribution network loss abnormality identification in this step represents a specific physical device, which is a line in this embodiment. The ontology contains the number of entities required for constructing the knowledge graph, and the ontology = {entity 1, entity 2, …, entity y}, such as line = {line 1, line 2, …, line y}. The form of each entity can be represented as (entity | entity corresponding identifier, associated attribute set of the identified entity), which is represented as (line 1 | Ningwan line, current) in this embodiment.

[0104] The selected entity capable of judging the network loss abnormality characteristics in this step is, for example, a transformer, a line, and the associated attributes are, for example, voltage, current, active power, etc. These entities and attributes are structured series data, which are saved in a source database. The source database fields are arranged into each table to represent the entity type, and each row represents the data form of different attributes of the same type of entity.

[0105] The method for constructing the primary knowledge graph of the distribution network according to the triple form in this step is to extract knowledge and its corresponding relationship from the source database according to the form of "entity-relation-entity", "entity-attribute-attribute value", and "entity-attribute-entity". For example, if the extracted type is "substation 1 and substation 2 are connected in series", the triple "substation 1-series-substation 2" is formed. If the extracted type is "voltage value of the transformer", the triple "transformer-voltage-voltage value" is formed. If the extracted type is "line 1 transmits current to line 2", the triple "line 1-transmit current-line 2" is formed. According to the saving method of the field of the source database, the primary knowledge graph of the distribution network is formed.

[0106] S102: Extract the features representing the network loss anomaly identification from the selected distribution network loss anomaly identification ontology, entity and attribute; in this embodiment, a word segmentation algorithm is used to process the unstructured data representing the network loss anomaly features, a TF-IDF algorithm is used to analyze the weight of the entities and the abnormal forms of the corresponding network loss contained in the word segmentation results, and the entities and network loss anomaly features with larger weights are selected to represent the network loss anomaly features;

[0107] The unstructured data processing method of the word segmentation algorithm representing the network loss anomaly features in this step is as follows: for example, "line transmission power imbalance causes network loss anomaly" is processed by the word segmentation algorithm to become "line / transmission power imbalance / causes / network loss / abnormal", based on the result after the word segmentation processing, the natural language can be composed into a triple similar to [line, transmission power imbalance, network loss anomaly].

[0108] The calculation method of the TF-IDF algorithm in this step is as formula (3), the larger the TFxIDF is, the larger the proportion of the word in the distribution network line loss anomaly reason is.

[0109]

[0110] In the formula:

[0111] TF i represents the word frequency of the word segmentation i;

[0112] IDF i represents the inverse file frequency of the word segmentation i;

[0113] n i represents the number of times that the word segmentation i appears;

[0114] represents the total number of times that all word segmentations appear;

[0115] t i represents the number of abnormal reasons for the selected word segmentation i;

[0116] d j represents the number of abnormal reasons containing the word segmentation i;

[0117] D represents the total number of all word segmentation abnormal reasons.

[0118] S103: Establish a network loss abnormality judgment rule according to the characteristics of network loss abnormality identification;

[0119] In this embodiment, based on the calculation results of the TF-IDF algorithm, the entity in the line loss abnormality reason and the entity, the entity and the abnormal feature are established to form a complete three tuple relationship, and are combined with the knowledge graph in step 3.1, then the distribution network line loss abnormality knowledge graph is formed. The specific reason of the abnormal feature is selected, which is greater than the set value, and the DG distribution network loss abnormality judgment rule is set. There is a sequence between the execution steps of the judgment rule, which is arranged in the form of three tuples according to this sequence, then a complete network loss abnormality identification knowledge graph is formed;

[0120] The word segmentation algorithm and TF-IDF calculation processing are performed on the plurality of distribution network loss abnormality feature samples, and the abnormal features with high network loss ratio are analyzed and the corresponding abnormal judgment rules are set. In this embodiment, according to the results of TF-IDF calculation processing, the abnormal features with a proportion greater than 6% of network loss abnormality features are selected, that is, two abnormal features of "unstable voltage" and "unbalanced current". According to the industry common sense, the judgment rules of the abnormal features "unstable voltage" and "unbalanced current" are "certain phase voltage is lower than 78% rated voltage or three-phase full loss voltage" and "transmission power is lower than 90% normal transmission power", as shown in Table 3. According to the actual demand, abnormal features with a proportion greater than other percentages can also be selected, which is not limited by the present application.

