A smart monitoring system and method based on power line carrier communication

Abstract

This invention discloses an intelligent monitoring system and method based on power line carrier communication, belonging to the field of power monitoring technology. Under ideal free-space conditions, this invention calculates the theoretical signal transmission strength; introduces an environmental correction coefficient to correct the theoretical signal transmission strength; clusters potential monitoring points within the tunnel to optimize the selection of monitoring points, obtaining a set of monitoring points; monitors the signal transmission strength of antennas within the tunnel and compares it with the theoretical signal transmission strength to obtain a set of transmission strength differences; defines a graph structure to determine the power line carrier communication coverage nodes corresponding to each monitoring point and binds them to the transmission strength differences; calculates the threshold range of transmission strength differences associated with all nodes and marks potential abnormal candidate nodes; analyzes the node neighborhood connectivity characteristics; and traces the propagation direction of abnormal differences along the transmission path of the power line carrier communication to determine abnormal nodes, forming a set of abnormal nodes in the power line carrier communication.

Description

Technical Field

[0001] This invention relates to the field of power regulation technology, specifically to an intelligent regulation system and method based on power line carrier communication. Background Technology

[0002] In enclosed, narrow spaces such as tunnels, power line carrier communication is widely used for equipment status monitoring and data transmission due to its advantages such as requiring no additional wiring and adaptability to complex environments. However, the complex environment inside tunnels makes antenna signal propagation susceptible to interference, leading to fluctuations in communication quality. Furthermore, the numerous nodes and complex topology of power line carrier communication networks make it difficult for traditional monitoring methods to accurately detect signal anomalies and locate faulty nodes, posing potential risks to tunnel operational safety.

[0003] Existing technologies often employ uniform or empirically-based point-of-sale (POS) deployment methods without tailoring optimization to antenna radiation characteristics and tunnel structures. This results in redundant monitoring in some areas and missing monitoring in critical areas, failing to comprehensively reflect the true signal status. Anomaly detection of communication network nodes relies on manual investigation or simple threshold judgments, neglecting to consider inter-node connectivity and signal transmission patterns. This makes it difficult to quickly locate core faulty nodes and differentiate between node-specific faults and conduction-related anomalies. Treating signal monitoring, network topology, and node status as independent modules without establishing a multi-dimensional data correlation mechanism prevents the identification of systemic anomalies from a global perspective, resulting in insufficient overall and forward-looking oversight. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent monitoring system and method based on power line carrier communication to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] In a first aspect, this application provides an intelligent monitoring method based on power line carrier communication, comprising the following steps:

[0007] The antenna's operating parameters and environmental parameters are collected, and the antenna's radiated power is calculated. Based on the antenna's radiated power, under ideal free space conditions, the theoretical signal transmission intensity is calculated according to the inverse square law. Based on the influence of the tunnel's geometry on signal propagation, an environmental correction coefficient is introduced to correct the theoretical signal transmission intensity.

[0008] Antenna and tunnel features are acquired, potential monitoring points within the tunnel are clustered, and the selection of monitoring points is optimized to obtain a set of monitoring points. Based on the set of monitoring points, the signal transmission intensity of the antenna within the tunnel is monitored and compared with the theoretical signal transmission intensity to obtain a set of transmission intensity differences.

[0009] The graph structure is defined by taking each node in the power line carrier communication network as the vertex and the communication connection relationship between the nodes as the edge. Based on the physical location of the monitoring point set, the power line carrier communication coverage node corresponding to each monitoring point is determined, and the transmission intensity difference of the monitoring point is bound to the vertex attribute of the corresponding node. The threshold range of the transmission intensity difference associated with all nodes is calculated, and potential abnormal candidate nodes are marked.

[0010] When defining the graph structure, all functional nodes involved in the power line carrier communication system within the tunnel are acquired, including antenna communication modules, signal relay nodes, data forwarding nodes, and control terminal nodes. Each node is assigned a unique identifier for identification. The actual communication links between nodes are examined one by one to determine which nodes have direct power line carrier signal transmission relationships, recording the direction of the connections to form a node connection list. Each node is mapped to a vertex of the graph, and the communication connections between nodes are mapped to edges. For example, if node A and node B have bidirectional communication, A and B are represented by two vertices in the graph and connected by an undirected edge; if it is unidirectional transmission, a directed edge is used to represent the transmission direction. Attributes describing the connection characteristics are added to each edge. Vertices and edges are combined according to their actual connection relationships to form the initial graph structure.

[0011] For potential abnormal candidate nodes, analyze the node's neighborhood connectivity characteristics; trace the propagation direction of abnormal differences along the power line carrier communication transmission path to determine abnormal nodes; integrate all vertices determined to be abnormal to form a set of abnormal nodes in power line carrier communication.

[0012] In conjunction with the first aspect, in a first embodiment of the first aspect of this application, the calculation of the theoretical signal transmission intensity based on the antenna radiated power under ideal free space conditions, according to the inverse square law, includes:

[0013] Antenna radiated power represents the total energy radiated into space by the antenna under ideal conditions. The propagation environment is set as ideal free space, specifically where there are no obstacles, reflectors, or electromagnetic interference. During signal propagation, energy attenuation only occurs due to spatial diffusion, with no other additional losses. The antenna is used as a point radiation source, and the radiated signal propagates uniformly in all directions in the form of a spherical wave.

[0014] In ideal free space, the wavefront of a spherical wave is a complete sphere, with energy uniformly distributed on the surface and unimpeded propagation in all directions along a straight line. The total power radiated by the antenna is uniformly distributed on a sphere with the antenna as its center and the propagation distance as its radius. The theoretical signal transmission intensity at a certain distance is specifically the proportion of radiated power allocated to the spherical area corresponding to that distance. The target distance for calculating the theoretical signal transmission intensity is determined as the straight-line distance from the antenna to the monitoring point. According to the inverse square law, the spherical area at that distance is proportional to the square of the distance. Dividing the antenna radiated power by the spherical area yields the signal energy per unit area at that distance, which is taken as the theoretical signal transmission intensity.

[0015] In conjunction with the first aspect, in the second embodiment of the first aspect of this application, the influence of the tunnel geometry on signal propagation is addressed by introducing an environmental correction coefficient to correct the theoretical signal transmission strength, including:

[0016] The tunnel's geometry includes its length, cross-sectional shape, diameter, curvature, and whether it branches. Analysis of historical data from similar tunnels reveals the correlation between geometric factors and signal attenuation, reflecting signal energy loss under different geometric structures. Using ideal free-space conditions as a benchmark, the environmental correction coefficient is 1, representing no additional attenuation. Based on this correlation, an adjustment rule for the environmental correction coefficient is established: when geometric factors causing signal attenuation exist, the environmental correction coefficient is less than 1, and the more severe the attenuation, the smaller the environmental correction coefficient.

