Electric power communication network evaluation method based on multi-source data fusion

Abstract

The invention belongs to the technical field of electric power communication network evaluation, and particularly discloses an electric power communication network evaluation method based on multi-source data fusion, which comprises the following steps: calculating node connectivity, link robustness, network efficiency and link redundancy indexes in a logic layer; in the physical layer, the actual laying path efficiency, the coil margin index, the fiber core margin, the common trench risk factor and the common cable vulnerability index are calculated; constructing a logic-physical coupling evaluation model, and identifying a hidden coupling relationship through matrix analysis; and performing weighted fusion on evaluation results of the logic layer, the physical layer and the coupling layer, and outputting a network comprehensive score and a health level. According to the invention, multi-dimensional, full-link and systematic evaluation is carried out on the operation mode of the electric power communication network, so that the scientificity of network planning and the pertinence of operation and maintenance management are improved. According to the method, the problems that an existing electric power communication network evaluation method is insufficient in physical consideration and logic-physical coupling relation identification is insufficient are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power communication network evaluation, and particularly relates to a power communication network evaluation method based on multi-source data fusion. Background Art

[0002] As a key infrastructure supporting the safe and stable operation of the power system and the efficient transmission of information, the scientific evaluation of the operation mode of the power communication network is of great significance for improving network planning efficiency, enhancing system reliability, and ensuring high availability. With the continuous acceleration of the digital and intelligent processes in the power industry, the power communication network is facing the challenges of continuous expansion of business scale and increasing requirements for real-time performance. Traditional evaluation methods are difficult to meet the network design and operation and maintenance needs under the new situation.

[0003] From the perspective of the research and development process, early domestic and foreign research mainly focused on the modeling and analysis of network topological structures. Based on graph theory and complex network theory, indicators such as node degree distribution, connectivity, and path redundancy were used to evaluate network functionality, providing theoretical support for the preliminary design and structural optimization of communication networks. Although such methods have a certain representativeness at the topological level, they pay less attention to the physical layer factors involved in actual deployment, and the applicability of the evaluation results in engineering practice is limited.

[0004] In recent years, the evaluation of power communication networks has gradually expanded towards system stability, reliability, and risk control. Some research has simulated the performance changes of the network under different working conditions by constructing network efficiency models, fault propagation models, and redundant structure models to quantify the robustness and invulnerability of the system. At the same time, some scholars have proposed to refine evaluation indicators into multiple levels such as the topological layer, device layer, and operation layer to achieve multi-dimensional comprehensive evaluation. However, most of the above methods are still limited to logical structure analysis and are difficult to reflect the real operating state of the network at the physical level. In particular, the consideration of key operation and maintenance factors such as the physical routing of optical cables, co-gutter / co-cable relationships, fiber core redundancy, and pipeline resource occupancy is significantly insufficient. In current practical applications, the operation quality of the power communication network is highly coupled with the physical deployment characteristics, and the mapping relationship between the logical topology and the physical path has a decisive impact on the reliability, maintainability, and scalability of the network. Most existing evaluation methods ignore the coupling analysis between the logical layer and the physical layer, resulting in deviations between the evaluation results and the actual operation of the network, thus affecting the scientific nature of the planning scheme and the effectiveness of the operation and maintenance strategy. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems of insufficient consideration of the physical layer and inadequate identification of the logical-physical coupling relationship in existing power communication network evaluation methods, and to propose a power communication network evaluation method based on multi-source data fusion.

[0006] The technical solution of the present invention is as follows: A power communication network evaluation method based on multi-source data fusion, comprising the following steps: Collect the logical topology structure information and physical layer deployment information of the power communication network; According to the logical topology structure information, at the logical level, calculate the node connectivity, link robustness, network efficiency value, and link redundancy; Calculate the comprehensive logical evaluation score based on the node connectivity, link robustness, network efficiency value, and link redundancy; According to the physical layer deployment information, at the physical level, calculate the actual laying path efficiency, the spare capacity index, the fiber core richness, the common trench risk factor, and the common cable vulnerability index; Calculate the comprehensive physical evaluation score based on the actual laying path efficiency, the spare capacity index, the fiber core richness, the common trench risk factor, and the common cable vulnerability index; Construct a logical redundancy matrix and a physical coupling matrix, and quantify the true redundancy through matrix analysis to obtain the logical-physical coupling evaluation score; Perform weighted summation on the comprehensive logical evaluation score, the comprehensive physical evaluation score, and the logical-physical coupling evaluation score to obtain the network health score, and judge the power communication network health level according to the network health score to complete the power communication network evaluation.