[0121]

[0122] Table 3 TF-IDF calculation results and network loss abnormality feature judgment rules

[0123] S104: According to the sequence between the execution steps of the network loss abnormality judgment rule, arrange it in the form of three tuples, to form a distribution network loss abnormality identification knowledge graph, and take IEEE28 node DG distribution system as an example, part of the knowledge graph is as follows Figure 5 ;

[0124] S105: According to the size of the possibility of network loss abnormality, search the network loss abnormal line of step 3.3 DG distribution network loss abnormality knowledge graph, determine the network loss abnormal line and carry out artificial checking process.

[0125] In this step, the K-means clustering algorithm is used to sort the size of the network loss abnormality possibility. This algorithm divides the line into three categories according to the line loss rate, that is, large, normal and small. The lines classified as large and small are more likely to be network loss abnormal lines, so the abnormal feature judgment should be carried out on these two types of lines first.

[0126] The judgment method of the line loss abnormal line with large line loss rate in this step is based on the judgment standard of the power company for the distribution network line loss abnormality, that is, when the line loss rate exceeds 10%, it is considered as line loss abnormality. For the line whose line loss rate exceeds 10%, find and judge the abnormal characteristics according to its attributes and abnormal judgment rules in the knowledge graph.

[0127] The line loss abnormal line with small line loss rate (line loss rate less than 10%) is subjected to K-means clustering analysis in this step, so as to Figure 4 The equivalent DG distribution system of IEEE 28 nodes is shown in FIG. 1. First, the line loss rates of 28 lines randomly extracted from the data set obtained from the system are taken as "cluster centers", as shown in FIG. 2. Figure 6 Then, the distances of other sample points to the three cluster centers are calculated according to formula (4), and the sample points are divided into the cluster where the cluster center with the nearest distance is located; then the average line loss rate of each cluster is taken as a new "cluster center", and the above steps are repeated until the "cluster center" no longer changes, as shown in FIG. 3. Figure 7 Finally, the lines can be divided into five categories according to the line loss rate, that is, large, relatively large, medium, relatively small and small. The line loss rate of the line is large or small, and the abnormal characteristics are judged according to the knowledge graph. The judgment result is shown in Table 4.

[0128] d ij = |l i -l j | Formula (4)

[0129] In the formula:

[0130] d ij represents the relative distance of the line loss rate of the line i, j;

[0131] l i , l j represent the line loss rate of the line i, j.

[0132] Class Line No. Large 2、3、6、23 Large 1、5、11、12、14、20、25 Medium 8、13、19、21 Small 4、10、15、17、18、22、24 Small 7、9、16、26、27、28

[0133] Table 4 K-means clustering results of IEEE 28 node distribution network topology graph

[0134] S2, search the pure power node and / or pure load node in the data information chain table, calculate the power contribution coefficient matrix of each pure power node to each node and / or the power extraction coefficient matrix of each pure load node to each node in the node topology graph according to the data information chain table; obtain the occupation coefficient matrix of each pure power node / or each pure load node to the power of each line according to the power contribution coefficient matrix of each pure power node to each node and / or the power extraction coefficient matrix of each pure load node to each node;

[0135] (1) Using the principle of downstream tracing in network topology, combined with data information chain table, starting with a pure power node, search for new downstream pure power nodes one by one, mark the searched pure power nodes and the power outflow branches associated therewith, and the marked pure power nodes do not participate in the subsequent search;

