Visualization method of power optical communication network based on improved Louvain algorithm

Abstract

The electric power optical communication network visualization method based on the improved Louvain algorithm relates to the field of operation and maintenance management technology of the electric power optical communication network, solves the problems that the community structure to which the node belongs is ignored in the existing electric power optical communication network, resulting in inconsistent or misplaced node positions within the community, and then leading to low efficiency and misjudgment that cannot meet the existing application needs. The method can automatically detect the community structure in the network, divide the device nodes in the network into closely related communities, introduce the use of the node degree to select the seed node when dividing the community, and set two thresholds. On the one hand, it can reduce the number of candidate nodes and avoid useless comparisons, thereby improving the efficiency of the algorithm; on the other hand, it can exclude nodes with larger degrees, prevent the divided community from being too large, improve the quality of the algorithm's community division results, and thus reveal the internal relevance of the network. The present invention can significantly improve the reliability and efficiency of network operation and maintenance.

Description

Technical Field

[0001] The present invention relates to the technical field of power optical communication network operation and maintenance management, and in particular to a power optical communication network visualization method based on an improved Louvain algorithm. Background Art

[0002] Effective management of the power optical communication network is an important guarantee for supporting the operation of smart grids. As a solid foundation for the information construction of the industry, power communication not only provides efficient and reliable communication capabilities, but also provides strong support for the operation and management of the power grid. However, with the rapid expansion of the scale of power optical communication networks, the difficulty of management has also increased. Traditional power optical communication network management methods mostly rely on data presentation in text and table form. Managers need to monitor and maintain the network by reading a large number of data reports and logs. Although this method can provide detailed information, it has limitations in quickly locating problems and grasping the overall network status. Especially when faced with complex network structures and large-scale data, traditional methods may lead to inefficiency and misjudgment. Obviously, traditional management methods can no longer meet existing needs.

[0003] Existing technology: He Zhifang, Zhang Xiuli, Zhu Yanfang, Xu Limei, Liu Yu. Research on multi-path method of power grid business transmission based on force-directed algorithm [J]. Modern Industrial Economy and Informatization, 2021, 11(05). This technology uses the traditional force-directed layout algorithm to visualize the power grid business transmission path, but ignores the community structure to which the nodes in the network belong, which may cause inconsistent or misplaced node positions within the community, and will face efficiency and scalability issues when the scale of the power grid expands.

[0004] Therefore, the present invention proposes a visualization method for electric power optical communication network based on the improved Louvain algorithm. By introducing network community division technology and visualization technology, the electric power optical communication network is divided into communities, and the network topology and business data are visualized, so that managers can clearly understand the connection relationship between transmission sites, equipment and infrastructure, and perform operation and maintenance management more efficiently. Summary of the invention

[0005] The present invention provides a visualization method for power optical communication network based on an improved Louvain algorithm to solve the problems that the community structure to which the nodes belong is ignored in the existing power optical communication network, resulting in inconsistent or misplaced node positions within the community, and when the scale of the power grid is expanded, it will face efficiency and scalability problems, which will lead to low efficiency and misjudgment and cannot meet existing application needs.

[0006] The electric power optical communication network visualization method based on the improved Louvain algorithm is implemented by the following steps:

[0007] Step 1: Based on the optical communication equipment management system, the data of the service transmission equipment of the power optical communication network and the connection relationship data between the starting station equipment and the ending station equipment are obtained, and the obtained data set is used as the service transmission equipment data set; based on the service type, the service data set is obtained from the dispatch telephone service system, the video service system, and the relay protection service system for storage, and the obtained data set is used as the service data set;

[0008] The service transmission equipment data set includes equipment model, starting station equipment, and ending station equipment, constituting a service transmission set;

[0009] Step 2: Based on the service transmission equipment data set obtained in step 1, the improved Louvain algorithm is used to perform community division of the power optical communication network to obtain the community division result C′. The specific process is as follows:

[0010] Step 21: Take the device as a node, determine the connection relationship between the nodes, draw a network diagram, and add attribute information as needed to obtain a preliminary community division result C;

[0011] Step 22: Compress the community division result C obtained in step 21 to obtain a new network graph G′(V′,E′); where V′ is the new node set {V1′,V2′...,V r ′}, E′ is the new edge set {E1′,E′2...,E′ r};

[0012] Step 2: Select seed nodes based on the node degree for the new network graph G'(V', E'), define two thresholds as the upper and lower limits of the node degree, and the seed node set P is expressed as follows:

[0013]

[0014] In the formula, deg(v s ) is the node v s g is the average node degree, q is the standard deviation of the node degree, c and d are both constants;

[0015] Step 24: For non-seed node v′ o , v′ o ∈V′-P,s≤nr, then they are assigned to the corresponding new community. When all non-seed nodes v′ o The community it belongs to no longer changes, and the community division result C′ is output;

[0016] Step 3: Apply the community division result C′ to the attribute-constrained force-directed layout algorithm to layout the network graph in two-dimensional or three-dimensional space, and use visualization tools to visualize the layout results.

