Node anomaly diagnosis method and system based on carrier time-frequency synchronization network cluster

Abstract

The invention discloses a node anomaly diagnosis method and system based on a carrier time-frequency synchronization network cluster, and the method comprises the steps: collecting time-frequency node real-time measurement information based on a carrier time-frequency network; dividing the time-frequency node real-time measurement information into a characteristic quantity and an abnormal type; randomly selecting a part of data from the time-frequency node real-time measurement information as a verification set, taking the other part of data as a training set, and learning a decision tree from the training set based on a decision tree generation algorithm C4.5; and evaluating the performance of the decision tree in the verification set according to the accuracy and the recall rate, if the performance cannot reach a set threshold, pruning the decision tree, and retraining the pruned decision tree until the performance of the decision tree reaches the set threshold or the number of iterations exceeds a preset upper limit.

Description

Technical Field

[0001] The present invention relates to the field of measurement, and more specifically, to a method and system for diagnosing node anomalies in a carrier-based time-frequency synchronization network cluster. Background Art

[0002] At present, power line carrier communication technology has been fully used in the State Grid substation area. The metering devices use carrier communication to collect data and perform remote fee control operations. The actual operation data statistics show that the communication quality is stable and reliable, and the substation carrier communication technology has matured. In order to solve the problem that many metering devices in the substation area lack effective time synchronization means, the time accuracy error is large, and it cannot meet the needs of the new power system, the current research and construction of the time-frequency network based on carrier technology for online time-frequency value transmission and online time-frequency performance monitoring of metering devices in the substation area is underway. However, there is no relevant research on the real-time diagnosis technology of abnormalities of carrier time-frequency networks and time-frequency nodes. This method proposes an abnormality diagnosis method based on big data based on the characteristics of time-frequency networks and time-frequency nodes, which plays a key role in effectively improving the operation and maintenance efficiency of future carrier-based time-frequency transmission networks. Summary of the invention

[0003] According to the present invention, a method and system for diagnosing node anomalies in a carrier-based time-frequency synchronization network cluster are provided to address the deficiencies of existing network anomaly diagnosis technologies and the technical problems of the actual situation in which carrier communication is used for time-frequency synchronization in station areas.

[0004] According to a first aspect of the present invention, a method for diagnosing node abnormality in a carrier-based time-frequency synchronization network cluster is provided, comprising:

[0005] Determine the carrier-based time and frequency synchronization network for the substation area, and collect real-time measurement information of time and frequency nodes based on the carrier-based time and frequency network;

[0006] The real-time measurement information of the time-frequency node is divided into a characteristic quantity and an abnormal type, the characteristic quantity of the real-time measurement information of the time-frequency node is extracted, and the characteristic value of the time-frequency node is obtained, and the range of the abnormal node is determined by analyzing the characteristic value of the time-frequency node, and the abnormal type of the time-frequency node is divided by analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster;

[0007] Randomly select a portion of data from the real-time measurement information of the time-frequency nodes as a validation set, and the remaining portion of data as a training set, and learn a decision tree from the training set based on the decision tree generation algorithm C4.5;

[0008] The performance of the decision tree in the validation set is evaluated based on precision and recall. If the performance cannot reach the set threshold, the decision tree is pruned and retrained until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

[0009] Optionally, determining a carrier-based station area time and frequency synchronization network, and collecting time and frequency node real-time measurement information based on the carrier time and frequency network, includes:

[0010] Determine that the area time and frequency synchronization network includes multiple terminal nodes, namely, a carrier master clock node and a plurality of carrier slave clock nodes;

[0011] Based on the power carrier network, the terminal nodes are topologically related, and then a synchronization system is established in the entire power carrier network. In this topological relationship: T-GM is the main node of the network and is the master clock; T-TC is a relay node, which is both a slave clock and a transparent clock; T-SC is the terminal node and is a slave clock;

[0012] The carrier is carried out through the T-GM master node, and time and frequency synchronization is performed on other nodes in the network. At the same time, real-time measurement information of the time and frequency nodes is collected. The real-time measurement information of the time and frequency nodes includes: the device number, the upper device number, the lower device number, the frequency error, the time error, the network delay, the last time and frequency synchronization time and the ambient temperature.

[0013] Optionally, extracting the characteristic quantity of the real-time measurement information of the time-frequency node to obtain the characteristic value of the time-frequency node includes:

[0014] The time-frequency node characteristic values ​​include: frequency temperature correction coefficient, normalized frequency difference after node temperature correction, node normalized frequency difference mean, node normalized frequency difference standard deviation, node frequency deviation aging rate, node time difference mean, node time difference standard deviation, frequency deviation with upper-level node, time deviation with upper-level node, frequency deviation with lower-level node, time deviation with upper-level node, node network delay mean and node network delay standard deviation;

[0015] According to the accurate frequency value fyi of the master clock, the frequency difference measurement value fci of the slave clock, and the real-time temperature measurement value Ti, (fci / fyi) = k1*Ti+k2*Ti 2 +k3*Ti 3 , perform fitting, and obtain the frequency temperature correction coefficient K(k1, k2, k3) online;

[0016] Based on the frequency difference measurement value fci and the temperature measurement value Ti, the main node frequency measurement value fyi, and the node temperature correction coefficient K, according to fc = [fci-fyi*(1+k1*Ti+k2*Ti2+k3*Ti3)] / fyi, the normalized frequency difference fc after node temperature correction is obtained;

[0017] The normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time are averaged to obtain the node normalized frequency difference mean fm;

[0018] The standard deviation of the normalized frequency difference fc measurement results after multiple node temperature corrections within a period of time is calculated to obtain the node normalized frequency difference standard deviation fs;

[0019] The normalized frequency deviation mean fmi of the node after temperature correction within a period of time is fitted according to fmi=λ*T+c to obtain the node frequency deviation aging rate λ;

[0020] The average of multiple time difference measurement results of the node within a period of time is obtained to obtain the node time difference mean ΔTm;

[0021] The standard deviation of the node time difference measurement results over a period of time is calculated to obtain the node time difference standard deviation ΔTs;

[0022] Subtract the normalized frequency difference of the node from the normalized frequency difference of the upper node in the same period to obtain the frequency deviation fuc between the node and the upper node;

[0023] Subtract the time deviation of the superior node from the node time deviation to get the time deviation ΔTuc between the node and the superior node;

[0024] Subtract the normalized frequency difference of the node from the normalized frequency difference of the subordinate node in the same period to obtain the frequency deviation fnc between the node and the subordinate node;

[0025] Subtract the node time deviation from the subordinate node time deviation to obtain the time deviation ΔTnc between the node and the subordinate node;

[0026] The average network delay measurement value of the node within a period of time is obtained to obtain the node network delay mean value ΔTnm;

[0027] The standard deviation of the node network delay measurement value within a period of time is calculated to obtain the node network delay standard deviation ΔTns.

[0028] Optionally, the abnormal node range is determined by analyzing the characteristic values ​​of the time-frequency nodes, and the abnormal types of the time-frequency nodes are divided by analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster, including:

[0029] By analyzing the characteristic values ​​of the time-frequency nodes, the range of abnormal nodes is determined. By analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the single-node synchronization network delay instability failure of the time-frequency parameters of adjacent nodes in the cluster, the time-frequency node abnormality types are divided into single-node time-frequency synchronization data packet loss failure, single-node synchronization network networking instability failure, single-node time-frequency performance attenuation exceeding threshold failure, single-node time-frequency performance mutation failure, regional time-frequency synchronization data packet loss failure, and regional synchronization network delay instability failure;

[0030] Single-node time-frequency synchronization data packet loss failure means that the node monitoring data can be collected, but the node's last time is not updated. At the same time, the time-frequency characteristics of the subordinate node are consistent with that of the node, while the time-frequency characteristics of the superior node are normal.