[0136] The principle of downstream tracing based on network topology is to determine the associated branch and downstream node of the pure power node according to the power flow direction of the pure power node, and to remove it and its associated branch. Then, the same method is repeatedly used to remove all branches in the remaining network diagram until there is only a pure load node in the network diagram, and the analysis process is as shown in Figure 8

[0137] Each time a new pure power node is searched, the power contribution coefficient of each pure power node to each node is calculated and recorded in the pure power node power contribution coefficient matrix of each node. In this embodiment, according to the proportional sharing principle, the power source occupies the same proportion in the power of each line downstream of the node, so the occupation coefficient matrix of each power source to each line can be obtained;

[0138] In the proportional sharing principle, the total inflow power is the sum of the net injection power of the generator and the inflow power of the line, and the total outflow power is the sum of the power drawn by the load and the outflow power of the line, and then the total inflow power is equal to the total outflow power, which is equal to the node power flow.

[0139] The elements in the pure power node power contribution coefficient matrix of each node reflect the contribution proportion of the node where each pure power source is located to the total injection power of each node. Assuming that the initial node of line k is i and the terminal node is j, the power source G i The calculation formula of the power contribution coefficient of node j is as follows:

[0140]

[0141] In the formula:

[0142] L k represents the active power flowing through line k;

[0143] N j represents the total active power flowing into node j;

[0144] k∈-j represents the branch associated with node j.

[0145] ​The power contribution coefficient matrix of each pure power supply node to each node can obtain the occupation coefficient matrix of each pure power supply node to each line power. The occupation coefficient represents the occupation share of each pure power supply node to the power flowing on each line, and further obtains the occupation coefficient matrix of each pure power supply node to each line power. Assuming that there are n nodes, m lines and g power supplies in the system, the power contribution coefficient matrix α of each pure power supply node to each node is shown in formula (6).

[0146]

[0147] In the formula:

[0148] The power contribution coefficient value of power supply g to node n is represented.

[0149] G g The power supply numbered g is represented.

[0150]

[0151] In the formula:

[0152] G i The power supply numbered i is represented.

[0153] The power contribution coefficient of power supply G i To line k is represented.

[0154] The power contribution coefficient of power supply G i To node i is represented.

[0155]

[0156] In the formula:

[0157] G i The power supply numbered i is represented.

[0158] The occupation coefficient of power supply g to line k n is represented.

[0159] The power contribution coefficient of power supply G i To node i is represented.

[0160] (2) Based on the network topology inverse flow tracking method combined with the data information linked list, a pure load node is started to search new upstream pure load nodes one by one, and the searched pure load nodes and the power outflow branches associated therewith are marked. The marked pure load nodes do not participate in the subsequent search.

[0161] The principle of the backtracking in this step is to trace the power absorbed by the "pure load" along the associated lines to its upstream nodes, and to eliminate the "pure load" node and its associated branches, i.e. in the matrix, the corresponding row of the "pure load" node 4 and the columns corresponding to the elements with value 1 in the corresponding row are deleted, thus obtaining a subgraph. The above method is repeated to find the "pure load" node and delete it and its associated branches in the remaining subgraph until all branches are deleted, and the remaining nodes will be "pure power" nodes. The specific steps are shown in the following table. Figure 9 Figure 9

[0162] Each time a new pure load node is searched, the power consumption coefficient of each pure load node to each node is calculated and recorded in the power consumption coefficient matrix of each pure load node to each node. Similarly, according to the power consumption coefficient matrix of each pure load node to each node, the power occupation coefficient matrix of each pure load node to each line can be obtained. The pure load node is a node that only contains a load;

[0163] The elements in the power consumption coefficient matrix of each pure load node to each node reflect the proportion of the total power consumption of the load in each pure load node to each node. In this embodiment, the load F j The power consumption coefficient of the load F to node j is:

[0164]

[0165] wherein:

[0166] represents the power consumption coefficient of the load F j to node j;

[0167] L k represents the active power flowing through line k;

[0168] N i represents the total active power flowing into node i;

[0169] f represents the number of loads in the system;

[0170] i represents the upstream node of j.