[0017] Based on the current tunnel geometry, and in accordance with the established geometric factors and adjustment rules, the corresponding environmental correction coefficient is calculated. The calculated environmental correction coefficient is then multiplied by the theoretical signal transmission strength for correction.

[0018] In conjunction with the first aspect, in the third embodiment of the first aspect of this application, the step of acquiring antenna features and tunnel features, clustering potential monitoring points within the tunnel, optimizing the selection of monitoring points, and obtaining a set of monitoring points includes:

[0019] The process involves acquiring antenna features, including radiation direction, coverage area, and signal strength distribution patterns; acquiring tunnel features, including tunnel length, width, location of curved sections, and material; pre-setting potential monitoring points within the tunnel to cover all areas; determining the similarity of signal characteristics at different potential monitoring points based on the signal strength distribution patterns in the antenna features; using the DBSCAN algorithm to group potential monitoring points with similar signal characteristics into the same cluster; monitoring points in the same cluster exhibit consistent signal performance, representing the signal characteristics of that area; and finally, dividing the potential monitoring points into several clusters through clustering.

[0020] The goal of reinforcement learning is to select monitoring points from clusters obtained through clustering, so that the monitoring points cover all key signal areas in the tunnel with the minimum number of points. The signal environment inside the tunnel is used as the reinforcement learning environment, each cluster is a selectable state, and selecting a monitoring point in a certain cluster is an action. The reinforcement learning agent starts from the initial state and gradually selects monitoring points in different clusters. After each selection, a corresponding reward or penalty is given based on whether the coverage of the selected monitoring point is comprehensive and whether it can effectively reflect signal changes. Through continuous trial and adjustment of the selection strategy, the agent gradually learns the optimal selection method, selects a suitable number of monitoring points with reasonable distribution, and forms a monitoring point set.

[0021] In conjunction with the first aspect, in the fourth embodiment of the first aspect of this application, the step of monitoring the signal transmission strength of the antenna inside the tunnel based on the set of monitoring points and comparing it with the theoretical signal transmission strength to obtain a set of transmission strength difference values ​​includes:

[0022] Based on the set of monitoring points, a corresponding external signal monitoring device is installed at each monitoring point location within the tunnel. The external monitoring device is activated to monitor the antenna signal transmission strength of each monitoring point within the tunnel. The theoretical signal transmission strength corresponding to each monitoring point is retrieved and correlated with the location of each monitoring point. The actual signal transmission strength of each monitoring point is compared with the corresponding theoretical signal transmission strength at each time point. For each time point, the actual signal strength of the monitoring point is subtracted from the theoretical signal strength to obtain the transmission strength difference at that moment. The transmission strength differences of all monitoring points at each time point are integrated and sorted according to the monitoring point location and time point order to obtain a set of transmission strength difference values.

[0023] In conjunction with the first aspect, in the fifth embodiment of the first aspect of this application, the step of determining the power line carrier communication coverage node corresponding to each monitoring point based on the physical location of the monitoring point set, and binding the transmission strength difference of the monitoring points to the vertex attributes of the corresponding node, includes:

[0024] The physical location of each monitoring point is compared with the coverage area of ​​each node to determine which node's coverage area the monitoring point belongs to. When a monitoring point is located within the coverage area of ​​a certain node, and that node is the closest node to the monitoring point and has the most stable signal connection, that node is identified as the power line carrier communication coverage node corresponding to that monitoring point. For monitoring points located in areas where the coverage areas of multiple nodes overlap, the node with the best signal transmission quality is selected as the power line carrier communication coverage node by filtering historical communication data. For each power line carrier communication coverage node, the difference in transmission intensity among all its corresponding monitoring points is collected and added as a new attribute to the vertex attributes of that node in the graph structure.

[0025] In conjunction with the first aspect, in the sixth embodiment of the first aspect of this application, the step of calculating the threshold range of the emission intensity differences associated with all nodes and marking potential abnormal candidate nodes includes:

[0026] Extract the emission intensity difference associated with all nodes from the graph structure, and set a threshold range based on the overall distribution characteristics of the emission intensity difference. For each node, obtain the emission intensity difference of all monitoring points corresponding to it, and determine whether each difference is within the threshold range. When some or all of the differences of a node exceed the threshold range, mark the node as a potential abnormal candidate node.

[0027] In conjunction with the first aspect, in the seventh embodiment of the first aspect of this application, the step of analyzing the node neighborhood connectivity characteristics for potential abnormal candidate nodes includes:

[0028] Determine the neighboring nodes of each potential anomaly candidate node, specifically all nodes directly connected to that potential anomaly candidate node through edges; query whether each neighboring node is a potential anomaly candidate node, record the emission intensity difference of the neighboring nodes, and form a list of neighboring node states;

[0029] Based on the state list of neighboring nodes, the connection attributes of edges between potential anomaly candidate nodes and neighboring nodes, as well as the labeling information of potential anomaly candidate nodes, are collected to obtain a dataset. The support and confidence indices in association rule mining are used to analyze the association strength between connection attributes and labeling information in the dataset. When the support and confidence of a certain connection attribute and labeling information are both higher than a set threshold, and the neighboring node is in the communication transmission direction between the potential anomaly candidate node and the neighboring node, it is determined that the connection attribute has a strong matching relationship with the neighboring node.

[0030] In conjunction with the first aspect, in the eighth embodiment of the first aspect of this application, the step of tracing the propagation direction of abnormal differences along the transmission path of power line carrier communication and determining abnormal nodes includes:

[0031] The main transmission path of data in the power line carrier communication network is determined, specifically the connection sequence of the main nodes from the signal source to the data endpoint. The node connection relationships of each branch path are recorded to form a transmission path map. Based on the transmission path map, for nodes with strong matching relationships, their positions in the main transmission path are obtained. When a node is the starting node of a certain path segment, and its abnormal difference value appears earlier than that of the downstream node, and the abnormal difference value of the downstream node is consistent with that node, the node is identified as a potential starting point for abnormal propagation. Starting from the potential starting point, the abnormality of the downstream nodes is checked sequentially along the transmission path. When the abnormal difference value of the downstream node gradually weakens with the increase of distance from the starting point, and the abnormality of each node has a strong matching connection attribute with the preceding node, it is confirmed that the abnormal difference value is propagated along the path, and the abnormal propagation path is recorded, including the propagation direction and the order in which the abnormality of each node appears.