[0031] The single-node synchronization network delay instability fault refers to the node network delay standard deviation exceeding the threshold;

[0032] The unstable fault of single-node synchronous network refers to the node being offline;

[0033] The single-node time-frequency performance degradation exceeding threshold fault means that the node frequency deviation aging rate exceeds the threshold, while the time-frequency characteristics of the upper-level node are normal;

[0034] A single-node time-frequency performance sudden change fault refers to a comprehensive assessment of the changes in the node normalized frequency difference mean, node normalized frequency difference standard deviation, node time difference mean, and node time difference standard deviation. The assessment result exceeds the threshold, and the time-frequency characteristics of the upper-level node are normal, and the time-frequency performance of the lower-level node is consistent with that of the node.

[0035] Regional time-frequency synchronization data packet loss failure refers to data packet loss failures at multiple nodes in the station area;

[0036] The regional synchronization network delay instability fault refers to the network delay instability faults in multiple nodes in the station area;

[0037] The unstable fault of regional synchronous network refers to the situation that multiple nodes in the area are offline;

[0038] Batch quality failure of time-frequency equipment refers to the failure of multiple devices in operation to have performance degradation exceeding the threshold or sudden change in time-frequency performance.

[0039] Optionally, the performance of the decision tree in the validation set is evaluated based on precision and recall. If the performance does not reach a set threshold, the decision tree is pruned, and the pruned decision tree is retrained until the performance of the decision tree reaches a set threshold or the number of iterations exceeds a preset upper limit, including:

[0040] The performance of the decision tree in the validation set is evaluated based on precision and recall. If the performance cannot reach the set threshold, a hyperparameter probability distribution is defined to prune the decision tree, initialize the Markov chain Monte Carlo chain, and consider the feature quantity of the real-time measurement information of the time-frequency nodes to ensure that the pruning strategy is optimized for the actual application scenario.

[0041] Sampling the pruning strategies, evaluating the theoretical performance of each pruning strategy on the validation set, and updating the hyperparameter probability distribution based on the evaluation performance, so that the pruning strategy is gradually optimized;

[0042] The pruned decision tree is retrained with the updated hyperparameters, trained on the training set, and its performance is evaluated on the validation set. The above process is repeated until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

[0043] According to another aspect of the present invention, there is also provided a node abnormality diagnosis system in a carrier-based time-frequency synchronization network cluster, comprising:

[0044] Determine the real-time measurement information module, which is used to determine the carrier-based station area time and frequency synchronization network and collect the real-time measurement information of the time and frequency nodes based on the carrier time and frequency network;

[0045] A real-time measurement information classification module is used to classify the real-time measurement information of the time-frequency node into characteristic quantities and abnormal types, extract the characteristic quantities of the real-time measurement information of the time-frequency node, obtain the characteristic values ​​of the time-frequency node, determine the range of abnormal nodes by analyzing the characteristic values ​​of the time-frequency node, and classify the abnormal types of the time-frequency node by analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster;

[0046] A learning decision tree module is used to randomly select a part of the data from the real-time measurement information of the time-frequency node as a validation set and the rest of the data as a training set, and learn a decision tree from the training set based on the decision tree generation algorithm C4.5;

[0047] The validation set performance evaluation module is used to evaluate the performance of the decision tree in the validation set based on precision and recall. If the performance cannot reach the set threshold, the decision tree is pruned and retrained until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

[0048] Optionally, determining a real-time measurement information module includes:

[0049] Determine the time-frequency synchronization network submodule, which is used to determine that the station area time-frequency synchronization network includes multiple terminal nodes, namely, a carrier master clock node and several carrier slave clock nodes;

[0050] The topology relationship establishment submodule is used to establish a topology relationship between each terminal node based on the power carrier network, and then establish a synchronization system in the entire power carrier network. In this topology relationship: T-GM is the network master node and the master clock; T-TC is a relay node, which is both a slave clock and a transparent clock; T-SC is the terminal node and a slave clock;

[0051] The submodule for collecting real-time measurement information of time and frequency nodes is used to carry out carrier through the T-GM master node, synchronize the time and frequency of other nodes in the network, and collect real-time measurement information of time and frequency nodes. The real-time measurement information of time and frequency nodes includes: the device number, the upper device number, the lower device number, the frequency error, the time error, the network delay, the last time and frequency synchronization time and the ambient temperature.

[0052] Optionally, the real-time measurement information module is divided into:

[0053] The time-frequency node characteristic values ​​include: frequency temperature correction coefficient, normalized frequency difference after node temperature correction, node normalized frequency difference mean, node normalized frequency difference standard deviation, node frequency deviation aging rate, node time difference mean, node time difference standard deviation, frequency deviation with upper-level node, time deviation with upper-level node, frequency deviation with lower-level node, time deviation with upper-level node, node network delay mean and node network delay standard deviation;

[0054] The frequency temperature correction coefficient submodule is used to calculate the temperature correction coefficient according to the accurate frequency value fyi of the master clock, the frequency difference measurement value fci of the slave clock, and the real-time temperature measurement value Ti, according to (fci / fyi)=k1*Ti+k2*Ti 2 +k3*Ti 3 , perform fitting, and obtain the frequency temperature correction coefficient K(k1, k2, k3) online;

[0055] The submodule for obtaining the normalized frequency difference after node temperature correction is used to obtain the normalized frequency difference fc after node temperature correction based on the frequency difference measurement value fci and the temperature measurement value Ti, the main node frequency measurement value fyi, and the node temperature correction coefficient K according to fc=[fci-fyi*(1+k1*Ti+k2*Ti2+k3*Ti3)] / fyi;

[0056] A node normalized frequency difference mean submodule is obtained, which is used to average the normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time to obtain the node normalized frequency difference mean fm;

[0057] The node normalized frequency difference standard deviation submodule is used to obtain the standard deviation of the normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time to obtain the node normalized frequency difference standard deviation fs;

[0058] The node frequency deviation aging rate submodule is obtained, which is used to fit the normalized frequency deviation mean fmi of the node after temperature correction within a period of time according to fmi=λ*T+c to obtain the node frequency deviation aging rate λ;

[0059] The node time difference mean submodule is used to average multiple time difference measurement results of the node within a period of time to obtain the node time difference mean ΔTm;

[0060] The node time difference standard submodule is used to obtain the standard deviation of the node time difference measurement results for multiple times within a period of time to obtain the node time difference standard deviation ΔTs;

[0061] The submodule for obtaining the frequency deviation between the node and the upper node is used to obtain the frequency deviation fuc between the node and the upper node by subtracting the normalized frequency difference of the upper node at the same period from the normalized frequency difference of the node;

[0062] The submodule for obtaining the time deviation between the node and the upper node is used to obtain the time deviation ΔTuc between the node and the upper node by subtracting the time deviation of the upper node in the same period from the node time deviation;

[0063] The submodule for obtaining the frequency deviation between the node and the lower level node is used to obtain the frequency deviation fnc between the node and the lower level node by subtracting the normalized frequency difference of the lower level node in the same period from the normalized frequency difference of the node;

[0064] A submodule for obtaining a time deviation between a node and a subordinate node is used to obtain a time deviation ΔTnc between a node and a subordinate node by subtracting the time deviation of the subordinate node in the same period from the node time deviation;

[0065] The node network delay mean submodule is used to average the node network delay measurement values ​​within a period of time to obtain the node network delay mean ΔTnm;

[0066] The node network delay standard deviation submodule is used to calculate the standard deviation of the node network delay measurement value within a period of time to obtain the node network delay standard deviation ΔTns.