[0171] Therefore, the power consumption coefficient of the load F j to node i is:

[0172]

[0173] wherein:

[0174] k∈+i represents the set of outgoing lines associated with node i.

[0175] ​​According to formula (10), each element in the power consumption coefficient matrix of the load to the node can be calculated one by one, so as to obtain the power consumption coefficient matrix λ of the load to each node.

[0176]

[0177] In the formula, λfn represents the power consumption coefficient value of the load numbered f to the node numbered n.

[0178] In the formula, λfn represents the power consumption coefficient value of the load numbered f to the node numbered n.

[0179] F f In the formula, Fj represents the load numbered j.

[0180] According to the proportional sharing principle, the occupation coefficient matrix of each pure load node to the power of each line can be obtained according to the power consumption coefficient matrix of each pure load node to each node. The load F j The occupation coefficient matrix of line k As shown in formula (12), the coefficient matrix δ is expressed as formula (13).

[0181]

[0182] In the formula, δfk represents the occupation coefficient of the load numbered f to the line numbered k.

[0183] F i In the formula, Fj represents the load numbered j.

[0184] In the formula, Fj represents the load numbered j. j The power consumption coefficient matrix of node j;

[0185] Wherein j is the downstream node of line k.

[0186]

[0187] In the formula, δfk represents the occupation coefficient of the load numbered f to the line numbered k.

[0188] In the formula, δfk represents the occupation coefficient of the load numbered f to the line numbered k. n

[0189] S3, according to the line active loss in the data information chain table and the occupation coefficient matrix of each pure power node and / or each pure load node to the power of each line, the total amount of network loss allocation borne by each pure power node and / or each pure load node in the distribution network is obtained.

[0190] (1) According to the power flow calculation, the loss of each line and the use proportion of the power source to the line The product can obtain the network loss allocation amount borne by the power source Gi to the line k, and then the network loss allocation amount of each line is superimposed and summed to obtain the total amount of network loss allocation borne by the power source Gi ​For:

[0191]

[0192] The step is described in Figure 8 The four-node topology graph of the distributed power supply in I shown in the figure is taken as an example, and the line part A in the contribution factor matrix of the DG is calculated by combining the principle of downstream tracking in network topology l , A l can be solved by the ratio of the power flowing through the line k to the total power of the upstream node i of the line k, where node i is the upstream node of line k. Figure 8 A l is calculated as:

[0193]

[0194] The power contribution coefficient matrix a of the DG to each downstream pure power supply node connected thereto can be calculated by combining formulas (5) and (6), and the contribution factor matrix a can be calculated by formula (16):

[0195]

[0196] In the formula:

[0197] ki indicates that k is an upstream node or a downstream node of i;

[0198] A jn a nk indicates the contribution of power supply k to the total power of node i through node n and line L jn .

[0199] Figure 8 In the four-node topology graph of the distributed power supply in I shown in the figure, the calculation formula of the power contribution coefficient matrix of each pure power supply node to each node is:

[0200]

[0201] According to formulas (7) and (8), β=A l B, then Figure 8 In the four-node topology graph of the distributed power supply in I shown in the figure, the calculation formula of the power contribution coefficient matrix of each pure power supply node to each line is as follows:

[0202]

[0203] The power supply power diagonal matrix P G is as formula (19):

[0204]

[0205] Based on equations (18) and (19), the contribution power matrix of DG to each line can be obtained and expressed as follows after normalization:

[0206]

[0207] Depend on Figure 8 It can be seen that distributed power source G1 alone bears the network losses of lines a and b, while the network losses of lines c, d, and e are shared by G1 and G2. Let the network losses of the five lines be P. a,loss P b,loss P c,loss P d,loss P e,loss , Figure 9 The total network loss P borne by power sources G1 and G2 in the four-node distributed power topology diagram shown in Figure I is as follows: G1,loss and P G2,loss for:

[0208]

[0209] Based on the same calculation method, the results of allocating network losses to distributed power sources connected to the distribution network in the IEEE 28-node topology diagram are shown in Table 5, with units of kW.