[0032] In the anomaly propagation path, when a node is the initial source of the abnormal difference and has the largest abnormal difference magnitude, that node is determined to be a core anomaly node. The condition for determining the initial source is that the node has the earliest occurrence time of the anomaly and no upstream node propagates the anomaly to that node through a strong matching connection. In the downstream transmission path of the core anomaly node, nodes that have abnormal differences due to strong matching connections with the core anomaly node or a preceding anomaly node, and that do not have independent anomaly triggering factors, are determined to be derived anomaly nodes. For potential anomaly candidate nodes, when there is no strong matching relationship between their connection attributes and those of neighboring nodes, and no upstream node propagates the anomaly to that node in the transmission path, and their abnormal difference is unrelated to other nodes, they are determined to be self-faulting nodes. Their anomalies are caused by their own hardware or local environmental problems and are unrelated to propagation.

[0033] Secondly, this application provides an intelligent monitoring system based on power line carrier communication, comprising:

[0034] The theoretical signal transmission strength calculation module includes: an antenna radiated power calculation unit, a theoretical signal transmission strength calculation unit, and a theoretical signal transmission strength correction unit. The antenna radiated power calculation unit collects the antenna's operating parameters and environmental parameters to calculate the antenna radiated power. The theoretical signal transmission strength calculation unit, based on the antenna radiated power, calculates the theoretical signal transmission strength under ideal free-space conditions according to the inverse square law. The theoretical signal transmission strength correction unit, based on the influence of the tunnel's geometry on signal propagation, introduces an environmental correction coefficient to correct the theoretical signal transmission strength.

[0035] The transmission intensity difference calculation module includes a monitoring point generation unit and a transmission intensity difference calculation unit. The monitoring point generation unit acquires antenna features and tunnel features, clusters potential monitoring points in the tunnel, optimizes the selection of monitoring points, and obtains a set of monitoring points. The transmission intensity difference calculation unit monitors the signal transmission intensity of the antenna in the tunnel based on the set of monitoring points, compares it with the theoretical signal transmission intensity, and obtains a set of transmission intensity differences.

[0036] The potential anomaly marking module includes: a graph structure definition unit, a transmission intensity difference binding unit, and a potential anomaly marking unit. The graph structure definition unit defines the graph structure using each node in the power line carrier communication network as a vertex and the communication connections between nodes as edges. The transmission intensity difference binding unit determines the power line carrier communication coverage node corresponding to each monitoring point based on the physical location of the monitoring point set, and binds the transmission intensity difference of the monitoring point to the vertex attributes of the corresponding node. The potential anomaly marking unit calculates the threshold range of the transmission intensity differences associated with all nodes and marks potential anomaly candidate nodes.

[0037] The abnormal node set generation module includes: a neighborhood connectivity characteristic analysis unit, an abnormal node determination unit, and an abnormal node set generation unit. The neighborhood connectivity characteristic analysis unit analyzes the neighborhood connectivity characteristics of potential abnormal candidate nodes. The abnormal node determination unit tracks the propagation direction of abnormal differences along the transmission path of power line carrier communication and determines abnormal nodes. The abnormal node set generation unit integrates all vertices determined to be abnormal to form an abnormal node set for power line carrier communication.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] 1. This application selects monitoring points through clustering and optimization algorithms, and combines antenna characteristics and tunnel structure to achieve on-demand point deployment, which reduces redundancy, lowers data acquisition costs, and improves monitoring efficiency while ensuring comprehensive signal coverage.

[0040] 2. This application analyzes the neighborhood connectivity characteristics of nodes based on graph structure analysis and combines signal transmission path tracing to identify the source of anomalies. It can quickly distinguish core abnormal nodes, derived abnormal nodes, and self-faulty nodes, achieving accurate fault location and shortening troubleshooting time.

[0041] 3. This application establishes a correlation mechanism between signal monitoring, network topology and node status, integrating scattered data into a global view, which can identify systemic anomaly risks, provide decision support for preventive maintenance of tunnel communication networks, and enhance the initiative and comprehensiveness of supervision. Attached Figure Description

[0042] Figure 1 This is a schematic diagram illustrating the steps of an intelligent monitoring method based on power line carrier communication according to the present invention.

[0043] Figure 2 This is a system structure diagram of an intelligent monitoring system based on power line carrier communication according to the present invention. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Example: Figures 1-2 As shown, the present invention provides a technical solution.

[0046] like Figure 1 As shown, this application provides an intelligent monitoring method based on power line carrier communication, including the following steps:

[0047] Step S100: Collect the antenna's operating parameters and environmental parameters, and calculate the antenna's radiated power; based on the antenna's radiated power, under ideal free space conditions, calculate the theoretical signal transmission intensity according to the inverse square law; based on the influence of the tunnel's geometry on signal propagation, introduce an environmental correction coefficient to correct the theoretical signal transmission intensity;

[0048] Specifically, antenna radiated power represents the total energy radiated into space by the antenna under ideal conditions. The propagation environment is set as ideal free space, specifically where there are no obstacles, reflectors, or electromagnetic interference. During signal propagation, energy attenuation only occurs due to spatial diffusion, with no other additional losses. The antenna is used as a point radiation source, and the radiated signal propagates uniformly in all directions in the form of a spherical wave.

[0049] In ideal free space, the wavefront of a spherical wave is a complete sphere, with energy uniformly distributed on the surface and unimpeded propagation in all directions along a straight line. The total power radiated by the antenna is uniformly distributed on a sphere with the antenna as its center and the propagation distance as its radius. The theoretical signal transmission intensity at a certain distance is specifically the proportion of radiated power allocated to the spherical area corresponding to that distance. The target distance for calculating the theoretical signal transmission intensity is determined as the straight-line distance from the antenna to the monitoring point. According to the inverse square law, the spherical area at that distance is proportional to the square of the distance. Dividing the antenna radiated power by the spherical area yields the signal energy per unit area at that distance, which is taken as the theoretical signal transmission intensity.

[0050] Furthermore, the tunnel's geometry includes its length, cross-sectional shape, diameter, curvature, and whether it branches. By analyzing historical data from similar tunnels, the correlation between geometric factors and signal attenuation is obtained, reflecting the signal energy loss under different geometric structures. Using the signal propagation state under ideal free space conditions as a benchmark, the environmental correction coefficient is 1, representing no additional attenuation. Based on this correlation, an adjustment rule for the environmental correction coefficient is established: when geometric factors causing signal attenuation exist, the environmental correction coefficient is less than 1, and the more severe the attenuation, the smaller the environmental correction coefficient.

