Network parameter configuration verification method and system for electric power plant station

Abstract

The invention discloses a network parameter configuration verification method and system for an electric power plant station, and particularly relates to the technical field of parameter configuration verification. A hierarchical network model of a device layer, a subnet layer and a system layer is constructed, consistency verification is carried out on static parameters in combination with a rule base, and basic correctness of network configuration is verified; then, dynamic behaviors are simulated through a simulation tool, forwarding path consistency and time delay deviation in different communication scenes are analyzed, and the stability of a communication path in an abnormal condition is evaluated; when instability is detected, the layering and partitioning model is used for gradually backtracking and positioning abnormity, error configuration items are accurately identified from a system layer to an equipment layer, correction suggestions are automatically generated based on the network configuration management system, configuration updating is executed, and the correction effect is verified again and simulated; accurate positioning and rapid repairing of a network configuration problem are realized, and reliability and stability of a power plant station communication network in a complex environment are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of parameter configuration verification, and in particular to a network parameter configuration verification method and system for a power plant. Background Art

[0002] The network parameter configuration verification of power plants refers to the inspection and verification of the communication network parameters of power plants in the power system to ensure the correctness, integrity and consistency of the configuration, thereby ensuring the normal operation of the communication network. This technology is particularly important for the automation and stable operation of power systems, especially in smart grids and automated sites. In the prior art, power plants usually rely on communication networks to transmit various data, such as protection information, monitoring data and control instructions. These network parameters include IP addresses, gateways, port numbers, routing strategies, etc. Since the communication network of power plants involves a large number of devices (such as protection devices, measurement and control devices, intelligent terminals, etc.) and complex topological structures, network configuration errors may lead to communication failures or malfunctions. Traditional verification methods rely on manual inspection or simple tool support, but are prone to missed detection or false detection, especially when faced with complex configurations. To solve this problem, more and more systems are introducing automated verification tools to improve verification efficiency and accuracy by comparing configuration files, simulating communication scenarios or real-time monitoring.

[0003] For example, in a smart power plant, a communication network for protection devices and measurement and control devices is configured, requiring the IP address of each device to be within a specific network segment and not conflicting. Through the network parameter configuration verification tool, the system can automatically scan the configuration files of all devices, check whether the IP address meets the specifications, and verify whether the gateway settings are correct. If a device is configured with an incorrect subnet mask or conflicts with the IP address of another device, the verification tool can quickly discover and prompt corrections. In addition, when configuring a virtual local area network (VLAN), the verification tool can verify whether the device is correctly assigned to the predetermined VLAN, thereby avoiding unnecessary network broadcasts or the risk of data loss.

[0004] The prior art has the following deficiencies:

[0005] In some complex networks, there may be a need for multiple subnets to communicate. Although there is no conflict in the IP configuration within each subnet, the routing policy across subnets may lead to hidden conflicts due to incorrect static routing settings, routing redundancy, or undetected loops. For example, the communication data between two subnets is accidentally forwarded through a third-party intermediate device (such as a switch or router), forming an unnecessary network broadcast storm. In addition, due to data flow anomalies caused by routing loops or broadcast storms, network parameter configuration verification tools may detect unexpected traffic paths or a large number of duplicate data packets. This may cause the tool to mistakenly determine that there is a conflict in the network configuration (false positive), or fail to detect critical problems (missed negative) due to failure to capture the actual conflict path. Summary of the invention

[0006] The object of the present invention is to provide a method and system for verifying network parameter configuration of a power plant to solve the deficiencies in the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a network parameter configuration verification method for a power plant, comprising the following steps:

[0008] S1: Based on the topological structure of the power plant communication network, a hierarchical and partitioned network model is constructed, wherein the network model includes a device layer, a subnet layer, and a system layer;

[0009] S2: Based on the constructed network model, the rule base is used to perform consistency check on static parameters, including checking whether the network parameters comply with the allocation rules, whether there is redundancy or loop in the routing table, and whether the device is assigned to the correct VLAN, so as to verify the correctness of the network configuration;

[0010] S3: After completing the static verification, simulate the dynamic behavior of the actual network operation, generate data flow paths under different communication scenarios through simulation tools, determine the consistency of the forwarding paths of the simulated data flow under different communication scenarios, and compare the real-time collected network traffic with the simulated path to determine the delay deviation between the real-time traffic and the simulated data;

[0011] S4: Evaluate the stability of the communication path under abnormal conditions based on the forwarding path consistency of the simulated data flow in different communication scenarios and the delay deviation between the real-time traffic and the simulated data;

[0012] S5: When the communication path becomes unstable, the detected anomaly is gradually traced back and located in combination with the hierarchical partitioning model. The abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer. Finally, the wrong parameter configuration item is accurately located at the device layer.

[0013] S6: After the anomaly is located, a correction suggestion is generated, including a configuration adjustment plan and impact analysis. In combination with the network configuration management system, the configuration update is performed through the automated correction function, and static verification and dynamic simulation are re-run to verify the effectiveness of the correction.

[0014] Preferably, in S1, the device layer defines the IP address, subnet mask, gateway and port number of each device, the subnet layer describes the routing strategy and communication rules of each subnet, and the system layer is used to model cross-subnet routing tables, global topology and protocol interactions.

[0015] Preferably, in S3, after analyzing the forwarding path consistency of the simulated data flow in different communication scenarios, a forwarding path anomaly index is generated, and the method for obtaining the forwarding path anomaly index is:

[0016] The communication network of the power plant is represented as a directed graph G = (V, E), where: V represents the node set, represents the network equipment; E is the edge set, represents the connection relationship between devices, and the edge weight w(e) represents the forwarding cost of the connection; the expected path P is calculated from the source node s to the target node t using the shortest path algorithm expected , the path is represented as a sequence of nodes: P expected =[s, v1, v2, ..., t]; where each node on the path represents the forwarding order of the data flow in order, and the simulation path is obtained by outputting the simulation path P from the simulation tool. simulated , expressed as: P simulated = [s, v′1, v′2, ..., t]; Calculate the edit distance D (P expected , P simulated ), the edit distance represents the number of operations required to transform one path into another; the operations include: insertion: adding a node to the path; deletion: removing a node from the path; replacement: replacing a node in the path with another node;

[0017] Let P expected The length is n, P simulated The length is m, and the dynamic programming matrix D(i,j) is defined as: c(i,j)=0if P expected [i]=P simulated [j], otherwise c(i,j)=1; D(n,m) is the path edit distance, and the path deviation PD is calculated, which is defined as the ratio of the edit distance to the expected path length, and the expression is: Among them, |P expected | is the number of nodes in the expected path; the forwarding path anomaly index is defined as a comprehensive index weighted by path deviation, taking into account edge weight deviation: Where MKH is the forwarding path anomaly index, is the sum of the weights of the edges in the simulated path that are not in the expected path, is the sum of the edge weights in the expected path, and w(e) is the weight of each edge in the path.

