Optical network performance optimization solution acquisition and fault diagnosis method, device and system

Abstract

The embodiment of the present application provides a method, device and system for obtaining an optical network performance optimization solution and diagnosing faults. The method includes: collecting operation logs, performance indicators and alarm information of various types of equipment in the optical network according to a pre-set data collection frequency, as well as collecting optical network topology data, network-level performance data, equipment configuration information and computer room environment data; cleaning and preprocessing all collected data; inputting all processed data into a pre-trained performance optimization model, a pre-built expert system and / or a pre-trained fault prediction model, generating an optical network performance optimization solution through the performance optimization model, and predicting the potential fault type, possible fault occurrence time and fault location of the optical network through the expert system and / or the fault prediction model. The embodiment of the present application can not only obtain a better optical network performance optimization solution, but also improve the fault location speed and accuracy.

Description

Technical Field

[0001] The present application relates to the field of optical network operation and maintenance technology, and in particular to an optical network performance optimization solution acquisition and fault diagnosis method, device and system. Background Art

[0002] With the rapid development of information technology, optical networks play a vital role in the field of communications and are widely used in telecommunications, Internet data centers, enterprise networks, etc. As the scale of optical networks continues to expand, their complexity is also increasing, and traditional operation and maintenance methods face many challenges.

[0003] During the operation and maintenance of optical networks, a huge amount of data is generated, including equipment performance data, alarm data, traffic data, configuration data, etc. These data contain rich information, but due to the huge amount of data, diverse data types, and fast data generation speed, traditional data analysis tools and methods are difficult to effectively process and mine them. Operation and maintenance personnel can often only judge network faults and performance problems based on limited experience and simple statistical analysis, resulting in inaccurate fault location, long fault repair time, poor performance optimization, and difficulty in achieving refined management and optimization of optical networks.

[0004] In addition, optical network failures are diverse and complex, and may be caused by a variety of factors such as equipment hardware failure, software vulnerabilities, optical link damage, configuration errors, traffic anomalies, etc. Different failures have different manifestations and impact ranges. Traditional fault analysis and location methods are prone to generate a large number of false alarms and vulnerabilities, and cannot accurately identify the root cause of the fault, further increasing the difficulty and cost of operation and maintenance.

[0005] Therefore, the present application provides an optical network performance optimization solution acquisition and fault diagnosis method to solve the above problems. Summary of the invention

[0006] The embodiment of the present application provides a method for obtaining an optical network performance optimization solution and diagnosing faults, which can not only obtain a better optical network performance optimization solution, but also improve the fault locating speed and positioning accuracy, thereby improving the operation and maintenance efficiency and reliability of the optical network, reducing the operation and maintenance difficulty and cost, and providing a strong guarantee for the stable operation of the optical network.

[0007] The present application provides a method for obtaining an optical network performance optimization solution and diagnosing a fault, the method comprising: Collect the operation logs, performance indicators and alarm information of various devices in the optical network according to the pre-set data collection frequency, as well as collect the optical network topology data, network-level performance data, device configuration information and computer room environment data; Clean all collected data to remove noise data, duplicate data, and obviously erroneous data, and use data completion algorithms to reasonably fill in missing data; Standardize all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format; All data obtained after the standardization process are input into a pre-trained performance optimization model, a pre-built expert system and / or a pre-trained fault prediction model, the correlation between different network parameters is analyzed by the performance optimization model, an optical network performance optimization plan is generated based on the correlation, and the potential fault type, possible fault occurrence time and fault occurrence location of the optical network are predicted by the expert system and / or the fault prediction model.

[0008] In the optical network performance optimization solution acquisition and fault diagnosis method described in the embodiment of the present application, after predicting the potential fault type, possible fault time and fault location of the optical network by the expert system and / or the fault prediction model, it also includes: Generate corresponding fault handling suggestions, and send the potential fault type of the optical network, the time when the fault may occur, the location where the fault occurs, and the fault handling suggestions to the operation and maintenance personnel or a pre-developed intelligent operation and maintenance assistant for processing by the operation and maintenance personnel or the intelligent operation and maintenance assistant.

[0009] In the optical network performance optimization solution acquisition and fault diagnosis method described in the embodiment of the present application, after generating the optical network performance optimization solution based on the association relationship, and predicting the potential fault type, possible fault occurrence time and fault occurrence location of the optical network by the expert system and / or the fault prediction model, it also includes: Generate an operation and maintenance report of the optical network regularly and present the operation and maintenance report in a visual manner, wherein the content of the operation and maintenance report includes but is not limited to a summary of the operation status of the optical network, a statistical analysis of faults, and a trend of performance indicators.

