Remote fault diagnosis method and system based on FTTR network, and storage medium

Abstract

The invention relates to the technical field of network communication, in particular to a remote fault diagnosis method and system based on an FTTR network and a storage medium. The method comprises the following steps that a monitoring module used for monitoring the network state is deployed on a node of each FTTR network, each monitoring module is provided with a data acquisition unit and a state analysis unit, and each node collects network state data in real time through the monitoring module. The network state data comprises monitoring network delay, packet loss rate, equipment temperature and bandwidth utilization rate, and the network state data is preliminarily analyzed through the data acquisition unit and is compared with historical data to determine the change trend of the current state. According to the invention, through a distributed management mode, a single-point fault problem in traditional centralized management is avoided. Each FTTR node can independently monitor the state of itself and work cooperatively with other nodes, even if a certain node breaks down, other nodes can still work normally, and the stability of the whole network is ensured.

Description

Technical Field

[0001] The present invention relates to the field of network communication technology, and in particular to a remote fault diagnosis method, system and storage medium based on FTTR network. Background Art

[0002] Fiber to the Room (FTTR) network is a high-speed fiber network that directly routes fiber to every room in the user's home, providing high-bandwidth, low-latency network connections. With the development of smart homes and the Internet of Things, the demand for FTTR networks is increasing, and the requirements for their stability are also increasing. Remote monitoring and fault diagnosis systems are key technologies to ensure the stable operation of FTTR networks. Remote monitoring systems are systems that monitor and manage equipment in real time through the network, which can help administrators discover and solve problems in a timely manner and improve system reliability. Fault diagnosis technology analyzes system failures and finds the root cause of the failure, thereby quickly and effectively solving the problem.

[0003] At present, most remote monitoring and fault diagnosis systems of FTTR networks adopt a centralized management mode, that is, the status information of all network devices (such as optical modems, routers, etc.) is collected to the central node, which collects data and performs fault analysis. Usually, the central node will regularly collect the operating data of each device and judge the health of the system by analyzing the data. If an abnormality is detected in the equipment, the central node will notify the administrator to repair it, or trigger the preset repair strategy through automated tools. Although the existing solutions can ensure the stable operation of the FTTR network to a certain extent, there are still some problems:

[0004] First, the centralized management model is prone to single point failure problems. If the central node fails, the management and monitoring of the entire network will be affected. Second, the existing system can usually only monitor the status of the network passively and cannot actively discover potential fault hazards. In addition, the existing fault diagnosis technology often relies on experienced administrators and lacks intelligence, and cannot quickly and accurately diagnose the root cause of the fault. Summary of the invention

[0005] The present invention provides a remote fault diagnosis method, system and storage medium based on FTTR network to solve the single point failure risk and passive monitoring problems in the existing centralized management mode, and improve the fault diagnosis and processing efficiency through intelligent means, while ensuring the scalability of the system in a large-scale network environment.

[0006] The technical solution of the present invention to solve the above technical problems is as follows:

[0007] On the one hand, a remote fault diagnosis method based on an FTTR network is provided, the method comprising the following steps:

[0008] A monitoring module for monitoring the network status is deployed on each node of the FTTR network. Each monitoring module is equipped with a data acquisition unit and a status analysis unit. Each node collects network status data in real time through the monitoring module.

[0009] Network status data includes monitoring network delay, packet loss rate, device temperature and bandwidth utilization. The data acquisition unit preliminarily analyzes the network status data and compares it with historical data to determine the change trend of the current status.

[0010] Each node publishes the status information of the current state to a shared topic through the message queue system, synchronizes with other nodes, and is used to form a global status view;

[0011] Analyze the synchronized global state view using unsupervised learning algorithms to identify abnormal patterns for detecting extreme fluctuations in traffic and persistent high latency behavior;

[0012] Use supervised learning models to analyze historical fault data, extract features, and analyze abnormal patterns based on the extracted features to predict potential faults and issue early warnings;

[0013] When an early warning is triggered, a deep learning model is used to analyze the current fault, infer the cause of the fault, generate and execute repair operations.

[0014] Furthermore, the state analysis unit adopts a dual threshold model for analysis and prediction of change trends. The expression of the dual threshold model is:

[0015] Among them, ΔM(t) is the state change rate, and T1 and T2 are thresholds.

[0016] Furthermore, unsupervised learning algorithms include clustering methods based on dynamic discriminant boundaries, expressed as:

[0017] Where D is the data point x i To cluster center c j The distance is , and p is the adjustment parameter.

[0018] Furthermore, each node publishes the status information of the current state to a shared topic through a message queue system. The steps of synchronization with other nodes include:

[0019] Each node publishes JSON format status information to the global shared topic through the message queue system every 30 seconds, and after receiving it, it integrates other node information to generate a global status table.

[0020] Furthermore, the supervised learning model is based on an ensemble learning method, including the use of multi-model collaboration, which is expressed as:

[0021] f(x)=Σ k w k ·h k (x), where f(x) is the total fault prediction, h k (x) is the output of each sub-model, w k is the weight corresponding to the sub-model.

[0022] Furthermore, the fault prediction also adopts the time series analysis method, using the autoregressive integrated moving average model, the expression of the autoregressive integrated moving average model is:

[0023] Among them, Y(t) represents the observed value of the time series at time t, c represents the mean of the time series, φ i (i=1,2,…,p) is the autoregressive coefficient, which indicates the influence of the Y(t) value at the past i time points on the current Y(t). Y(ti) indicates the observed value of the time series at time ti. θ j (j=1,2,…,q) represents the impact of the error term ε at the past j time points on the current Y(t), ε(tj) represents the random error term of the time series at time tj, and ε(t) represents the random error term of the time series at time t.

