Electrical equipment operation state abnormity identification method based on association hybrid reasoning

Through the improved DLG algorithm mining of the association rule, the problem of poor model universality and high operating cost in the abnormal identification of electrical equipment operating status is solved, effective identification and real-time monitoring of abnormal status of electrical equipment is realized, and the accuracy and fault tolerance of identification are improved. .

CN118965093APending Publication Date: 2024-11-15HUADIAN ELECTRIC POWER SCI INST CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202410924202.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-10
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The existing methods for identifying abnormal states of electrical equipment have problems such as poor model universality, high operating costs and poor learning ability when training large-scale samples, making it difficult to effectively identify abnormal states in complex systems.

Method used

Using the DLG algorithm based on association rule mining, through association rule data mining and attribute reduction, a reduction decision table and association rule interactive mining are formed to identify the abnormal operating status of electrical equipment, and the association graph structure is used to reduce IO costs and improve identification. efficiency.

Benefits of technology

It realizes effective identification of abnormal operating status of electrical equipment at a lower cost, improves the accuracy and real-time identification, has high fault tolerance and practicality, and can be incomplete in a certain power grid scale and protection operation information. Accurately identify abnormal states.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118965093A_ABST
    Figure CN118965093A_ABST
Patent Text Reader

Abstract

The invention relates to the field of electrical equipment operation state abnormity identification methods, in particular to an electrical equipment operation state abnormity identification method based on association hybrid reasoning, which introduces an association rule data mining DLG algorithm into electrical equipment operation state abnormity identification so as to protect as condition attributes and abnormal areas as decision attributes. Various abnormal conditions are investigated, an original decision table is established, attribute reduction is performed by utilizing association rule mining, interactive mining is performed by modifying a threshold value, an optimal attribute reduction combination is directly extracted, and interactive mining is performed by utilizing a reduction decision table formed by the optimal attribute reduction combination and association rules. And carrying out abnormity identification on the abnormal operation state information of the electrical equipment under various conditions. According to the method, the abnormal operation state of the equipment is reasoned and identified by utilizing the reduction decision table formed by the optimal attribute reduction combination and the association rule interactive mining, so that the effectiveness and the real-time performance of identifying the abnormal operation state of the electrical equipment at relatively low cost are realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of methods for identifying abnormal operation status of electrical equipment, and in particular to a method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning. Background Art

[0002] With the continuous expansion of the scale of power grid, when the operating status of electrical equipment is abnormal, a large amount of rich abnormal information will be sent to the dispatching center. How to effectively extract information from massive abnormal data and how to timely identify abnormal operating status from the information have become urgent problems to be solved. At present, most of the methods for identifying abnormal operating status of electrical equipment remain in the manual identification stage. Some electrical equipment operating status abnormality identification systems that have been put into operation are difficult to guarantee fault tolerance when the abnormal information is uncertain. Therefore, the identification of abnormal operating status of electrical equipment has always been a hot topic in theoretical and applied research.

[0003] At present, the main status and problems of abnormal operation status identification methods of electrical equipment at home and abroad are summarized as follows:

[0004] 1) The abnormal operation status identification method based on physical models usually includes the analysis of the dynamic characteristics of electrical equipment. By exploring the corresponding physical properties, degradation and other laws, a model describing the state identification is established to predict the development trend of the future operation status. The prediction accuracy of the equipment with obvious abnormal operation status is high. The semi-physical model is improved, and the K-means clustering method is combined to segment the related objects of the predicted state. The function is established based on the segment distance to achieve accurate prediction of the operation status of electrical equipment. However, due to the different influencing factors of the composition, structure and operation status of different electrical equipment, the model needs to be constantly changed according to the characteristics of the equipment, which increases the cost and complexity. The model-based identification method is usually used in systems with simple structures and easy-to-quantify mechanisms, which is very limited in practical applications.