[0017] Beneficial effects of the present invention:

[0018] (1) In the method described in the present invention, an improved Louvain algorithm is proposed that can automatically detect the community structure in the network and divide the device nodes in the network into closely related communities. When dividing the community, the degree of the node is introduced to select the seed node, and two thresholds are set. On the one hand, the number of candidate nodes can be reduced and useless comparisons can be avoided, thereby improving the efficiency of the algorithm; on the other hand, nodes with larger degrees can be excluded to prevent the divided communities from being too large, thereby improving the quality of the algorithm's community division results, thereby revealing the internal correlation of the network.

[0019] (2) The force-directed layout algorithm based on attribute constraints can present the nodes and links in the network in an intuitive way on a two-dimensional plane. The distance and position relationship between nodes can reflect the relationship between the nodes, making the structure and characteristics of the network clear at a glance, which helps managers to fully understand the network situation and clearly understand the connection relationship between transmission sites, equipment and infrastructure, so as to better plan, configure and troubleshoot. Visualizing a large amount of business data by module according to business type can provide managers with a more intuitive understanding of the data, understand the business operation of the network, improve decision-making efficiency and accuracy, and better understand the business situation.

[0020] (3) In the present invention, the combination of the improved Louvain algorithm and the force-directed layout algorithm based on attribute constraints has strong scalability in network analysis and can clearly display the community structure and node layout of the network when facing a complex network. In addition, users can perform interactive operations such as zooming, panning, and selecting nodes as needed to better understand the details and changes of the network.

[0021] In summary, for a large network such as the electric power optical communication network, by visually analyzing the community structure, node layout and business data of the electric power optical communication network, the network structure and network business data can be displayed intuitively in front of managers, which can help find bottlenecks, weak links and optimization space in the network, provide decision support and improve network management strategies. Compared with the operation and maintenance management method of reading a large amount of log data and manually analyzing data, it can significantly improve the reliability and efficiency of network operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The present invention provides a flowchart of a method for visualizing a power optical communication network based on an improved Louvain algorithm.

[0023] Figure 2 This is a flow chart of the improved Louvain algorithm for dividing the power optical communication network community in the present invention.

[0024] Figure 3 The non-seed node v is allocated in the present invention o ′Flowchart of .

[0025] Figure 4 The present invention uses a force-directed layout algorithm with attribute constraints to realize a visualization flow chart of the layout of a power optical communication network diagram. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be described clearly and completely below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, other embodiments obtained by ordinary technicians in this field without creative work should all fall within the scope of protection of this specification.

[0027] Combination Figures 1 to 4 This embodiment describes a method for visualizing an electric power optical communication network based on an improved Louvain algorithm. The core of this method is the combination of two key technologies: one is an improved Louvain algorithm, which is used to efficiently and clearly identify and divide the community structure of the electric power optical communication network; the other is an attribute-constrained force-directed layout algorithm, which is used to visualize the topology and business data of the network, so that managers can fully understand the network situation and perform operation and maintenance management more efficiently. This implementation method is implemented by the following steps:

[0028] Step 1: Based on the optical communication equipment management system, the data of the service transmission equipment of the power optical communication network and the connection relationship data between the starting equipment and the terminating equipment are obtained, and the obtained data set is recorded as the service transmission equipment data set; based on the service type, the service data set is obtained from the dispatch telephone service system, the video service system, the relay protection service system and other service systems for storage, and the obtained data set is recorded as the service data set;

[0029] Specifically, the service transmission device data set specifically includes: device model, starting station device, and terminating point device; forming a service transmission set M = {[1,3], [5,7]......[a,b]}, wherein a in [a,b] represents the starting station device number, and b represents the terminating station device number;

[0030] Step 2: Based on the service transmission equipment data set obtained in step 1, the power optical communication network community is divided based on the proposed improved Louvain algorithm, which is specifically achieved through the following steps:

[0031] First, the basic definition of graph is given;

[0032] Let G be a network graph, denoted as G(V,E), where V is a set of nodes {V1,V2...,V n}, E is the set of edges {E1,E2...,E n}.