[0007] Preferably, the node connectivity is obtained by calculating the network average degree, and the calculation formula of the network average degree is:

[0008] Among them, represents the network average degree, that is, the node connectivity, represents the total number of network nodes, represents the th node in the network, represents the node 's degree, that is, the number of connection links of the node ; The calculation formula of the link robustness is:

[0009] Among them, represents the link robustness, represents the total number of network edges, represents the measured k-edge connectivity, and there is , the measured k-edge connectivity is obtained through Monte Carlo simulation, represents the theoretical maximum connectivity, and there is , represents the minimum value, represents the natural base logarithmic function with a certain base; The calculation formula for the network efficiency value is:

[0010] Where, represents the network efficiency value, represents node to node the number of hops of the shortest path, When it approaches 1, it means that the number of hops of the shortest path between all node pairs in the network is close to 1, that is, the network is a fully connected graph and the communication efficiency is the highest; when is close to 0, it means that the communication efficiency between nodes in the network is low, and there are many long paths or communication bottlenecks; The calculation formula for the link redundancy is:

[0011] Where, represents the link redundancy, is the number of redundant links, is the total number of links.

[0012] Preferably, the calculation formula for the comprehensive score of logical evaluation is:

[0013] Where, represents the comprehensive score of logical evaluation, , , and respectively represent the weights of node connectivity , link robustness coefficient , network efficiency value and link redundancy , and there is .

[0014] Preferably, the calculation formula for the actual laying path efficiency is:

[0015] Where, represents the actual laying path efficiency, represents the total number of evaluated paths, represents the th logical design path length, represents the th actual physical path length; The calculation formula for the disk surplus index is:

[0016] Among them, represents the disk surplus index, represents the unused length of the optical cable redundancy, represents the total laid length of the optical cable; The calculation formula for the core redundancy is:

[0017] Among them, represents the core redundancy, represents the total number of cores of the optical cable, represents the number of occupied cores; The calculation formula for the co - trench risk factor is:

[0018] Among them, represents the co - trench risk factor, is the length of the optical cable laid in the co - trench, is the total laid length of the optical cable, is the risk weight of the trench type; The calculation formula for the co - cable vulnerability index is:

[0019] Among them, represents the co - cable vulnerability index, represents the number of links sharing the same cable, represents the total number of logical links, represents the natural logarithm base, represents the length influence coefficient, represents the maximum continuous length of the co - cable section.

[0020] Preferably, the calculation formula for the comprehensive physical evaluation score is:

[0021] Among them, represents the comprehensive physical evaluation score, represents the actual laying path efficiency, represents the disk surplus index, represents the core redundancy, represents the normalized co - trench risk factor, represents the normalized co - cable vulnerability index, , , , and respectively represent the actual laying path efficiency , the disk surplus index , fiber core redundancy , the normalized co-trench risk factor and the normalized co-cable vulnerability index weights.

[0022] Preferably, the formula for calculating the normalized co-trench risk factor is:

[0023] where represents the hyperbolic tangent function; By normalizing the co-trench risk factor a negative correlation is achieved where the larger the risk value, the lower the standardized score.

[0024] Preferably, the formula for calculating the normalized co-cable vulnerability index is:

[0025] By normalizing the co-cable vulnerability index, the medium and low risk intervals are monitored intensively to avoid over-sensitivity in the high value interval.

[0026] Preferably, the formula for calculating the true redundancy is:

[0027] where represents the true redundancy, , represents the logical-physical coupling evaluation score, represents the logical redundancy matrix, and the logical redundancy matrix The element in it represents the existence of redundant links, represents the physical coupling matrix, and the physical coupling matrix The element represents the link and the link have co-trench / co-cable, represents the Hadamard product, represents the matrix Frobenius norm.