[0010] In the optical network performance optimization solution acquisition and fault diagnosis method described in the embodiment of the present application, after the operation and maintenance report of the optical network is regularly generated and the operation and maintenance report is presented in a visual manner, it also includes: In response to data query, analysis task customization and model parameter adjustment operations performed by operation and maintenance personnel through a visual interactive interface, the corresponding data is queried, customized tasks are analyzed, and corresponding model parameters are adjusted.

[0011] In the optical network performance optimization solution acquisition and fault diagnosis method described in the embodiment of the present application, after generating the optical network performance optimization solution based on the association relationship, it also includes: Determining whether it is necessary to notify the operation and maintenance personnel to optimize the performance of the optical network according to the optical network performance optimization plan according to the content of the optical network performance optimization plan; If not, the optical network performance is optimized according to the optical network performance optimization solution through the developed intelligent operation and maintenance assistant.

[0012] In the optical network performance optimization solution acquisition and fault diagnosis method described in the embodiment of the present application, after the standardization of all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format, it also includes: All data obtained after standardization processing are stored in a memory database, a distributed storage system or a cloud storage service according to preset rules, and a data index structure is established to record the storage location, data format and data source of each data. When the distributed storage system is used for storage, the data to be stored is stored in a distributed manner on multiple storage nodes of the distributed storage system.

[0013] In the optical network performance optimization solution acquisition and fault diagnosis method described in the embodiment of the present application, after the standardization of all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format, it also includes: All data obtained after the standardization process are input into a pre-trained fault diagnosis model, and the type of optical network fault and the location where the fault occurs are determined by the fault diagnosis model.

[0014] An embodiment of the present application also provides an optical network performance optimization solution acquisition and fault diagnosis device, the device comprising a processor and a memory, the memory storing a computer program, and the processor calling the computer program stored in the memory to execute the optical network performance optimization solution acquisition and fault diagnosis method described in any of the above embodiments.

[0015] An embodiment of the present application also provides an optical network performance optimization solution acquisition and fault diagnosis system, and the optical network performance optimization solution acquisition and fault diagnosis system includes the optical network performance optimization solution acquisition and fault diagnosis device described in the above embodiment.

[0016] An embodiment of the present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is run on a computer, the computer executes the optical network performance optimization solution acquisition and fault diagnosis method described in any of the above embodiments.

[0017] The optical network performance optimization solution acquisition and fault diagnosis method provided in the embodiment of the present application first collects the operation logs, performance indicators and alarm information of various devices in the optical network according to a preset data collection frequency, and collects optical network topology data, network-level performance data, equipment configuration information and computer room environment data, and then cleans all the collected data to remove noise data, duplicate data and obviously erroneous data, and uses a data completion algorithm to reasonably fill in the missing data, and then standardizes all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format, and finally converts the data after standardization. All the data obtained are input into a pre-trained performance optimization model, a pre-built expert system and / or a pre-trained fault prediction model. The performance optimization model is used to analyze the correlation between different network parameters, and an optical network performance optimization plan is generated based on the correlation. The expert system and / or the fault prediction model is used to predict the potential fault type, possible fault occurrence time and fault location of the optical network. This not only can obtain a better optical network performance optimization plan, but also can improve the fault locating speed and positioning accuracy, thereby improving the operation and maintenance efficiency and reliability of the optical network, reducing the operation and maintenance difficulty and cost, and providing a strong guarantee for the stable operation of the optical network. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application, and those skilled in the art can obtain other drawings based on these drawings without creative work.

[0019] Figure 1 A flowchart of a method for obtaining an optical network performance optimization solution and diagnosing a fault in an embodiment of the present application.

[0020] Figure 2 A schematic diagram of the structure of an optical network performance optimization solution acquisition and fault diagnosis device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0022] An embodiment of the present application provides an optical network performance optimization solution acquisition and fault diagnosis method, which can be applied to an optical network performance optimization solution acquisition and fault diagnosis system.