[0051] Based on the current tunnel geometry, and in accordance with the established geometric factors and adjustment rules, the corresponding environmental correction coefficient is calculated. The calculated environmental correction coefficient is then multiplied by the theoretical signal transmission strength for correction.

[0052] In one specific embodiment, a circular tunnel with a length of 1000 meters and a cross-sectional diameter of 5 meters is selected, with two 90° bends (located at 300 meters and 700 meters respectively), and the tunnel wall is made of reinforced concrete; a certain type of directional antenna with an operating frequency of 500 kHz is used and deployed at the tunnel starting point (0 meters).

[0053] Antenna operating parameters: operating voltage 220V, operating current 0.5A, power conversion efficiency 90%, antenna inherent efficiency 85%.

[0054] Calculate the input power: 220V × 0.5A = 110W. After correction, the input power is 110W × 90% = 99W.

[0055] Environmental parameters collected: temperature 25℃, humidity 60% inside the tunnel, no strong electromagnetic interference, environmental efficiency correction factor 0.95, spatial attenuation influence factor 0.9 (because the antenna is 1 meter away from the tunnel wall, the material has little effect on signal attenuation).

[0056] Calculate the antenna radiated power: 99W × 85% × 0.95 × 0.9 ≈ 68.5W.

[0057] Selected target monitoring points: located at 100 meters (straight section), 350 meters (just past the first curve), and 800 meters (after the second curve) in the tunnel, with straight-line distances to the antenna of 100 meters, 350 meters, and 800 meters, respectively.

[0058] At 100 meters: the spherical area is 4π×(100m)²≈125664m², and the theoretical signal transmission intensity is 68.5W÷125664m²≈0.000545W / m².

[0059] At 350 meters: the spherical area is 4π×(350m)²≈1539380m², and the theoretical signal transmission strength is 68.5W÷1539380m²≈0.0000445W / m².

[0060] At 800 meters: the spherical area is 4π×(800m)²≈8042477m², and the theoretical signal transmission intensity is 68.5W÷8042477m²≈0.0000085W / m².

[0061] Straight section (100 meters): No bends, uniform cross-section, minimal impact of geometry on signal attenuation, environmental correction factor 0.9 (slight attenuation). Corrected signal strength = 0.000545 W / m² × 0.9 ≈ 0.00049 W / m².

[0062] Just past the bend (350 meters): The bend causes signal diffraction loss, with an environmental correction factor of 0.6. The corrected signal strength = 0.0000445W / m² × 0.6 ≈ 0.0000267W / m².

[0063] After the second bend (800 meters): Double bend + increased distance, severe signal attenuation, environmental correction factor 0.3. Corrected signal strength = 0.0000085W / m² × 0.3 ≈ 0.00000255W / m².

[0064] Step S200: Obtain antenna features and tunnel features, cluster potential monitoring points in the tunnel, optimize the selection of monitoring points, and obtain a set of monitoring points; based on the set of monitoring points, monitor the signal transmission intensity of the antenna in the tunnel, compare it with the theoretical signal transmission intensity, and obtain a set of transmission intensity difference values.

[0065] Specifically, antenna characteristics are acquired, including radiation direction, coverage area, and signal strength distribution patterns; tunnel characteristics are acquired, including tunnel length, width, location of curved sections, and material; potential monitoring points are pre-set within the tunnel, covering all areas of the tunnel; based on the signal strength distribution patterns in the antenna characteristics, it is determined whether the signal characteristics at different potential monitoring points are similar; using the DBSCAN algorithm, potential monitoring points with similar signal characteristics are grouped into the same cluster, and monitoring points in the same cluster have consistent signal performance, representing the signal characteristics of that area; through clustering, the potential monitoring points are divided into several clusters;

[0066] The goal of reinforcement learning is to select monitoring points from clusters obtained through clustering, so that the monitoring points cover all key signal areas in the tunnel with the minimum number of points. The signal environment inside the tunnel is used as the reinforcement learning environment, each cluster is a selectable state, and selecting a monitoring point in a certain cluster is an action. The reinforcement learning agent starts from the initial state and gradually selects monitoring points in different clusters. After each selection, a corresponding reward or penalty is given based on whether the coverage of the selected monitoring point is comprehensive and whether it can effectively reflect signal changes. Through continuous trial and adjustment of the selection strategy, the agent gradually learns the optimal selection method, selects a suitable number of monitoring points with reasonable distribution, and forms a monitoring point set.

[0067] Furthermore, based on the set of monitoring points, corresponding external signal monitoring equipment is installed at each monitoring point location within the tunnel; the external monitoring equipment is activated to monitor the antenna signal transmission strength of each monitoring point within the tunnel; the theoretical signal transmission strength corresponding to each monitoring point is retrieved and correlated with the location of each monitoring point; the actual signal transmission strength of each monitoring point is compared with the corresponding theoretical signal transmission strength at each time point; for each time point, the actual signal strength of the monitoring point is subtracted from the theoretical signal strength to obtain the transmission strength difference at that moment; the transmission strength differences of all monitoring points at each time point are integrated and sorted according to the monitoring point location and time point order to obtain a set of transmission strength difference values.

[0068] In one specific embodiment, antenna characteristics are obtained: the radiation direction is the tunnel extension direction (0°-180°), the coverage range is 0-1000 meters, and the signal strength decreases exponentially with increasing distance (the attenuation is slower in straight sections than in curved sections). Tunnel characteristics are obtained: same as in step S100 (1000-meter circular tunnel, 5-meter diameter, with 90° bends at 300 meters and 700 meters, made of reinforced concrete). Potential monitoring points are obtained: within the 0-1000-meter range of the tunnel, one monitoring point is preset every 10 meters, for a total of 101 points, numbered P0 (0 meters), P10 (10 meters)...P1000 (1000 meters).

[0069] The similarity indices were based on signal strength attenuation trend and the degree of influence from curved sections. The clustering results are as follows:

[0070] Cluster 1 (Straight Line Segment I): Includes P0-P290 (0-290 meters), with a total of 30 monitoring points. The signal characteristics are gradual attenuation (intensity decreases by 30% every 100 meters).

[0071] Cluster 2 (around the curved section I): includes P300-P390 (300-390 meters), with a total of 10 monitoring points. The signal characteristics are severe attenuation (intensity decreases by 15% every 10 meters).

[0072] Cluster 3 (Straight Line Segment II): Includes P400-P690 (400-690 meters), with a total of 30 monitoring points. The signal characteristics are gradual attenuation (intensity decreases by 35% every 100 meters).