[0210]

[0211] Table 5. Line Loss DG Allocation

[0212] (2) The principle of reverse tracing in network topology is used to reduce line loss. The proportion of load on this line Multiplying them together gives the load F. j For the network loss allocation of line k, the load F is obtained by summing the network loss allocation of each line using the same method. j Total network loss sharing for:

[0213]

[0214] The reverse tracing method described in this step uses... Figure 9 Taking a four-node distributed generation topology in I as an example, to solve for the power user's share of network loss, first calculate the line portion C in the contribution factor matrix. l Due to vector P l The order of P is the same, and P l =C l P, C l The formula for calculating non-zero elements in C is: (C l ) ji = Power flowing through line j / Total power flowing through node i Pi Where node i is the downstream node of line j, Figure 9 C l for

[0215]

[0216] Let P L For load power vectors (if node i is not a load node, then P...), Li =0), P is the total power vector flowing through the node, P L If P has the same reverse tracking order, then the load power absorption factor matrix λ of the total power flowing through the upstream node can be defined as P = λP. L In matrix λ, the upper triangular matrix elements reflecting the power drawn by the downstream nodes from the load of a node are zero. The elements in matrix λ can be calculated using the following formula.

[0217]

[0218] In the formula k i Let node k be a downstream node of i; j∈i be the node that draws power from node i through line j; λ mk For each element already calculated in the λ matrix, C represents the power extraction factor of load k on the total power flowing through node m; jm λ mk Let be the power drawn by load k from node i through node m and line j. In this example, it can be calculated row by row according to equation (24).

[0219]

[0220] In the formula:

[0221] λ ik The elements that have been calculated in the λ matrix represent the power extraction factor of load k on the total power flowing through node i;

[0222] Because of P l =C l P, and P = λP l , can be obtained

[0223] P l =C l λP L =K L P L Equation (26)

[0224] K L This is the load power extraction factor matrix for the line. Since the calculation process is similar to that of downstream tracking, it can be obtained as follows:

[0225]

[0226] Equivalent load diagonal matrix PL It can be expressed as the following formula:

[0227]

[0228] Therefore, we can obtain Figure 9 In the distributed generation four-node topology diagram shown in Figure I, the power contribution matrix of each pure load node to each node is as follows:

[0229]

[0230] From the absorption matrix of equation (29), it can be seen that load L3 and load L4 jointly bear the losses of lines a and d; load L4 bears all the losses of lines b, c, and e. Therefore... Figure 1 The total network loss P borne by all loads L3 and L4 in the distributed generation four-node topology diagram shown in Figure I is... Lk,loss It can be represented as:

[0231]

[0232] For the IEEE 28-node topology, the distribution of power load on line losses is shown in Table 6, in kW.

[0233]

[0234]

[0235] Table 6. Line Loss and Electricity Load Allocation

[0236] The method proposed in this invention is compared and analyzed with the improved power flow tracing method and the improved average network loss coefficient method. The comparison results are shown in Table 7 below.

[0237]

[0238] Table 7. Network loss rates borne by the power grid and loads before and after the introduction of DG.

[0239] Based on the simulation results of the 28-node diagram above, the effectiveness of the proposed method for allocating network losses in a distribution network with distributed power sources based on network topology is verified.

[0240] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0241] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, the computer program comprising program instructions, the program instructions being executed by a processor to implement each step of the method of the present application, which will not be repeated here.

