Method for identifying key area of urban electric power communication network

Abstract

The invention discloses an urban power communication network key area identification method, relates to the field of power communication, and solves the problems that an existing power communication network key area identification method cannot accurately identify a key area in a power communication network and only considers a single node and the like. The method comprises the following steps: constructing an urban power communication network model; determining a damaged area; calculating a site importance degree, a tower node importance degree and an edge importance degree through an entropy weight method, and obtaining a candidate region comprehensive importance degree; and carrying out candidate region initialization by adopting a combined algorithm of a black widow optimization algorithm and a colistia algorithm, calculating a fitness value of the candidate region according to the comprehensive importance of the candidate region, carrying out position updating and the like, and finally obtaining the position of the global optimal candidate region, namely the position of the key region. According to the method, the accuracy and effectiveness of key region identification and subsequent analysis are ensured, and a more effective solution is provided for maintenance and optimization of complex networks such as an electric power communication network and the like.

Description

Technical Field

[0001] The present invention relates to the field of electric power communications, and in particular to a method for identifying key areas of an urban electric power communication network. Background Art

[0002] With the acceleration of urbanization, the demand for urban electricity continues to grow, and the role of power communication networks in urban power systems has become increasingly important. As an important part of the smart grid, the urban power communication network undertakes key tasks such as real-time monitoring, dispatching, load management, and fault diagnosis of urban power equipment. With the rapid development of smart grids, the urban power communication network is no longer a simple information transmission channel, but a core platform that supports the efficient operation and stable power supply of the power system. However, with the increase in the load of the urban power communication network, the complexity of the network structure, and the threat of natural disasters and external malicious attacks, the vulnerability of the urban power communication network has gradually emerged, especially in terms of the reliability of key nodes and key areas. The robustness of the power communication network has become an important factor affecting the stability of the urban power system.

[0003] In fact, the vulnerability of urban power communication networks has been exposed many times in history.

[0005] Therefore, identifying vulnerable areas of the power communication network is a necessary measure to ensure the normal operation of the smart grid. This experiment aims to use innovative algorithms to more comprehensively and accurately evaluate and improve the anti-destruction capability of the urban power communication network by constructing a complex network model including tower nodes and real nodes. Summary of the invention

[0006] The present invention provides a method for identifying key areas in an urban power communication network to solve the problems that the existing key area identification method of the power communication network fails to accurately identify the key areas in the power communication network and only considers a single node.

[0007] A method for identifying key areas of an urban power communication network is implemented by the following steps:

[0008] Step 1: Construct an urban power communication network model G;

[0009] Step 2: Determine the damaged area based on the urban power communication network model G constructed in step 1;

[0010] Step 3: According to the damaged area determined in step 2, the site importance, tower node importance and edge importance are calculated by entropy weight method, and the comprehensive importance of the candidate area is obtained according to the site importance, tower node importance and edge importance;

[0011] Step 4: Use the combined algorithm of the Black Widow optimization algorithm and the slime mold algorithm to initialize the candidate regions, and calculate the fitness value of the candidate regions according to the comprehensive importance of the candidate regions obtained in Step 3;

[0012] Step 5: Sort the fitness values of the candidate regions in Step 4 to obtain new candidate regions, calculate the weight of position update according to the fitness of the new candidate regions, and use the position update weight to calculate the moving direction and step size of the new candidate regions in the search space;

[0013] Step 6: Update the position;

[0014] For each candidate region in the new candidate regions, use a random probability z to determine whether to randomly re-initialize the position of the candidate region. If rand < z, re-initialize the position of the candidate region; otherwise, determine the search strategy through the search ratio p*er; p is the calculation probability, which is a threshold for controlling the position update strategy;

[0015] er = min(1, max(0, 1 - diversity)), where diversity is the ratio of the standard deviation to the mean of the fitness value vector F;

[0016] If rand < p*er, use the Black Widow optimization algorithm to search for the globally optimal candidate region; otherwise, use the slime mold algorithm to search for the globally optimal candidate region; finally, the position of the globally optimal candidate region is the position of the key region.