[0018] Preferably, in S3, a delay deviation fluctuation index is generated after analyzing the transmission delay of the real-time traffic and the expected delay deviation fluctuation of the simulation path. The method for obtaining the delay deviation fluctuation index is:

[0019] Assume the real-time collected delay is T real-time =[t1, t2, ..., t N ], the expected delay of simulation is T simulated =[s1,s2,...,s N ]; Calculate the delay deviation δ N =|t N -s N |, and construct the corresponding sequence: ΔT = [δ1, δ2, ..., δ N ],δ i =t i -s i , i = 1, 2, ..., N; the sequence will be used as the input of wavelet transform, and the time delay deviation sequence ΔT will be decomposed by wavelet, and ΔT will be decomposed into multi-level low-frequency components and high-frequency components. The expression is: Low frequency component A j Used to capture the long-term trend of deviation, D k Used to capture short-term fluctuations in deviations; for each layer of high-frequency components D k Calculate energy: Where: D k [i] is the high frequency component D k The i-th value in N k Yes D k The total energy E of the delay deviation sequence is calculated by total , the expression is: The energy of the high-frequency component is compared with the total energy to calculate the delay deviation fluctuation index, which is expressed as: GBH is the delay deviation fluctuation index.

[0020] Preferably, in S4, based on the forwarding path consistency of the simulated data flow in different communication scenarios and the delay deviation between the real-time traffic and the simulated data, the stability of the communication path under abnormal conditions is evaluated, specifically:

[0021] The forwarding path anomaly index and the delay deviation fluctuation index are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the stability value label of the communication path under abnormal conditions as the prediction target, and takes minimizing the sum of prediction errors of the stability value labels of all communication paths under abnormal conditions as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The stability value of the communication path under abnormal conditions is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0022] Preferably, in S4, the acquired stability value of the communication path under abnormal conditions is compared with a stability value reference threshold preset according to historical data; if the stability value of the communication path under abnormal conditions is greater than or equal to the preset stability value reference threshold, it means that the stability of the communication path under abnormal conditions is high, and no warning signal is generated at this time, indicating that the path can maintain reliable communication under abnormal conditions; if the stability value of the communication path under abnormal conditions is less than the preset stability value reference threshold, it means that the stability of the communication path under abnormal conditions is low, and a warning signal is generated at this time, indicating that the path is easily affected under abnormal conditions and the communication path is unstable.

[0023] Preferably, in S5, when the communication path becomes unstable, the detected anomaly is gradually traced back and located in combination with the hierarchical partition model, the abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer, and finally the wrong parameter configuration item is accurately located at the device layer, specifically:

[0024] Map the communication path stability value S to the distribution of the subnet to which it belongs, and calculate the average stability value S of each subnet 子网 , the expression is: Among them, N 路径 is the number of communication paths within the subnet, S i is the stability value of the ith path. If the average stability value of a subnet is less than the preset average stability threshold, the subnet is marked as an abnormal subnet.

[0025] In the abnormal subnet, identify the device that causes path instability. In the subnet, for all communication paths in which each device participates, calculate the device's impact index I 设备 , the expression is: Among them, N 路径，设备 is the number of paths that the device participates in; S j is the stability value of the jth path, S threshold is the reference threshold of the stability value. If the impact index of a device exceeds the device impact threshold, the device is considered to be abnormal.

[0026] Locate the specific wrong configuration item in the abnormal device and check each network configuration parameter P of the device. k , calculate its impact on the path stability value S (P k , S), the expression is: Among them, Cov(P k , S) is the parameter P k Covariance with stability value S; σ(P k ), σ(S) are P k and the standard deviation of S; when C(P k , S) is close to -1 or 1, indicating that the parameter is strongly correlated with the stability value and is an abnormal parameter, so it is marked as an abnormal configuration item.

[0027] The present invention also provides a network parameter configuration verification system for a power plant, including a network modeling module, a static parameter verification module, a flow comparison module, a stability evaluation module, an abnormal backtracking positioning module, and a correction and verification module:

[0028] Network modeling module: Based on the topological structure of the power plant communication network, a hierarchical and partitioned network model is constructed, and the network model includes a device layer, a subnet layer, and a system layer;

[0029] Static parameter verification module: Based on the constructed network model, the rule base is used to perform consistency verification on static parameters, including checking whether the network parameters comply with the allocation rules, whether there is redundancy or loop in the routing table, and whether the device is assigned to the correct VLAN, so as to verify the correctness of the network configuration;

[0030] Traffic comparison module: After completing the static verification, simulate the dynamic behavior of the actual network operation, generate data flow paths under different communication scenarios through simulation tools, determine the consistency of the forwarding paths of the simulated data flow under different communication scenarios, and compare the real-time collected network traffic with the simulated path to determine the delay deviation between the real-time traffic and the simulated data;

[0031] Stability evaluation module: Based on the forwarding path consistency of simulated data flow in different communication scenarios and the delay deviation between real-time traffic and simulated data, the stability of the communication path under abnormal conditions is evaluated;

[0032] Abnormal backtracking and positioning module: When the communication path becomes unstable, the detected abnormality is gradually backtracked and positioned in combination with the hierarchical partitioning model. The abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer. Finally, the wrong parameter configuration item is accurately located at the device layer.

[0033] Correction and verification module: After the anomaly is located, correction suggestions are generated, including configuration adjustment plans and impact analysis. Combined with the network configuration management system, configuration updates are performed through automated correction functions, and static checks and dynamic simulations are re-run to verify the effectiveness of the corrections.