[0067] Optionally, the real-time measurement information module is divided into:

[0068] The abnormal type classification submodule is used to determine the range of abnormal nodes by analyzing the characteristic values ​​of the time-frequency nodes, and to classify the abnormal types of time-frequency nodes into single-node time-frequency synchronization data packet loss failure, single-node synchronization network networking instability failure, single-node time-frequency performance attenuation exceeding threshold failure, single-node time-frequency performance mutation failure, regional time-frequency synchronization data packet loss failure, and regional synchronization network delay instability failure by analyzing the evolution law of single-node time-frequency parameters and the evolution law of single-node synchronization network delay instability failure of adjacent nodes in the cluster;

[0069] Single-node time-frequency synchronization data packet loss failure means that the node monitoring data can be collected, but the node's last time is not updated. At the same time, the time-frequency characteristics of the subordinate node are consistent with that of the node, while the time-frequency characteristics of the superior node are normal.

[0070] The single-node synchronization network delay instability fault refers to the node network delay standard deviation exceeding the threshold;

[0071] The unstable fault of single-node synchronous network refers to the node being offline;

[0072] The single-node time-frequency performance degradation exceeding threshold fault means that the node frequency deviation aging rate exceeds the threshold, while the time-frequency characteristics of the upper-level node are normal;

[0073] A single-node time-frequency performance sudden change fault refers to a comprehensive assessment of the changes in the node normalized frequency difference mean, node normalized frequency difference standard deviation, node time difference mean, and node time difference standard deviation. The assessment result exceeds the threshold, and the time-frequency characteristics of the upper-level node are normal, and the time-frequency performance of the lower-level node is consistent with that of the node.

[0074] Regional time-frequency synchronization data packet loss failure refers to data packet loss failures at multiple nodes in the station area;

[0075] The regional synchronization network delay instability fault refers to the network delay instability faults in multiple nodes in the station area;

[0076] The unstable fault of regional synchronous network refers to the situation that multiple nodes in the area are offline;

[0077] Batch quality failure of time-frequency equipment refers to the failure of multiple devices in operation to have performance degradation exceeding the threshold or sudden change in time-frequency performance.

[0078] Optionally, evaluate the validation set performance module, including:

[0079] The decision tree pruning submodule is used to evaluate the performance of the decision tree in the validation set based on precision and recall. If the performance cannot reach the set threshold, the hyperparameter probability distribution is defined to prune the decision tree, initialize the Markov chain Monte Carlo chain, and consider the characteristic quantity of the real-time measurement information of the time-frequency node to ensure that the pruning strategy is optimized for the actual application scenario;

[0080] A submodule for updating the probability distribution of hyperparameters is used to sample the pruning strategies, evaluate the theoretical performance of each pruning strategy on the validation set, and update the probability distribution of hyperparameters based on the evaluation performance, so that the pruning strategies are gradually optimized;

[0081] The retraining decision tree submodule is used to retrain the pruned decision tree with the updated hyperparameters, train the decision tree on the training set, evaluate its performance on the validation set, and repeat the above process until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

[0082] Therefore, the cause of abnormal communication network in the substation can be quickly diagnosed, and the abnormal diagnosis result can be given by the problem feature value. The diagnostic effect is significant in the pilot substation where this algorithm is deployed. The accuracy of intelligent diagnosis in this solution can improve the detection rate and hit rate of abnormal communication network diagnosis in the substation according to the time of accumulating problem features and the continuous training of the model for new problem data sets, which greatly reduces the time and economic cost of manual investigation and on-site problem location. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] A more complete understanding of exemplary embodiments of the present invention may be obtained by referring to the following drawings:

[0084] Figure 1 A schematic diagram of a process for diagnosing node anomalies in a carrier-based time-frequency synchronization network cluster according to the present embodiment;

[0085] Figure 2 A schematic diagram of establishing a topological relationship between the terminal nodes described in this implementation mode;

[0086] Figure 3 This is a schematic diagram of a node abnormality diagnosis system in a carrier-based time-frequency synchronization network cluster described in this embodiment. DETAILED DESCRIPTION

[0087] Now, exemplary embodiments of the present invention are described with reference to the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely and to fully convey the scope of the present invention to those skilled in the art. The terms used in the exemplary embodiments shown in the accompanying drawings are not intended to limit the present invention. In the accompanying drawings, the same units / elements are marked with the same reference numerals.

[0088] Unless otherwise specified, the terms (including technical terms) used herein have the commonly understood meanings to those skilled in the art. In addition, it is understood that the terms defined in commonly used dictionaries should be understood to have the same meanings as those in the context of the relevant fields, and should not be understood as idealized or overly formal meanings.

[0089] According to a first aspect of the present invention, a method 100 for diagnosing node abnormality in a carrier-based time-frequency synchronization network cluster is provided. Figure 1 As shown, the method 100 includes:

[0090] S101: Determine a carrier-based station area time and frequency synchronization network, and collect real-time measurement information of time and frequency nodes based on the carrier time and frequency network;

[0091] S102: Divide the real-time measurement information of the time-frequency node into characteristic quantities and abnormal types, extract the characteristic quantities of the real-time measurement information of the time-frequency node, obtain characteristic values ​​of the time-frequency node, determine the range of abnormal nodes by analyzing the characteristic values ​​of the time-frequency node, and classify the abnormal types of the time-frequency nodes by analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster;

[0092] S103: Randomly select a part of the data from the real-time measurement information of the time-frequency node as a validation set, and the rest of the data as a training set, and learn a decision tree from the training set based on the decision tree generation algorithm C4.5;

[0093] S104: Evaluate the performance of the decision tree in the validation set based on precision and recall. If the performance cannot reach the set threshold, prune the decision tree and retrain the pruned decision tree until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

[0094] Specifically, the carrier-based substation time and frequency synchronization network includes a carrier master clock node and several carrier slave clock nodes. Based on the power carrier network, the terminal nodes are topologically related, and then a synchronization system is established in the entire network. Figure 2 In this topology, T-GM is the master node of the network and is the master clock; T-TC is a relay node and is both a slave clock and a transparent clock; T-SC is the end node and is a slave clock.

[0095] T-GM synchronizes the time and frequency of other nodes in the network through the carrier, and collects real-time measurement information of the time and frequency parameters of each node.

[0096] The specific analysis process and methods are as follows:

[0097] I) Collect the real-time measurement results of each node in the substation area. The specific data items are shown in Table 1.

[0098] Table 1 Real-time measurement data items of time-frequency nodes

[0099] Collection item number Collection item name 1 This equipment number 2 Parent device number 3 Subordinate device number 4 Frequency Error 5 Time error 6 Network delay 7 Last time and frequency synchronization time 8 Ambient temperature

[0100] II) Based on the real-time measurement results, the feature quantities in Table 2 are extracted and analyzed, and these feature quantities are used as inputs of the abnormality diagnosis algorithm.

[0101] Table 2 Time-frequency node eigenvalues

[0102] Feature Number Feature Name 1 Frequency Temperature Correction Coefficient 2 Normalized frequency difference after node temperature correction 3 Node normalized frequency difference mean 4 Node normalized frequency difference standard deviation 5 Node frequency deviation aging rate 6 Mean time difference of nodes 7 Node time difference standard deviation 8 Frequency deviation from the upper node 9 Time deviation with the parent node 10 Frequency deviation from lower-level nodes 11 Time deviation with lower-level nodes 12 Average node network delay 13 Node network delay standard deviation

[0103] Frequency temperature correction coefficient K: According to the accurate frequency value fyi of the master clock, the frequency difference measurement value fci of the slave clock, and the real-time temperature measurement value Ti, fitting is performed according to (fci / fyi)=k1*Ti+k2*Ti2+k3*Ti3 to obtain the frequency correction coefficient K(k1, k2, k3) online.