[0017] Advantages of the present invention:

[0018] The recognition method of the present invention can provide strong theoretical support and technical means for the planning, maintenance and emergency handling of power communication systems by deeply studying the topological structure, service transmission requirements of power communication networks, and the importance of different nodes and links, ensuring that the power communication network can still maintain high reliability and stability under various adverse conditions and meeting the strict requirements of modern society for power supply.

[0019] In the recognition method of the present invention, by constructing a complex network model including pole tower nodes and real nodes, using intelligent optimization algorithms to identify key regions in the network, and sequentially damaging these key regions according to the importance order through simulation to comprehensively evaluate the impact of damage behavior on network efficiency, and at the same time, by reasonably setting the search distance mechanism to avoid evaluation deviation caused by excessive overlapping area of key regions, and by adding pole tower nodes, it is also possible to more accurately estimate the cost of the lowest complete repair network based on this, ensuring the accuracy and effectiveness of key region recognition and subsequent analysis, and providing a more effective solution for the maintenance and optimization of complex networks such as power communication networks. Description of the Drawings

[0020] Figure 1 It is a flow chart of a method for identifying key areas of an urban power communication network according to the present invention;

[0021] Figure 2 A schematic diagram of the concept of the damage circle in the method of the present invention;

[0022] Figure 3 This is a schematic diagram of a partial power communication network model in a certain city;

[0023] Figure 4 The damage degree (network efficiency) trend diagram of the power communication network at different radii;

[0024] Figure 5 To intercept Figure 4 Damage degree trend chart at r = 0.08;

[0025] Figure 6 This is a graph comparing the network efficiency of this method with existing key node identification methods. DETAILED DESCRIPTION

[0026] Step 1: In the urban power network structure, it is necessary to calculate the weight of the importance of tower nodes. Tower nodes play a key role in connecting and supporting the physical network architecture. However, tower node failure may only affect a small number of lines directly connected to it, unlike key substation node failure or main transmission line failure, which will cause large-scale power outages or network splits. First, this paper conducts spatial modeling based on the power communication network in a certain city. The urban power communication sites and towers are abstracted as network nodes, and the optical cable communication links are abstracted as network edges. Due to the mutual transmission between services, the urban power communication network can be abstracted as a weighted undirected graph.

[0027] Each site in the urban power communication network represents nine different types of communication sites. Based on the physical length of each communication link, it is divided into several sections by towers, and the corresponding tower nodes are abstracted on the communication link.

[0028] Therefore, a power communication network model G has the following expression:

[0029] According to the direct mapping of sites and tower nodes containing longitude and latitude coordinate information, the urban power communication network model is defined as G(V, V', E, W).

[0030] V is the site set, that is, V = {v 1 ,v 2 ,…,v i …v n},v irepresents the ith site, and the number of sites is n. The properties of a site can be expressed as: Vn = (ID, LO, LA), where ID is the node number, LO is the longitude of the node, and LA is the latitude of the node.

[0031] E is the set of edges corresponding to the communication links in the power communication network, that is, E = {e i,j |v i ,v j ∈V,i≠j,e i,j =e ji},e i,j Represents the site v of the power communication network model G i With site v j The edge between. In the set E, if e i,j =1, it means that at site v i With site v j There is an edge connection if e i,j =0, it means that at site v i With site v j There are no edge connections.

[0032] V' is the set of tower nodes, if e i,j =1, then e ij =1 to divide the tower nodes. The tower nodes can be represented by a set: V' = {V' i,j |V' i,j ∈e i,j}in, Represents at site v i With site v j The kth ij Tower nodes; k ij For the side i,j The number of nodes on the upper tower;

[0033] k ij =L ij / (Ls-1). L ij is the physical length of the edge in the network, and Ls is the segmentation length of the tower node.

[0034] Step 2: Determine the circular space damage form and damage method, and determine the damage area.