[0028] Preferably, the formula for calculating the network health score is:

[0029] where represents the network health score, represents the weight of the comprehensive logical evaluation score weights, Represents the comprehensive score of physical assessment of the weight, Represents the logical - physical coupling assessment score of the weight.

[0030] Preferably, the determination of the power communication network health level according to the network health score is specifically as follows: The power communication network health level is excellent, and the network health score is: ; The power communication network health level is good, and the network health score is: ; The power communication network health level is qualified, and the network health score is: ; The power communication network health level is sub - healthy, and the network health score is: ; The power communication network health level is at risk, and the network health score is: .

[0031] The beneficial effects of the present invention are as follows: 1. The present invention incorporates the logical topology structure, physical routing information and their coupling relationship of the power communication network into a unified evaluation framework, breaking through the limitations of traditional methods that only start from the service layer or the logical layer, realizing multi - dimensional, multi - level and full - element coverage of the network operation state, and significantly improving the comprehensiveness and accuracy of the evaluation.

[0032] 2. By evaluating key physical - layer indicators such as actual routing efficiency, disk space remaining, fiber core richness, common trench risk and common cable vulnerability, the present invention accurately reflects the engineering status of the power communication network in actual deployment and operation and maintenance, makes up for the lack of understanding of the physical layer in existing methods, and enhances the guiding ability of the model in actual network construction and optimization.

[0033] 3. By constructing a redundancy effectiveness verification model, the present invention quantifies the non - consistency and potential risks between the logical topology structure and the physical routing, effectively identifies the hidden dangers of "redundancy failure" caused by common cables and common trenches, and improves the accuracy and depth of the fault - tolerance analysis of the power communication network.

[0034] 4. The evaluation method of the present invention has a strong engineering implementation basis, can be quickly deployed in combination with existing data platforms such as GIS systems, fiber optic resource databases, and network topology maps, is applicable to multiple business links such as the planning and design, capacity expansion and upgrade, operation monitoring and fault diagnosis of the power communication network, and has good versatility and promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1The following shows a power communication network evaluation method based on multi-source data fusion provided in Embodiment 1 of the present invention.

[0036] Figure 2 The following shows a schematic diagram of the topology logic and physical path matching of a certain power communication subnet in a certain city provided in Embodiment 2 of the present invention. Detailed implementation manners

[0037] Now, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the drawings are merely exemplary, intended to illustrate the principles and spirit of the present invention, and not to limit the scope of the present invention.

[0038] Embodiment 1: As Figure 1 shown, a power communication network evaluation method based on multi-source data fusion includes the following steps: S1. Collect the logical topology structure information and physical layer deployment information of the power communication network; S2. According to the logical topology structure information, at the logical level, calculate node connectivity, link robustness, network efficiency value, and link redundancy; S3. Calculate the comprehensive logical evaluation score based on node connectivity, link robustness, network efficiency value, and link redundancy; S4. According to the physical layer deployment information, at the physical level, calculate the actual laying path efficiency, spare capacity index, fiber core richness, common trench risk factor, and common cable vulnerability index; S5. Calculate the comprehensive physical evaluation score based on the actual laying path efficiency, spare capacity index, fiber core richness, common trench risk factor, and common cable vulnerability index; S6. Construct a logical redundancy matrix and a physical coupling matrix, and quantify the true redundancy through matrix analysis to obtain the logical-physical coupling evaluation score; S7. Perform weighted summation on the comprehensive logical evaluation score, the comprehensive physical evaluation score, and the logical-physical coupling evaluation score to obtain the network health score, and judge the power communication network health level according to the network health score to complete the power communication network evaluation.

[0039] In this embodiment, the logical topology structure of the power communication network is modeled, and indicators such as node connectivity, link robustness, network efficiency, and link redundancy are calculated to reflect the connectivity, stability, and transmission performance of the logical structure.

[0040] In this embodiment, the balance of node connections, i.e., node connectivity, is evaluated by calculating the network average degree, which can effectively eliminate the influence of network scale differences on the evaluation results. The higher the average degree, the more links each node in the network is connected to, the stronger the connectivity of the overall network, and the higher the information transfer efficiency between nodes. The calculation formula is as follows:

[0041] where, represents the network average degree, represents the total number of network nodes, represents the th node in the network, represents node 's degree (number of connected links). First, traverse all nodes, count the number of connected links of each node, then sum up the degrees of all nodes, and divide the sum by the total number of network nodes to obtain the network average degree.