[0104] Normalized frequency difference fc after node temperature correction: based on the frequency difference measurement value fci and temperature measurement value Ti, the main node frequency measurement value fyi, and the node temperature correction coefficient K, according to fc = [fci-fyi*(1+k1*Ti+k2*Ti2+k3*Ti3)] / fyi.

[0105] Node normalized frequency difference mean fm: fm is obtained by averaging multiple fc measurement results of the node within a period of time.

[0106] Node normalized frequency difference standard deviation fs: fs is obtained by calculating the standard deviation of multiple fc measurement results of the node within a period of time.

[0107] Node frequency deviation aging rate λ: The normalized frequency deviation mean fmi of the node after temperature correction within a period of time is fitted according to fmi=λ*T+c to obtain λ.

[0108] Node time difference mean ΔTm: ΔTm is obtained by averaging multiple time difference measurement results of the node within a period of time.

[0109] Node time difference standard deviation ΔTs: ΔTs is obtained by calculating the standard deviation of multiple time difference measurement results of the node within a period of time.

[0110] Frequency deviation fuc from the upper node: fuc is obtained by subtracting the normalized frequency difference of the node from the normalized frequency difference of the upper node at the same period.

[0111] Time deviation ΔTuc with the upper-level node: ΔTuc is obtained by subtracting the time deviation of the upper-level node in the same period from the node time deviation.

[0112] Frequency deviation fnc between the nodes and the lower level: fuc is obtained by subtracting the normalized frequency difference of the node from the normalized frequency difference of the lower level node in the same period.

[0113] Time deviation ΔTnc between the upper node and the lower node: ΔTuc is obtained by subtracting the time deviation of the lower node in the same period from the node time deviation.

[0114] Average node network delay ΔTnm: ΔTnm is obtained by averaging the network delay measurements of the node over a period of time.

[0115] Node network delay standard deviation ΔTns: Calculate the standard deviation of the node's network delay measurement value over a period of time to get ΔTns.

[0116] III) Based on the abnormal situations of the time-frequency network, the abnormal types are divided, as shown in Table 3.

[0117] Table 3 Classification of abnormal types

[0118]

[0119]

[0120] Single-node time-frequency synchronization data packet loss failure: The node monitoring data can be collected, but the node's last time is not updated. At the same time, the time-frequency characteristics of the subordinate node are consistent with this node, while the time-frequency characteristics of the superior node are normal.

[0121] Single-node synchronization network delay unstable failure: The standard deviation of the node network delay exceeds the threshold.

[0122] Single-node synchronous network unstable fault: The node is offline.

[0123] Single-node time-frequency performance degradation exceeding threshold fault: The node frequency deviation aging rate exceeds the threshold, while the time-frequency characteristics of the upper-level node are normal.

[0124] Single-node time-frequency performance sudden change failure: Comprehensively evaluate the changes through the node normalized frequency difference mean, node normalized frequency difference standard deviation, node time difference mean, and node time difference standard deviation. The evaluation result exceeds the threshold. At the same time, the time-frequency characteristics of the upper-level node are normal, and the time-frequency performance of the lower-level node is consistent with that of this node.

[0125] Regional time and frequency synchronization data packet loss failure: Data packet loss failure occurs at multiple nodes in the substation area.

[0126] Regional synchronization network delay and unstable failure: There are network delay and unstable failures at multiple nodes in the area.

[0127] Unstable regional synchronization network failure: multiple nodes in the area are offline.

[0128] Batch quality failure of time-frequency equipment: Multiple devices in operation have performance degradation exceeding the threshold or time-frequency performance mutation failure.

[0129] IV) Randomly select a portion of the data as the validation set and the rest as the training set, and use the decision tree generation algorithm C4.5 to learn a decision tree from the training data.

[0130] Based on the real-time measurement results of each node in the collection area, they are divided into feature quantities and abnormal types for analysis. First, a part of the data is randomly selected as the validation set, and the rest is used as the training set. The decision tree generation algorithm C4.5 is used to learn a decision tree from the training data. The feature quantities are shown in Table 2, including the frequency temperature correction coefficient, the normalized frequency difference after node temperature correction, etc.; the abnormal types are shown in Table 3, including single-node time-frequency synchronization data packet loss failure, regional synchronization network delay instability failure, etc.

[0131] V) Use precision and recall to evaluate the performance of the decision tree on the validation set. If the performance does not reach the threshold, prune the decision tree, define a hyperparameter probability distribution for pruning, and initialize the Markov chain Monte Carlo chain.

[0132] The performance of the decision tree on the validation set is evaluated using precision and recall. If the performance does not reach the set threshold, the decision tree needs to be pruned and optimized. During pruning, a hyperparameter probability distribution is defined and the Markov chain Monte Carlo chain is initialized to improve the classification effect of the decision tree. During pruning, the feature quantities in Table 2 are mainly considered, such as the mean of the normalized frequency difference of the nodes and the standard deviation of the node time difference, to ensure that the pruning strategy is optimized for the actual application scenario.

[0133] VI) Use this Markov chain Monte Carlo chain to sample the pruning strategies and calculate the theoretical performance of each strategy on the validation set. Use the evaluation results to update the hyperparameter probability distribution using the Bayesian formula.

[0134] The Markov Chain Monte Carlo method is used to sample pruning strategies and calculate the theoretical performance of each strategy on the validation set. Based on the evaluation results, the Bayesian formula is used to update the probability distribution of hyperparameters, so that the pruning strategy is gradually optimized to improve the robustness and generalization ability of the model. This process realizes the dynamic adjustment of the pruning strategy through a comprehensive analysis of the feature quantity, ensuring the adaptability of the model in a complex time-frequency network environment.

[0135] VII) Use the pruned decision tree with updated hyperparameter values ​​to train the decision tree on the training set, evaluate the performance on the validation set, and repeat the above process VI) until the performance reaches a threshold or the number of iterations exceeds the set number.

[0136] The pruned decision tree is retrained using the updated hyperparameters, and the decision tree is trained on the training set and its performance is evaluated on the validation set. The above process is repeated until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit. This cyclic process ensures the accuracy and stability of the decision tree in the diagnosis of time-frequency node anomalies, and improves the accuracy and reliability of anomaly detection through the optimized pruning strategy.

[0137] Therefore, the cause of abnormal communication network in the substation can be quickly diagnosed, and the abnormal diagnosis result can be given by the problem feature value. The diagnostic effect is significant in the pilot substation where this algorithm is deployed. The accuracy of intelligent diagnosis in this solution can improve the detection rate and hit rate of abnormal communication network diagnosis in the substation according to the time of accumulating problem features and the continuous training of the model for new problem data sets, which greatly reduces the time and economic cost of manual investigation and on-site problem location.

[0138] Optionally, determining a carrier-based station area time and frequency synchronization network, and collecting time and frequency node real-time measurement information based on the carrier time and frequency network, includes:

[0139] Determine that the area time and frequency synchronization network includes multiple terminal nodes, namely, a carrier master clock node and a plurality of carrier slave clock nodes;

[0140] Based on the power carrier network, the terminal nodes are topologically related, and then a synchronization system is established in the entire power carrier network. In this topological relationship: T-GM is the main node of the network and is the master clock; T-TC is a relay node, which is both a slave clock and a transparent clock; T-SC is the terminal node and is a slave clock;

[0141] The carrier is carried out through the T-GM master node, and time and frequency synchronization is performed on other nodes in the network. At the same time, real-time measurement information of the time and frequency nodes is collected. The real-time measurement information of the time and frequency nodes includes: the device number, the upper device number, the lower device number, the frequency error, the time error, the network delay, the last time and frequency synchronization time and the ambient temperature.