[0005] 2) The abnormal operation state recognition method based on support vector machine is a binary classification model with simple structure, fast initialization speed, strong generalization ability, and can effectively solve nonlinear problems. The variational mode method is used to decompose the power load curve, and then the SVM method is used for state recognition and the results are optimized by the sparrow search algorithm, which improves the recognition accuracy of short-term power load. The SVM model uses the inner product kernel function to reduce the computational complexity and has certain robustness, but the operating cost is high when training large-scale samples, and improper selection of the kernel function will have a greater impact on the results, so it has certain limitations.

[0006] 3) An abnormal operation status identification method based on artificial neural network. Neural network is one of the important research contents of modern intelligent technology. It is based on network topology knowledge, has strong nonlinearity and high robustness, and can handle complex logical systems. The model combining genetic algorithm and BP neural network is used to predict the vibration operation trend of the unit, and the accuracy of identifying future abnormal operation status is improved by predicting vibration signals. Artificial neural network has high flexibility and strong information adaptive ability, but compared with deep learning algorithm, its learning ability is not strong enough, and it will be gradually replaced by deep learning algorithm in the future.

[0007] At present, the main problems of the abnormal operation status identification methods of electrical equipment are concentrated on the poor universality of the model, high operating cost when training large-scale samples, and poor learning ability. Summary of the invention

[0008] The present invention provides a method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning, applies the Apriori algorithm improved based on association rule mining, namely the DLG algorithm, to abnormal operation status of electrical equipment, and utilizes a simplified decision table formed by the optimal attribute simplification combination and interactive mining of association rules to infer and identify abnormal operation status of equipment, thereby achieving effectiveness and real-time performance in identifying abnormal operation status of electrical equipment at a relatively low cost.

[0009] The present invention introduces the association rule data mining DLG algorithm into the abnormal identification of the operating status of electrical equipment. First, protection is used as the conditional attribute and the abnormal area is used as the decision attribute. Various abnormal situations are examined and an original decision table is established. Then, association rule mining is used to simplify the attributes. Interactive mining is performed by modifying the threshold to directly extract the best attribute simplification combination. Then, the simplified decision table formed by the best attribute simplification combination and the association rule interactive mining are used to perform abnormal identification on the abnormal information of various situations of the electrical equipment.

[0010] In order to achieve the purpose of the present invention, the technical solution adopted is: a method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning, comprising:

[0011] The abnormal operation status information of electrical equipment is obtained, and the association rule data mining DLG algorithm is introduced into the abnormal operation status identification of electrical equipment. First, protection is used as the condition attribute and abnormal area is used as the decision attribute. Various abnormal situations are examined and the original decision table is established. Then, the association rule mining is used to simplify the attributes. Interactive mining is performed by modifying the threshold to directly extract the best attribute simplification combination. Then, the simplified decision table formed by the best attribute simplification combination and the association rule interactive mining are used to identify the abnormal operation status information of electrical equipment in various situations. The association rule mining is used to discover the valuable data patterns with correlation relationships hidden in the massive data, and the area where the electrical equipment operation status is abnormal is obtained.

[0012] As an optimization solution of the present invention, the association rule data mining DLG algorithm includes:

[0013] 1) Generate frequent 1-item set stage: generate frequent 1-item set and record relevant information;

[0014] 2) Construction of association graphs to show the associations between frequent items;

[0015] 3) Generate k frequent item sets based on the construction of association graph.

[0016] As an optimization solution of the present invention, using association rule mining to perform attribute simplification includes: I is a text set describing abnormal operation status, the elements of which are called terms, D is a set of causal relationships between abnormalities, T is a set of terms, and Each abnormal causal relationship has a unique identifier, denoted as TID. Let X be the set of items in I. If Then T is said to contain X; an association rule is an implication of X=>Y, where And X∩Y=φ.

[0017] As an optimization scheme of the present invention, there are three indicators for measuring association rules, namely support, confidence and relevance, which means: the support of the association rule X=>Y represents the proportion of tuples containing X∪Y in the causal set D, the confidence of the association rule X=>Y represents the ratio of the number of tuples containing X∪Y in the causal set D to the number of tuples containing X, and the relevance means whether X and Y are positively correlated or negatively correlated.