[0024] Furthermore, a deep learning model is used to analyze the current fault based on the deep learning pattern matching process of the long short-term memory network (LSTM), and the parameters are optimized through error back propagation to quickly respond to faults similar to those in the past;

[0025] Execute repair operations to generate the optimal repair path based on the fault mode library, and automatically adjust parameters or trigger maintenance prompts in conjunction with the device API.

[0026] Furthermore, after the fault is repaired, the actual effect data is recorded and fed back to the deep learning model to adjust and optimize the parameter library to improve the decision-making efficiency of subsequent operations.

[0027] On the other hand, a remote fault diagnosis system based on an FTTR network is provided, the system comprising:

[0028] Data collection module, deployed at each FTTR node, for real-time collection of network status information;

[0029] The status analysis module obtains the network status data provided by the data acquisition module, and performs preliminary analysis and distributed synchronization on the network status data;

[0030] Anomaly detection module, which uses unsupervised learning algorithms to detect abnormal patterns and identify potential network problems;

[0031] The fault prediction module uses supervised learning algorithms to predict possible network risks and provide early warning signals;

[0032] Fault diagnosis module, which analyzes fault causes based on deep learning algorithms and performs fault pattern matching;

[0033] The repair suggestion and execution module generates the optimal repair suggestion and performs automatic repair or triggers maintenance prompts through the system interface.

[0034] On the other hand, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed on an electronic device, it executes the above-mentioned remote fault diagnosis method based on the FTTR network.

[0035] The beneficial effects of the present invention are:

[0036] The present invention avoids the single point failure problem in traditional centralized management through a distributed management mode. Each FTTR node can independently monitor its own status and work with other nodes. Even if a node fails, other nodes can still work normally to ensure the stability of the entire network. The present invention introduces unsupervised learning and supervised learning models to achieve real-time analysis and prediction of network status, and can actively identify potential network anomalies and hidden faults. Before a fault occurs, the system can send out an early warning signal and take measures in advance, reducing the impact of sudden faults on users and improving the network's fault prevention capabilities. In addition, through a deep learning model, the system can quickly and accurately locate the root cause of the fault and provide intelligent repair suggestions based on historical data. For common faults, the system can automatically repair, reducing the need for manual intervention and improving the speed of fault handling. For complex faults, the system can provide an optimized repair solution, greatly shortening the troubleshooting and repair time, and ensuring the rapid recovery of the network. Using a distributed computing framework and a distributed storage architecture, the system can efficiently process massive monitoring data from multiple FTTR nodes, and maintain rapid response and analysis capabilities even in a high-concurrency environment. Through the storage mechanism of the distributed database, the system ensures high data availability, and can quickly retrieve and analyze historical data, improving the efficiency of data processing and query. The distributed architecture and efficient data processing methods make the system have good scalability and can adapt to the rapid growth of smart home and IoT devices in the future. The system can easily expand the node scale without affecting the overall performance, thereby ensuring efficient operation in a large-scale network environment. Since the system can handle faults through intelligent diagnosis and automatic repair mechanisms, it reduces dependence on manual intervention, reduces operation and maintenance costs and manpower requirements. In addition, the system has a high degree of automation, which can effectively reduce the workload of network administrators. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of the remote fault diagnosis method based on FTTR network of the present invention;

[0038] Figure 2 is a node operation flow chart of the present invention;

[0039] Figure 3 is a distributed state synchronization diagram of the present invention;

[0040] Figure 4 is the AI-driven active monitoring and early warning flow chart of the present invention;

[0041] Figure 5 is a flow chart of the intelligent fault diagnosis and repair system of the present invention;

[0042] Figure 6 is a flow chart of efficient large-scale data processing of the present invention;

[0043] Figure 7 It is the overall workflow diagram of the present invention. DETAILED DESCRIPTION

[0044] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments.

[0045] In the embodiments of the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0046] like Figures 1 to 7 As shown, the present invention provides the following preferred embodiments:

[0047] Embodiment 1

[0048] In order to solve the problem of accuracy and efficiency of remote fault diagnosis in FTTR network, this embodiment proposes a remote fault diagnosis method based on FTTR network. Specifically, the method includes the following steps:

[0049] S100. A monitoring module for monitoring the network status is deployed on each node of the FTTR network. Each monitoring module is provided with a data acquisition unit and a status analysis unit. Each node collects network status data in real time through the monitoring module.

[0050] S200, network status data includes monitoring network delay, packet loss rate, device temperature and bandwidth utilization, preliminarily analyzing the network status data through the data acquisition unit, and comparing historical data to determine the change trend of the current status.

[0051] S300. Each node publishes the status information of the current state to a shared topic through a message queue system, and synchronizes with other nodes to form a global status view.

[0052] S400,uses unsupervised learning algorithms to analyze the synchronized global state view and,identify abnormal patterns to detect extreme fluctuations in traffic and,persistent high latency behavior.

[0053] S500, use supervised learning model to analyze historical fault data, extract features, and analyze abnormal patterns in combination with extracted features to predict potential faults and issue warnings.

[0054] S600: When the warning is triggered, the deep learning model is used to analyze the current fault, infer the cause of the fault, generate and execute repair operations.

[0055] In this embodiment, a monitoring module is deployed on each FTTR network node. The monitoring module includes a data acquisition unit and a status analysis unit, which are responsible for collecting network status data in real time. It should be understood that the sources of these data include network delay, packet loss rate, device temperature and bandwidth utilization, and this information is preliminarily analyzed by the data acquisition unit. The analysis method involves comparison with historical data to determine the change trend of the current status. It should be noted that the identification of the change trend is used to quickly respond to abnormal conditions in network operation.

[0056] Furthermore, each node publishes the current status information to a shared topic through the deployed message queue system, and synchronizes with other nodes to form a global status view. It can be understood that this design enables the distributed network architecture to share real-time data, thereby having a unified understanding of the operating status of the entire network.