[0042] In this embodiment, link robustness is used to evaluate whether the network remains connected after removing a certain number of links. To improve the efficiency of connectivity evaluation for large-scale networks, the embodiment of the present invention uses Monte Carlo simulation to replace the traditional exhaustive method. Compared with the high computational cost of traversing all link removal combinations one by one, the Monte Carlo method approximates the real result through multiple simulations by randomly sampling some link removal scenarios. This method significantly reduces the computational complexity while ensuring the evaluation accuracy, and is especially suitable for the rapid evaluation of large-scale power communication networks. The link robustness calculation formula is as follows:

[0043] where, represents link robustness, represents the total number of network edges, represents the measured k-edge connectivity, and there is (obtained through Monte Carlo simulation), represents the theoretical maximum connectivity, and there is , represents the logarithmic function with the natural base ; The specific process of the Monte Carlo algorithm simulation is as follows: Initialization: Set the removal counter m = 0; Loop test: Randomly select an unmarked link; virtually remove the link and check the network connectivity; if it remains connected, set the removal counter m + 1 = 1; otherwise record k = m + 1; Monte Carlo optimization: Repeat the loop test 1000 times for sampling, take the mode value of k, and obtain the measured k-edge connectivity.

[0044] In this embodiment, the network efficiency value is currently planned to be evaluated at the logical level, and the shortest path hop count is used to evaluate the efficiency of the power communication network. Specifically, the network efficiency evaluation reflects the convenience of communication between all node pairs in the network. First, the shortest path hop count from node to is obtained through Dijkstra (or other shortest path algorithms), and then the communication efficiencies of all node pairs are summed up, and finally normalized. The specific calculation formula is:

[0045] where, represents the network efficiency value, represents the total number of network nodes; When it approaches 1, it means that the shortest path hop count between all node pairs in the network is close to 1, that is, the network is a fully connected graph and the communication efficiency is the highest. When approaches 0, it means that the communication efficiency between nodes in the network is low, and there may be more long paths or communication bottlenecks.

[0046] In this embodiment, the link redundancy is an important indicator to measure the survivability and reliability of the power communication network. Checking whether each link has an alternative backup link helps to evaluate the connectivity and recovery ability of the network in case of failures or attacks. A higher link redundancy means that when a certain link fails, there is still a viable communication path to ensure that critical data can be transmitted without affecting the normal operation of the power dispatching and control system. It is calculated by the following formula:

[0047] where, represents the link redundancy, is the number of redundant links (which can be identified through topology analysis), is the total number of links.

[0048] In this embodiment, the comprehensive score of the logical evaluation reflects the overall performance at the logical level by quantifying the connectivity, robustness, and efficiency characteristics of the network topology. It combines four core indicators: node connectivity, link robustness, network efficiency, and link redundancy, and the weights are evenly distributed to comprehensively characterize the topology characteristics. The calculation formula is:

[0049] where, represents the comprehensive score of the logical level of the power communication network, and the subscript represents the logical level, , , and ​respectively represent the node connectivity index (normalized value, range [0, 1]), the link robustness coefficient (normalized value, range [0, 1]), the network efficiency value (normalized value, range [0, 1]), and the link redundancy (normalized value, range [0, 1]) of weights, there are .

[0050] In this embodiment, based on the GIS system and on-site measurement data, operation and maintenance related indicators such as the actual laying path efficiency, the coiling margin, the fiber core surplus, the common trench risk factor, and the common cable vulnerability are extracted to evaluate the rationality and physical stability of the actual deployment path.

[0051] In this embodiment, the efficiency of the actual laying path is evaluated by calculating the difference value between the topological straight-line distance and the actual laying length between nodes. When the difference value is too large, on the one hand, it will cause a relatively high laying cost, and at the same time, the long operation line will reduce the network efficiency; the specific calculation formula is:[[]]

[0052] Among them, represents the actual laying path efficiency index, represents the total number of paths to be evaluated, represents the th logical design path length (km, taken from the network planning drawing), represents the th actual physical path length (km, on-site measurement value).