[0073] Cluster 4 (around the curved section II): includes P700-P790 (700-790 meters), a total of 10 monitoring points, the signal characteristics are severe attenuation (intensity decreases by 20% every 10 meters).

[0074] Cluster 5 (Straight Line Segment III): Includes P800-P1000 (800-1000 meters), with a total of 21 monitoring points. The signal characteristics are gradual attenuation (intensity decreases by 40% every 100 meters).

[0075] The goal of reinforcement learning is to cover 5 clusters, with ≤10 monitoring points and an anomaly detection rate ≥90%. The agent initially selects the central node of each cluster (e.g., P150, P350, etc.), and adjusts it through 5 rounds of iteration (removing redundant points and adding points around key bends). The final set of monitoring points is: P100 (cluster 1), P300 (cluster 2), P350 (cluster 2), P500 (cluster 3), P700 (cluster 4), P750 (cluster 4), P800 (cluster 5), and P900 (cluster 5), a total of 8 points, covering all clusters with higher density around bends.

[0076] Actual signal monitoring data (data selected at a specific moment) is as follows:

[0077] P100: 0.00048 W / m² (theoretical value 0.00049 W / m²), difference = -0.00001 W / m².

[0078] P300: 0.000020W / m² (theoretical value 0.0000267W / m²), difference = -0.0000067W / m².

[0079] P350: 0.000025W / m² (theoretical value 0.0000267W / m²), difference = -0.0000017W / m².

[0080] P500: 0.000018W / m² (theoretical value 0.000020W / m², corrected by cluster 3), difference = -0.000002W / m².

[0081] P700: 0.0000015W / m² (theoretical value 0.00000255W / m²), difference = -0.00000105W / m².

[0082] P750: 0.0000012W / m² (theoretical value 0.0000022W / m², corrected by cluster 4), difference = -0.000001W / m².

[0083] P800: 0.0000025W / m² (theoretical value 0.00000255W / m²), difference = -0.00000005W / m².

[0084] P900: 0.0000010W / m² (theoretical value 0.0000012W / m², corrected by cluster 5), difference = -0.0000002W / m².

[0085] The set of emission intensity differences is arranged in order of monitoring points as [-0.00001, -0.0000067, -0.0000017, -0.000002, -0.00000105, -0.000001, -0.0000005, -0.0000002] (unit: W / m²), with the timestamp uniformly set to "T10:00:00".

[0086] Step S300: Define the graph structure by taking each node in the power line carrier communication network as the vertex of the graph and the communication connection relationship between the nodes as the edge of the graph; determine the power line carrier communication coverage node corresponding to each monitoring point according to the physical location of the monitoring point set, and bind the transmission intensity difference of the monitoring point to the vertex attribute of the corresponding node; calculate the threshold range of the transmission intensity difference associated with all nodes, and mark potential abnormal candidate nodes.

[0087] Specifically, the physical location of each monitoring point is compared with the coverage area of ​​each node to determine which node's coverage area the monitoring point belongs to. When a monitoring point is located within the coverage area of ​​a certain node, and that node is the closest to the monitoring point and has the most stable signal connection, that node is identified as the power line carrier communication coverage node corresponding to that monitoring point. For monitoring points located in areas where the coverage areas of multiple nodes overlap, the node with the best signal transmission quality is selected as the power line carrier communication coverage node through historical communication data. For each power line carrier communication coverage node, the difference in transmission intensity among all its corresponding monitoring points is collected and added as a new attribute to the vertex attributes of that node in the graph structure.

[0088] Furthermore, the emission intensity difference associated with all nodes is extracted from the graph structure, and a threshold range is set based on the overall distribution characteristics of the emission intensity difference. For each node, the emission intensity difference of all monitoring points corresponding to it is obtained, and it is determined one by one whether it is within the threshold range. When some or all of the differences of a node exceed the threshold range, the node is marked as a potential abnormal candidate node.

[0089] In one specific embodiment, five power line carrier communication nodes are deployed within the tunnel: N1 (0-300 meters), N2 (300-500 meters), N3 (500-700 meters), N4 (700-900 meters), and N5 (900-1000 meters). The nodes are connected by power lines, forming a unidirectional communication path N1→N2→N3→N4→N5 (the edge attributes are all "medium stability," with a historical interruption frequency of <1 time / day). Using the five nodes as vertices and the communication connections between nodes as edges, the initial vertex attributes include coverage area and hardware model, while the edge attributes include transmission direction and stability.

[0090] N1 covers 0-300 meters, N2 covers 300-500 meters, N3 covers 500-700 meters, N4 covers 700-900 meters, and N5 covers 900-1000 meters. The matching results are as follows: P100 (100 meters) → N1, P300 (300 meters) → N2 (closer to N2, signal connection stability 92% higher than N1's 85%), P350 (350 meters) → N2, P500 (500 meters) → N3, P700 (700 meters) → N4, P750 (750 meters) → N4, P800 (800 meters) → N4, P900 (900 meters) → N5.

[0091] N1 vertex attribute added: Coverage monitoring point difference set [P100: -0.00001W / m²].

[0092] N2 vertex attribute added: Coverage monitoring point difference set [P300: -0.0000067W / m², P350: -0.0000017W / m²].

[0093] N3 vertex attribute added: Coverage monitoring point difference set [P500: -0.000002W / m²].

[0094] N4 vertex attribute added: Coverage monitoring point difference set [P700: -0.00000105W / m², P750: -0.000001W / m², P800: -0.00000005W / m²].

[0095] N5 vertex attribute added: Coverage monitoring point difference set [P900: -0.0000002W / m²].

[0096] The emission intensity difference across all nodes ranges from -0.00001 W / m² to -0.00000005 W / m², with a mean of -0.000003 W / m², exhibiting minimal fluctuation. Based on these distribution characteristics, a threshold range of [-0.000008 W / m², -0.00000001 W / m²] is set (encompassing over 95% of normal differences).

[0097] Node difference judgment: N1: Difference -0.00001W / m², exceeding the lower limit (-0.000008W / m²). N2: Differences -0.0000067W / m² and -0.0000017W / m², both within the range. N3: Difference -0.000002W / m², within the range. N4: Differences -0.00000105W / m², -0.000001W / m², and -0.00000005W / m², all within the range. N5: Difference -0.0000002W / m², within the range. The difference in N1 exceeds the threshold range and is marked as a potential abnormal candidate node.

[0098] Step S400: For potential abnormal candidate nodes, analyze the node neighborhood connectivity characteristics; trace the propagation direction of abnormal differences along the power line carrier communication transmission path to determine abnormal nodes; integrate all vertices determined to be abnormal to form a set of abnormal nodes for power line carrier communication.