[0033] There are many problems with the traditional Louvain algorithm, mainly including the iterative execution in stage 2 until the modularity value Q no longer changes, and then selecting the division corresponding to the maximum modularity value Q as the final community division result, resulting in slow running time, excessive merging of large communities, and a large number of small communities, which makes it impossible to quickly and clearly display network data information. Therefore, for large networks such as power optical communication networks, the traditional Louvain algorithm cannot meet its requirements for community division quality and algorithm efficiency. Therefore, in this implementation, improvements are made to the traditional Louvain algorithm. The improvement is mainly to set two thresholds when selecting seed nodes, as the lower limit and the upper limit respectively. On the one hand, the upper limit can be used to exclude nodes with smaller degrees, and the upper limit can be used to exclude nodes with larger degrees, so as to prevent the community from being too large and difficult to find a more detailed community structure, thereby improving the quality of the division; on the other hand, the number of candidate nodes for the seed node is reduced, thereby improving the running time of the algorithm. Figure 2 As shown, the specific steps include:

[0034] Step 21: Take the device as a node, determine the connection relationship between the nodes, draw a network diagram, add attribute information as needed, execute stage 1 of the Louvain algorithm, and obtain a preliminary community division result C;

[0035] Each device node is classified into a community. For each node v i (i=1,2,...,n), assuming that its neighbor nodes belong to N communities, node v i Add them to N communities one by one, and calculate the change in modularity value ΔQ before and after adding according to the following formula t (t=1,2,...,N), if maxΔQ t >0, then the node v i Assign to maxΔQ t The corresponding community, otherwise the node v i The community you belong to remains the same;

[0036]

[0037] Among them, ∑ in is the sum of all internal edge weights in community A(A∈N); Σ tot is the sum of the edge weights of all nodes adjacent to the community A; k i are all adjacent nodes v iThe sum of the edge weights of k i,in are all slave nodes v i The sum of the edge weights adjacent to the nodes in community A; m is the sum of the weights of all edges in the network.

[0038] Step 22: Compress the community division result C to obtain a new network graph G′(V′, E′);

[0039] The purpose of compressing the community division result C is to simplify the network structure and make the subsequent community detection steps more efficient, which generally includes node merging and edge reconstruction; where V′ is the new node set {V1′, V2′…, V r ′}, E′ is the new edge set {E1′,E′2…,E′ r}, where r≤n.

[0040] Step 23: For the network graph G′(V′,E′), obtain the seed node set P according to formula (2);

[0041] In this implementation, seed nodes are selected based on the node degree, and two thresholds are defined as the upper and lower limits of the node degree, respectively. This can reduce the number of candidate nodes, avoid iterative execution of stage 2 of the traditional Louvain algorithm, and significantly improve the algorithm efficiency.

[0042]

[0043] Among them, deg(v s ) represents node v s g represents the average node degree, q represents the standard deviation of the node degree, and c and d are both constants.

[0044] Step 24: For non-seed node v′ o (v′ o ∈V′-P,s≤nr), assigned to the corresponding new community;

[0045] like Figure 3 As shown, for each non-seed node v′ o , the non-seed node v′ o Add it to the T (T<N) communities to which its neighboring nodes belong in turn, and calculate the change in modularity value ΔQ′ before and after adding u (u=1,2,…,T), if the community contains a seed node, then add the community to set D1, otherwise add it to set D2. It should be noted that D1 is the community set containing seed nodes, and D2 is the community set not containing seed nodes. Calculate ΔQ′ in set D1 u The maximum value of is recorded as maxΔQ′. If maxΔQ′>0, the non-seed node v′ oAssign to the community corresponding to maxΔQ′; otherwise calculate ΔQ′ in set D2 u The maximum value of is recorded as maxΔQ″. If maxΔQ″>0, the non-seed node v′ o Assign to the community corresponding to maxΔQ″; if it does not meet the above two conditions, then the non-seed node v′ o The community you belong to remains the same.

[0046] Step 25: Repeat step 24 until all non-seed nodes v′ o The community you belong to no longer changes;

[0047] Step 26: Output the community division result C′ and the algorithm ends.