[0018] As an optimization scheme of the present invention, the simplified decision table and interactive mining of association rules formed by the optimal attribute simplification combination refers to: utilizing the DLG algorithm based on the association graph structure to reduce the number of scans of the transaction database, thereby reducing the IO cost of the association hybrid reasoning process, and each reasonable association rule in the result of the association hybrid reasoning corresponds to a simplified combination containing an attribute.

[0019] As an optimization scheme of the present invention, the use of association rule mining to discover valuable data patterns with associated relationships hidden in massive data includes: establishing a decision table input database based on the abnormal operating status information of electrical equipment to form a data source as a transaction table for association rule mining processing, and then using the DLG algorithm to obtain frequent item sets, and performing interactive mining by modifying the association rule parameters to determine the appropriate support threshold, confidence threshold and relevance threshold, and based on the extracted strong association rules, the area where the electrical equipment operating status is abnormal can be inferred.

[0020] As an optimization solution of the present invention, the area where the abnormal operation status of the electrical equipment can be inferred based on the extracted strong association rules includes: given a set of abnormal operation status information of the electrical equipment C = {C 1 ,C 2 ,...C m} and the abnormal operation status area S = {Sec1,Sec2,...,SecN}, the abnormal identification association rule is the implication of X=>Y. If the expression of the association rule obtained by the algorithm is {C 1 ,C 2 ,C 3 =>Sec2}, which means that the abnormal area identified by the operating status information of these electrical devices is Sec2.

[0021] The present invention has positive effects: 1) In order to solve the problems of high cost and poor model universality in the process of identifying abnormal operating conditions of electrical equipment, the present invention utilizes association rule mining to discover valuable data patterns with associated relationships hidden in massive data, which has important practical value for the generation of identification decisions. Under the conditions of a certain power grid scale and incomplete protection action information, this method has high accuracy in abnormal identification, good fault tolerance and strong practicality.

[0022] 2) When the present invention uses association rule mining to simplify attributes, it can directly obtain the best simplified transaction table by modifying the threshold for interactive mining, eliminating the computational complexity of the traditional algorithm that uses rough sets to simplify attributes and then uses the best attribute simplification combination selection method based on average information.

[0023] 3) The present invention can quickly obtain correct abnormal operation status identification results when the abnormal information is accurate and complete, and the missing or erroneous abnormal information is not a core attribute, and the calculation results show that the accuracy of the identification is high. When the core attribute information is partially lost or erroneous, the correct identification rate is not high, but when the probability of abnormality in the set abnormal area is not the highest, it ranks second, which has an auxiliary effect on the decision-making of the dispatcher, indicating that the method has a certain fault tolerance.

[0024] 4) The present invention utilizes the DLG algorithm, whose association graph structure reduces two possible overheads of the Apriori series algorithms, namely, the possible need to generate a large number of candidate sets and the need to repeatedly scan the database. The DLG algorithm checks a large candidate set through pattern matching, thereby effectively reducing the IO cost of the abnormality identification process and improving the computational efficiency of the abnormality identification model, which is also extremely important for improving the effectiveness and real-time performance of power grid fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0026] Figure 1 A flowchart of identifying abnormal operating status of electrical equipment based on associative hybrid reasoning provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] The terms "first" and "second" and the like in the specification and claims of the present invention and the above drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but may include steps or units that are not listed.