[0034] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0035] 1. The present invention divides the network into device layer, subnet layer and system layer through a hierarchical and partitioned network model, and comprehensively verifies static parameters in combination with a rule base to ensure the correctness of the network configuration foundation; at the same time, through dynamic behavior simulation and real-time traffic analysis, the forwarding path consistency and delay deviation are evaluated, and the stability value of the communication path under abnormal conditions is further quantified. This method can accurately identify potential unstable paths, generate stability assessment reports, and automatically diagnose and repair unstable paths.

[0036] 2. The present invention introduces a polynomial regression model based on machine learning, takes the forwarding path anomaly index and the delay deviation fluctuation index as feature vectors, establishes a path stability prediction model, and realizes efficient prediction of communication path stability and abnormal warning; combined with the hierarchical partitioning model, it gradually traces back from the system layer to the device layer to locate anomalies, accurately identifies erroneous network configuration items, and uses the network configuration management system to perform automatic corrections. After the correction is completed, the correction effect is rechecked and verified to form a complete closed-loop processing mechanism. The present invention improves the efficiency and accuracy of network anomaly diagnosis and repair, significantly enhances the reliability and stability of power plant communication networks under complex conditions, and provides strong support for the safe operation of smart grids. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0038] Figure 1 The figure is a flow chart of the method of the present invention.

[0039] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Example 1, please refer to Figure 1 As shown, the network parameter configuration verification method for a power plant described in this embodiment includes the following steps:

[0042] S1: Based on the topological structure of the power plant communication network, a hierarchical and partitioned network model is constructed, wherein the network model includes a device layer, a subnet layer, and a system layer;

[0043] S2: Based on the constructed network model, the rule base is used to perform consistency check on static parameters, including checking whether the network parameters comply with the allocation rules, whether there is redundancy or loop in the routing table, and whether the device is assigned to the correct VLAN, so as to verify the correctness of the network configuration;

[0044] S3: After completing the static verification, simulate the dynamic behavior of the actual network operation, generate data flow paths under different communication scenarios through simulation tools, determine the consistency of the forwarding paths of the simulated data flow under different communication scenarios, and compare the real-time collected network traffic with the simulated path to determine the delay deviation between the real-time traffic and the simulated data;

[0045] S4: Evaluate the stability of the communication path under abnormal conditions based on the forwarding path consistency of the simulated data flow in different communication scenarios and the delay deviation between the real-time traffic and the simulated data;

[0046] S5: When the communication path becomes unstable, the detected anomaly is gradually traced back and located in combination with the hierarchical partitioning model. The abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer. Finally, the wrong parameter configuration item is accurately located at the device layer.

[0047] S6: After the anomaly is located, a correction suggestion is generated, including a configuration adjustment plan and impact analysis. In combination with the network configuration management system, the configuration update is performed through the automated correction function, and static verification and dynamic simulation are re-run to verify the effectiveness of the correction.

[0048] In S1, based on the topological structure of the power plant communication network, a hierarchical and partitioned network model is constructed, and the network model includes a device layer, a subnet layer, and a system layer, specifically:

[0049] The device layer covers the basic network configuration parameters of all physical devices, including but not limited to: Device 1: protection device, IP address: 192.168.1.10, subnet mask: 255.255.255.0, gateway: 192.168.1.1, port number: 502 (Modbus TCP protocol); Device 2: measurement and control device, IP address: 192.168.2.20, subnet mask: 255.255.255.0, gateway: 192.168.2.1, port number: 2404 (IEC 60870-5-104 protocol); Device 3: intelligent terminal, IP address: 192.168.3.30, subnet mask: 255.255.255.0, gateway: 192.168.3.1, port number: 102 (IEC 61850 protocol MMS channel); the device layer collects the above device parameters through the automatic discovery tool and compares them with the predefined network rule library to ensure that the allocation of IP addresses and gateways is correct and there is no conflict.

[0050] At the subnet layer, the devices are grouped into three logical subnets, and the communication rules within each subnet are defined: Subnet 1 (protection subnet), including protection devices and switches related to protection. Internal communication rules: All devices are only allowed to send protection signals through the TCP / IP protocol, and broadcast traffic is prohibited. Routing strategy: External traffic is uniformly forwarded through the gateway 192.168.1.1. Subnet 2 (monitoring subnet) contains measurement and control devices and related servers. Internal communication rules: Devices are allowed to publish monitoring data to the server, and the maximum traffic limit of UDP protocol data is 5Mbps. Routing strategy: Inter-subnet communication is configured to the target subnet 192.168.3.0 / 24 through a static routing table. Subnet 3 (control subnet) contains intelligent terminals and dispatching devices. Internal communication rules: MMS and GOOSE messages of the IEC 61850 protocol are allowed, and all GOOSE messages must be transmitted first. Routing strategy: Dynamic routing (OSPF) is supported to ensure high reliability and low latency of critical data flows. The subnet layer generates a communication rule table through a network topology analysis tool and verifies whether the rules meet the set security and performance requirements.

[0051] The system layer is responsible for describing the network topology of the entire plant and the cross-subnet communication path: Global topology model: Subnet 1, Subnet 2, and Subnet 3 are connected to the router through the main switch, and the router connects the plant dispatch center and the remote control center. The router supports dynamic routing protocols (such as OSPF) and static routing table configuration to achieve communication between subnets and external networks. Routing table: Static routing table: 192.168.1.0 / 24→Gateway 192.168.1.1; 192.168.2.0 / 24→Gateway 192.168.2.1; 192.168.3.0 / 24→Gateway 192.168.3.1. Dynamic routing table: The optimal path is calculated through the OSPF protocol. When the default gateway of a subnet fails, it is automatically adjusted to the backup path. Support the priority transmission strategy of MMS messages and GOOSE messages of the IEC 61850 protocol to ensure the real-time performance of key data flows. The monitoring subnet and the protection subnet are relayed through the control subnet, and the traffic path is: subnet 1 → subnet 3 → subnet 2. The system layer integrates the routing strategies and communication rules of each subnet to generate a global model of cross-subnet communication to guide dynamic simulation and parameter verification.

[0052] By simulating the gateway failure of the measurement and control device, verify whether the traffic of subnet 2 can reach the external dispatch center through the dynamic routing relay of subnet 3. Check whether the protection subnet causes broadcast traffic to overflow to the control subnet or monitoring subnet due to configuration errors. Verify whether the system layer ensures that GOOSE messages can still be transmitted first in high-load scenarios to avoid delays.