[0048] Step 3: Apply the community division results to the attribute-constrained force-directed layout algorithm to layout the network graph in two-dimensional or three-dimensional space, and use visualization tools (such as D3.js) to visualize the layout results to present the overall network structure and community structure relationship;

[0049] like Figure 4 As shown in the figure, the community division result C′ obtained by Louvain algorithm adds spring force and repulsive force constraints to adjacent communities and adjacent nodes on the basis of the traditional FR (Fruchterman-Reingold) algorithm; adjacent communities here refer to communities with different attributes, and adjacent nodes refer to nodes existing in the same attribute community;

[0050] Step 31: According to the input community division result C′, randomly initialize the positions of all nodes in the community division result C′;

[0051] Step 32: Calculate the spring force and repulsive force between adjacent communities according to formula (3) (7);

[0052] Specifically, the spring force of the adjacent community Acting on two attribute communities connected by an edge, the calculation formula is as follows:

[0053]

[0054] in, is the edge for calculating the spring force, d(v1,v2) is the Euclidean distance between the node seeds of adjacent communities, l(v1,v2) is the ideal distance between the seed nodes of adjacent communities, is the elastic coefficient of this side, and the formula is as follows:

[0055]

[0056] Among them, deg(v1) and deg(v2) are the degrees of the adjacent community seed nodes respectively.

[0057] Because the number of nodes in communities with different attributes varies, the rational distance l(v1,v2) between communities is as follows:

[0058]

[0059] Among them, l0 is the basic distance between communities; n is the number of all nodes; is the number of nodes contained in the community with v1 and v2 as seed nodes respectively; l max is the maximum distance from the edge, which depends on the size of the layout area and is calculated as follows:

[0060]

[0061] Among them, w and h are the width and height of the layout area respectively.

[0062] The repulsive force is the force that the attribute community is repelled by others. The repulsive force calculation formula between communities with v1 and v2 as seed nodes is as follows:

[0063] f r (v1,v2)=s(v1,v2)·x (7)

[0064] Among them, s(v1,v2) is the repulsion strength between communities with v1 and v2 as seed nodes, and x is a multi-segment function. The formula is as follows:

[0065]

[0066] Here, k is a coefficient between 0 and 1.

[0067] Step 33: Calculate the spring force and repulsion between adjacent nodes according to formula (9) (12); Adjacent nodes v in the same attribute community i and node v j The spring force f a (v i ,v j ) acts on a node connected by an edge, and the calculation formula is as follows:

[0068]

[0069] Among them, d(v i ,v j ) is the adjacent node v i and node v j The Euclidean distance between i ,v j ) is the adjacent node v i and node v j The ideal distance is calculated as follows:

[0070]

[0071] Among them, l max is the maximum distance of the edge, and n is the number of all nodes; is the adjacent node v i and node v j The number of nodes in the community, and They are the adjacent nodes v i and node v j degree max is the maximum degree of all nodes, For connecting node v i With v j The elastic coefficient of the edge is calculated as follows:

[0072]

[0073] Node v j Subject to node v i The repulsive force calculation formula is as follows:

[0074] f r (v i ,v j )=s(v i ,v j )·y (12)

[0075] Among them, s(v i ,v j ) is the adjacent node v i and node v j The repulsion strength between them, y is a multi-segment function, and the formula is as follows:

[0076]

[0077] Among them, t is a coefficient between 0 and 1.

[0078] Step 34: Calculate the new position of the node according to the forces obtained in steps 32 and 33, and update the node position;

[0079] Step 35: If the positions of all nodes are considered to be stable after the specified number of layout iterations, the position of each node is output, the algorithm ends, and the final new network layout graph is obtained;

[0080] Step 4: Based on the business data set obtained in step 1, call the graphics library Echarts.js and D3.js for data visualization based on the front-end underlying drawing technology SVG technology and Canvas technology;

[0081] In this implementation, there are more than ten types of services in the power optical communication network, including dispatching data network services, dispatching automation services, distribution automation services, dispatching telephone services, administrative telephone services, video conferencing services, relay protection services, safety automatic device services, integrated data network services, video services and other services. These business data correspond to specific business systems, including dispatching data network business systems, dispatching automation business systems, distribution automation business systems, dispatching telephone business systems, administrative telephone business systems, video conferencing business systems, relay protection business systems, safety automatic device business systems, integrated data network business systems, video business systems and other business systems. The data set management in these systems is scattered and the information is not interoperable. In this implementation, the business data obtained from these business systems is stored in the same database for unified management.

[0082] Step 5: Design interface interactive operations to support functions such as zooming in and out and selecting nodes to achieve dynamic visualization of the power optical communication network.