[0029] like Figure 1 As shown, the present invention discloses an abnormal identification method for the operation state of electrical equipment based on associated hybrid reasoning, including: obtaining abnormal operation state information of electrical equipment, introducing an association rule data mining DLG algorithm into the abnormal identification of the operation state of electrical equipment, firstly taking protection as a condition attribute and abnormal area as a decision attribute, examining various abnormal situations and establishing an original decision table, then using association rule mining to simplify attributes, performing interactive mining by modifying thresholds, directly extracting the best attribute simplification combination, and then using the simplified decision table formed by the best attribute simplification combination and the association rule interactive mining to perform abnormal identification for abnormal operation state information of various situations of electrical equipment; using association rule mining to discover valuable data patterns with associated relationships hidden in massive data, and obtaining the area where the operation state of the electrical equipment is abnormal

[0030] The technical solution of the present invention includes:

[0031] 1. Association rule mining. The object of association rule mining is generally a data set of things. Let I be a set of binary characters, where the elements are called items, and the set of items is called an itemset. An itemset containing k items is called a k-itemset. The frequency of occurrence of an itemset is the number of transactions containing the itemset, which is referred to as the frequency, support or count of the itemset. If the itemset meets the minimum support (set by the user or domain expert), it is called a frequent itemset. The problem of mining association rules is to generate association rules whose support and confidence are respectively greater than the minimum support (minsupp) and the minimum confidence (minconf).

[0032] 2. DLG algorithm, DLG is based on the association graph structure, by reducing the number of scans of the transaction database, thereby effectively reducing the IO cost of the mining process. A reasonable association rule in the results of association mining corresponds to a reduced combination containing an attribute. The DLG algorithm has three steps. The first step is the frequent 1-item set generation stage, which generates frequent 1-item sets and records related information; the second step is the construction of the association graph to show the association between frequent items; the last step is the frequent k-item set (k>2) generation stage, which generates k frequent item sets based on the construction of the association graph.

[0033] 3. The process of identifying abnormal operation status of electrical equipment based on association hybrid reasoning. Using the above concept, according to the problem of identifying abnormal operation status of electrical equipment to be solved by the present invention, firstly, an association rule is defined: given a set of abnormal operation status information of electrical equipment C = {C 1 ,C 2 ,...C m} and the abnormal operation state area S = {Sec1,Sec2,...,SecN}, and further determine the implied expression of the abnormal identification association rule. If the expression of the association rule obtained by the algorithm is {C 1 ,C 2 ,C 3 =>Sec2}, which means that the abnormal area identified by the operating status information of these electrical devices is Sec2.

[0034] Embodiment 1

[0035] In combination with the problem of identifying abnormal operation status of electrical equipment to be solved by the present invention, firstly, an association rule is defined. Given a set of abnormal operation status information of electrical equipment C = {C 1 ,C 2 ,...C m} and abnormal operation state area S = {Sec 1 ,Sec 2 ,...,Sec N}, the anomaly identification association rule is an implication of the form X=>Y, where X and Y are the patterns of C and S respectively.

[0036] Assume that the expression of the association rule obtained by the algorithm is {C 1 ,C 2 ,C 3 =>Sec2}, which means that the abnormal area identified by the operating status information of these electrical devices is Sec2.

[0037] The algorithm flow of the present invention is as follows Figure 1 As shown, firstly, a decision table input database is established according to the abnormal operation status information of the electrical equipment to form a data source, which is used as a transaction table for association rule mining processing. Then, the implemented DLG algorithm is used to obtain its frequent item sets. Interactive mining is performed by modifying the threshold (association rule parameters) to determine the appropriate support, confidence and relevance thresholds. If the rules are reasonable, several association rules that meet the strong association rules are extracted. According to the attribute correspondence in the extracted strong association rules that meet the conditions, the area where the electrical equipment has abnormal operation status can be inferred.

[0038] Embodiment 2

[0039] Corresponding to the process of identifying abnormal operation status of electrical equipment based on association hybrid reasoning disclosed in the first embodiment of the present invention, the second embodiment of the present invention further provides the basic principle of association rules, and the method includes:

[0040] Let I be a set of words describing abnormal operation status, where the elements are called items, and let D be the set of causal relationships T (transaction) between abnormalities, where T is a set of items, and Each abnormal causal relationship has a unique identifier, denoted as TID. Let X be the set of items in I. If Then T is said to contain X. An association rule is an implication of the form X=>Y, where And X∩Y=φ, there are three important indicators for measuring association rules, namely support, confidence and relevance. For the convenience of calculation, values ​​between 0%-100% are used to correspond to its support and confidence.