[0053] S2: Based on the constructed network model, the rule base is used to perform consistency checks on static parameters, including checking whether the network parameters comply with the allocation rules, whether there is redundancy or loops in the routing table, and whether the device is assigned to the correct VLAN, so as to verify the correctness of the network configuration.

[0054] The network parameters of each device in the device layer are checked item by item through the rule base. Check rule example: The IP address must be within the specified subnet range (such as the range of subnet 1 is 192.168.1.0 / 24). The gateway must be set to the subnet gateway that matches the device IP (such as 192.168.1.1). The subnet mask must comply with the standard configuration (such as 255.255.255.0). For example: Device 1 (protection device): The IP address is 192.168.1.10, and the gateway is 192.168.1.1, which complies with the rules; Device 2 (measurement and control device): The IP address is configured as 192.168.3.25, and the gateway is 192.168.2.1, but it should belong to the range of subnet 2 (192.168.2.0 / 24). The check tool prompts that the gateway setting is wrong and needs to be corrected to 192.168.3.1.

[0055] Analyze the routing tables at the subnet layer and the system layer to ensure the integrity of the routing configuration and that there are no conflicts. Verification rule example: Each subnet in the static routing table must have a unique destination address (no duplication or conflict); detect whether there is a loop between subnets to avoid routing falling into an infinite loop; check whether the routes generated by the dynamic routing protocol conflict with the static routes. For example: At the system layer, the verification tool found that there are two static routes for 192.168.2.0 / 24: Route 1: 192.168.2.0 / 24 → Gateway 192.168.2.1; Route 2: 192.168.2.0 / 24 → Gateway 192.168.3.1. The tool determines that it is a redundant configuration and prompts to keep Route 1 and remove Route 2. In the dynamic routing check, the routing table generated by the OSPF protocol is compared with the static routing table to confirm that there are no conflicting paths.

[0056] At the subnet layer, according to the predefined VLAN configuration rules, verify whether the device is correctly assigned to the corresponding VLAN, and check whether the communication rules between VLANs meet the specifications. Verification rule example: The VLAN ID of each device must match its logical subnet; the communication traffic between the protection subnet and the monitoring subnet must be isolated through a specific trunk VLAN. For example: Device 1 (protection device) should belong to VLAN 10 (protection subnet), but the verification tool detects that it is configured as VLAN 20 (monitoring subnet), prompting that the mapping between the device and the subnet is incorrect. The VLAN communication rule verification found that the broadcast traffic of the protection subnet was accidentally transmitted to the monitoring subnet. The verification tool prompts that the VLAN division rules of the switch should be adjusted to limit cross-subnet broadcasts.

[0057] The gateway configuration of the measurement and control device is wrong, and the gateway needs to be adjusted to match its actual subnet. The static routing table has redundant configuration, and after optimization, the only valid path is retained to eliminate potential conflicts. The VLAN allocation of the protection device is wrong, and it needs to be re-divided to ensure the correctness of subnet isolation and communication rules.

[0058] S3: After completing the static verification, simulate the dynamic behavior of the actual network operation, generate data flow paths under different communication scenarios through simulation tools, determine the consistency of the forwarding paths of the simulated data flow under different communication scenarios, and compare the real-time collected network traffic with the simulated path to determine the delay deviation between the real-time traffic and the simulated data.

[0059] According to the actual operation requirements of the power plant, three typical communication scenarios are set: Scenario 1: The protection device sends fault information to the dispatching center; Trigger condition: an abnormality occurs in a certain device (such as voltage exceeding the limit). Expected data flow path: protection subnet (device 1) → router → dispatching center (remote server). Scenario 2: The measurement and control device uploads real-time data to the monitoring server; Trigger condition: the measurement and control device periodically uploads monitoring data (such as power, frequency). Expected data flow path: monitoring subnet (device 2) → monitoring server (internal subnet). Scenario 3: The dispatching system sends control instructions to the intelligent terminal; Trigger condition: the dispatching center sends a load adjustment instruction. Expected data flow path: dispatching center → router → control subnet (device 3).

[0060] Based on the hierarchical partitioned network model, use simulation tools (such as Mininet or network simulator NS3) to simulate the above communication scenario and generate the actual forwarding path of the data flow:

[0061] The header information of the simulated data packet includes the source address, destination address, protocol type (such as TCP, UDP) and priority field; the data flow path is identified by the number of hops between devices and the network devices (switches, routers) on the path.

[0062] Compare the actual path generated by the simulation with the expected path to determine whether the forwarding behavior is consistent. Scenario 1 result: Simulation path: Device 1 (192.168.1.10) → Subnet 1 switch → Router (192.168.1.1) → Dispatch Center (10.0.0.1). Comparison result: The path is completely consistent, and the simulation verification passed. Scenario 2 result: Simulation path: Device 2 (192.168.2.20) → Subnet 2 switch → Monitoring server (192.168.2.100). Comparison result: The path is completely consistent, but the simulation tool detects that UDP protocol traffic is occasionally mislabeled as low priority. The QoS settings need to be adjusted to ensure that real-time data is transmitted first. Scenario 3 result: Simulation path: Dispatch Center (10.0.0.1) → Router (192.168.3.1) → Subnet 3 switch → Device 3 (192.168.3.30). Comparison results: The paths are inconsistent. The simulation tool found that the router mistakenly forwarded the data flow to the switch of subnet 1, resulting in communication failure. The reason for the analysis is that the static routing configuration lacks the entry of subnet 3, and the relevant routing configuration should be supplemented.

[0063] After analyzing the forwarding path consistency of the simulated data flow in different communication scenarios, a forwarding path anomaly index is generated. The forwarding path anomaly index is obtained as follows:

[0064] The communication network of the power plant is represented as a directed graph G = (V, E), where: V represents a node set, represents network devices (such as switches, routers, terminal devices, etc.). E is an edge set, which represents the connection relationship between devices, and the edge weight w(e) represents the forwarding cost of the connection (such as delay, bandwidth consumption or number of hops). Use the shortest path algorithm (such as Dijkstra algorithm or Floyd-Warshall algorithm) to calculate the expected path P from the source node s to the target node t. expected , the path is represented as a sequence of nodes: P expected =[s, v1, v2, ..., t]; where each node on the path represents the forwarding order of the data flow in sequence, and the simulation path is obtained:

[0065] Output simulation path P from simulation tool simulated , expressed as: P simulated =[s,v1 ′ , v2 ′ , ..., t]; Calculate the edit distance D(P expeced , P simulated ), the edit distance represents the number of operations required to transform one path into another. The operations include: insertion: adding a node to the path; deletion: removing a node from the path; replacement: replacing a node in the path with another node.