[0099] Specifically, determine the neighboring nodes of each potential anomaly candidate node, which are all nodes directly connected to the potential anomaly candidate node through edges; query whether each neighboring node is a potential anomaly candidate node, record the emission intensity difference of the neighboring nodes, and form a list of neighboring node states.

[0100] Based on the state list of neighboring nodes, the connection attributes of edges between potential anomaly candidate nodes and neighboring nodes, as well as the labeling information of potential anomaly candidate nodes, are collected to obtain a dataset. The support and confidence indices in association rule mining are used to analyze the association strength between connection attributes and labeling information in the dataset. When the support and confidence of a certain connection attribute and labeling information are both higher than a set threshold, and the neighboring node is in the communication transmission direction between the potential anomaly candidate node and the neighboring node, it is determined that the connection attribute has a strong matching relationship with the neighboring node.

[0101] Furthermore, the main transmission path of data in the power line carrier communication network is determined, specifically the connection sequence of the main nodes from the signal source to the data endpoint. The node connection relationships of each branch path are recorded to form a transmission path map. Based on the transmission path map, for nodes with strong matching relationships, their positions in the main transmission path are obtained. When a node is the starting node of a certain path segment, and its abnormal difference value appears earlier than that of the downstream node, and the abnormal difference value of the downstream node is consistent with that node, the node is identified as a potential starting point for abnormal propagation. Starting from the potential starting point, the abnormality of the downstream nodes is checked sequentially along the transmission path. When the abnormal difference value of the downstream node gradually weakens with the increase of distance from the starting point, and the abnormality of each node has a strong matching connection attribute with the preceding node, it is confirmed that the abnormal difference value is propagated along the path, and the abnormal propagation path is recorded, including the propagation direction and the order in which the abnormality of each node appears.

[0102] In the anomaly propagation path, when a node is the initial source of the abnormal difference and has the largest abnormal difference magnitude, that node is determined to be a core anomaly node. The condition for determining the initial source is that the node has the earliest occurrence time of the anomaly and no upstream node propagates the anomaly to that node through a strong matching connection. In the downstream transmission path of the core anomaly node, nodes that have abnormal differences due to strong matching connections with the core anomaly node or a preceding anomaly node, and that do not have independent anomaly triggering factors, are determined to be derived anomaly nodes. For potential anomaly candidate nodes, when there is no strong matching relationship between their connection attributes and those of neighboring nodes, and no upstream node propagates the anomaly to that node in the transmission path, and their abnormal difference is unrelated to other nodes, they are determined to be self-faulting nodes. Their anomalies are caused by their own hardware or local environmental problems and are unrelated to propagation.

[0103] In one specific embodiment, the potential anomalous candidate node is N1 (already marked). N1's direct neighbor is N2 (connected by only one edge from N1 to N2). N2 is not a potential anomalous candidate node, with emission intensity differences of -0.0000067 W / m² and -0.0000017 W / m² (both within the threshold range). The edge attribute between N1 and N2 is "moderate stability," with a historical outage frequency of 0.5 times / day, and the transmission direction is unidirectional (N1 to N2). Data sets of "moderate connection stability" and "anomalous node" are collected, and the calculated support is 20% (below the set threshold of 30%) and the confidence is 60% (below the set threshold of 70%). Therefore, it is determined that the connection attribute between N1 and N2 and the anomalous node have no strong matching relationship.

[0104] The main transmission path is: N1→N2→N3→N4→N5. The abnormal difference value of N1 is -0.00001W / m² (occurring at T10:00:00). The abnormal difference value of downstream node N2 is normal at T10:00:00, and N2 has no abnormality in the following 30 minutes. The differences of N3, N4, and N5 are all stable within the threshold range in the same time period, without any abnormal propagation trend of attenuation with distance. No abnormal difference value propagation along the transmission path was found, and the abnormality of N1 is an isolated phenomenon.

[0105] N1 exhibits the largest abnormal difference (-0.00001 W / m²), but an upstream check reveals no nodes (N1 is the starting point of the transmission path), and no abnormality propagates downstream, failing to meet the "initial source of abnormality propagation" condition. Therefore, the core abnormal node is excluded. Downstream nodes N2-N5 show no abnormal differences and have no strong matching connection with N1, thus no derivative abnormal nodes exist. N1 has no strong matching connection with neighboring node N2, and no upstream node propagates abnormality in the transmission path. Its abnormal difference (-0.00001 W / m²) is unrelated to other nodes (other node differences are all within the threshold). Furthermore, an inspection of N1's hardware log reveals local circuit aging records (an independent abnormality triggering factor). Therefore, N1 is determined to be a self-faulting node.

[0106] like Figure 2 As shown, this application provides an intelligent monitoring system based on power line carrier communication, comprising:

[0107] The theoretical signal transmission strength calculation module includes: an antenna radiated power calculation unit, a theoretical signal transmission strength calculation unit, and a theoretical signal transmission strength correction unit. The antenna radiated power calculation unit collects the antenna's operating parameters and environmental parameters to calculate the antenna radiated power. The theoretical signal transmission strength calculation unit, based on the antenna radiated power, calculates the theoretical signal transmission strength under ideal free-space conditions according to the inverse square law. The theoretical signal transmission strength correction unit, based on the influence of the tunnel's geometry on signal propagation, introduces an environmental correction coefficient to correct the theoretical signal transmission strength.

[0108] The transmission intensity difference calculation module includes a monitoring point generation unit and a transmission intensity difference calculation unit. The monitoring point generation unit acquires antenna features and tunnel features, clusters potential monitoring points in the tunnel, optimizes the selection of monitoring points, and obtains a set of monitoring points. The transmission intensity difference calculation unit monitors the signal transmission intensity of the antenna in the tunnel based on the set of monitoring points, compares it with the theoretical signal transmission intensity, and obtains a set of transmission intensity differences.

[0109] The potential anomaly marking module includes: a graph structure definition unit, a transmission intensity difference binding unit, and a potential anomaly marking unit. The graph structure definition unit defines the graph structure using each node in the power line carrier communication network as a vertex and the communication connections between nodes as edges. The transmission intensity difference binding unit determines the power line carrier communication coverage node corresponding to each monitoring point based on the physical location of the monitoring point set, and binds the transmission intensity difference of the monitoring point to the vertex attributes of the corresponding node. The potential anomaly marking unit calculates the threshold range of the transmission intensity differences associated with all nodes and marks potential anomaly candidate nodes.