[0242] The computer readable storage medium can be an internal storage unit of the data transmission device or the computer device according to any one of the foregoing embodiments, for example, a hard disk or a memory of the computer device. The computer readable storage medium can also be an external storage device of the computer device, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card and the like.

[0243] Further, the computer readable storage medium can include both the internal storage unit and the external storage device of the computer device. The computer readable storage medium is used to store the computer program and other programs and data required by the computer device. The computer readable storage medium can also be used to temporarily store data to be output or data that has been output.

[0244] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage and the like) containing computer-usable program code.

[0245] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the method, device (system) and computer program product according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device implemented in the flowcharts and / or block diagrams. Figure 1 The device for implementing the function specified in one flow or multiple flows and / or blocks. Figure 1 The device for implementing the function specified in one flow or multiple flows and / or blocks.

[0246] These computer program instructions can 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 instructions which implement the Figure 1 function specified in the flow or flows and / or blocks Figure 1 of the block or blocks.

[0247] These computer program instructions can 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 that are executed on the computer or other programmable apparatus provide steps for implementing the Figure 1 function specified in the flow or flows and / or blocks ​ of the block or blocks.

[0248] The specific embodiments of the present application described herein are illustrative, and not restrictive, of the application. Many variations of these specific embodiments can become apparent to those of ordinary skill in the art upon reading the foregoing description, and the general principles defined herein can be applied to other variations without departing from the scope of the application. The described embodiments are merely representative and were chosen for purposes of illustration and description.

[0249] The description herein of any particular aspects is not intended to be exhaustive or to be necessarily assumed to be the only possible design of the application. Rather, the scope of the application is to be accorded the broadest interpretation so as to encompass all meaningful adaptations, equivalents, and equivalents of the elements covered by the following claims. Numerous specific aspects have been discussed herein for purposes of illustration. These specific aspects are not intended to limit the scope of the application, which is to be measured by the claims.

Claims

1. A method for loss allocation in a power distribution network with distributed generation, characterized in that, It comprises the following steps: S1, obtaining a data information linked list according to a power flow result of a node topology graph under full grid connection of distributed power sources; S2, searching for pure power source nodes and / or pure load nodes in the data information linked list, calculating a power contribution coefficient matrix of each pure power source node to each node and / or a power consumption coefficient matrix of each pure load node to each node according to the data information linked list, and obtaining an occupation coefficient matrix of each pure power source node and / or each pure load node to power of each line according to the power contribution coefficient matrix of each pure power source node to each node and / or the power consumption coefficient matrix of each pure load node to each node; S3, obtaining a total amount of network loss allocation borne by each pure power source node and / or each pure load node in the distribution network according to line active loss and the occupation coefficient matrix of each pure power source node and / or each pure load node to power of each line in the data information linked list; The node topology graph under full grid connection of the distributed power sources comprises a distributed power source four-node topology graph, and the calculation method of the power contribution coefficient matrix of each pure power source node to each node in the distributed power source four-node topology graph comprises: ； The calculation method of the occupation coefficient matrix of each pure power source node to power of each line comprises: ； In the formula, B represents the power contribution coefficient matrix of each pure power source node to each node; and β represents the occupation coefficient matrix of each pure power source node to power of each line. The step S1 further comprises the following steps: S101, selecting a distribution network loss abnormality identification ontology, entity and attribute based on the electrical parameters in the data information linked list; P a , P b , P c , P d , P e : represents the active power flowing through all lines a, b, c, d, e in the four-node topology of the distributed power supply; P G1 , P G2 : denotes the active power flowing into all power sources G1 and G2.