[0057] Furthermore, in the process of identifying abnormal patterns, this embodiment uses an unsupervised learning algorithm to analyze the global state view. It should be understood that the algorithm identifies and monitors extreme fluctuations in traffic and persistent high latency behavior, thereby providing abnormal warnings. Through this abnormal detection mechanism, abnormal network conditions can be identified at an early stage to prevent the spread of problems.

[0058] Furthermore, supervised learning models are used to analyze and extract features from historical fault data, so as to combine these features to conduct a more accurate analysis of abnormal patterns. This process is used to predict potential faults and issue early warnings. It can be understood that the supervised learning model establishes the ability to identify potential problems in the future through the analysis of past faults, which provides a theoretical basis for subsequent diagnosis.

[0059] When an alert is triggered, a deep learning model is used to analyze the current fault, infer the cause of the fault, and generate repair actions. It is important to understand that deep learning models can learn complex associations from current and historical data to accurately identify and handle faults.

[0060] The benefit of this embodiment is that a complete remote fault diagnosis method is formed by integrating monitoring, data analysis, anomaly detection and repair functions. This method effectively improves the operational stability and maintenance efficiency of the FTTR network in a complex environment and provides an efficient and reliable network management solution. Through such a system architecture, operators can quickly identify and respond to network problems, thereby reducing the impact on user services and enhancing user experience.

[0061] Embodiment 2

[0062] In order to solve the problem of accuracy of state change detection and reliability of trend analysis in FTTR network, this embodiment further optimizes the state analysis unit and adopts a dual threshold model to analyze and predict the change trend of network state. The expression of the dual threshold model is:

[0063]

[0064] It can be understood that the model sets two different thresholds T1 and T2 by analyzing the state change rate ΔM(t) to provide more accurate state monitoring. In this dual-threshold model, the state function S(t) is defined as follows: when ΔM(t) is greater than the upper threshold T1, the function output is 1, which indicates a significant state change, suggesting a possible abnormal situation; and when ΔM(t) is less than the lower threshold T2, the function output is 0, indicating that the network state is within the normal range of change. The state changes in the middle area are considered to be uncertain or within the normal fluctuation range, and need to be further analyzed in combination with other technical features and context.

[0065] Furthermore, ΔM(t), as a dynamically changing value, is calculated based on the continuous changes in network delay, packet loss rate, device temperature, and bandwidth utilization provided by the data acquisition unit. It should be understood that the selection of appropriate T1 and T2 values ​​is closely related to the node's environment, historical data, and the specific needs of the network operator. Therefore, T1 and T2 are not fixed, but are gradually adjusted based on historical operating data and expected sensitivity to adapt to changing network conditions.

[0066] In addition, the dual threshold model of this embodiment can effectively filter out false alarms caused by short-term fluctuations in terms of mechanism, which can be understood as avoiding false triggering caused by instantaneous changes that are not abnormal. This improves the accuracy of network management and fault diagnosis. In addition, the elastic design of this threshold setting can improve the adaptability of anomaly detection in different scenarios in real-time monitoring, providing a strong guarantee for the complexity and diversity of the network.

[0067] By sensitively tracking the above-mentioned changing trends, the model provides a simple and effective monitoring method for the network. This is not limited to the collection and simple analysis of existing data, but also provides accurate predictions of possible future state changes, thereby enhancing the ability of fault warning and prevention. In specific applications, it can be combined with information from other monitoring and analysis modules in the network to form a more comprehensive network operation status assessment.

[0068] The benefit of this embodiment is that by adopting such a model with dual thresholds, the state changes in the FTTR network can be analyzed more flexibly and accurately. Especially in complex and changeable network environments, the dual threshold model provides a balanced solution: while ensuring sensitivity, it can effectively reduce the false alarm rate. This capability plays an important role in ensuring the stability and reliable operation of the FTTR network. Through this embodiment, operators can accurately identify trends and potential failure points over a longer time span, and thus exert the efficiency of operation and maintenance in maintenance and response strategies.

[0069] Embodiment 3

[0070] In order to solve the accuracy problem of anomaly detection and fault identification in FTTR networks, this embodiment introduces a clustering method based on dynamic judgment boundaries in the unsupervised learning algorithm, and its expression is:

[0071]

[0072] The expression of this embodiment realizes the data point x i To cluster center c j The dynamic distance calculation of p is a parameter. It should be understood that the clustering method of dynamic judgment boundary can automatically adjust the clustering boundary according to the dynamic changes of input data to adapt to the newly emerged abnormal situation.

[0073] Furthermore, the unsupervised learning algorithm involves a mathematical dynamic distance formula. By adjusting p, the compactness and sensitivity of clustering can be flexibly changed under different network environment conditions, thereby improving the ability to identify abnormal samples. It can still maintain a high adaptability when facing different amounts and types of network data.

[0074] In practical applications, unsupervised learning algorithms monitor changes in network status data and calculate the distance between each data point and multiple possible cluster centers at the current moment. The distance D(x i ,C) is not only a distance in the geometric sense, but also a measure of the significant difference between abnormal samples and normal samples. This enables the system to classify abnormal data into a new or existing abnormal category in a timely manner after detecting it, thereby improving the efficiency of fault identification.

[0075] In addition, the flexibility of the clustering method of this embodiment in terms of boundary adjustment is also reflected in its adaptive learning ability for new data sets. Specifically, when a new set of abnormal data points appears in the network, it can automatically identify that these points do not belong to any existing normal clusters, and calculate these points according to the self-adjustment mechanism to form a new abnormal cluster. This dynamic adaptability enables the system to automatically learn without human intervention when faced with a completely new abnormal pattern, thereby improving the intelligence level of the entire remote fault diagnosis method.