[0053] In this embodiment, the coiling margin (Fiber Slack) is mainly used to measure the redundancy ability and physical reliability of the optical fiber path. Sufficient coiling margin can ensure quick communication restoration in case of network failures, improving the reliability and robustness of the network. The coiling margin index calculation formula is:[[]]

[0054] Among them, represents the coiling margin index, the total length represents the unused length of the optical cable redundancy (m, measured value of the joint box), the coiling margin represents the total laying length of the optical cable (m, project acceptance data).

[0055] In this embodiment, the fiber core surplus reflects the remaining situation of the available fiber cores in the optical cable, directly determining the feasibility of future service growth or new link construction. In case of link failures or optical fiber damage, the spare fiber cores can be used for quick switching to ensure uninterrupted communication and improve the robustness of the power communication network. The calculation formula is:[[]]

[0056] Among them, represents the core redundancy, and the total number of cores : The total number of cores in the optical cable (standard values: 24 / 48 / 96), the used cores : The number of occupied cores (OTDR test result).

[0057] In this embodiment, the common trench risk factor measures the potential risk when the optical cable in the power communication network shares the same trench with other power, telecommunications or municipal facilities. If the optical cable shares the same trench with high-voltage cables or gas pipelines, external construction, natural disasters or equipment failures may affect multiple key infrastructures simultaneously, resulting in cascading failures. The calculation formula is:

[0058] Among them, represents the common trench risk factor, is the length of the optical cable laid in the common trench (km, statistically analyzed by the GIS system), is the total length of the laid optical cable (km, statistically analyzed by the GIS system), is the risk weight of the trench type (direct burial: 1.0, pipeline: 0.6, aerial: 0.3).

[0059] For the above common trench risk factor perform normalization to achieve a negative correlation where the larger the risk value, the lower the standardized score. When the proportion of the common trench length is greater than 30%, the score drops rapidly. When the proportion of the common trench length is less than 10%, the score changes gently, retaining a reasonable discrimination for low-risk areas. The normalization formula is:

[0060] Among them, represents the common trench risk factor after normalization, represents the hyperbolic tangent function.

[0061] In this embodiment, the common cable vulnerability index measures the overall vulnerability when multiple communication links share the same optical cable, that is, the Single Point of Failure (SPOF) risk. If multiple key communication links are carried in one optical cable, the damage of the optical cable may cause a large-scale communication interruption, affecting power grid dispatching and control. The calculation formula is:

[0062] Among them, represents the common cable vulnerability index, represents the number of links sharing the same cable, represents the total number of logical links, represents the natural base represents the length influence coefficient (default 0.02 / km), represents the maximum continuous length of the coaxial cable section (km).

[0063] For the above coaxial cable vulnerability index perform normalization: In the actual network, the length of the coaxial cable usually follows a long-tailed distribution (there are serious coaxial cables in a few links). By standardizing the key monitoring of the medium and low-risk intervals, the over-sensitivity in the high-value interval is avoided, so that the high coaxial cable links will not completely dominate the evaluation results.

[0064]

[0065] Among them, represents the coaxial cable vulnerability index after normalization.

[0066] In this embodiment, in the evaluation method at the physical level, a weighted summation method similar to that at the logical level is used for calculation. However, different from the equal-weight processing before, different weights are given to each index at this time to obtain the comprehensive physical evaluation score :

[0067] Among them, , , , and respectively represent the actual laying path efficiency , the coiling surplus index , the fiber core richness of the network , the normalized co-trench risk factor and the normalized coaxial cable vulnerability index weights. Compared with the co-trench risk factor and the coaxial cable vulnerability index that affect network security, the weights of the indicators related to maintenance and scalability (such as physical length, fiber core richness, coiling surplus) are lower. This weight setting reflects the priority of network security to ensure that the evaluation results are more in line with actual needs and focus on the factors that have a greater impact on the stability of the power communication network. For example: = = = 0.1, while = = 0.35.