[0035] In this embodiment, the importance of setting the damage area can be expressed as a triple W=(W V ,W E ,W V' ), where W V =(w V1 ,w V2 ,...w Vi ...,w Vn) is the importance of the site, W E =(w E1 ,w E2 ,...w Ei ...,w En ) is the importance of the edge, W V' =(w V'1 ,w V'2 ,...w V'i ...,w V'n ) is the importance of the tower node.

[0036] The damage form used is: use a circular area to damage the power communication network model G, and set the coordinates of the center of the damaged area to (lo 0 ,la 0 ), the damage radius is r, then the damage area can be defined as D = {(lo i ,la i )|(la i ,la 0 ) 2 +(lo i ,lo 0 ) 2 ≤r 2}, where lo and la are the longitude and latitude of the damaged area respectively; lo 0 and la 0 are the longitude and latitude of the initial population respectively; la i With lo i For site v i The longitude and latitude coordinates of the area; that is, all sites, edges and tower nodes in this area must be damaged, and the damaged sites, edges and tower nodes must be removed from the model.

[0037] The damage method adopted is: obtain the location information of each site in the power communication network model G, traverse the location of each site, and calculate the Euclidean distance dv between the site and the center of the damage area. If dv≤r, the site is considered to be located in the critical area. Traverse all sites in the critical area and delete the site and its associated edges. Traverse each edge and calculate the Euclidean distance de between the edge and the center of the critical area. If de≤r, the edge is considered to cross the critical area, and the deleted edge and site are added to the site removal table remove_node and the edge removal table removed_edges list.

[0038] Step 3: Based on the damaged area determined in step 2, define the calculation rules for the importance of the damaged area, calculate the importance of the site, the importance of the tower node, the importance of the edge, and calculate the comprehensive importance of the area through the entropy weight method.

[0039] In this implementation, the importance of sites and edges in the damaged area is calculated from three aspects: topology, service, and voltage level.

[0040] Step 31, calculating the importance of sites in the damaged area from three aspects: topology, service, and site level;

[0041] Step 311: Obtain the topological importance set of the site TI by calculating the betweenness centrality = (C B (v 1 ),C B (v 2 ),...C B (v i )...,C B (v n )), C B (v i ) is site v i The topological importance of .

[0042]

[0043] Among them, v k and v j are any two different i Site. jk Indicates that from site v j To site v k The total number of shortest paths, σ jk (v i ) indicates that from site v j To site v k and passes through site v i The number of shortest paths.

[0044] Step 312: Obtain the service importance vector BI of the site by calculation = (BI (v 1 ), BI(v 2 ), ..BI(v i )..., BI(v n )), where BI(v i ) is site v i business importance.

[0045]

[0046] Where T is the site v i The number of business types. BW is the weight of the business type. bn(v i ) is site v i The number of each type of business, M is the total number of business types of all sites.

[0047] Step 313, taking into account the grid level factors, including 500kv substation, 220kv substation, 66kv substation, user communication station, provincial secondary unit communication station, municipal secondary unit communication station, provincial company communication station, power plant, and municipal company communication station. They are divided into 9 levels according to their importance ranking, with the most important station having an importance value of 9 and the lowest importance value being 1.

[0048] In this implementation, the importance of the site W is calculated based on the entropy weight method. v The specific calculation process of is:

[0049] The original data is standardized, and the standardization formula is as follows: Among them, x cd is the cth site at the dth index (C B (v i ), BI(v i ), voltage level). cd * is the standardized data of the cth node on the dth index, and the standardized data matrix is ​​obtained, x cd * ∈(0,1). max(x d ) and min(x d ) are the minimum and maximum values ​​of the dth indicator.

[0050] After standardization, calculate the proportion of the site under each indicator p cd , the calculation formula is: Get the proportion matrix, where: p cd is the proportion of the cth node under the dth index, It is the sum of the standardized values ​​of all solutions under the dth indicator.

[0051] Then calculate the information entropy value H of each site d , the entropy value calculation formula is as follows:

[0052] in: is a constant and n is the number of sites.