[0066] Let P expected The length is n, P simulated The length is m, and the dynamic programming matrix D(i,j) is defined as: c(i,j)=0if P expected [i]=P simulated [j], otherwise c(i,j)=1; D(n,m) is the path edit distance, and the path deviation PD is calculated, which is defined as the ratio of the edit distance to the expected path length, and the expression is: Among them, |P expected | is the number of nodes in the expected path; the forwarding path anomaly index is defined as a comprehensive index weighted by path deviation, taking into account edge weight deviation: Where MKH is the forwarding path anomaly index, is the sum of the weights of the edges in the simulated path that are not in the expected path, is the sum of the edge weights in the expected path, and w(e) is the weight of each edge in the path (such as delay, number of hops, bandwidth occupancy).

[0067] The larger the forwarding path anomaly index is, the less consistent the actual forwarding path of the simulated data flow is with the expected path. This may indicate serious configuration errors or abnormal behavior in the network, such as path deviation, missing routing table configuration, instability of dynamic routing, or data packets being mistakenly forwarded to irrelevant nodes. In addition, deviations in edge weights (such as latency, hop count, or bandwidth consumption) may also significantly increase the anomaly index, further reflecting that the actual network performance is far below expectations. A high anomaly index usually indicates that the reliability and efficiency of communications may be seriously affected, and it is necessary to prioritize troubleshooting of problematic devices and configuration items.

[0068] On the contrary, the smaller the forwarding path anomaly index is, the more consistent the actual forwarding path of the simulated data flow is with the expected path, and the network is in good operation. When the path deviation is close to zero, it means that the data packets pass through each node strictly according to the designed forwarding order, and a small weight deviation means that the transmission performance (such as delay and bandwidth utilization) is in line with expectations. The low anomaly index reflects the high reliability and stability of the communication network, indicating that the existing configuration and routing strategy are suitable for the current scenario and no major adjustments are required. In this case, the network can effectively support key tasks such as real-time monitoring, protection and control of power plants.

[0069] Use traffic collection tools (such as NetFlow, sFlow, or Wireshark) to monitor the transmission of packets between network devices in real time and record key information: source address and destination address (IP or MAC). Transport protocol (such as TCP, UDP). Packet size (Bytes). Timestamp (accurate to milliseconds). Get expected traffic information from the simulation path generated by the simulation tool (such as Mininet or NS3), including: forwarding path (device sequence) for each data flow. Transport protocol, source and destination address, and expected traffic size. Expected latency and bandwidth utilization.

[0070] Compare the actual forwarding path of real-time traffic with the simulated path: Path consistency: Check whether the actual path of each data flow is included in the expected path; Path deviation: If the data flow deviates from the expected path, record the node sequence of the deviation and the corresponding traffic proportion. Calculation formula: Path deviation PD: PD = number of deviated nodes / number of nodes in the simulated path. The higher the deviation, the greater the deviation between the actual transmission of the path and the simulation result. Compare the real-time collected traffic size with the simulated traffic size to determine the deviation of the transmission data: Traffic difference: ΔF = |Freal-time-Fsimulated|; where Freal-time is the real-time traffic size and Fsimulated is the simulated traffic size. Compare the transmission delay of real-time traffic with the expected delay of the simulated path, and calculate the delay deviation ΔT: ΔT = Treal-time-Tsimulated; Compare the bandwidth occupancy of real-time and simulation to determine whether there is insufficient or over-utilized bandwidth. Combined with the comparison results, analyze the deviation between real-time traffic and simulated data: If there are unexpected forwarding nodes in the actual path, it may be unexpected forwarding caused by routing configuration errors or network failures. If the real-time traffic is significantly different from the simulated traffic, it may be caused by packet loss, retransmission, or traffic bursts. If the deviation of latency or bandwidth utilization exceeds the tolerance range, it may reflect network congestion, device overload, or dynamic routing instability.

[0071] The delay deviation fluctuation index is generated by analyzing the transmission delay of real-time traffic and the expected delay deviation fluctuation of the simulation path. The delay deviation fluctuation index is obtained as follows:

[0072] Assume the real-time collected delay is T real-time =[t1, t2, ..., t N ], the expected delay of simulation is T simulated =[s1,s2,...,s N ]; Calculate the delay deviation δ N =|t N -s N |, and construct the corresponding sequence: ΔT = [δ1, δ2, ..., δ N ], δ i =t i -s i , i = 1, 2, ..., N; the sequence will be used as the input of wavelet transform. Decompose the time delay deviation sequence ΔT by wavelet, and select the appropriate wavelet basis function (such as Daubechies wavelet, Haar wavelet, etc.). Decompose ΔT into multi-level low-frequency components and high-frequency components, and the expression is: Low frequency component A j Used to capture the long-term trend of deviation, D k It is used to capture short-term fluctuations of deviation and reflect the jitter characteristics of deviation.

[0073] For each layer of high frequency component D k Calculate energy: Where: D k [i] is the high frequency component D k The i-th value in N k Yes D k The total energy E of the delay deviation sequence is calculated by total , the expression is: The energy of the high-frequency component is compared with the total energy to calculate the delay deviation fluctuation index, which is expressed as: GBH is the delay deviation fluctuation index.

[0074] A high volatility index indicates that short-term delay fluctuations account for a large proportion of the total delay deviation, which means that the transmission performance of the network is significantly affected by dynamic factors (such as burst traffic, network congestion, equipment overload, etc.). In this case, the delay performance of the communication path is difficult to predict, and the data stream may be subject to significant jitter and uneven transmission delays, resulting in reduced real-time and reliability of key tasks. For data transmission scenarios that require high real-time performance, such as power plants, a high volatility index indicates that the path is prone to instability under abnormal conditions, and it is necessary to optimize traffic scheduling or increase network resource redundancy to enhance stability.