[0083] The improved Louvain algorithm described in this embodiment is used for simulation experiments, which shows that the improved Louvain algorithm can improve the quality and speed of network community division compared with the traditional community division algorithm. Commonly used measurement indicators for community division algorithms include modularity and running time; the higher the modularity value, the better the quality of community division. Therefore, this simulation experiment uses three data sets with different network sizes, namely Facebook1 (148 nodes, 1692 edges), Facebook2 (1893 nodes, 13835 edges) and Hamsterster (2000 nodes, 16098 edges); the traditional Louvain algorithm is used to compare with the method of this embodiment, as shown in Table 1. Table 1 is the comparison data of the simulation experiment conducted by the improved Louvain algorithm of the present invention.

[0084] Table 1

[0085]

[0086]

[0087] Among them, compared with the improved Louvain algorithm, the modularity value Q obtained by the traditional Louvain algorithm is smaller and the running time is longer. This is mainly because the improved Louvain algorithm in this paper sets two thresholds when selecting seed nodes, which serve as the lower limit and the upper limit respectively. On the one hand, the upper limit can be used to exclude nodes with smaller degrees, and the upper limit can be used to exclude nodes with larger degrees, so as to prevent the community from being too large and making it difficult to find a more detailed community structure, thereby improving the quality of the division; on the other hand, the number of candidate nodes for the seed nodes is reduced, thereby improving the running time of the algorithm.

[0088] The method described in this embodiment introduces network community division technology and visualization technology into the management of the electric power optical communication network. The community discovery algorithm and the network layout visualization algorithm are combined to present the electric power optical communication network in a graphical manner, thereby intuitively revealing the community structure, topology and connection relationship of the network. Managers can clearly understand the connection relationship between transmission sites, equipment and infrastructure, so as to better plan, configure and troubleshoot. Compared with traditional methods, the visualization management method not only helps managers to fully grasp the network situation, but also quickly discover and solve problems, and improve operation and maintenance efficiency. In addition, the business types of electric power optical communication networks are becoming increasingly rich, resulting in an increase in business data and an increase in processing pressure. Traditional manual analysis methods are difficult to meet the requirements of efficiency and accuracy. The present invention introduces data visualization and data analysis technology to better handle massive amounts of business data. By visually displaying business data, managers can intuitively observe the operating status, abnormal conditions and performance indicators of the network, so as to make adjustments and decisions in a timely manner.

[0089] In summary, the network topology is visualized and integrated with the display of business data. This allows managers to not only see the network structure, but also understand the business operation of the network, quickly find problems and take corresponding measures. This will make the operation and management of the power optical communication network more systematic, standardized and scientific, adapt to the development requirements of the smart grid, and improve the operation and maintenance management efficiency of the power optical communication network.

[0090] The above-described embodiments are only preferred implementation modes of the present invention, but they should not be understood as limiting the patent scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0091] This specification and the accompanying drawings are merely exemplary illustrations of the present invention and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present invention. Obviously, those skilled in the art may make various changes and modifications without departing from the scope of the present invention.