[0053] According to the calculated entropy value, calculate the weight w of each indicator d , the formula is: Finally, the comprehensive score of each site is calculated by weight and standardized data, that is, the importance of the cth site w vc =w 1 *p c1 +w 2 *p c2 +...w d *pcd ...+w n *p cn .

[0054] Step 32: Calculate the importance of edges in the damage area from three aspects: topology, service, and site level;

[0055] Step 321: Calculate the topological importance of the edge, the business importance of the edge, and the voltage level of the edge from the perspective of edge betweenness according to the entropy weight method; i and v j The edge ij , where edge betweenness centrality C B The calculation formula for (e) is:

[0056]

[0057] where v m and v n are any two sites in the network. mn From site v m To site v n The total number of shortest paths, σ mn (e) is from site v m To site v n The shortest path includes edge e ij The number of

[0058] Step 322: assign different weights W to different service types u according to the criticality of the service and the impact on the network. u Then the business importance of the edge B(e) can be expressed as:

[0059]

[0060] Among them, U is the number of edge business types, w u is the weight coefficient of the u-th type of business, n e,u is the number of the u-th business type on the edge.

[0061] Step 323: assign different weights according to the voltage level of the power transmission line corresponding to the communication link. 220 =0.6; 66kV lines are generally used to supply power to larger user groups or small substations, and are less important, so they are given a weight of W 66 =0.3; 10kV lines are mainly used to distribute power to end users locally, so they are relatively less important and are given a weight of W 10 =0.1.

[0062] In this implementation, the specific process of calculating the importance of edges based on the entropy weight method is as follows:

[0063] The original data is standardized, and the standardization formula is as follows: Among them, x cd The cth edge has the dth index (C B (v i ), BI(v i ), voltage level). cd * is the standardized data of the cth edge at the dth index, and the standardized data matrix is ​​obtained, x cd * ∈(0,1). max(x d ) and min(x d ) are the minimum and maximum values ​​of the dth indicator.

[0064] After standardization, calculate the proportion of edges under each indicator p cd , the calculation formula is: Get the proportion matrix, where: p cd is the proportion of the cth edge under the dth index, It is the sum of the standardized values ​​of all solutions under the dth indicator.

[0065] Then calculate the information entropy value H of each edge d , the entropy value calculation formula is as follows: in: is a constant and n is the number of sites.

[0066] According to the calculated entropy value, calculate the weight w of each indicator d , the formula is: Finally, the comprehensive score of each edge is calculated by weight and standardized data, that is, the importance of the edge. As shown in the following formula.

[0067] w Ec =w 11 *p c1 +w 22 *p c2 +...w dd *p cd ...+w nn *p cn

[0068] Step 33: Calculate the tower node V′ based on the importance of the two adjacent sites and the importance of the edge where they are located ij The importance of W V′i , the formula is as follows:

[0069]

[0070] Since the importance of edges and sites is much greater than that of tower nodes, the importance of edges and sites will be given greater weights during calculation, and the importance of tower nodes will be given a smaller weight. The weights of tower node importance, site importance, and edge importance are And meet at the same time and set up

[0071] Step 34: Calculate the comprehensive importance of the candidate area based on the importance of the site obtained in step 31, the importance of the edge in step 32, and the importance of the tower node in step 33.

[0072] Assume that each site v i Has an importance w Vi , tower node V′ ij Has an importance w V'i , in the entire spatial topological network, an irregular candidate region D(C, r) is generated by Tent mapping. The center coordinates of the candidate region are represented by C, which is the set of central node positions of the candidate region;

[0073] Where C = (c 1 ,c 2 ,...c ρ ...,c pop ),c ρ is the ρth candidate region, pop is the candidate region population size, c ρ ∈[lb,ub], lb and ub are the lower and upper bounds of the candidate region respectively, lb∈

[0074] [max(lo),min(lo)],ub∈[max(la),min(la)], each candidate region D(c ρ ,r) can be expressed as:

[0075]

[0076] Among them, the tower node, site, and edge weights are W 1 , W 2 , W 3 .