[0142] Optionally, extracting the characteristic quantity of the real-time measurement information of the time-frequency node to obtain the characteristic value of the time-frequency node includes:

[0143] The time-frequency node characteristic values ​​include: frequency temperature correction coefficient, normalized frequency difference after node temperature correction, node normalized frequency difference mean, node normalized frequency difference standard deviation, node frequency deviation aging rate, node time difference mean, node time difference standard deviation, frequency deviation with upper-level node, time deviation with upper-level node, frequency deviation with lower-level node, time deviation with upper-level node, node network delay mean and node network delay standard deviation;

[0144] According to the accurate frequency value fyi of the master clock, the frequency difference measurement value fci of the slave clock, and the real-time temperature measurement value Ti, (fci / fyi) = k1*Ti+k2*Ti 2 +k3*Ti 3 , perform fitting, and obtain the frequency temperature correction coefficient K(k1, k2, k3) online;

[0145] Based on the frequency difference measurement value fci and the temperature measurement value Ti, the main node frequency measurement value fyi, and the node temperature correction coefficient K, according to fc = [fci-fyi*(1+k1*Ti+k2*Ti2+k3*Ti3)] / fyi, the normalized frequency difference fc after node temperature correction is obtained;

[0146] The normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time are averaged to obtain the node normalized frequency difference mean fm;

[0147] The standard deviation of the normalized frequency difference fc measurement results after multiple node temperature corrections within a period of time is calculated to obtain the node normalized frequency difference standard deviation fs;

[0148] The normalized frequency deviation mean fmi of the node after temperature correction within a period of time is fitted according to fmi=λ*T+c to obtain the node frequency deviation aging rate λ;

[0149] The average of multiple time difference measurement results of the node within a period of time is obtained to obtain the node time difference mean ΔTm;

[0150] The standard deviation of the node time difference measurement results over a period of time is calculated to obtain the node time difference standard deviation ΔTs;

[0151] Subtract the normalized frequency difference of the node from the normalized frequency difference of the upper node in the same period to obtain the frequency deviation fuc between the node and the upper node;

[0152] Subtract the time deviation of the superior node from the node time deviation to get the time deviation ΔTuc between the node and the superior node;

[0153] Subtract the normalized frequency difference of the node from the normalized frequency difference of the subordinate node in the same period to obtain the frequency deviation fnc between the node and the subordinate node;

[0154] Subtract the node time deviation from the subordinate node time deviation to obtain the time deviation ΔTnc between the node and the subordinate node;

[0155] The average network delay measurement value of the node within a period of time is obtained to obtain the node network delay mean value ΔTnm;

[0156] The standard deviation of the node network delay measurement value within a period of time is calculated to obtain the node network delay standard deviation ΔTns.

[0157] Optionally, the abnormal node range is determined by analyzing the characteristic values ​​of the time-frequency nodes, and the abnormal types of the time-frequency nodes are divided by analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster, including:

[0158] By analyzing the characteristic values ​​of the time-frequency nodes, the range of abnormal nodes is determined. By analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the single-node synchronization network delay instability failure of the time-frequency parameters of adjacent nodes in the cluster, the time-frequency node abnormality types are divided into single-node time-frequency synchronization data packet loss failure, single-node synchronization network networking instability failure, single-node time-frequency performance attenuation exceeding threshold failure, single-node time-frequency performance mutation failure, regional time-frequency synchronization data packet loss failure, and regional synchronization network delay instability failure;

[0159] Single-node time-frequency synchronization data packet loss failure means that the node monitoring data can be collected, but the node's last time is not updated. At the same time, the time-frequency characteristics of the subordinate node are consistent with that of the node, while the time-frequency characteristics of the superior node are normal.

[0160] The single-node synchronization network delay instability fault refers to the node network delay standard deviation exceeding the threshold;

[0161] The unstable fault of single-node synchronous network refers to the node being offline;

[0162] The single-node time-frequency performance degradation exceeding threshold fault means that the node frequency deviation aging rate exceeds the threshold, while the time-frequency characteristics of the upper-level node are normal;

[0163] A single-node time-frequency performance sudden change fault refers to a comprehensive assessment of the changes in the node normalized frequency difference mean, node normalized frequency difference standard deviation, node time difference mean, and node time difference standard deviation. The assessment result exceeds the threshold, and the time-frequency characteristics of the upper-level node are normal, and the time-frequency performance of the lower-level node is consistent with that of the node.

[0164] Regional time-frequency synchronization data packet loss failure refers to data packet loss failures at multiple nodes in the station area;

[0165] The regional synchronization network delay instability fault refers to the network delay instability faults in multiple nodes in the station area;

[0166] The unstable fault of regional synchronous network refers to the situation that multiple nodes in the area are offline;

[0167] Batch quality failure of time-frequency equipment refers to the failure of multiple devices in operation to have performance degradation exceeding the threshold or sudden change in time-frequency performance.

[0168] Optionally, the performance of the decision tree in the validation set is evaluated based on precision and recall. If the performance does not reach a set threshold, the decision tree is pruned, and the pruned decision tree is retrained until the performance of the decision tree reaches a set threshold or the number of iterations exceeds a preset upper limit, including:

[0169] The performance of the decision tree in the validation set is evaluated based on precision and recall. If the performance cannot reach the set threshold, a hyperparameter probability distribution is defined to prune the decision tree, initialize the Markov chain Monte Carlo chain, and consider the feature quantity of the real-time measurement information of the time-frequency nodes to ensure that the pruning strategy is optimized for the actual application scenario.

[0170] Sampling the pruning strategies, evaluating the theoretical performance of each pruning strategy on the validation set, and updating the hyperparameter probability distribution based on the evaluation performance, so that the pruning strategy is gradually optimized;

[0171] The pruned decision tree is retrained with the updated hyperparameters, trained on the training set, and its performance is evaluated on the validation set. The above process is repeated until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

[0172] Therefore, the cause of abnormal communication network in the substation can be quickly diagnosed, and the abnormal diagnosis result can be given by the problem feature value. The diagnostic effect is significant in the pilot substation where this algorithm is deployed. The accuracy of intelligent diagnosis in this solution can improve the detection rate and hit rate of abnormal communication network diagnosis in the substation according to the time of accumulating problem features and the continuous training of the model for new problem data sets, which greatly reduces the time and economic cost of manual investigation and on-site problem location.

[0173] According to another aspect of the present invention, a node abnormality diagnosis system 300 in a carrier frequency synchronization network cluster is provided. Figure 3 As shown, the system 300 includes:

[0174] A real-time measurement information determination module 310 is used to determine a carrier-based station area time-frequency synchronization network and collect real-time measurement information of time-frequency nodes based on the carrier time-frequency network;

[0175] The real-time measurement information classification module 320 is used to classify the real-time measurement information of the time-frequency node into characteristic quantities and abnormal types, extract the characteristic quantities of the real-time measurement information of the time-frequency node, obtain the characteristic values ​​of the time-frequency node, determine the range of abnormal nodes by analyzing the characteristic values ​​of the time-frequency node, and classify the abnormal types of the time-frequency nodes by analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster;

[0176] A learning decision tree module 330 is used to randomly select a portion of data from the real-time measurement information of the time-frequency nodes as a validation set and the remaining portion of data as a training set, and learn a decision tree from the training set based on a decision tree generation algorithm C4.5;

[0177] The module 340 for evaluating the performance of the validation set is used to evaluate the performance of the decision tree in the validation set based on the precision and recall rate. If the performance cannot reach the set threshold, the decision tree is pruned and the pruned decision tree is retrained until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

[0178] Optionally, determining a real-time measurement information module includes:

[0179] Determine the time-frequency synchronization network submodule, which is used to determine that the station area time-frequency synchronization network includes multiple terminal nodes, namely, a carrier master clock node and several carrier slave clock nodes;

[0180] The topology relationship establishment submodule is used to establish a topology relationship between each terminal node based on the power carrier network, and then establish a synchronization system in the entire power carrier network. In this topology relationship: T-GM is the network master node and the master clock; T-TC is a relay node, which is both a slave clock and a transparent clock; T-SC is the terminal node and a slave clock;

[0181] The submodule for collecting real-time measurement information of time and frequency nodes is used to carry out carrier through the T-GM master node, synchronize the time and frequency of other nodes in the network, and collect real-time measurement information of time and frequency nodes. The real-time measurement information of time and frequency nodes includes: the device number, the upper device number, the lower device number, the frequency error, the time error, the network delay, the last time and frequency synchronization time and the ambient temperature.