[0075] A low volatility index indicates that short-term fluctuations contribute less to the total delay deviation, indicating that the network path has a high anti-interference ability when facing dynamic changes (such as traffic peaks or device switching). At this time, the transmission delay of the data stream is closer to the expected value, there is less jitter, and the path shows good robustness and consistency. For the transmission of mission-critical data, such as protection signals and control instructions for power systems, a low volatility index indicates that the communication path can remain stable under abnormal conditions, meet real-time and reliability requirements, and no major adjustments are required.

[0076] S4: Evaluate the stability of the communication path under abnormal conditions based on the forwarding path consistency of the simulated data flow in different communication scenarios and the delay deviation between the real-time traffic and the simulated data.

[0077] The forwarding path anomaly index and the delay deviation fluctuation index are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the stability value label of the communication path under abnormal conditions as the prediction target, and takes minimizing the sum of prediction errors of the stability value labels of all communication paths under abnormal conditions as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The stability value of the communication path under abnormal conditions is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0078] The method for obtaining the stability value of the communication path under abnormal conditions is: from the comprehensive feature vector training data of the trained machine learning model, obtain the corresponding function expression: S = F (MKH, GBH); where F is the output function of the model, MKH is the forwarding path anomaly index, GBH is the delay deviation fluctuation index, and S is the stability value of the communication path under abnormal conditions.

[0079] The obtained stability value of the communication path under abnormal conditions is compared with the stability value reference threshold preset according to historical data. If the stability value of the communication path under abnormal conditions is greater than or equal to the preset stability value reference threshold, it means that the stability of the communication path under abnormal conditions is high, and no warning signal is generated at this time, indicating that the path can maintain reliable communication under abnormal conditions; if the stability value of the communication path under abnormal conditions is less than the preset stability value reference threshold, it means that the stability of the communication path under abnormal conditions is low, and a warning signal is generated at this time, indicating that the path is easily affected under abnormal conditions and the communication path is unstable.

[0080] S5: When the communication path becomes unstable, the detected anomaly is gradually traced back and located in combination with the hierarchical partitioning model. The abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer. Finally, the wrong parameter configuration item is accurately located at the device layer.

[0081] Based on the global network topology, locate the subnet where the anomaly occurs. Abnormal path distribution calculation:

[0082] Map the communication path stability value S to the distribution of the subnet to which it belongs, and calculate the average stability value S of each subnet 子网 , the expression is: Among them, N 路径 is the number of communication paths within the subnet, S i is the stability value of the ith path. If the average stability value of a subnet is less than the preset average stability threshold, the subnet is marked as an abnormal subnet.

[0083] In the abnormal subnet, identify the device that causes path instability. In the subnet, for all communication paths in which each device participates, calculate the device's impact index I 设备 , the expression is: Among them, N 路径，设备 is the number of paths that the device participates in; S j is the stability value of the jth path, S threshold is the reference threshold of the stability value. If the impact index of a device exceeds the device impact threshold, the device is considered abnormal.

[0084] Locate specific incorrect configuration items (such as IP address, subnet mask, routing table items, etc.) in the abnormal device. Correlation analysis between configuration parameters and abnormal paths:

[0085] For each network configuration parameter P of the device k , calculate its impact on the path stability value S (P k , S), the expression is: Among them, Cov(P k , S) is the parameter P k Covariance with stability value S; σ(P k ), σ(S) are P k and the standard deviation of S;

[0086] When C(P k , S) is close to -1 or 1, indicating that the parameter is strongly correlated with the stability value, is an abnormal parameter, and has a significant impact on the path stability value, so it is marked as an abnormal configuration item.

[0087] S6: After the anomaly is located, a correction suggestion is generated, including a configuration adjustment plan and impact analysis. In combination with the network configuration management system, the configuration update is performed through the automated correction function, and static verification and dynamic simulation are re-run to verify the effectiveness of the correction.

[0088] Based on the results of the abnormal location, specific correction suggestions are made for the incorrect configuration items: If the routing table entry of a device points to the wrong gateway, it is recommended to modify it to the correct gateway address. For example, correct 192.168.1.254 to 192.168.1.1. If the device is assigned to the wrong VLAN, for example, the protection device is configured to the monitoring VLAN, it is recommended to adjust the VLAN configuration and assign the device to the correct VLAN (such as VLAN 10). If the IP address of the device is repeated with other devices, it is recommended to reallocate the unused IP address and update the relevant routing table.

[0089] Before adjusting the configuration, analyze the possible impact of the correction to avoid causing new problems to other network components or data flows: Adjusting the routing table may cause changes in the paths of some data flows. It is necessary to ensure that the new paths do not introduce delays or bandwidth bottlenecks. Modifying the VLAN configuration may affect the broadcast range of the data. It is necessary to verify whether the new configuration meets the isolation requirements. Adjusting the IP address or gateway configuration may cause temporary communication interruption. It is necessary to evaluate the adjustment time window and try to avoid operations during peak traffic periods. The generated correction suggestions include the detailed steps of the configuration adjustment, the possible scope of impact, and the expected effect after the correction, forming a complete correction plan.

[0090] Implement automated corrections through the NCM system and generate automated configuration scripts (such as CLI commands or API calls) based on the correction suggestions. Send the correction scripts to the target device through NCM to complete the parameter adjustment. NCM collects device status feedback to confirm whether the configuration update is successful. After the correction is completed, immediately verify the basic operating status of the device: check whether the routing table and VLAN configuration meet the correction requirements; confirm that the device communication is normal, and there is no connection interruption or data loss.

[0091] Re-run the static verification tool to check the consistency and integrity of the revised configuration: confirm whether the newly configured IP address, routing table, and VLAN meet the predefined specifications; detect whether there are new configuration conflicts or omissions. When the static verification result passes, it means that the basic correctness of the configuration is guaranteed. Run the dynamic behavior simulation tool to verify the actual impact of the correction on network performance: check whether the corrected data flow path is consistent with expectations; evaluate whether the correction improves latency deviation or bandwidth utilization to ensure that performance meets the requirements; simulate different communication scenarios to confirm whether the correction eliminates the original anomalies and does not introduce new problems.