[0077] Step 4: Initialize using the combination of the Black Widow Optimization Algorithm and the Slime Mold Algorithm (BW0A-SMA) algorithm; In the initial stage of key area identification, candidate area initialization has an important impact on search performance.

[0078] For the solution space B(c ρ,r), under the condition of given initial radius r = 0.08, set the starting position parameter c of the candidate region center in the network space 0 =(144.01,44.54) and the candidate region population size pop = 150, and the random probability z = 0.03. The candidate region D(C,r) follows the following rules:

[0079]

[0080] Where A is the control parameter that makes the position distribution of each candidate region more random and diverse, and c f is the position of the f-th candidate region.

[0081] Step 5: Calculate the fitness value of the candidate region;

[0082] The fitness value of the candidate region is calculated by the comprehensive importance of the candidate region calculated in step 3, and the fitness function F(c ρ ,r)=-W(c ρ ,r). And sort D(C,r) according to the fitness value to get the new area DI(C,r); the specific process is:

[0083] Calculate each candidate region D(c ρ ,r)’s importance value W(c ρ ,r), each candidate region D(c ρ ,r), through the fitness function F(c ρ ,r)=-W(c ρ ,r) calculates its fitness value, and sorts D(C,r) according to the fitness value to obtain the new candidate region DI(C,r). The lower the fitness value, the more critical the candidate region is, so the algorithm will give priority to the candidate region with low fitness value.

[0084] Step 6: Get the new candidate region DI(C, r) according to the fitness ranking obtained in step 5, and calculate the weight WI of the position update based on the fitness of the new candidate region DI(C, r). Guide the moving direction and step length of the candidate region in the search space.

[0085] Step 61: For DI(C, r), in the θth iteration (θ∈(0, MaxIter-1)), first calculate the fitness difference sf of DI(C, r), sf=fitness_best-fitness_worst, where fitness_worst is the worst fitness value of DI(C, r), fitness_worst=max(F(c ρ, r)), fitness_best is the optimal fitness value of DI(C, r), fitness_best = min(F(c ρ , r)).

[0086] Step 62: For DI(C, r), during the search process, according to the fitness difference sf, use WI to guide the moving direction and step size of the candidate region in the search space.

[0087]

[0088] In the formula, fitnessI is the fitness value; α is a parameter to control the weight adjustment amplitude; ∈ is a constant to prevent the denominator from being zero.

[0089] Step Seven: Position update;

[0090] For each candidate region D(cρ, r), it is determined by the random probability z whether to randomly re-initialize the position of the candidate region. If rand < z, re-initialize the individual position;

[0091] Otherwise, if rand ≥ z, the search strategy is determined by the search ratio p*er;

[0092] When z ≤ rand < p*er, for the candidate region with fitness close to the global optimum, use the black widow optimization algorithm for search. The algorithm position update formula is as follows:

[0093]

[0094] When rand ≥ p*er, use the slime mold algorithm for global search,

[0095]

[0096] Among them, c(θ + 1) is the position of the candidate region after the (θ + 1)-th iteration, c b (θ) is the optimal position of the current fitness value, m is a random number between [0.4, 0.9], β is a random number within [-1, 1], c A (θ), c B (θ) is the position of the randomly selected A, B candidate regions. c(θ) is the position of the current candidate region, vb is a random number [-a, a], a is a constant, rand, rand1 are random numbers in (0, 1).

[0097] In this implementation, p is a threshold value for calculating the probability and controlling the location update strategy, p = tanh(|fitness_best-fitnessI|), where fitness_best is the current global optimal fitness value, and the tanh function maps the fitness difference to the (0, 1) interval. er = min(1, max(0, 1-diversity)), Where σ(F(c ρ , r)) is the standard deviation of the fitness value vector F, μ(F(c ρ , r)) is the mean of the fitness value vector F.

[0098] After the position is updated, the algorithm searches for the position of the global optimal candidate area, which is the key area position.

[0099] In this implementation, it also includes detecting the effect of the key area identification algorithm on the urban power communication network, defining evaluation indicators, network efficiency, and analyzing the network after the key area is damaged.