[0182] Optionally, the real-time measurement information module is divided into:

[0183] The time-frequency node characteristic values ​​include: frequency temperature correction coefficient, normalized frequency difference after node temperature correction, node normalized frequency difference mean, node normalized frequency difference standard deviation, node frequency deviation aging rate, node time difference mean, node time difference standard deviation, frequency deviation with upper-level node, time deviation with upper-level node, frequency deviation with lower-level node, time deviation with upper-level node, node network delay mean and node network delay standard deviation;

[0184] The frequency temperature correction coefficient submodule is used to calculate the temperature correction coefficient according to the accurate frequency value fyi of the master clock, the frequency difference measurement value fci of the slave clock, and the real-time temperature measurement value Ti, according to (fci / fyi)=k1*Ti+k2*Ti 2 +k3*Ti 3 , perform fitting, and obtain the frequency temperature correction coefficient K(k1, k2, k3) online;

[0185] The submodule for obtaining the normalized frequency difference after node temperature correction is used to obtain the normalized frequency difference fc after node temperature correction based on the frequency difference measurement value fci and the temperature measurement value Ti, the main node frequency measurement value fyi, and the node temperature correction coefficient K according to fc=[fci-fyi*(1+k1*Ti+k2*Ti2+k3*Ti3)] / fyi;

[0186] A node normalized frequency difference mean submodule is obtained, which is used to average the normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time to obtain the node normalized frequency difference mean fm;

[0187] The node normalized frequency difference standard deviation submodule is used to obtain the standard deviation of the normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time to obtain the node normalized frequency difference standard deviation fs;

[0188] The node frequency deviation aging rate submodule is obtained, which is used to fit the normalized frequency deviation mean fmi of the node after temperature correction within a period of time according to fmi=λ*T+c to obtain the node frequency deviation aging rate λ;

[0189] The node time difference mean submodule is used to average multiple time difference measurement results of the node within a period of time to obtain the node time difference mean ΔTm;

[0190] The node time difference standard submodule is used to obtain the standard deviation of the node time difference measurement results for multiple times within a period of time to obtain the node time difference standard deviation ΔTs;

[0191] The submodule for obtaining the frequency deviation between the node and the upper node is used to obtain the frequency deviation fuc between the node and the upper node by subtracting the normalized frequency difference of the upper node at the same period from the normalized frequency difference of the node;

[0192] The submodule for obtaining the time deviation between the node and the upper node is used to obtain the time deviation ΔTuc between the node and the upper node by subtracting the time deviation of the upper node in the same period from the node time deviation;

[0193] The submodule for obtaining the frequency deviation between the node and the lower level node is used to obtain the frequency deviation fnc between the node and the lower level node by subtracting the normalized frequency difference of the lower level node in the same period from the normalized frequency difference of the node;

[0194] A submodule for obtaining a time deviation between a node and a subordinate node is used to obtain a time deviation ΔTnc between a node and a subordinate node by subtracting the time deviation of the subordinate node in the same period from the node time deviation;

[0195] The node network delay mean submodule is used to average the node network delay measurement values ​​within a period of time to obtain the node network delay mean ΔTnm;

[0196] The node network delay standard deviation submodule is used to calculate the standard deviation of the node network delay measurement value within a period of time to obtain the node network delay standard deviation ΔTns.

[0197] Optionally, the real-time measurement information module is divided into:

[0198] The abnormal type classification submodule is used to determine the range of abnormal nodes by analyzing the characteristic values ​​of the time-frequency nodes, and to classify the abnormal types of time-frequency nodes into single-node time-frequency synchronization data packet loss failure, single-node synchronization network networking instability failure, single-node time-frequency performance attenuation exceeding threshold failure, single-node time-frequency performance mutation failure, regional time-frequency synchronization data packet loss failure, and regional synchronization network delay instability failure by analyzing the evolution law of single-node time-frequency parameters and the evolution law of single-node synchronization network delay instability failure of adjacent nodes in the cluster;

[0199] Single-node time-frequency synchronization data packet loss failure means that the node monitoring data can be collected, but the node's last time is not updated. At the same time, the time-frequency characteristics of the subordinate node are consistent with that of the node, while the time-frequency characteristics of the superior node are normal.

[0200] The single-node synchronization network delay instability fault refers to the node network delay standard deviation exceeding the threshold;

[0201] The unstable fault of single-node synchronous network refers to the node being offline;

[0202] The single-node time-frequency performance degradation exceeding threshold fault means that the node frequency deviation aging rate exceeds the threshold, while the time-frequency characteristics of the upper-level node are normal;

[0203] A single-node time-frequency performance sudden change fault refers to a comprehensive assessment of the changes in the node normalized frequency difference mean, node normalized frequency difference standard deviation, node time difference mean, and node time difference standard deviation. The assessment result exceeds the threshold, and the time-frequency characteristics of the upper-level node are normal, and the time-frequency performance of the lower-level node is consistent with that of the node.

[0204] Regional time-frequency synchronization data packet loss failure refers to data packet loss failures at multiple nodes in the station area;

[0205] The regional synchronization network delay instability fault refers to the network delay instability faults in multiple nodes in the station area;

[0206] The unstable fault of regional synchronous network refers to the situation that multiple nodes in the area are offline;

[0207] Batch quality failure of time-frequency equipment refers to the failure of multiple devices in operation to have performance degradation exceeding the threshold or sudden change in time-frequency performance.

[0208] Optionally, evaluate the validation set performance module, including:

[0209] The decision tree pruning submodule is used to evaluate the performance of the decision tree in the validation set based on precision and recall. If the performance cannot reach the set threshold, the hyperparameter probability distribution is defined to prune the decision tree, initialize the Markov chain Monte Carlo chain, and consider the characteristic quantity of the real-time measurement information of the time-frequency node to ensure that the pruning strategy is optimized for the actual application scenario;

[0210] A submodule for updating the probability distribution of hyperparameters is used to sample the pruning strategies, evaluate the theoretical performance of each pruning strategy on the validation set, and update the probability distribution of hyperparameters based on the evaluation performance, so that the pruning strategies are gradually optimized;

[0211] The retraining decision tree submodule is used to retrain the pruned decision tree with the updated hyperparameters, train the decision tree on the training set, evaluate its performance on the validation set, and repeat the above process until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

[0212] A node abnormality diagnosis system 300 in a carrier-based time-frequency synchronization network cluster according to an embodiment of the present invention corresponds to a node abnormality diagnosis method 100 in a carrier-based time-frequency synchronization network cluster according to another embodiment of the present invention, and will not be described in detail here.