[0076] It is important to note that when introducing this clustering algorithm in the FTTR network, it is necessary to combine it with historical data and trend analysis. This not only helps to identify new abnormal patterns, but also improves future prediction models with the help of the identified abnormal data features. In this way, in-depth support for abnormal event identification can be provided in a wider range of application contexts.

[0077] Through the implementation of this embodiment, the operation and maintenance personnel of the FTTR network can achieve accurate anomaly detection and fault prediction by relying only on automated algorithms, which usually requires a lot of manual analysis. This improves the efficiency of daily network management while also reducing the response time to faults. In addition, the use of dynamic judgment boundaries effectively enhances the recognition accuracy of fault modes under variable network conditions, allowing the system to continuously adapt to changing network anomalies, thereby maintaining the stability of the FTTR network.

[0078] Embodiment 4

[0079] In order to solve the real-time and consistency issues of the state information synchronization of each node in the FTTR network, this embodiment further refines the specific steps for each node to publish the current state information to the shared topic and synchronize it through the message queue system. In this process, each node publishes the state information in JSON format to the global shared topic through the embedded message queue system at a fixed period, that is, every 30 seconds. This design ensures that the state information of all nodes in the network can be shared and synchronized in a timely and orderly manner, thereby providing a reliable data basis for generating the global state table.

[0080] Specifically, the JSON format was selected as the transmission format for status information because of its lightweight and easy-to-parse characteristics. The status information of each node includes but is not limited to key indicators such as network latency, packet loss rate, device temperature, and bandwidth utilization. This information is encapsulated in a JSON object and transmitted through a message queue system. It should be understood that the message queue system not only provides a reliable message delivery mechanism, but also ensures the orderliness and non-loss of messages.

[0081] Furthermore, the message queue system is designed to support high concurrency and high throughput to cope with the large number of nodes and massive amounts of data that may exist in the FTTR network. While each node publishes its own status information, it also receives the status information of other nodes from the message queue. This process is bidirectional, ensuring full synchronization of information. After receiving the status information of all relevant nodes, each node will integrate this information according to the preset logic to generate a global status table.

[0082] It is important to understand that the generation of the global state table is not just a simple data aggregation, it also involves some data processing and cleaning steps. For example, for duplicate or erroneous data, the node will filter and correct it according to the built-in logic to ensure the accuracy and consistency of the global state table. In addition, in order to improve the system's response speed and reduce resource consumption, the node will use incremental updates when integrating information, that is, only update the changed parts instead of regenerating the entire state table.

[0083] Furthermore, considering the complexity and variability of the FTTR network, the implementation of the message queue system also includes fault recovery and message retry mechanisms. If a node fails to successfully publish or receive status information within the specified time window, the system will automatically trigger the retry mechanism to ensure the integrity of the information. It should be understood that this mechanism is conducive to maintaining the real-time and consistency of the entire network status, especially in the case of network fluctuations or temporary failures.

[0084] In addition, this embodiment also optimizes the load balancing strategy of the message queue system to prevent performance bottlenecks during network peaks or large-scale data transmission. By dynamically adjusting the load distribution of the message queue, the system can smoothly cope with various network conditions and ensure the efficiency and stability of state information synchronization.

[0085] The benefit of this embodiment is that by using a fixed period and JSON format to publish and synchronize status information, combined with the high reliability and high concurrency characteristics of the message queue system, each node in the FTTR network can share and update the global status table in real time and accurately. This mechanism not only improves the accuracy and response speed of network status monitoring, but also enhances the fault tolerance and stability of the entire network. Through this embodiment, the FTTR network can maintain efficient operation in a complex and changing environment, providing strong support for network operation and management.

[0086] Embodiment 5

[0087] In order to solve the problem of accuracy and stability of fault diagnosis in FTTR networks, this embodiment further optimizes the construction method of the supervised learning model and adopts a multi-model collaboration method based on ensemble learning. This method integrates the outputs of multiple sub-models and assigns corresponding weights according to the performance of each sub-model to achieve more accurate prediction of faults. Specifically, the multi-model collaboration in this embodiment is achieved through the following expression:

[0088] f(x)=∑ k w k ·h k (x)

[0089] Among them, f(x) represents the final total fault prediction result, h k (x) is the output of the kth sub-model, w k is the weight corresponding to the sub-model. Through this mechanism, this embodiment can provide a more adaptable and stable fault diagnosis capability when facing a complex and changeable network environment.

[0090] Furthermore, the sub-model h k The selection of (x) in this embodiment includes a variety of machine learning algorithms, such as decision trees, support vector machines, neural networks, etc. These algorithms each have different advantages and applicable scenarios. By integrating these diverse sub-models, the system can analyze and predict the network status from multiple angles. It should be understood that different sub-models may show different accuracy when dealing with different types of faults. For example, a decision tree model may perform well when dealing with discrete features, while a neural network has advantages when dealing with complex nonlinear relationships. Therefore, multi-model collaboration can provide more balanced and comprehensive diagnostic results in various situations.

[0091] In practical applications, the output h of each sub-model k (x) is first standardized to ensure that the outputs of different models are comparable and consistent. Then, the system dynamically adjusts the corresponding weight w of each sub-model based on its performance on historical data. k . Weight w k The determination of is obtained by evaluating the error of the sub-model on the training set and the validation set. It can be understood that the sub-model with better performance will be given a higher weight, while the sub-model with poorer performance will be given a relatively lower weight. This mechanism ensures that the final integrated prediction f(x) can maximize the advantages of each sub-model, thereby improving the accuracy of the overall prediction.

[0092] Furthermore, in order to ensure the real-time and adaptability of the model, this embodiment also introduces an online learning mechanism. Specifically, when a new fault mode or abnormal data appears in the network, the system will automatically trigger the retraining and weight adjustment of the model. By continuously updating the weights and outputs of the sub-models, the system can maintain a high diagnostic accuracy in a dynamically changing network environment. In addition, in order to prevent overfitting of the model and improve generalization ability, this embodiment also uses cross-validation and regularization techniques during the training process. These techniques can achieve a good balance between model complexity and prediction accuracy, thereby further enhancing the practicality of the system.