[0100] During the regional damage process, the network efficiency is recorded and the network efficiency related data is output.

[0101] Network efficiency calculation method:

[0102]

[0103] Among them, d ij For site v i and v j The shortest path between .

[0104] Specific implementation method 2: Combination Figures 2 to 6 This embodiment is described as an example of a method for identifying key areas of an urban power communication network described in Specific Implementation Method 1: To illustrate the process of the method for identifying key areas based on damaged areas, Figure 2 This is an example of regional damage. The black nodes are sites, the gray nodes are tower nodes, and the dotted area is the damage circle, i.e. the damage area. The anti-destruction analysis of part of the power optical cable network in a city in Jilin Province is used as an example:

[0105] 1. Construct a power communication network model;

[0106] The substations at all levels, communication stations and communication sites in each power plant in the power communication network of a certain urban area of ​​a city in Jilin Province are abstracted as sites. The displayed geographical location coordinates of each communication site correspond to the position in the model space; the connections between the communication sites are abstracted as undirected edges; based on the physical length of each communication link, it is divided into several sections by towers, and the corresponding tower nodes are abstracted on the communication link.

[0107] Construct a partial power communication network model G(V,V',E,W) of a city in Jilin Province, such as Figure 3 As shown, Figure 3 This is a schematic diagram of a partial power communication network model in a certain city. The network includes 13 sites, 17 optical cables and 171 tower nodes. By counting the length of each link L and the number of tower nodes si on the link, Ls = 0.22 is calculated.

[0108] 2. According to step 3 in the first specific implementation method, the importance of the site, the importance of the edge and the importance of the tower node are calculated, and the comprehensive importance of the candidate area is calculated by comprehensively considering these three factors.

[0109] 3. Set the method parameters according to step 4 in the specific implementation method 1, set the population size pop = 150, and set the starting position parameter c of the candidate region center in the network space 0 =(144.01,44.54).

[0110] 4. Read the network and coordinate information of the power communication network and establish the corresponding power communication network model. Each slime mold represents an expected damage center in the actual space. Perform circular damage simulation on the position of each individual to calculate the individual fitness F(c ρ ,r), and then iteratively calculate the individual moving direction and step size in space by calculating the updated weight WI. Each iteration determines the search strategy according to the search ratio p*er, so as to update the position, so that the individual gradually approaches the key area of ​​the power communication network and finally reaches the key area.

[0111] 5. Analysis of key areas of power communication network; set the initial five radii r = 0.05, 0.08, 0.1, 0.12, 0.15, select the optimal radius for experiment, and conduct experiments on the five initial radii, select the top k individuals among the individuals, calculate their network efficiency decline, and then select the optimal radius r for key area identification experiment, and obtain the damage degree change trend of the power communication network under different radii, such as Figure 4 shown.

[0112] At the same time, it was found that the network had a lower degree of damage when r<0.08. Compared with the original network efficiency of 0.5021, when r=0.05, the network efficiency after damaging the key area was only less than and close to 0.5. Although the key area in the entire network can be found, the impact on the entire network is relatively small due to the small radius. When r>0.08, as the radius increases, the damage effect is also weakened. When r≥0.12, due to the large radius, it is easy to cause most of the network to fail when damaging the network, resulting in a change in the nature of the network, and the network efficiency will be higher even if the damage range is too large.

[0113] Depend on Figure 5 The chart shows how the network efficiency changes with the ranking of key areas after the damage area when the radius is 0.08. It can be seen from the chart that when the key area is not damaged, the network efficiency is close to 0.5, which is at a relatively high level. Once the key area is damaged, the network efficiency will immediately drop to about 0.47, which shows that the first-ranked area plays a vital role in network efficiency. In contrast, when other ranked areas are damaged, the change in network efficiency is relatively small.