[0110] The abnormal node set generation module includes: a neighborhood connectivity characteristic analysis unit, an abnormal node determination unit, and an abnormal node set generation unit. The neighborhood connectivity characteristic analysis unit analyzes the neighborhood connectivity characteristics of potential abnormal candidate nodes. The abnormal node determination unit tracks the propagation direction of abnormal differences along the transmission path of power line carrier communication and determines abnormal nodes. The abnormal node set generation unit integrates all vertices determined to be abnormal to form an abnormal node set for power line carrier communication.

[0111] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A smart monitoring method based on power line carrier communication, characterized in that, Includes the following steps: Collect the antenna's operating parameters and environmental parameters, and calculate the antenna's radiated power; based on the antenna's radiated power, under ideal free space conditions, calculate the theoretical signal transmission intensity according to the inverse square law; Based on the influence of tunnel geometry on signal propagation, an environmental correction coefficient is introduced to correct the theoretical signal transmission strength. Antenna and tunnel features are acquired, potential monitoring points within the tunnel are clustered, and the selection of monitoring points is optimized to obtain a set of monitoring points. Based on the set of monitoring points, the signal transmission strength of the antenna inside the tunnel is monitored and compared with the theoretical signal transmission strength to obtain a set of transmission strength difference values; The graph structure is defined by taking each node in the power line carrier communication network as a vertex and the communication connection between nodes as edges. Based on the physical location of the monitoring point set, the power line carrier communication coverage node corresponding to each monitoring point is determined, and the transmission intensity difference of the monitoring points is bound to the vertex attribute of the corresponding node. Calculate the threshold range of the emission intensity difference associated with all nodes and mark potential abnormal candidate nodes; For potential abnormal candidate nodes, analyze the node's neighborhood connectivity characteristics; trace the propagation direction of the abnormal difference along the power line carrier communication transmission path to determine the abnormal node; integrate all vertices determined to be abnormal to form a set of abnormal nodes in power line carrier communication. The process of tracing the propagation direction of abnormal differences along the power line carrier communication transmission path and determining abnormal nodes includes: The main transmission path of data in the power line carrier communication network is determined, specifically the connection sequence of the main nodes from the signal source to the data endpoint. The node connection relationships of each branch path are recorded to form a transmission path map. Based on the transmission path map, for nodes with strong matching relationships, their positions in the main transmission path are obtained. When a node is the starting node of a certain path segment, and its abnormal difference value appears earlier than that of the downstream node, and the abnormal difference value of the downstream node is consistent with that node, the node is identified as a potential starting point for abnormal propagation. Starting from the potential starting point, the abnormality of the downstream nodes is checked sequentially along the transmission path. When the abnormal difference value of the downstream node gradually weakens with the increase of distance from the starting point, and the abnormality of each node has a strong matching connection attribute with the preceding node, it is confirmed that the abnormal difference value is propagated along the path, and the abnormal propagation path is recorded, including the propagation direction and the order in which the abnormality of each node appears. In the anomaly propagation path, when a node is the initial source of the abnormal difference and has the largest abnormal difference magnitude, that node is determined to be a core anomaly node. The condition for determining the initial source is that the node has the earliest occurrence time of the anomaly and no upstream node propagates the anomaly to that node through a strong matching connection. In the downstream transmission path of the core anomaly node, nodes that have abnormal differences due to strong matching connections with the core anomaly node or a preceding anomaly node, and that do not have independent anomaly triggering factors, are determined to be derived anomaly nodes. For potential anomaly candidate nodes, when there is no strong matching relationship between their connection attributes and those of neighboring nodes, and no upstream node propagates the anomaly to that node in the transmission path, and their abnormal difference is unrelated to other nodes, they are determined to be self-faulting nodes. Their anomalies are caused by their own hardware or local environmental problems and are unrelated to propagation.

2. The intelligent monitoring method based on power line carrier communication according to claim 1, characterized in that, The calculation of the theoretical signal transmission intensity based on antenna radiated power under ideal free space conditions, according to the inverse square law, includes: Antenna radiated power represents the total energy radiated into space by the antenna under ideal conditions. The propagation environment is set as ideal free space, specifically where there are no obstacles, reflectors, or electromagnetic interference. During signal propagation, energy attenuation only occurs due to spatial diffusion, with no other additional losses. The antenna is used as a point radiation source, and the radiated signal propagates uniformly in all directions in the form of a spherical wave. In ideal free space, the wavefront of a spherical wave is a complete sphere, with energy uniformly distributed on the surface and unimpeded propagation in all directions along a straight line. The total power radiated by the antenna is uniformly distributed on a sphere with the antenna as its center and the propagation distance as its radius. The theoretical signal transmission intensity at a certain distance is specifically the proportion of radiated power allocated to the spherical area corresponding to that distance. The target distance for calculating the theoretical signal transmission intensity is determined as the straight-line distance from the antenna to the monitoring point. According to the inverse square law, the spherical area at that distance is proportional to the square of the distance. Dividing the antenna radiated power by the spherical area yields the signal energy per unit area at that distance, which is taken as the theoretical signal transmission intensity.

3. The intelligent monitoring method based on power line carrier communication according to claim 1, characterized in that, The influence of the tunnel-based geometry on signal propagation is addressed by introducing an environmental correction coefficient to adjust the theoretical signal transmission strength, including: The tunnel's geometry includes its length, cross-sectional shape, diameter, curvature, and whether it branches. Analysis of historical data from similar tunnels reveals the correlation between geometric factors and signal attenuation, reflecting signal energy loss under different geometric structures. Using ideal free-space conditions as a benchmark, the environmental correction coefficient is 1, representing no additional attenuation. Based on this correlation, an adjustment rule for the environmental correction coefficient is established: when geometric factors causing signal attenuation exist, the environmental correction coefficient is less than 1, and the more severe the attenuation, the smaller the environmental correction coefficient. Based on the current tunnel geometry, and in accordance with the established geometric factors and adjustment rules, the corresponding environmental correction coefficient is calculated. The calculated environmental correction coefficient is then multiplied by the theoretical signal transmission strength for correction.