[0092] Determine whether the correction is successful based on the results of static verification and dynamic simulation: Correction success: If the verification and simulation results are as expected, it means that the correction plan is effective and the configuration adjustment is complete. Correction failure: If the anomaly still exists, re-analyze the problem point, generate new correction suggestions and repeat the process. Record the correction process and its verification results in the network operation and maintenance system as a reference for future fault diagnosis and optimization.

[0093] In this embodiment, based on the topological structure of the power plant communication network, a hierarchical and partitioned network model is constructed, including the device layer, the subnet layer and the system layer. On this basis, the network parameters are statically checked using a rule base to check the consistency and integrity of the configuration rules. After the static check is completed, the dynamic behavior of the network is simulated by a simulation tool to generate data flow paths under different communication scenarios, and the real-time collected traffic is compared with the simulated path to analyze the consistency of the forwarding path and the delay deviation. The stability value of the communication path under abnormal conditions is comprehensively evaluated. When the path becomes unstable, the hierarchical and partitioned model is used to gradually trace back and locate the abnormal subnet from the system layer to the abnormal device at the subnet layer, and finally the wrong configuration item is accurately found at the device layer. After the positioning is completed, correction suggestions are generated, including configuration adjustment and impact analysis, and automatic correction is performed through the network configuration management system, and static verification and dynamic simulation are re-performed to verify the effectiveness of the correction, so as to ensure the final correctness of the network configuration and the stability of the communication path.

[0094] Example 2, please refer to Figure 2As shown, the network parameter configuration verification system for a power plant described in this embodiment includes a network modeling module, a static parameter verification module, a flow comparison module, a stability evaluation module, an abnormal backtracking positioning module, and a correction and verification module:

[0095] Network modeling module: Based on the topological structure of the power plant communication network, a hierarchical and partitioned network model is constructed, and the network model includes a device layer, a subnet layer, and a system layer;

[0096] Static parameter verification module: Based on the constructed network model, the rule base is used to perform consistency verification on static parameters, including checking whether the network parameters comply with the allocation rules, whether there is redundancy or loop in the routing table, and whether the device is assigned to the correct VLAN, so as to verify the correctness of the network configuration;

[0097] Traffic comparison module: After completing the static verification, simulate the dynamic behavior of the actual network operation, generate data flow paths under different communication scenarios through simulation tools, determine the consistency of the forwarding paths of the simulated data flow under different communication scenarios, and compare the real-time collected network traffic with the simulated path to determine the delay deviation between the real-time traffic and the simulated data;

[0098] Stability evaluation module: Based on the forwarding path consistency of simulated data flow in different communication scenarios and the delay deviation between real-time traffic and simulated data, the stability of the communication path under abnormal conditions is evaluated;

[0099] Abnormal backtracking and positioning module: When the communication path becomes unstable, the detected abnormality is gradually backtracked and positioned in combination with the hierarchical partitioning model. The abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer. Finally, the wrong parameter configuration item is accurately located at the device layer.

[0100] Correction and verification module: After the anomaly is located, correction suggestions are generated, including configuration adjustment plans and impact analysis. Combined with the network configuration management system, configuration updates are performed through automated correction functions, and static checks and dynamic simulations are re-run to verify the effectiveness of the corrections.

[0101] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0102] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0103] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0104] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for verifying network parameter configuration of a power plant, characterized in that: The following steps are involved: S1: Based on the topological structure of the power plant communication network, a hierarchical and partitioned network model is constructed, wherein the network model includes a device layer, a subnet layer, and a system layer; S2: Based on the constructed network model, the rule base is used to perform consistency check on static parameters, including checking whether the network parameters comply with the allocation rules, whether there is redundancy or loop in the routing table, and whether the device is assigned to the correct VLAN, so as to verify the correctness of the network configuration; S3: After completing the static verification, simulate the dynamic behavior of the actual network operation, generate data flow paths under different communication scenarios through simulation tools, determine the consistency of the forwarding paths of the simulated data flow under different communication scenarios, and compare the real-time collected network traffic with the simulated path to determine the delay deviation between the real-time traffic and the simulated data; S4: Evaluate the stability of the communication path under abnormal conditions based on the forwarding path consistency of the simulated data flow in different communication scenarios and the delay deviation between the real-time traffic and the simulated data; S5: When the communication path becomes unstable, the detected anomaly is gradually traced back and located in combination with the hierarchical partitioning model. The abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer. Finally, the wrong parameter configuration item is accurately located at the device layer. S6: After the anomaly is located, a correction suggestion is generated, including a configuration adjustment plan and impact analysis. In combination with the network configuration management system, the configuration update is performed through the automated correction function, and static verification and dynamic simulation are re-run to verify the effectiveness of the correction.

2. A method for verifying network parameter configuration of a power plant according to claim 1, characterized in that: In S1, the device layer defines the IP address, subnet mask, gateway and port number of each device, the subnet layer describes the routing strategy and communication rules of each subnet, and the system layer is used to model the routing table, global topology and protocol interaction across subnets.

3. A method for verifying network parameter configuration for a power plant according to claim 1, characterized in that: In S3, the forwarding path consistency of the simulated data flow in different communication scenarios is analyzed to generate a forwarding path anomaly index. The forwarding path anomaly index is obtained by: The communication network of the power plant is represented as a directed graph G = (V, E), where: V represents the node set, represents the network equipment; E is the edge set, represents the connection relationship between devices, and the edge weight w(e) represents the forwarding cost of the connection; the expected path P is calculated from the source node s to the target node t using the shortest path algorithm expected , the path is represented as a sequence of nodes: P expected =[s, v1, v2, ..., t]; where each node on the path represents the forwarding order of the data flow in order, and the simulation path is obtained by outputting the simulation path P from the simulation tool. simulated , expressed as: P simulated = [s, v′1, v′2, ..., t]; Calculate the edit distance D (P expected , P simulated ), the edit distance represents the number of operations required to transform one path into another; the operations include: insertion: adding a node to the path; deletion: removing a node from the path; replacement: replacing a node in the path with another node; Let P expected The length is n, P simulated The length is m, and the dynamic programming matrix D(i,j) is defined as: c(i,j)=0if P expected [i]=P simulated [j], otherwise c(i,j)=1; D(n,m) is the path edit distance, and the path deviation PD is calculated, which is defined as the ratio of the edit distance to the expected path length, and the expression is: Among them, |P expected | is the number of nodes in the expected path; the forwarding path anomaly index is defined as a comprehensive index weighted by path deviation, taking into account edge weight deviation: Where MKH is the forwarding path anomaly index, is the sum of the weights of the edges in the simulated path that are not in the expected path, is the sum of the edge weights in the expected path, and w(e) is the weight of each edge in the path.