[0114] Table 1

[0115]

[0116] As shown in Table 1, after finding the key area, the tower nodes are numbered in sequence after the site according to the tower node number. These nodes are nodes in the key area. These nodes have important connection functions in the network. For example, node 1 is connected to multiple nodes. 1,2 ,e 1,3 ,e 1,4 ,e 1,12 , its removal will destroy the network connection structure, thus affecting network efficiency. Observing the node information, the sites corresponding to these removed nodes have different business importance and site levels. For example, v 1 The corresponding service importance is 0.5 and the site level is 220kv. These properties indicate that the node may play an important role in the service transmission and structural support of the network, and its damage will have a significant impact on the network, which is consistent with the characteristics of nodes in key areas.

[0117] The comparison results of the method described in this embodiment and the existing key area identification method around key nodes are as follows: Figure 6 As shown in the figure, it can be seen that whether it is the network efficiency after the comparison experimental identification method damages the area when the radius is 0.08 (dashed line) or the network efficiency after the key area is damaged when the radius is 0.08 (solid line), the network efficiency shows an overall upward trend after the area is damaged. From the experimental results, after damaging the key area found by the algorithm in this paper, the network efficiency is 10% higher than the comparison algorithm. This is because the site v in this area 1 and edge 1,2 ,e 1,3 ,e 1,4 ,e 1,12In the network, it undertakes important functions such as business transmission, information exchange and network structure support. After the key area found by the comparison algorithm is damaged, due to its relatively low importance in the network (judged only based on the node importance), the network efficiency after damage is 0.50, and the impact on the overall network efficiency is relatively small, and the network efficiency decreases less than the key area found by the algorithm in this paper. This further proves the effectiveness of the method of the present invention in finding key areas, and can find key areas that have a greater and more comprehensive impact on network efficiency.

[0118] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0119] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for identifying key areas of an urban power communication network, characterized in that: This method is implemented by the following steps: Step 1: Construct an urban power communication network model G; Step 2: Determine the damage area according to the urban power communication network model G constructed in Step 1; Step 3: According to the damage area determined in Step 2, calculate the importance of stations, the importance of pole and tower nodes, and the importance of edges by the entropy weight method, and obtain the comprehensive importance of the candidate area according to the importance of stations, the importance of pole and tower nodes, and the importance of edges; Step 4: Use the combined algorithm of the black widow optimization algorithm and the slime mold algorithm to initialize the candidate area, and calculate the fitness value of the candidate area according to the comprehensive importance of the candidate area obtained in Step 3; Step 5: Sort the fitness values of the candidate areas described in Step 4 to obtain new candidate areas, calculate the weight of position update according to the fitness of the new candidate areas, and use the position update weight to calculate the moving direction and step size of the new candidate areas in the search space; Step 6: Position update; For each candidate area in the new candidate area, use a random probability z to determine whether to randomly re-initialize the position of the candidate area. If rand < z, re-initialize the position of the candidate area; otherwise, determine the search strategy through the search ratio p * er; p is the calculation probability, which is a threshold for controlling the position update strategy; er = min(1, max(0, 1 - diversity)), where diversity is the ratio of the standard deviation to the mean of the fitness value vector F; If rand < p * er, use the black widow optimization algorithm to search for the globally optimal candidate area; otherwise, use the slime mold algorithm to search for the globally optimal candidate area; rand is a random number taken from (0, 1); finally, the position of the globally optimal candidate area obtained is the position of the key area.

2. A method for identifying key areas of an urban power communication network according to claim 1, characterized in that: In Step 1, according to the topological structure and spatial information of the power communication network, construct a power communication network model G(V, V’, E, W); where: V={v1,v2,…,v i ,…,v n } is the site set of the power communication network model G, and the number of nodes is n. E={e i,j |v i ,v j ∈V,i≠j,e i,j =e ji } is the set of edges corresponding to the communication links in the power communication network, e i,j Represents the site v of the power communication network model G i With site v j The edge between; in set E if e i,j =1, then at site v i With site v j There is an edge connection if e i,j = 0, then at site v i With site v j There are no edge connections; V'={V' i,j |V' i,j ∈e i,j } is the tower node set, where: The k ij For the side i,j The number of nodes on the upper tower; k ij =L ij / (Ls-1); Ls is the split length of the tower node, L ij is the physical length of the edge in the network.