[0041] The support (s) of the association rule X=>Y represents the proportion of tuples containing X∪Y in the causal set D, which can be expressed in terms of probability theory as follows:

[0042] s(X∪Y)=Pr(X∪Y)

[0043] The confidence (α) of the association rule X=>Y represents the ratio of the number of tuples containing X∪Y to ​​the number of tuples containing X in the causal set D, which can be expressed in probability theory as:

[0044]

[0045] The correlation between X and Y (σ) is expressed in probability theory as:

[0046]

[0047] If Pr(X∪Y)=Pr(X)Pr(Y) holds, it means that the appearance of pattern X is independent of pattern Y. If the correlation is greater than 1, it proves that patterns X and Y are positively correlated. If the correlation is less than 1, it proves that patterns X and Y are negatively correlated.

[0048] Based on the above three indicators, if the association rule X=>Y meets the following three conditions, it is called a strong association rule:

[0049] s(X∪Y)≥s min

[0050] α(X∪Y)≥α min

[0051] σ X,Y ≥σ min

[0052] Among them, here s min , α min >0,σ min >1, all of which can be customized data thresholds. In identifying abnormal operating status of electrical equipment, what we want to obtain is a strong association rule that satisfies the above three formulas.

[0053] Embodiment 3

[0054] Corresponding to the process of identifying abnormal operation status of electrical equipment based on associative hybrid reasoning disclosed in the first embodiment of the present invention, the third embodiment of the present invention further provides a DLG algorithm, which includes:

[0055] The DLG algorithm is an improved algorithm based on the classic Apriori algorithm. DLG is based on the association graph structure. By reducing the number of scans of the transaction database, it effectively reduces the IO cost of the association hybrid reasoning process. Each reasonable association rule in the result of association hybrid reasoning corresponds to a simplified combination containing an attribute.

[0056] The specific implementation steps of the DLG algorithm are as follows:

[0057] Generate frequent 1-item set stage: The DLG algorithm first scans the database to count its support and creates a bit vector for each item. The length of each bit vector is the number of things in the database. The bit vector associated with item i is represented as BVi. If an item i appears in the jth transaction, the jth bit of BVi is set to 1, otherwise the jth bit of BVi is set to 0. The number of 1s in BVi is equal to the number of things supporting item i, that is, the support of item i.

[0058] Construction of association graph phase: Item set {i 1 ,i 2 ,...,i k The support of} is the number of 1s in BVi1∧BVi2∧...∧BVik. After the first step, there is no need to scan the database again. In the association graph construction stage, DLG constructs an association graph to represent the association between items. If the number of 1s in BVi∧Bvj(i<j) is not less than the minimum support, a directed edge from item i to item j is constructed, and the item set {i,j} is a frequent 2-item set.

[0059] Frequent item set generation stage: Based on the association graph, frequent k-item sets (k>2) can be generated. In the association graph construction stage, frequent 2-item sets L2 are generated. In the frequent item set generation stage, the DLG algorithm generates frequent k2-item sets Lk (k>2). The last item of the k2-item set is used to expand the item set to a (k+1)2-item set. Let {i 1 ,i 2 ,...,i k} is a frequent k2-itemset if there is a k To item j, the item set {i 1 ,i 2 ,...,i k} is expanded to k+1 itemsets. If the number of 1s in BVi1∧BVi2∧...∧BVik∧BVj is not less than the minimum support, then {i 1 ,i 2 ,...,i k ,j} is a frequent k+1 item set. If frequent k item sets cannot be generated further, the DLG algorithm terminates.