[0213] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0214] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0215] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0216] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0217] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0218] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A method for diagnosing node anomalies in a carrier-based time-frequency synchronization network cluster, characterized in that: include: Determine the carrier-based time and frequency synchronization network for the substation area, and collect real-time measurement information of the time and frequency nodes based on the carrier-based time and frequency network; The real-time measurement information of the time-frequency node is divided into a characteristic quantity and an abnormal type, the characteristic quantity of the real-time measurement information of the time-frequency node is extracted, and the characteristic value of the time-frequency node is obtained, and the range of the abnormal node is determined by analyzing the characteristic value of the time-frequency node, and the abnormal type of the time-frequency node is divided by analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster; Randomly select a portion of data from the real-time measurement information of the time-frequency nodes as a validation set, and the remaining portion of data as a training set, and learn a decision tree from the training set based on the decision tree generation algorithm C4.5; The performance of the decision tree in the validation set is evaluated based on precision and recall. If the performance cannot reach the set threshold, the decision tree is pruned and retrained until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

2. The method according to claim 1, characterized in that: Determine the carrier-based time and frequency synchronization network for the substation area, and collect real-time measurement information of time and frequency nodes based on the carrier time and frequency network, including: Determine that the area time and frequency synchronization network includes multiple terminal nodes, namely, a carrier master clock node and a plurality of carrier slave clock nodes; Based on the power carrier network, the terminal nodes are topologically related, and then a synchronization system is established in the entire power carrier network. In this topological relationship: T-GM is the main node of the network and is the master clock; T-TC is a relay node, which is both a slave clock and a transparent clock; T-SC is the terminal node and is a slave clock; The carrier is carried out through the T-GM master node, and time and frequency synchronization is performed on other nodes in the network. At the same time, real-time measurement information of the time and frequency nodes is collected. The real-time measurement information of the time and frequency nodes includes: the device number, the upper device number, the lower device number, the frequency error, the time error, the network delay, the last time and frequency synchronization time and the ambient temperature.

3. The method according to claim 2, characterized in that Extracting the characteristic quantity of the real-time measurement information of the time-frequency node to obtain the characteristic value of the time-frequency node includes: The time-frequency node characteristic values ​​include: frequency temperature correction coefficient, normalized frequency difference after node temperature correction, node normalized frequency difference mean, node normalized frequency difference standard deviation, node frequency deviation aging rate, node time difference mean, node time difference standard deviation, frequency deviation with upper node, time deviation with upper node, frequency deviation with lower node, time deviation with upper node, node network delay mean and node network delay standard deviation; According to the accurate frequency value fyi of the master clock, the frequency difference measurement value fci of the slave clock, and the real-time temperature measurement value Ti, (fci / fyi) = k1*Ti+k2*Ti 2 +k3*Ti 3 , perform fitting, and obtain the frequency temperature correction coefficient K(k1, k2, k3) online; Based on the frequency difference measurement value fci and the temperature measurement value Ti, the main node frequency measurement value fyi, and the node temperature correction coefficient K, according to fc = [fci-fyi*(1+k1*Ti+k2*Ti2+k3*Ti3)] / fyi, the normalized frequency difference fc after node temperature correction is obtained; The normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time are averaged to obtain the node normalized frequency difference mean fm; The standard deviation of the normalized frequency difference fc measurement results after multiple node temperature corrections within a period of time is calculated to obtain the node normalized frequency difference standard deviation fs; The normalized frequency deviation mean fmi of the node after temperature correction within a period of time is fitted according to fmi=λ*T+c to obtain the node frequency deviation aging rate λ; The average of multiple time difference measurement results of the node within a period of time is obtained to obtain the node time difference mean ΔTm; The standard deviation of the node time difference measurement results over a period of time is calculated to obtain the node time difference standard deviation ΔTs; Subtract the normalized frequency difference of the node from the normalized frequency difference of the upper node in the same period to obtain the frequency deviation fuc between the node and the upper node; Subtract the time deviation of the superior node from the node time deviation to get the time deviation ΔTuc between the node and the superior node; Subtract the normalized frequency difference of the node from the normalized frequency difference of the subordinate node in the same period to obtain the frequency deviation fnc between the node and the subordinate node; Subtract the node time deviation from the subordinate node time deviation to obtain the time deviation ΔTnc between the node and the subordinate node; The average network delay measurement value of the node within a period of time is obtained to obtain the node network delay mean value ΔTnm; The standard deviation of the node network delay measurement value within a period of time is calculated to obtain the node network delay standard deviation ΔTns.

4. The method according to claim 1, characterized in that By analyzing the characteristic values ​​of the time-frequency nodes, the range of abnormal nodes is determined. By analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster, the abnormal types of the time-frequency nodes are divided, including: By analyzing the characteristic values ​​of the time-frequency nodes, the range of abnormal nodes is determined. By analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the single-node synchronization network delay instability failure of the time-frequency parameters of adjacent nodes in the cluster, the time-frequency node abnormality types are divided into single-node time-frequency synchronization data packet loss failure, single-node synchronization network networking instability failure, single-node time-frequency performance attenuation exceeding threshold failure, single-node time-frequency performance mutation failure, regional time-frequency synchronization data packet loss failure, and regional synchronization network delay instability failure; Single-node time-frequency synchronization data packet loss failure means that the node monitoring data can be collected, but the node's last time is not updated. At the same time, the time-frequency characteristics of the subordinate node are consistent with that of the node, while the time-frequency characteristics of the superior node are normal. The single-node synchronization network delay instability fault refers to the node network delay standard deviation exceeding the threshold; The unstable fault of single-node synchronous network refers to the node being offline; The single-node time-frequency performance degradation exceeding threshold fault means that the node frequency deviation aging rate exceeds the threshold, while the time-frequency characteristics of the upper-level node are normal; A single-node time-frequency performance sudden change fault refers to a comprehensive assessment of the changes in the node normalized frequency difference mean, node normalized frequency difference standard deviation, node time difference mean, and node time difference standard deviation. The assessment result exceeds the threshold, and the time-frequency characteristics of the upper-level node are normal, and the time-frequency performance of the lower-level node is consistent with that of the node. Regional time-frequency synchronization data packet loss failure refers to data packet loss failures at multiple nodes in the station area; The regional synchronization network delay instability fault refers to the network delay instability faults in multiple nodes in the station area; The unstable fault of regional synchronous network refers to the situation that multiple nodes in the area are offline; Batch quality failure of time-frequency equipment refers to the failure of multiple devices in operation to have performance degradation exceeding the threshold or sudden change in time-frequency performance.

5. The method according to claim 1, characterized in that The performance of the decision tree in the validation set is evaluated based on precision and recall. If the performance does not reach the set threshold, the decision tree is pruned and retrained until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit, including: The performance of the decision tree in the validation set is evaluated based on precision and recall. If the performance cannot reach the set threshold, a hyperparameter probability distribution is defined to prune the decision tree, initialize the Markov chain Monte Carlo chain, and consider the feature quantity of the real-time measurement information of the time-frequency nodes to ensure that the pruning strategy is optimized for the actual application scenario. Sampling the pruning strategies, evaluating the theoretical performance of each pruning strategy on the validation set, and updating the hyperparameter probability distribution based on the evaluation performance, so that the pruning strategy is gradually optimized; The pruned decision tree is retrained with the updated hyperparameters, trained on the training set, and its performance is evaluated on the validation set. The above process is repeated until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