3. A method for identifying key areas of an urban power communication network according to claim 1, characterized in that: In Step 2, the specific process of determining the damage area is: A circular area is used to damage the power communication network model G. The center coordinates of the damaged area are set to (lo0, la0) and the damage radius is r. The damaged area is defined as D = {(lo i ,la i )|(la i ,la0) 2 + (lo i ,lo0) 2 ≤r 2 }, where lo and la are the longitude and latitude of the candidate region, lo0 and la0 are the longitude and latitude of the initial population, la i With lo i For site v i The latitude and longitude coordinates of .

4. A method for identifying key areas of an urban power communication network according to claim 3, characterized in that: In Step 2, the damage method used is: obtain the position information of each station in the power communication network model G, traverse the positions of each station, and calculate the Euclidean distance dv between the station and the center of the damage area; If dv ≤ r, it is considered that the station is located within the key area; traverse all stations located within the key area, delete the stations and their associated edges; traverse each edge, and calculate the Euclidean distance de between the edge and the center of the damage area; If de ≤ r, it is considered that the edge crosses the key area, and add the deleted edges and stations to the station deletion table remove_node and the edge deletion table removed_edges list; 5. A method for identifying key areas of an urban power communication network according to claim 4, characterized in that: In Step 3, generate an irregular candidate area D(C, r) through Tent mapping. The center coordinates of the candidate area are represented by C, which is the set of the positions of the central nodes of the candidate area; C=(c1,c2,...c ρ ...,c pop ),c ρ is the ρth candidate region, pop is the candidate region population size, c ρ ∈[lb,ub],lb∈[max(lo),min(lo)],ub∈[max(la),min(la)], lb and ub are the lower and upper bounds of the candidate region respectively; each candidate region D(c ρ ,r) is as follows: Among them, W1, W2, and W3 are the tower node, site, and edge weights respectively; W V is the importance of the site, W E is the importance of the edge, W V' is the importance of the tower node.

6. A method for identifying key areas of an urban power communication network according to claim 5, characterized in that: In Step 4, it is set that the candidate area D(C, r) follows the following rules: In the formula, A is a control parameter that makes the location distribution of each candidate area more random and diverse, and c f is the position of the f-th candidate region; For each candidate region D(c ρ ,r) through the fitness function F(c ρ ,r)=-W(c ρ ,r) calculate its fitness value.

7. A method for identifying key areas of an urban power communication network according to claim 6, characterized in that: The specific process of Step 5 is: Step 5-1: Sort the fitness obtained in Step 5 to obtain a new candidate area DI(C, r), and calculate the position update weight WI according to the fitness value of the new candidate area DI(C, r); Step 52: For the new candidate region DI(C, r), in the θth iteration, first calculate the fitness difference sf of DI(C, r), sf = fitness_best-fitness_worst; where fitness_worst is the worst fitness value of DI(C, r), fitness_worst = max(F(c ρ ,r)), fitness_best is the optimal fitness value of DI(C,r), fitness_best=min(F(c ρ ,r));θ∈(0,MaxIter-1); Step Five Three. According to the fitness difference sf, use the weight WI of position update to guide the moving direction and step size of the candidate region in the search space; which is expressed by the following formula: In the formula, α is a parameter to control the weight adjustment amplitude; ∈ is a constant.

8. A method for identifying key areas of an urban power communication network according to claim 7, characterized in that: In Step Six, when rand < p * er, for the candidate region with fitness close to the global optimum, use the black widow optimization algorithm for search; which is expressed by the following formula: When rand ≥ p * er, use the slime mold algorithm for global search, which is expressed by the following formula: Among them, c(θ+1) is the candidate region position after the θ+1th iteration, c b (θ) is the optimal position of the current fitness value, m is a random number between [0.4, 0.9], β is a random number between [-1, 1], c A (θ),c B (θ) is the position of the randomly selected candidate region A, B. c(θ) is the position of the current candidate region, vb is a random number [-a, a], and a is a constant.

Images