[0068] In this embodiment, through the path matching degree and redundancy effectiveness matrix analysis, identify the co-trench / coaxial cable problems existing in the physical deployment of the logical link, prevent the occurrence of the "false redundancy" phenomenon, and ensure the authenticity and reliability of the evaluation.

[0069] In the evaluation process of power communication networks, traditional redundancy evaluation methods usually assume that redundant links in the logical layer are completely independent in the physical layer. That is, if there are multiple available paths in the logical topology, the network is considered to have a high redundancy. However, in reality, there may be cases of common trenches (sharing the same pipeline) or common cables (sharing the same optical cable) in the physical layer, which can lead to implicit coupling of seemingly independent logical links in the physical layer.

[0070] When a failure occurs in the physical layer, these common trench / cable links may fail simultaneously, thus forming the problem of "false redundancy", that is, redundant links in the logical layer cannot actually provide true independent backup capabilities. Identify implicit coupling through matrix analysis to solve the misjudgment problem of "false redundancy" in traditional methods. Construct a logical redundancy matrix and a physical coupling matrix respectively, and quantify the true redundancy through matrix analysis:

[0071] Among them, represents the true redundancy, represents the logical redundancy matrix (N×N). The element in the logical redundancy matrix indicates that there is a redundant relationship (such as a backup path) between link and in the logical topology. indicates no redundant relationship. represents the physical coupling matrix (N×N). The element in the physical coupling matrix indicates that there is a common trench / cable between link and link . represents the Hadamard product. represents the Frobenius norm of the matrix. Among them, represents the true redundancy, represents the logical redundancy matrix (N×N), indicates the existence of redundant links, represents the physical coupling matrix (N×N), indicates that there is a common trench / cable between link i and j, represents the Hadamard product (element-wise multiplication) to identify the situation where links are both redundant and physically coupled. represents the Frobenius norm of the matrix, which is the square root of the sum of the squares of all elements of the matrix, measuring the "total strength" or "total number". The true redundancy measures how many of the logical redundant links are "contaminated", that is, they are not actually independent in the physical layer. Therefore, the true redundancy is a normalized index between [0,1], that is , Represents the logical-physical coupling evaluation score.

[0072] In this embodiment, the evaluation results of the logical layer, physical layer, and coupling layer are standardized and given corresponding weights, and finally a comprehensive evaluation score of the network operation mode is formed.

[0073] The embodiment of the present invention adopts a three-layer weighted average model to fuse the evaluation results of the three dimensions of logic, physics, and coupling according to preset weights to obtain the network health score :

[0074] Among them, represents the network health score, represents the comprehensive logical evaluation score weight, represents the comprehensive physical evaluation score weight, represents the logical-physical coupling evaluation score weight.

[0075] The weight distribution is shown in Table 1.

[0076] Table 1 Weight Distribution

[0077] Evaluation level division and operation and maintenance recommendation output: The network health status is classified according to the total evaluation score, and the corresponding inspection cycle, optimization direction, and rectification suggestions are output accordingly to assist the refined operation and maintenance of the power communication network. The specific classification criteria are shown in Table 2.

[0078] Table 2 Classification Criteria

[0079] Embodiment 2: On the basis of Embodiment 1, the embodiment of the present invention takes a typical power communication sub-network in a certain city as an example for a comprehensive evaluation test to verify the applicability of the method of the present invention in an actual scenario. The power communication sub-network includes several key communication nodes and links, has a certain scale and structural complexity, and is representative. The relevant network data sources include: Network topology structure information, derived from the modeling of the NetworkX library; Actual physical path information, derived from the GIS platform and construction acceptance data; Operation and maintenance index data, including optical cable routing, remaining length of the cable reel, core usage, etc.

[0080] As Figure 2 shown, it shows the superposition of the logical topology structure and the actual physical path of the power communication sub-network: The red dashed line represents the physical path formed during actual construction, which has a certain deviation from the planned path; The orange area is marked as the co-cable section, indicating that multiple logical links share the same cable; The blue area is marked as the co-trench section, indicating that the links share the same physical trench.

[0081] Taking nodes N16 to N12 as an example, although there are two paths at the logical layer, they share a single optical cable at the physical layer, presenting a potential single-point failure risk; for another example, the starting sections of N14 to N18 and N14 to N19 share a trench, indicating that they have co-risk conductivity when encountering external damage.