4. A method for verifying network parameter configuration for a power plant according to claim 3, characterized in that: In S3, the transmission delay of the real-time traffic and the expected delay deviation fluctuation of the simulation path are analyzed to generate a delay deviation fluctuation index. The delay deviation fluctuation index is obtained as follows: Assume the real-time collected delay is T real-time =[t1, t2, ..., t N ], the expected delay of simulation is T simulated =[s1,s2,...,s N ]; Calculate the delay deviation δ N =|t N -s N |, and construct the corresponding sequence: ΔT = [δ1, δ2, ..., δ N ], δ i =t i -s i , i = 1, 2, ..., N; The sequence will be used as the input of wavelet transform, and the time delay deviation sequence ΔT will be decomposed by wavelet, and ΔT will be decomposed into multi-level low-frequency components and high-frequency components. The expression is: Low frequency component A j Used to capture the long-term trend of deviation, D k Used to capture short-term fluctuations in deviations; For each layer of high frequency component D k Calculate energy: Where: D k [i] is the high frequency component D k The i-th value in N k Yes D k The total energy E of the delay deviation sequence is calculated by total , the expression is: The energy of the high-frequency component is compared with the total energy to calculate the delay deviation fluctuation index, which is expressed as: GBH is the delay deviation fluctuation index.

5. A method for verifying network parameter configuration for a power plant according to claim 4, characterized in that: In S4, based on the forwarding path consistency of the simulated data flow in different communication scenarios and the delay deviation between the real-time traffic and the simulated data, the stability of the communication path under abnormal conditions is evaluated, specifically: The forwarding path anomaly index and the delay deviation fluctuation index are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the stability value label of the communication path under abnormal conditions as the prediction target, and takes minimizing the sum of prediction errors of the stability value labels of all communication paths under abnormal conditions as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The stability value of the communication path under abnormal conditions is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

6. A method for verifying network parameter configuration for a power plant according to claim 5, characterized in that: In S4, the acquired stability value of the communication path under abnormal conditions is compared with the stability value reference threshold preset according to historical data. If the stability value of the communication path under abnormal conditions is greater than or equal to the preset stability value reference threshold, it means that the stability of the communication path under abnormal conditions is high, and no warning signal is generated at this time, indicating that the path can maintain reliable communication under abnormal conditions; if the stability value of the communication path under abnormal conditions is less than the preset stability value reference threshold, it means that the stability of the communication path under abnormal conditions is low, and a warning signal is generated at this time, indicating that the path is easily affected under abnormal conditions and the communication path is unstable.

7. A method for verifying network parameter configuration for a power plant according to claim 6, characterized in that: In S5, when the communication path becomes unstable, the detected anomaly is gradually traced back and located in combination with the hierarchical partitioning model. The abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer. Finally, the wrong parameter configuration item is accurately located at the device layer, specifically: Map the communication path stability value S to the distribution of the subnet to which it belongs, and calculate the average stability value S of each subnet 子网 , the expression is: Among them, N 路径 is the number of communication paths within the subnet, S i is the stability value of the ith path. If the average stability value of a subnet is less than the preset average stability threshold, the subnet is marked as an abnormal subnet. In the abnormal subnet, identify the device that causes path instability. In the subnet, for all communication paths in which each device participates, calculate the device's impact index I 设备 , the expression is: Among them, N 路径，设备 is the number of paths that the device participates in; S j is the stability value of the jth path, S threshold is the reference threshold of the stability value. If the impact index of a device exceeds the device impact threshold, the device is considered to be abnormal. Locate the specific wrong configuration item in the abnormal device and check each network configuration parameter P of the device. k , calculate its impact on the path stability value S (P k , S), the expression is: Among them, Cov(P k , S) is the parameter P k Covariance with stability value S; σ(P k ), σ(S) are P k and the standard deviation of S; when C(P k , S) is close to -1 or 1, indicating that the parameter is strongly correlated with the stability value and is an abnormal parameter, so it is marked as an abnormal configuration item.

8. A network parameter configuration verification system for a power plant, used to implement a network parameter configuration verification method for a power plant according to any one of claims 1 to 7, characterized in that: It includes network modeling module, static parameter verification module, traffic comparison module, stability assessment module, abnormal backtracking location module and correction and verification module: Network modeling module: Based on the topological structure of the power plant communication network, a hierarchical and partitioned network model is constructed, and the network model includes a device layer, a subnet layer, and a system layer; Static parameter verification module: Based on the constructed network model, the rule base is used to perform consistency verification on static parameters, including checking whether the network parameters comply with the allocation rules, whether there is redundancy or loop in the routing table, and whether the device is assigned to the correct VLAN, so as to verify the correctness of the network configuration; Traffic comparison module: After completing the static verification, simulate the dynamic behavior of the actual network operation, generate data flow paths under different communication scenarios through simulation tools, determine the consistency of the forwarding paths of the simulated data flow under different communication scenarios, and compare the real-time collected network traffic with the simulated path to determine the delay deviation between the real-time traffic and the simulated data; Stability evaluation module: Based on the forwarding path consistency of simulated data flow in different communication scenarios and the delay deviation between real-time traffic and simulated data, the stability of the communication path under abnormal conditions is evaluated; Abnormal backtracking and positioning module: When the communication path becomes unstable, the detected abnormality is gradually backtracked and positioned in combination with the hierarchical partitioning model. The abnormal subnet is identified from the system layer, and the abnormal configuration of the specific device is confirmed from the subnet layer. Finally, the wrong parameter configuration item is accurately located at the device layer. Correction and verification module: After the anomaly is located, correction suggestions are generated, including configuration adjustment plans and impact analysis. Combined with the network configuration management system, configuration updates are performed through automated correction functions, and static checks and dynamic simulations are re-run to verify the effectiveness of the corrections.

Images