2. The method for cost allocation of power loss in a distribution network with distributed power sources as claimed in claim 1 wherein, S102, extracting a feature representing the network loss abnormality identification from the selected distribution network loss abnormality identification ontology, entity and attribute; S103, establishing a network loss abnormality judgment rule according to the feature of the network loss abnormality identification; S104, constructing a distribution network loss abnormality identification knowledge graph according to the network loss abnormality judgment rule; S105, identifying a network loss abnormal line according to the distribution network loss abnormality identification knowledge graph. In the step S2, the searched pure power source and / or pure load node does not participate in subsequent search of the data information linked list. The node topology graph under full grid connection of the distributed power sources comprises a distributed power source four-node topology graph, and the calculation method of the total amount of network loss allocation borne by each pure power source node in the distribution network in the distributed power source four-node topology graph comprises:

3. The method for cost allocation of power loss in a distribution network with distributed power sources as claimed in claim 1 wherein, In the formula, B represents the power contribution coefficient matrix of each pure power source node to each node; and β represents the occupation coefficient matrix of each pure power source node to power of each line.

4. The method for cost allocation of power loss in a distribution network with distributed power sources as claimed in claim 1 wherein, The node topology graph under full grid connection of the distributed power sources comprises a distributed power source four-node topology graph, and the calculation method of the power consumption coefficient matrix of each pure load node to each node in the distributed power source four-node topology graph comprises: ； In the formula, B represents the power contribution coefficient matrix of each pure power source node to each node; and β represents the occupation coefficient matrix of each pure power source node to power of each line. P G1,loss 、P G2,loss : represents the total amount of loss allocation undertaken by the pure power node where all power sources G1 and G2 are located; P a,loss , P b,loss , P c,loss , P d,loss , P e,loss : denotes the losses of all lines a, b, c, d, e.

5. The method for cost allocation of power loss in a distribution network with distributed power sources as claimed in claim 1 wherein, The node topology graph under full grid connection of the distributed power sources comprises a distributed power source four-node topology graph, and the calculation method of the total amount of network loss allocation borne by each pure load node in the distribution network in the distributed power source four-node topology graph comprises: ； In the formula, B represents the power contribution coefficient matrix of each pure power source node to each node; and β represents the occupation coefficient matrix of each pure power source node to power of each line. : Power draw coefficient matrix of each node for each node in the four-node topology of distributed power supply P L3 , P L4 : represents the active power flowing through all loads L3, L4 in the four-node topology of the distributed power supply.

6. The method for cost allocation of power loss in a distribution network with distributed power sources as claimed in claim 1 wherein, It comprises: ； a data acquisition module, a power contribution coefficient calculation module, an occupation coefficient calculation module and a total amount of network loss calculation module. P a,loss , P b,loss , P c,loss , P d,loss , P e,loss : represents the losses of all lines a, b, c, d, e; P L3,loss 、P L4,loss : represents the total amount of network loss allocation undertaken by all pure load nodes where loads L3 and L4 are located in the four-node topology of distributed power supply P L3 , P L4 : represents the active power flowing through all loads L3, L4 in the four-node topology of distributed power sources.

7. A system for loss allocation of a distribution network with distributed generation as claimed in claim 1 wherein, ​ ​ The data acquisition module is configured to obtain a data information linked list according to a power flow result of a node topology graph under full grid connection of the distributed power supply, the data information linked list including distributed power supply power, power consumption load power, branch power, branch active power loss, node and branch number; The power contribution coefficient calculation module is configured to search for pure power supply nodes and / or pure load nodes in the data information linked list, and calculate a power contribution coefficient matrix of each pure power supply node to each node and / or a power consumption coefficient matrix of each pure load node to each node according to the data information linked list; The occupation coefficient calculation module is configured to obtain an occupation coefficient matrix of each pure power supply node and / or each pure load node to power of each line according to the power contribution coefficient matrix of each pure power supply node to each node and / or the power consumption coefficient matrix of each pure load node to each node; The total network loss calculation module is configured to obtain a total amount of network loss borne by each pure power supply node and / or each pure load node in the distribution network according to the active power loss in the data information linked list and the occupation coefficient matrix of each pure power supply node and / or each pure load node to power of each line.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the steps of the network loss allocation method for the distribution network with distributed power supply according to any one of claims 1 to 6.

Images