[0082] Next, according to the logical layer evaluation method of the present invention, the following indicators are calculated for the subnet: Node connectivity quantification: Based on the average number of connections of network nodes, the calculated score is 0.50; Link robustness index: Calculated using the minimum edge connectivity normalization method, the score is 0.10; Network efficiency index: Calculated as the reciprocal mean of the shortest path hop counts between all node pairs, and the normalized score is 0.98; Link redundancy index: Calculated by normalizing the number of redundant links in the minimum spanning tree, the score is 0.53.

[0083] Using equal weights (0.25) to perform weighted calculation on the four indicators, the comprehensive score of the logical topology layer is obtained as:

[0084] For the physical layer, it is evaluated from three dimensions: routing efficiency, resource redundancy, and risk factors: Actual routing efficiency: The average score of the ratio of the logical path length to the physical path length is 0.473; Spare length index: After statistical processing of the spare length and normalization, the score is 0.326; Core abundance: Calculated based on the proportion of the remaining cores in the optical cable, the score is 0.741; Co-trench risk factor: After weighted processing by combining the trench type and the co-trench ratio, the score is 0.459; Co-cable vulnerability index: Calculated based on the number of co-cable links and the length influence coefficient index, the score is 0.648.

[0085] Using a weighted strategy (w5 = w6 = w7 = 0.1, w8 = w9 = 0.35), the comprehensive score of the physical layer is calculated as:

[0086] To evaluate the potential implicit coupling and failure correlation of logical links in physical deployments, a logical redundancy matrix and a physical coupling matrix are constructed, and the Hadamard product and Frobenius norm methods are used to calculate the true redundancy. The comprehensive score of the coupling layer is obtained as follows:

[0087] Based on the evaluation results of the above three dimensions and the multi-dimensional weighted fusion model proposed in the present invention, the final network health score is calculated:

[0088] According to the grading standard set in the present invention, the evaluation level corresponding to this score is the "sub-healthy" state. It is recommended that the operation and maintenance unit implement the following measures: Perform link reconstruction on the coaxial cable concentration area (such as N16–N12) to avoid potential single-point failures; Optimize the design of the common trench section to improve the fault isolation ability; Increase the spare redundancy and fiber core configuration of key links to enhance the network recoverability.

[0089] Those of ordinary skill in the art will realize that the embodiments described herein are for helping the reader understand the principles of the present invention. It should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations without departing from the essence of the present invention according to the technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. A power communication network evaluation method based on multi-source data fusion, characterized in that It includes the following steps: Collect the logical topology structure information and physical layer deployment information of the power communication network; According to the logical topology structure information, at the logical level, calculate the node connectivity, link robustness, network efficiency value, and link redundancy; Calculate the comprehensive logical evaluation score based on the node connectivity, link robustness, network efficiency value, and link redundancy; According to the physical layer deployment information, at the physical level, calculate the actual laying path efficiency, spare coil index, fiber core abundance, common trench risk factor, and common cable vulnerability index; Calculate the comprehensive physical evaluation score based on the actual laying path efficiency, spare coil index, fiber core abundance, common trench risk factor, and common cable vulnerability index; Construct a logical redundancy matrix and a physical coupling matrix, and quantify the true redundancy through matrix analysis to obtain the logical-physical coupling evaluation score; Perform a weighted sum of the comprehensive logical evaluation score, the comprehensive physical evaluation score, and the logical-physical coupling evaluation score to obtain the network health score, and judge the health level of the power communication network according to the network health score, completing the evaluation of the power communication network.

2. The power communication network evaluation method based on multi-source data fusion according to claim 1, wherein The node connectivity is obtained by calculating the network average degree, and the calculation formula for the network average degree is: Among them, represents the average degree of the network, that is, the node connectivity, represents the total number of network nodes, represents the th node in the network, represents node , and the degree of the node, that is, the number of connection links of node ; The calculation formula for the link robustness is: Among them, represents the link robustness, represents the total number of network edges, represents the measured k-edge connectivity, and there is , the measured k-edge connectivity is obtained through Monte Carlo simulation, represents the theoretical maximum connectivity, and there is , represents the minimum value, represents the natural base as the base of the logarithmic function; The calculation formula for the network efficiency value is: Among them, represents the network efficiency value, represents the node to node the shortest path hop count, When it approaches 1, it means that the shortest path hop count between all node pairs in the network is close to 1, that is, the network is a fully connected graph and the communication efficiency is the highest; when is close to 0, it means that the communication efficiency between nodes in the network is low, and there are many long paths or communication bottlenecks; The calculation formula for the link redundancy is: Among them, represents the link redundancy, is the number of redundant links, is the total number of links.