[0093] It is important to understand that the collaboration between sub-models is not limited to simple weighted summation. In some cases, the system can further improve the performance of the model through more complex fusion strategies (such as stacking or blending). These advanced fusion methods introduce meta-models to perform secondary learning on the output of sub-models, thereby obtaining better prediction results in certain specific scenarios.

[0094] The benefit of this embodiment is that by adopting a multi-model collaboration method based on ensemble learning, the fault diagnosis capability of the FTTR network has been significantly improved. This method can not only effectively integrate the advantages of multiple sub-models, but also dynamically adjust the weights of each sub-model according to the actual network conditions, thereby providing more accurate and reliable fault prediction. Through this embodiment, the FTTR network can maintain efficient and stable fault diagnosis capabilities when facing a complex and changeable operating environment, providing strong technical support for network operation, maintenance and management.

[0095] Embodiment 6

[0096] In order to solve the time series correlation and accuracy problems of fault prediction in FTTR networks, this embodiment further refines the time series analysis method in fault prediction and introduces the autoregressive integrated moving average model (ARIMA). The introduction of this model aims to improve the prediction ability of network faults by capturing the time series dependency structure in time series data. Specifically, the ARIMA model models the time series through the following expression:

[0097]

[0098] Among them, Y(t) represents the observed value of the time series at time t, c represents the mean of the time series, φ i (i=1,2,…,p) is the autoregressive coefficient, which indicates the influence of the Y(t) value at the past i time points on the current Y(t). Y(ti) indicates the observed value of the time series at time ti. θ j(j=1,2,…,q) represents the impact of the error term ε at the past j time points on the current Y(t), ε(tj) represents the random error term of the time series at time tj, and ε(t) represents the random error term of the time series at time t.

[0099] Furthermore, before applying the ARIMA model, the time series data in the FTTR network needs to be preprocessed. This includes the stationarity test and difference processing of the data. Specifically, if the time series data shows a non-stationary trend, it can be converted into a stationary series through the difference operation. It should be understood that stationarity is a prerequisite for applying the ARIMA model. Only when the data is stationary can the model accurately capture the temporal dependencies in the time series.

[0100] After determining the stationarity of the data, it is necessary to estimate and select the parameters p, d, and q of the model. Among them, p is the order of the autoregressive part, d is the number of differences, and q is the order of the moving average part. The selection of these parameters is usually analyzed and determined through the autocorrelation function (ACF) and partial autocorrelation function (PACF) graphs. By observing the characteristics of the ACF and PACF graphs, the most suitable parameter combination can be determined, thereby constructing the best ARIMA model.

[0101] Furthermore, during the model training process, the maximum likelihood estimation (MLE) method is used to estimate the parameters of the ARIMA model. This process enables the model to best fit the training data by maximizing the likelihood function. It should be understood that the model's fitting effect directly affects the accuracy of fault prediction, so multiple experiments and tuning are required during the training process to ensure the optimal performance of the model.

[0102] In order to improve the generalization ability of the model, this embodiment also introduces cross-validation and regularization techniques. Specifically, by performing sliding window segmentation on the time series data, multiple validation and evaluation of the model can be achieved. In addition, by introducing regularization terms, the model can be effectively prevented from overfitting, thereby maintaining a high prediction accuracy when facing new data.

[0103] Furthermore, considering the complexity and variability of the FTTR network, this embodiment also optimizes the online update mechanism of the model. Specifically, when new fault data or abnormal patterns appear in the network, the system will automatically trigger the retraining and parameter adjustment of the model. This mechanism ensures that the ARIMA model can adapt to changes in network status in real time and maintain high prediction performance.

[0104] It is important to understand that the application of the ARIMA model is not limited to the prediction of a single time series, but can also be combined with other features to perform multivariate time series analysis. For example, multiple key indicators in the network (such as bandwidth utilization, packet loss rate, latency, etc.) can be used as multivariate inputs to build a multivariate ARIMA model. By comprehensively considering multiple factors, the comprehensiveness of fault prediction can be further improved.

[0105] The benefit of this embodiment is that by introducing the autoregressive integrated moving average model, the fault prediction capability of the FTTR network is improved. The ARI MA model can capture the temporal dependency structure in the time series data and provide accurate fault prediction results through reasonable parameter selection and model optimization.

[0106] Embodiment 7

[0107] In order to solve the problems of real-time response and historical fault mode reproduction in fault diagnosis in FTTR networks, this embodiment further optimizes the application of deep learning models and introduces a deep learning pattern matching process based on long short-term memory (LSTM) networks. This process aims to analyze the current fault characteristics through the LSTM network and optimize the model parameters using the error back propagation mechanism, thereby achieving a rapid response to similar historical faults. At the same time, this embodiment also refines the fault repair operation, generates the optimal repair path through the fault mode library, and automatically adjusts parameters or triggers maintenance prompts in combination with the device API to ensure that the fault can be handled promptly and effectively.

[0108] Specifically, as a special recurrent neural network (RNN), the LSTM network has the advantage of processing and predicting time series data, especially when facing long-term dependency problems. In the fault diagnosis of FTTR networks, the LSTM network can effectively capture the time dependency of network status, and through its internal memory units and gating mechanisms, it can effectively learn and reproduce historical fault features. The input of the LSTM network includes multi-dimensional features of the current network status, such as bandwidth utilization, packet loss rate, delay, etc. These features are modeled in time series through the LSTM layer, and feature fusion and output are performed through the fully connected layer.