4. The intelligent monitoring method based on power line carrier communication according to claim 1, characterized in that, The process involves acquiring antenna and tunnel features, clustering potential monitoring points within the tunnel, optimizing the selection of monitoring points, and obtaining a set of monitoring points, including: The process involves acquiring antenna features, including radiation direction, coverage area, and signal strength distribution patterns; acquiring tunnel features, including tunnel length, width, location of curved sections, and material; pre-setting potential monitoring points within the tunnel to cover all areas; determining the similarity of signal characteristics at different potential monitoring points based on the signal strength distribution patterns in the antenna features; using the DBSCAN algorithm to group potential monitoring points with similar signal characteristics into the same cluster; monitoring points in the same cluster exhibit consistent signal performance, representing the signal characteristics of that area; and finally, dividing the potential monitoring points into several clusters through clustering. The goal of reinforcement learning is to select monitoring points from clusters obtained through clustering, so that the monitoring points cover all key signal areas in the tunnel with the minimum number of points. The signal environment inside the tunnel is used as the reinforcement learning environment, each cluster is a selectable state, and selecting a monitoring point in a certain cluster is an action. The reinforcement learning agent starts from the initial state and gradually selects monitoring points in different clusters. After each selection, a corresponding reward or penalty is given based on whether the coverage of the selected monitoring point is comprehensive and whether it can effectively reflect signal changes. Through continuous trial and adjustment of the selection strategy, the agent gradually learns the optimal selection method, selects a suitable number of monitoring points with reasonable distribution, and forms a monitoring point set.

5. The intelligent monitoring method based on power line carrier communication according to claim 1, characterized in that, The method involves monitoring the signal transmission strength of antennas within the tunnel based on a set of monitoring points, comparing this strength with the theoretical signal transmission strength, and obtaining a set of transmission strength difference values, including: Based on the set of monitoring points, a corresponding external signal monitoring device is installed at each monitoring point location within the tunnel. The external monitoring device is activated to monitor the antenna signal transmission strength of each monitoring point within the tunnel. The theoretical signal transmission strength corresponding to each monitoring point is retrieved and correlated with the location of each monitoring point. The actual signal transmission strength of each monitoring point is compared with the corresponding theoretical signal transmission strength at each time point. For each time point, the actual signal strength of the monitoring point is subtracted from the theoretical signal strength to obtain the transmission strength difference at that time point. The transmission strength differences of all monitoring points at each time point are integrated and sorted according to the monitoring point location and time point order to obtain a set of transmission strength difference values.

6. The intelligent monitoring method based on power line carrier communication according to claim 1, characterized in that, The step of determining the power line carrier communication coverage node corresponding to each monitoring point based on the physical location of the monitoring point set, and binding the transmission strength difference of the monitoring points to the vertex attributes of the corresponding node, includes: The physical location of each monitoring point is compared with the coverage area of ​​each node to determine which node's coverage area the monitoring point belongs to. When a monitoring point is located within the coverage area of ​​a certain node, and that node is the closest node to the monitoring point and has the most stable signal connection, that node is identified as the power line carrier communication coverage node corresponding to that monitoring point. For monitoring points located in areas where the coverage areas of multiple nodes overlap, the node with the best signal transmission quality is selected as the power line carrier communication coverage node by filtering historical communication data. For each power line carrier communication coverage node, the difference in transmission intensity among all its corresponding monitoring points is collected and added as a new attribute to the vertex attributes of that node in the graph structure.

7. The intelligent monitoring method based on power line carrier communication according to claim 1, characterized in that, The threshold range for calculating the emission intensity difference associated with all nodes, and marking potential anomalous candidate nodes, includes: Extract the emission intensity difference associated with all nodes from the graph structure, and set a threshold range based on the overall distribution characteristics of the emission intensity difference. For each node, obtain the emission intensity difference of all monitoring points corresponding to it, and determine whether each difference is within the threshold range. When some or all of the differences of a node exceed the threshold range, mark the node as a potential abnormal candidate node.

8. The intelligent monitoring method based on power line carrier communication according to claim 1, characterized in that, The analysis of the neighborhood connectivity characteristics of potential abnormal candidate nodes includes: Determine the neighboring nodes of each potential anomaly candidate node, specifically all nodes directly connected to that potential anomaly candidate node through edges; query whether each neighboring node is a potential anomaly candidate node, record the emission intensity difference of the neighboring nodes, and form a list of neighboring node states; Based on the state list of neighboring nodes, the connection attributes of edges between potential anomaly candidate nodes and neighboring nodes, as well as the labeling information of potential anomaly candidate nodes, are collected to obtain a dataset. The support and confidence indices in association rule mining are used to analyze the association strength between connection attributes and labeling information in the dataset. When the support and confidence of a certain connection attribute and labeling information are both higher than a set threshold, and the neighboring node is in the communication transmission direction between the potential anomaly candidate node and the neighboring node, it is determined that the connection attribute has a strong matching relationship with the neighboring node.

9. An intelligent monitoring system based on power line carrier communication, using the intelligent monitoring method based on power line carrier communication according to any one of claims 1-8, characterized in that, include: The theoretical signal transmission strength calculation module includes: an antenna radiated power calculation unit, a theoretical signal transmission strength calculation unit, and a theoretical signal transmission strength correction unit. The antenna radiated power calculation unit collects the antenna's operating parameters and environmental parameters to calculate the antenna radiated power. The theoretical signal transmission strength calculation unit, based on the antenna radiated power, calculates the theoretical signal transmission strength under ideal free-space conditions according to the inverse square law. The theoretical signal transmission strength correction unit, based on the influence of the tunnel's geometry on signal propagation, introduces an environmental correction coefficient to correct the theoretical signal transmission strength. The transmission intensity difference calculation module includes a monitoring point generation unit and a transmission intensity difference calculation unit. The monitoring point generation unit acquires antenna features and tunnel features, clusters potential monitoring points in the tunnel, optimizes the selection of monitoring points, and obtains a set of monitoring points. The transmission intensity difference calculation unit monitors the signal transmission intensity of the antenna in the tunnel based on the set of monitoring points, compares it with the theoretical signal transmission intensity, and obtains a set of transmission intensity differences. The potential anomaly marking module includes: a graph structure definition unit, a transmission intensity difference binding unit, and a potential anomaly marking unit. The graph structure definition unit defines the graph structure using each node in the power line carrier communication network as a vertex and the communication connections between nodes as edges. The transmission intensity difference binding unit determines the power line carrier communication coverage node corresponding to each monitoring point based on the physical location of the monitoring point set, and binds the transmission intensity difference of the monitoring point to the vertex attributes of the corresponding node. The potential anomaly marking unit calculates the threshold range of the transmission intensity differences associated with all nodes and marks potential anomaly candidate nodes. The abnormal node set generation module includes: a neighborhood connectivity characteristic analysis unit, an abnormal node determination unit, and an abnormal node set generation unit. The neighborhood connectivity characteristic analysis unit analyzes the neighborhood connectivity characteristics of potential abnormal candidate nodes. The abnormal node determination unit tracks the propagation direction of abnormal differences along the transmission path of power line carrier communication and determines abnormal nodes. The abnormal node set generation unit integrates all vertices determined to be abnormal to form an abnormal node set for power line carrier communication.

Images