3. The power communication network evaluation method based on multi-source data fusion according to claim 1, wherein The calculation formula for the comprehensive logical evaluation score is: Among them, represents the comprehensive score of logical evaluation, , , and respectively represent the weights of node connectivity , link robustness coefficient , network efficiency value and link redundancy , and there is .

4. The power communication network evaluation method based on multi-source data fusion according to claim 1, characterized in that The calculation formula for the actual laying path efficiency is: Among them, represents the actual laying path efficiency, represents the total number of paths evaluated, represents the length of the th logical design path, represents the length of the th actual physical path; The calculation formula for the spare coil index is: Among them, represents the disk surplus index, represents the unused length of the optical cable redundancy, represents the total laid length of the optical cable; The calculation formula for the fiber core abundance is: Among them, represents the core redundancy, represents the total number of cores in the optical cable, represents the number of occupied cores; The calculation formula for the common trench risk factor is: Among them, represents the co-groove risk factor, is the length of the optical cable laid in the co-groove, is the total length of the optical cable laid, is the risk weight of the channel type; The calculation formula for the common cable vulnerability index is: Among them, represents the coaxial cable vulnerability index, represents the number of links sharing the same coaxial cable, represents the total number of logical links, represents the natural base, represents the length influence coefficient, represents the maximum continuous length of the coaxial cable section.

5. The power communication network evaluation method based on multi-source data fusion according to claim 4, wherein, The calculation formula for the comprehensive physical evaluation score is: Among them, represents the comprehensive score of physical evaluation, represents the actual laying path efficiency, represents the remaining coil index, represents the fiber core abundance, represents the co-groove risk factor after normalization, represents the co-cable vulnerability index after normalization, , , , and respectively represent the weights of the actual laying path efficiency , the remaining coil index , the fiber core abundance of the network , the co-groove risk factor after normalization and the co-cable vulnerability index after normalization .

6. The method for evaluating a power communication network based on multi-source data fusion according to claim 5, wherein The normalized co-groove risk factor has the following calculation formula: Among them, represents the hyperbolic tangent function; By normalizing the co-gutter risk factors a negative correlation is achieved where the greater the risk value, the lower the standardized score.

7. The power communication network evaluation method based on multi-source data fusion according to claim 5, characterized in that The normalized coaxial cable vulnerability index is calculated by the following formula: By normalizing the common cable vulnerability index, focus on monitoring the medium and low risk intervals to avoid over-sensitivity in the high value interval.

8. The power communication network evaluation method based on multi-source data fusion according to claim 1, characterized in that The calculation formula for the true redundancy is: Among them, represents the true redundancy, , represents the logical-physical coupling evaluation score, represents the logical redundancy matrix, and the logical redundancy matrix The element in indicates the existence of redundant links, represents the physical coupling matrix, and the physical coupling matrix The element in represents the link and the link share a common trench / cable, represents the Hadamard product, represents the matrix Frobenius norm.

9. The power communication network evaluation method based on multi-source data fusion according to claim 1, wherein The calculation formula for the network health score is: Among them, represents the network health score, represents the comprehensive logic evaluation score weight, represents the comprehensive physical evaluation score weight, represents the logic-physical coupling evaluation score weight.

10. The power communication network evaluation method based on multi-source data fusion according to claim 9, wherein The specific method of judging the health level of the power communication network according to the network health score is: The health level of the power communication network is excellent, and the network health score is: ; The health level of the power communication network is good, and the network health score is: ; The health level of the power communication network is qualified, and the network health score is: ; The health level of the power communication network is sub-healthy, and the network health score is: ; The health level of the power communication network is at risk, and the network health score is: .

Images