[0109] Furthermore, the training process of the LSTM network achieves parameter optimization through the error back propagation mechanism. Specifically, the error between the model output and the actual fault label is measured by the loss function, and then the error is transmitted layer by layer through the back propagation algorithm, and the network weights are adjusted according to the gradient descent method. This process is repeated until the loss function of the model converges, thereby optimizing the LSTM network parameters. It should be understood that the error back propagation mechanism can effectively adjust the weights of the LSTM network, so that the model can respond quickly and make accurate diagnoses when faced with similar historical faults.

[0110] In order to further improve the accuracy and efficiency of fault diagnosis, this embodiment also introduces a fault mode library. The fault mode library stores the characteristics of historical faults and the corresponding repair operations. By matching the pattern with the current fault characteristics, similar faults can be quickly located and the optimal repair path can be generated. It should be understood that the construction of the fault mode library not only includes historical fault data, but also involves the extraction and classification of fault characteristics, which can be represented by the output of the LSTM network. By comparing the current fault characteristics with the historical data in the fault mode library, the system can quickly identify the most similar fault mode and process it according to the preset repair path.

[0111] Furthermore, the matching results of the fault mode library will be used to guide specific repair operations. According to the severity and type of the fault, the system will generate corresponding repair suggestions and automatically adjust relevant parameters in combination with the device API. For example, for bandwidth bottleneck problems, the system can adjust the transmission power and channel selection of the router through the API; for device overheating problems, the system can trigger fan acceleration and load balancing operations. These operations are implemented through the device API, ensuring that the repair measures are timely and effective. In addition, for faults that require manual intervention, the system will generate maintenance prompts to notify the operation and maintenance personnel to take corresponding measures.

[0112] It is important to understand that the use of device APIs is not limited to parameter adjustment, but also includes real-time monitoring and feedback of device status. By continuously monitoring the operating status of the device, the system can detect potential problems in a timely manner and perform preventive maintenance before failure occurs. This mechanism helps to improve the stability and reliability of the FTTR network and reduce network interruptions and performance degradation caused by equipment failure.

[0113] The benefit of this embodiment is that by introducing the LSTM-based deep learning pattern matching process, the fault diagnosis capability of the FTTR network has been significantly improved. The LSTM network can effectively capture the temporal dependency structure in the time series data and optimize the model parameters through the error back propagation mechanism, thereby achieving a rapid response to similar historical faults. At the same time, through the introduction of the fault mode library, the system can quickly locate similar faults and generate the optimal repair path, and realize automatic parameter adjustment and maintenance prompts in combination with the device API.

[0114] Embodiment 8

[0115] In order to solve the problem of lack of feedback mechanism after fault repair in FTTR network, which leads to the inability of deep learning model to continuously optimize, this embodiment further refines the effect feedback process after fault repair and introduces the recording and analysis mechanism of actual effect data. This mechanism aims to feed back the actual effect data after repair to the deep learning model to adjust and optimize the parameter library, thereby improving the decision-making efficiency of subsequent fault diagnosis and repair operations. Through this optimization, the fault diagnosis system of FTTR network can achieve self-learning and continuous improvement, and gradually improve its adaptability and accuracy in complex network environments.

[0116] Specifically, after the fault is repaired, the system automatically collects actual effect data related to the fault. This data includes the specific content of the repair operation, the network performance indicators after the repair, and any possible user feedback. For example, if the fault involves a bandwidth bottleneck, the actual effect data may include indicators such as bandwidth utilization, data transmission rate, and network latency after the repair. This data is acquired by the system through a preset data collection mechanism, and is initially cleaned and sorted to ensure the accuracy and consistency of the data.

[0117] Furthermore, the collected actual effect data will be fed back to the deep learning model for parameter adjustment and optimization of the model. Specifically, the deep learning model (such as the aforementioned LSTM network) not only optimizes the model parameters during the training phase through the error backpropagation mechanism, but also can continue to learn based on new data in actual applications. The actual effect data is used as a new training sample to update the weights and biases of the model so that the model can better adapt to new changes and new patterns in the network. It should be understood that this feedback mechanism makes the deep learning model no longer a static decision-making tool, but an intelligent system that can learn and improve dynamically.

[0118] In order to realize this feedback mechanism, the present embodiment also introduces the concept of parameter library. The parameter library is used to store all parameters and hyperparameters of the deep learning model, including the weight, bias, learning rate, optimizer state, etc. of the LSTM network. By feeding back the actual effect data to the parameter library, the system can fine-tune and optimize the parameters of the model, thereby gradually improving the prediction and decision-making capabilities of the model. It should be understood that the optimization of the parameter library does not only rely on single feedback data, but gradually achieves the overall optimization of the model by continuously collecting and analyzing a large amount of actual effect data.

[0119] Furthermore, the optimization process of the parameter library also involves the adjustment of hyperparameters. For example, hyperparameters such as the number of layers, number of hidden units, and learning rate of the LSTM network will affect the performance and convergence speed of the model. Through the feedback of actual effect data, the system can automatically adjust these hyperparameters to find the optimal model configuration. For example, if the model converges too slowly, the system may try to increase the learning rate; if the model is found to be overfitting, the system may try to reduce the number of layers or hidden units of the LSTM network. These adjustments are implemented through automated scripts to ensure the real-time and efficient optimization of the model.

[0120] It is important to understand that the feedback of actual effect data and the optimization of the parameter library are not only to improve the accuracy of single fault diagnosis, but more importantly to improve the generalization ability of the entire fault diagnosis system. Through continuous learning and optimization, the system can still maintain a high level of prediction and decision-making capabilities when facing new fault modes and complex network environments. In addition, the optimization of the parameter library also involves retraining historical data to ensure that the model can continuously learn and improve from historical data.