[0060] The purpose of the present invention is to solve the main problems of the method for identifying the abnormal operation status of electrical equipment, which are concentrated on the poor universality of the model, high operating cost when training large-scale samples, and poor learning ability. The Apriori algorithm improved based on association rule mining, namely the DLG algorithm, is applied to the abnormal identification of electrical equipment, and the simplified decision table formed by the optimal attribute simplification combination and the interactive mining of association rules are used to infer and identify the abnormal operation status of the equipment, so as to achieve the effectiveness and real-time performance of the abnormal operation status identification of electrical equipment at a lower cost.

Claims

1. A method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning, characterized in that: include: The abnormal operation status information of electrical equipment is obtained, and the association rule data mining DLG algorithm is introduced into the abnormal operation status identification of electrical equipment. First, protection is used as the condition attribute and abnormal area is used as the decision attribute. Various abnormal situations are examined and the original decision table is established. Then, the association rule mining is used to simplify the attributes. Interactive mining is performed by modifying the threshold to directly extract the best attribute simplification combination. Then, the simplified decision table formed by the best attribute simplification combination and the association rule interactive mining are used to identify the abnormal operation status information of electrical equipment in various situations. The association rule mining is used to discover the valuable data patterns with correlation relationships hidden in the massive data, and the area where the electrical equipment operation status is abnormal is obtained.

2. The method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning according to claim 1 is characterized in that: The association rule data mining DLG algorithm includes: 1) Generate frequent 1-item set stage: generate frequent 1-item set and record relevant information; 2) Construction of association graphs to show the associations between frequent items; 3) Generate k frequent item sets based on the construction of association graph.

3. The method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning according to claim 2 is characterized in that: The attribute reduction using association rule mining includes: I is a text set describing the abnormal operation status, the elements of which are called terms, D is a set of causal relationships between abnormalities, T is a set of terms, and Each abnormal causal relationship has a unique identifier, denoted as TID. Let X be the set of items in I. If Then T is said to contain X; an association rule is an implication of X=>Y, where And X∩Y=φ.

4. The method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning according to claim 3 is characterized in that: There are three indicators for measuring association rules, namely support, confidence and relevance. The support of the association rule X=>Y represents the proportion of tuples containing X∪Y in the causal set D. The confidence of the association rule X=>Y represents the ratio of the number of tuples containing X∪Y in the causal set D to the number of tuples containing X. The relevance refers to whether X and Y are positively correlated or negatively correlated.

5. The method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning according to claim 3 is characterized in that: The simplified decision table and interactive mining of association rules formed by the optimal attribute simplification combination refers to: using the DLG algorithm based on the association graph structure, reducing the number of scans of the transaction database, thereby reducing the IO cost of the association hybrid reasoning process, and each reasonable association rule in the result of the association hybrid reasoning corresponds to a simplified combination containing an attribute.

6. The method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning according to claim 1 is characterized in that: The method of using association rule mining to discover valuable data patterns with associated relationships hidden in massive data includes: establishing a decision table input database based on abnormal operating status information of electrical equipment to form a data source as a transaction table for association rule mining processing, and then using the DLG algorithm to obtain frequent item sets, and performing interactive mining by modifying association rule parameters to determine appropriate support thresholds, confidence thresholds, and relevance thresholds. Based on the extracted strong association rules, the area where the operating status of the electrical equipment is abnormal can be inferred.

7. The method for identifying abnormal operation status of electrical equipment based on associative hybrid reasoning according to claim 6 is characterized in that: The area where the abnormal operation status of the electrical equipment can be inferred based on the extracted strong association rules includes: given a set of abnormal operation status information of the electrical equipment C = {C1, C2, ...C m } and the abnormal operation status area S = {Sec1,Sec2,...,SecN}, the abnormal identification association rule is the implication of X=>Y. If the expression of the association rule obtained by the algorithm implementation is {C1,C2,C3=>Sec2}, it means that the abnormal area identified by the operation status information of these electrical equipment is Sec2.

Citation Information

Patent Citations

  • Fault diagnosis method based on intelligent power distribution network

    CN110765185A