Claims

1. A visualization method for power optical communication network based on improved Louvain algorithm, characterized by: The method is implemented by the following steps: Step 1: Based on the optical communication equipment management system, the data of the business transmission equipment of the power optical communication network and the connection relationship data between the starting station equipment and the ending station equipment are obtained, and the obtained data set is used as the business transmission equipment data set; based on the business type, the business data set is obtained from the dispatch telephone business system, the video business system and the relay protection business system for storage; The service transmission equipment data set includes the starting station equipment and the starting station equipment model, the terminating station equipment and the terminating station equipment model; Step 2: Based on the service transmission equipment data set obtained in step 1, the improved Louvain algorithm is used to perform community division of the power optical communication network to obtain the community division result C′. The specific process is as follows: Step 21: Take the device as a node, determine the connection relationship between nodes and draw a network diagram, and add attribute information as needed to obtain a preliminary community division result C; Step 22: Compress the community division result C obtained in step 21 to obtain a new network graph G′(V′,E′); where V′ is the new node set {v'1,v'2...,v' r }, E′ is the new edge set {e'1,e'2...,e' w }; Step 2: Select seed nodes based on the node degree for the new network graph G'(V', E'), define two thresholds as the upper and lower limits of the node degree, and the seed node set P is expressed as follows: Among them, s<r; In the formula, deg(v s ) is the seed node v s g is the average degree of the seed node, q is the standard deviation of the degree of the seed node, and c and d are both constants; Step 24: For non-seed node v′ o , where v′ o ∈V′-P,o≤rs, then they are assigned to the corresponding new community. When all non-seed nodes v′ o The community it belongs to no longer changes, and the community division result C′ is output; Step 3: Apply the community division result C′ to the attribute-constrained force-directed layout algorithm to layout the new network graph in two-dimensional or three-dimensional space, and use visualization tools to visualize the layout results; The specific process is: Step 31: Randomly initialize the positions of all nodes in the community division result C′, and calculate the spring force and repulsive force between adjacent communities; Spring force between neighboring communities The calculation formula is as follows: In the formula, To calculate the edge of the spring force, d(v1,v2) is the Euclidean distance between the seed nodes of adjacent communities, l(v1,v2) is the ideal distance between the seed nodes of adjacent communities, For edge The elastic modulus of The formula for calculating the repulsive force between communities with v1 and v2 as seed nodes is as follows: f r (v1,v2)=s(v1,v2)·x Among them, s(v1,v2) is the repulsion strength between communities with v1 and v2 as seed nodes, and x is a multi-segment function. The formula is as follows: Among them, k is a coefficient between 0 and 1; Step 32: Calculate the spring force and repulsive force between adjacent nodes; Adjacent nodes v in the same attribute community i and node v j The spring force f a (v i ,v j ) acts on a node connected by an edge, and the calculation formula is as follows: In the formula, d(v i ,v j ) is the adjacent node v i and node v j The Euclidean distance between i ,v j ) is the adjacent node v i and node v j The ideal distance Node v j Subject to node v i The repulsive force calculation formula is as follows: f r (v i ,v j )=s(v i ,v j )·y Among them, s(v i ,v j ) is the adjacent node v i and node v j The repulsion strength between them, y is a multi-segment function, and the formula is as follows: Among them, w is a coefficient between 0 and 1; Step 34: Calculate the new position of the node based on the spring force and repulsive force between adjacent communities obtained in step 31 and the spring force and repulsive force between adjacent nodes obtained in step 32, and update the node position to obtain a new network diagram layout.

2. The method for visualizing a power optical communication network based on an improved Louvain algorithm according to claim 1, characterized in that: After step three, step four is also included, in which, according to the business data set obtained in step one, a graphics library is called based on the front-end underlying drawing technology and Canvas technology to perform data visualization.

3. The method for visualizing a power optical communication network based on an improved Louvain algorithm according to claim 1, characterized in that: It also includes step five, designing a visualization interface for interactive operations, supporting the functions of zooming in and out and selecting nodes, and realizing dynamic visualization of the power optical communication network.

4. The method for visualizing a power optical communication network based on an improved Louvain algorithm according to claim 1, characterized in that: The specific process of step 21 is: Take the devices as nodes, determine the connection relationship between nodes and draw a network diagram, and add attribute information as needed, classify each node into a community, and for each node v i , assume that its neighbor nodes belong to N communities, and set node v i Add them to N communities one by one, and calculate the change in modularity value ΔQ before and after adding t , t=1,2,...,N; if the maximum change of modularity value maxΔQ t >0, then the node v i Assign to maxΔQ t The corresponding community, otherwise the node v i The community to which it belongs remains unchanged, and the preliminary community division result C is obtained; In the formula, ∑ in is the sum of all internal edge weights in community A, where A∈N; ∑ tot is the sum of the edge weights of all nodes adjacent to the community A; k i For node v i The sum of all adjacent edge weights; k i,in For all nodes v i The sum of the edge weights adjacent to the nodes in community A; m is the sum of the weights of all edges in the network graph.

5. The method for visualizing a power optical communication network based on an improved Louvain algorithm according to claim 1, characterized in that: The specific process of step 24 is as follows: The non-seed node v′ o Add to the T communities to which its neighboring nodes belong in sequence, T<N; calculate the change in modularity value ΔQ′ before and after adding u , u=1,2,...,T, if the community contains a seed node, then add the community to set D1, otherwise add it to set D2, and calculate the change in modularity value ΔQ′ in set D1 u The maximum value maxΔQ′, if maxΔQ′>0, then the non-seed node v′ o Assign to the community corresponding to maxΔQ′; otherwise calculate ΔQ′ in set D2 u The maximum value of is recorded as maxΔQ″. If maxΔQ″>0, the non-seed node v′ o Assign to the community corresponding to maxΔQ″; if it does not meet the above two conditions, then the non-seed node v′ o The community you belong to remains the same.

Images