[0121] Additional features of this embodiment also include dynamic updating of the fault mode library. The fault mode library not only stores historical fault features and repair operations, but also continuously updates and expands its content through feedback from actual effect data. Specifically, the actual effect data after each fault repair will be used to update the fault mode library, making the fault modes in the library richer and more comprehensive. This dynamic update mechanism ensures the real-time and effectiveness of the fault mode library, so that the system can provide more accurate and efficient repair suggestions when facing similar faults.

[0122] The benefit of this embodiment is that, by introducing the feedback mechanism of actual effect data, the fault diagnosis and repair system of the FTTR network achieves self-learning and continuous optimization. The deep learning model continuously adjusts and optimizes the parameter library through the feedback of actual effect data, thereby improving the decision-making efficiency of subsequent fault diagnosis and repair operations. At the same time, the dynamic update mechanism of the fault mode library enables the system to continuously expand and enrich its fault modes and provide more comprehensive and accurate repair suggestions.

[0123] Embodiment 9

[0124] In order to solve the problems of insufficient real-time, accuracy and automation of the fault diagnosis system in the FTTR network, this embodiment proposes a remote fault diagnosis system based on the FTTR network. The system integrates multiple functional modules such as data acquisition, status analysis, anomaly detection, fault prediction, fault diagnosis, and repair suggestions and execution, realizing the automation and intelligence of the whole process from data acquisition to fault handling. Through the close collaboration between the modules, the system can timely discover problems, predict risks, diagnose the causes of faults, and automatically perform repair operations or trigger maintenance prompts in a complex network environment.

[0125] Specifically, the data acquisition module in this embodiment is deployed at each FTTR node to obtain network status information in real time. These nodes are distributed throughout the FTTR network and can continuously monitor key network indicators (such as bandwidth utilization, latency, packet loss rate, etc.). The data acquisition module uses high-precision sensors and data acquisition algorithms to ensure that the acquired network status data has high accuracy and real-time performance. These data are then transmitted to the status analysis module for preliminary analysis and distributed synchronization. It should be understood that the distributed synchronization mechanism ensures data consistency between nodes, allowing the system to maintain a unified view of the network status on different nodes.

[0126] Furthermore, the state analysis module performs preliminary processing on the collected network state data, including data cleaning, outlier detection, and data normalization. These processing steps are intended to improve the quality and consistency of the data and provide a reliable data basis for subsequent anomaly detection and fault diagnosis. The state analysis module also uses distributed computing technology to process data on different nodes in parallel, thereby improving the efficiency and scalability of data analysis. It should be understood that the output results of the state analysis module will be provided as input to the anomaly detection module and the fault prediction module for deeper analysis.

[0127] Furthermore, the anomaly detection module in this embodiment uses an unsupervised learning algorithm to identify abnormal patterns from normal network behaviors. The unsupervised learning algorithm establishes a baseline behavior model by learning the normal patterns of network status data. When the actual network state deviates from the model, the system will identify potential network problems and mark them as abnormal events. The advantage of this unsupervised learning method is that it does not require a large amount of labeled data and can automatically discover new abnormal behaviors in the case of unknown fault modes. Furthermore, the output of the anomaly detection module will be passed to the fault prediction module for further risk prediction.

[0128] Furthermore, in this embodiment, the fault prediction module uses a supervised learning algorithm to predict possible network risks and provide early warning signals. The supervised learning algorithm is trained with historical fault data to learn the typical features and patterns before the fault occurs. In actual operation, the module can predict the possible future fault risks based on the current network status data and issue early warning signals in advance. These early warning signals help operation and maintenance personnel take timely measures to prevent further expansion and deterioration of the fault. It should be understood that the accuracy of the fault prediction module depends on high-quality training data and reasonable model selection. Therefore, in this embodiment, model selection and parameter optimization are an ongoing process.

[0129] Furthermore, the fault diagnosis module performs a detailed analysis of the faults that have occurred based on the deep learning algorithm and performs fault pattern matching. The deep learning algorithm can handle complex nonlinear relationships and is suitable for modeling multi-dimensional and multi-variable network status data. In this embodiment, the fault diagnosis module uses deep learning models such as long short-term memory networks (LSTM) to conduct in-depth analysis of the causes of the faults and perform pattern matching through the fault pattern library. The fault pattern library stores a large amount of historical fault data and corresponding repair operations. By comparing with the current fault characteristics, the system can quickly locate similar faults and analyze the root cause of the fault.

[0130] Furthermore, the repair suggestion and execution module is designed to generate the optimal repair suggestion, and automatically repair or trigger maintenance prompts through the system interface. Specifically, the repair suggestion and execution module generates the optimal repair path and operation suggestions based on the fault diagnosis results and the historical data in the fault mode library. If the fault can be automatically repaired through the system interface, such as adjusting router parameters, optimizing bandwidth allocation, etc., the system will automatically perform related operations; if the fault requires manual intervention, the system will generate detailed maintenance prompts to notify the operation and maintenance personnel to handle it. It should be understood that the degree of automation of the repair suggestion and execution module directly affects the efficiency and accuracy of fault handling. Therefore, in this embodiment, the stability and reliability of the system interface are fully considered and optimized.

[0131] Additional features of this embodiment include optimizing the data flow and control flow between modules. By introducing efficient data transmission protocols and distributed computing technology, the system can achieve efficient collaboration and data sharing between modules. In addition, the system also introduces a fault backtracking mechanism. When the fault processing is completed, the system automatically records the processing process and results, and feeds back to the fault prediction and diagnosis module for further optimization and adjustment of the model.

[0132] Embodiment 10

[0133] In order to solve the execution efficiency and reliability issues of the software implementation level when the remote fault diagnosis method in the FTTR network is actually deployed, this embodiment proposes a computer-readable storage medium. The storage medium stores a computer program, and when the program is executed on an electronic device, it can implement the remote fault diagnosis method based on the FTTR network in the previous embodiment. Through this optimization, the system can not only achieve efficient fault diagnosis and repair at the hardware level, but also ensure the stability and flexibility of program execution at the software level, thereby improving the overall performance of the entire fault diagnosis system.