6. A node abnormality diagnosis system in a carrier frequency synchronization network cluster, characterized in that: include: Determine the real-time measurement information module, which is used to determine the carrier-based station area time and frequency synchronization network and collect the real-time measurement information of the time and frequency nodes based on the carrier time and frequency network; A real-time measurement information classification module is used to classify the real-time measurement information of the time-frequency node into characteristic quantities and abnormal types, extract the characteristic quantities of the real-time measurement information of the time-frequency node, obtain the characteristic values ​​of the time-frequency node, determine the range of abnormal nodes by analyzing the characteristic values ​​of the time-frequency node, and classify the abnormal types of the time-frequency node by analyzing the evolution law of the time-frequency parameters of a single node and the evolution law of the time-frequency parameters of adjacent nodes in the cluster; A learning decision tree module is used to randomly select a part of the data from the real-time measurement information of the time-frequency node as a validation set and the rest of the data as a training set, and learn a decision tree from the training set based on the decision tree generation algorithm C4.5; The validation set performance evaluation module is used to evaluate the performance of the decision tree in the validation set based on precision and recall. If the performance cannot reach the set threshold, the decision tree is pruned and retrained until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

7. The system according to claim 6, characterized in that Determine the real-time measurement information module, including: Determine the time-frequency synchronization network submodule, which is used to determine that the station area time-frequency synchronization network includes multiple terminal nodes, namely, a carrier master clock node and several carrier slave clock nodes; The topology relationship establishment submodule is used to establish a topology relationship between each terminal node based on the power carrier network, and then establish a synchronization system in the entire power carrier network. In this topology relationship: T-GM is the network master node and the master clock; T-TC is a relay node, which is both a slave clock and a transparent clock; T-SC is the terminal node and a slave clock; The submodule for collecting real-time measurement information of time and frequency nodes is used to carry out carrier through the T-GM master node, synchronize the time and frequency of other nodes in the network, and collect real-time measurement information of time and frequency nodes. The real-time measurement information of time and frequency nodes includes: the device number, the upper device number, the lower device number, the frequency error, the time error, the network delay, the last time and frequency synchronization time and the ambient temperature.

8. The system according to claim 7, characterized in that Divide the real-time measurement information module, including: The time-frequency node characteristic values ​​include: frequency temperature correction coefficient, normalized frequency difference after node temperature correction, node normalized frequency difference mean, node normalized frequency difference standard deviation, node frequency deviation aging rate, node time difference mean, node time difference standard deviation, frequency deviation with upper node, time deviation with upper node, frequency deviation with lower node, time deviation with upper node, node network delay mean and node network delay standard deviation; The frequency temperature correction coefficient submodule is used to calculate the temperature correction coefficient according to the accurate frequency value fyi of the master clock, the frequency difference measurement value fci of the slave clock, and the real-time temperature measurement value Ti, according to (fci / fyi)=k1*Ti+k2*Ti 2 +k3*Ti 3 , perform fitting, and obtain the frequency temperature correction coefficient K(k1, k2, k3) online; The submodule for obtaining the normalized frequency difference after node temperature correction is used to obtain the normalized frequency difference fc after node temperature correction based on the frequency difference measurement value fci and the temperature measurement value Ti, the main node frequency measurement value fyi, and the node temperature correction coefficient K according to fc=[fci-fyi*(1+k1*Ti+k2*Ti2+k3*Ti3)] / fyi; A node normalized frequency difference mean submodule is obtained, which is used to average the normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time to obtain the node normalized frequency difference mean fm; The node normalized frequency difference standard deviation submodule is used to obtain the standard deviation of the normalized frequency difference fc measurement results of the node after multiple node temperature corrections within a period of time to obtain the node normalized frequency difference standard deviation fs; The node frequency deviation aging rate submodule is obtained, which is used to fit the normalized frequency deviation mean fmi of the node after temperature correction within a period of time according to fmi=λ*T+c to obtain the node frequency deviation aging rate λ; The node time difference mean submodule is used to average multiple time difference measurement results of the node within a period of time to obtain the node time difference mean ΔTm; The node time difference standard submodule is used to obtain the standard deviation of the node time difference measurement results for multiple times within a period of time to obtain the node time difference standard deviation ΔTs; The submodule for obtaining the frequency deviation between the node and the upper node is used to obtain the frequency deviation fuc between the node and the upper node by subtracting the normalized frequency difference of the upper node at the same period from the normalized frequency difference of the node; The submodule for obtaining the time deviation between the node and the upper node is used to obtain the time deviation ΔTuc between the node and the upper node by subtracting the time deviation of the upper node in the same period from the node time deviation; The submodule for obtaining the frequency deviation between the node and the lower level node is used to obtain the frequency deviation fnc between the node and the lower level node by subtracting the normalized frequency difference of the lower level node in the same period from the normalized frequency difference of the node; A submodule for obtaining a time deviation between a node and a subordinate node is used to obtain a time deviation ΔTnc between a node and a subordinate node by subtracting the time deviation of the subordinate node in the same period from the node time deviation; The node network delay mean submodule is used to average the node network delay measurement values ​​within a period of time to obtain the node network delay mean ΔTnm; The node network delay standard deviation submodule is used to calculate the standard deviation of the node network delay measurement value within a period of time to obtain the node network delay standard deviation ΔTns.

9. The system according to claim 6, characterized in that Divide the real-time measurement information module, including: The abnormal type classification submodule is used to determine the range of abnormal nodes by analyzing the characteristic values ​​of the time-frequency nodes, and to classify the abnormal types of time-frequency nodes into single-node time-frequency synchronization data packet loss failure, single-node synchronization network networking instability failure, single-node time-frequency performance attenuation exceeding threshold failure, single-node time-frequency performance mutation failure, regional time-frequency synchronization data packet loss failure, and regional synchronization network delay instability failure by analyzing the evolution law of single-node time-frequency parameters and the evolution law of single-node synchronization network delay instability failure of adjacent nodes in the cluster; Single-node time-frequency synchronization data packet loss failure means that the node monitoring data can be collected, but the node's last time is not updated. At the same time, the time-frequency characteristics of the subordinate node are consistent with that of the node, while the time-frequency characteristics of the superior node are normal. The single-node synchronization network delay instability fault refers to the node network delay standard deviation exceeding the threshold; The unstable fault of single-node synchronous network refers to the node being offline; The single-node time-frequency performance degradation exceeding threshold fault means that the node frequency deviation aging rate exceeds the threshold, while the time-frequency characteristics of the upper-level node are normal; A single-node time-frequency performance sudden change fault refers to a comprehensive assessment of the changes in the node normalized frequency difference mean, node normalized frequency difference standard deviation, node time difference mean, and node time difference standard deviation. The assessment result exceeds the threshold, and the time-frequency characteristics of the upper-level node are normal, and the time-frequency performance of the lower-level node is consistent with that of the node. Regional time-frequency synchronization data packet loss failure refers to data packet loss failures at multiple nodes in the station area; The regional synchronization network delay instability fault refers to the network delay instability faults in multiple nodes in the station area; The unstable fault of regional synchronous network refers to the situation that multiple nodes in the area are offline; Batch quality failure of time-frequency equipment refers to the failure of multiple devices in operation to have performance degradation exceeding the threshold or sudden change in time-frequency performance.

10. The system according to claim 6, characterized in that Evaluate the validation set performance module, including: The decision tree pruning submodule is used to evaluate the performance of the decision tree in the validation set based on precision and recall. If the performance cannot reach the set threshold, the hyperparameter probability distribution is defined to prune the decision tree, initialize the Markov chain Monte Carlo chain, and consider the characteristic quantity of the real-time measurement information of the time-frequency node to ensure that the pruning strategy is optimized for the actual application scenario; A submodule for updating the probability distribution of hyperparameters is used to sample the pruning strategies, evaluate the theoretical performance of each pruning strategy on the validation set, and update the probability distribution of hyperparameters based on the evaluation performance, so that the pruning strategies are gradually optimized; The retraining decision tree submodule is used to retrain the pruned decision tree with the updated hyperparameters, train the decision tree on the training set, evaluate its performance on the validation set, and repeat the above process until the performance of the decision tree reaches the set threshold or the number of iterations exceeds the preset upper limit.

Images