[0134] Specifically, the computer program in this embodiment is designed as a modular structure, and each module corresponds to a specific step in the fault diagnosis process. The additional features of this embodiment also include exception handling and logging mechanisms during program execution. In order to ensure the stability and reliability of the program, a variety of exception handling mechanisms are introduced into the system, which can capture and handle exceptions in a timely manner when errors occur in program execution to avoid program crashes or data loss. At the same time, a detailed logging mechanism is used to record each key step of program execution, including data acquisition, state analysis, anomaly detection, fault prediction, fault diagnosis, and repair execution. These logs not only help operation and maintenance personnel to troubleshoot, but also provide data support for subsequent program optimization and model updates.

[0135] The benefit of this embodiment is that by optimizing the execution process and module design of the computer program, the remote fault diagnosis system of the FTTR network can achieve efficient, stable and reliable execution at the software level. The introduction of modular program design and distributed computing technology improves the scalability and execution efficiency of the system. At the same time, the optimization of the exception handling and logging mechanism enables the system to maintain good maintainability in the face of complex network environments and sudden failures.

Claims

1. A remote fault diagnosis method based on FTTR network, characterized in that: The method comprises the following steps: A monitoring module for monitoring the network status is deployed on each node of the FTTR network, each monitoring module is provided with a data acquisition unit and a status analysis unit, and each node collects network status data in real time through the monitoring module; The network status data includes monitoring network delay, packet loss rate, device temperature and bandwidth utilization, and the network status data is preliminarily analyzed by the data acquisition unit and compared with historical data to determine the change trend of the current status; Each of the nodes publishes the status information of the current status to a shared topic through a message queue system, and synchronizes with other nodes to form a global status view; Analyzing and synchronizing the global state view using an unsupervised learning algorithm to identify abnormal patterns for detecting extreme fluctuations in traffic and sustained high latency behavior; Use a supervised learning model to analyze historical fault data, extract features, and analyze the abnormal pattern in combination with the extracted features to predict potential faults and issue warnings; When the warning is triggered, the deep learning model is used to analyze the current fault, infer the cause of the fault, generate and execute repair operations.

2. The remote fault diagnosis method based on FTTR network as claimed in claim 1, characterized in that: The state analysis unit adopts a dual threshold model for analyzing and predicting the change trend. The expression of the dual threshold model is: Among them, ΔM(t) is the state change rate, and T1 and T2 are thresholds.

3. The remote fault diagnosis method based on FTTR network as claimed in claim 2, characterized in that: The unsupervised learning algorithm includes a clustering method based on a dynamic discriminant boundary, which is expressed as: Where D is the data point x i To cluster center c j The distance is , and p is the adjustment parameter.

4. The remote fault diagnosis method based on FTTR network as claimed in claim 1, characterized in that: Each of the nodes publishes the status information of the current status to a shared topic through a message queue system, and the step of synchronizing with other nodes includes: Each of the nodes publishes JSON format status information to the global shared topic through the message queue system every 30 seconds, and integrates other node information to generate a global status table after receiving it.

5. The remote fault diagnosis method based on FTTR network as claimed in claim 1, characterized in that: The supervised learning model is based on an integrated learning method, including the use of multi-model collaboration, and the expression of the multi-model collaboration is: f(x)=Σ k w k ·h k (x), where f(x) is the total fault prediction, h k (x) is the output of each sub-model, w k is the weight corresponding to the sub-model.

6. The remote fault diagnosis method based on FTTR network as claimed in claim 5, characterized in that: The fault prediction also adopts time series analysis method, using an autoregressive integrated moving average model, the expression of the autoregressive integrated moving average model is: Among them, Y(t) represents the observed value of the time series at time t, c represents the mean of the time series, φ i (i=1,2,…,p) is the autoregressive coefficient, which indicates the influence of the Y(t) value at the past i time points on the current Y(t). Y(ti) indicates the observed value of the time series at time ti. θ j (j=1,2,…,q) represents the impact of the error term ε at the past j time points on the current Y(t), ε(tj) represents the random error term of the time series at time tj, and ε(t) represents the random error term of the time series at time t.

7. The remote fault diagnosis method based on FTTR network as claimed in claim 1, characterized in that: The deep learning model is used to analyze the current fault based on the deep learning pattern matching process of the long short-term memory network (LSTM), and the parameters are optimized through error back propagation to quickly respond to faults similar to the history; The execution repair operation generates an optimal repair path according to the fault mode library, and automatically adjusts parameters or triggers maintenance prompts in combination with the device API.

8. The remote fault diagnosis method based on FTTR network as claimed in claim 7, characterized in that: After the fault is repaired, the actual effect data is recorded and fed back to the deep learning model to adjust and optimize the parameter library to improve the decision-making efficiency of subsequent operations.

9. A remote fault diagnosis system based on FTTR network, characterized in that: The system comprises: Data collection module, deployed at each FTTR node, for real-time collection of network status information; A status analysis module, which obtains the network status data provided by the data acquisition module, and performs preliminary analysis and distributed synchronization on the network status data; Anomaly detection module, which uses unsupervised learning algorithms to detect abnormal patterns and identify potential network problems; The fault prediction module uses supervised learning algorithms to predict possible network risks and provide early warning signals; Fault diagnosis module, which analyzes fault causes based on deep learning algorithms and performs fault pattern matching; The repair suggestion and execution module generates the optimal repair suggestion and performs automatic repair or triggers maintenance prompts through the system interface.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed on an electronic device, it enables the computer program to execute the remote fault diagnosis method based on the FTTR network as claimed in any one of claims 1 to 8.

Images