[0023] See also Figure 1 , Figure 1 A flow chart of a method for obtaining an optical network performance optimization solution and diagnosing a fault is provided in an embodiment of the present application. The method for obtaining an optical network performance optimization solution and diagnosing a fault is applied to an optical network performance optimization solution obtaining and diagnosing a fault system, and the method may include the following steps: Step 101, collects operation logs, performance indicators and alarm information of various devices in the optical network according to a preset data collection frequency, and collects optical network topology data, network-level performance data, device configuration information and computer room environment data.

[0024] Among them, various types of equipment in the optical network, such as optical transmission equipment, optical switching equipment, and optical access equipment, can be connected to these devices through specially developed data acquisition interfaces to collect operation logs, performance indicators, and alarm information of various types of equipment. At the same time, it connects with the network management system (NMS) and element management system (EMS) to obtain network topology data, network-level performance data, and equipment configuration information. In addition, it connects to external environmental monitoring equipment to collect computer room environmental data (such as computer room temperature, computer room humidity, and computer room dust concentration, etc.) to provide a reference for analyzing the stability of optical network equipment. Specifically, develop various interface drivers specifically for optical network equipment, which can seamlessly connect optical transmission equipment (such as OTN equipment, DWDM equipment, etc.), optical switching equipment, optical access equipment, etc. from different manufacturers and models, and obtain their detailed internal operating data, including but not limited to optical module parameters (such as transmission power, receiving sensitivity), device port status (such as port traffic, port bit error rate), device alarm information (such as optical path interruption alarm, device overheating alarm), etc.; design efficient data interaction interfaces with network management systems (NMS) and element management systems (EMS) to obtain optical network topology information, device configuration information, and network-level performance data (such as network delay, network bandwidth utilization), etc. At the same time, it can receive control instructions from these management systems in real time to achieve collaborative work between data collection and network management; build interfaces with external environmental monitoring systems to collect relevant data on the environment in which optical network equipment is located, such as room temperature, humidity, dust concentration, etc. These data play an important auxiliary role in analyzing the stability and causes of failures of optical network equipment.

[0025] If the existing software-based data acquisition interface performs poorly in certain specific optical network environments, you can consider using specialized hardware data acquisition devices. These devices can be directly connected to key nodes in the optical network, achieve high-speed data acquisition and subsequent preliminary data processing through hardware circuits, and then transmit the processed data to the data storage layer. For example, for optical signal monitoring in high-speed optical transmission networks, a dedicated optical power monitoring card can be used, which can collect optical power data in real time, and perform simple filtering and digital processing to improve the accuracy and speed of data acquisition.

[0026] Among them, this step is carried out at the data collection layer. Before data collection, first deploy the corresponding collection software module or hardware device according to the type and interface specification of the equipment in the optical network. For example, for optical transmission equipment, install a special SNMP agent, configure the collection parameters (such as collection interval, data type, etc.), and start the data collection process. The collected data is transmitted to the data cleaning and preprocessing server through the network, and the data cleaning algorithm and preprocessing program are run on the server for subsequent data cleaning and data preprocessing.

[0027] Among them, traditional SNMP (Simple Network Management Protocol) and other protocols can be used for data collection. It should be noted that in addition to using traditional SNMP (Simple Network Management Protocol) and other protocols for data collection, emerging NETCONF (Network Configuration Protocol) or RESTCONF (Representational State Transfer Configuration Protocol) protocols can be used. These protocols may have higher efficiency and flexibility in handling large-scale data collection and device configuration management. For example, for optical network equipment that supports the NETCONF protocol, it is easier to obtain detailed configuration information and real-time operation data of the equipment, and remote configuration and management of the equipment can be achieved.

[0028] Among them, determine a reasonable data collection frequency, and use high-frequency collection for key performance data (such as optical power, bit error rate, etc.), such as collecting once every 10 seconds, to ensure that the slight changes in network status can be captured in time; for relatively stable equipment configuration information and topology information, use low-frequency collection, such as collecting once a day or when the network topology changes, to reduce unnecessary data transmission and storage overhead. Among them, an adaptive data collection frequency adjustment algorithm can be used to automatically and dynamically adjust the collection frequency of different types of data according to the real-time load, business type and historical data change rules of the optical network to ensure the timeliness and integrity of the data, while optimizing data collection resource consumption.

[0029] Among them, an accurate data synchronization mechanism is established to ensure that the data collected from different data sources are consistent and comparable on the time axis. The Network Time Protocol (NTP) or Global Positioning System (GPS) clock synchronization technology is used to accurately timestamp each collected data, so that accurate timing analysis and correlation analysis can be performed in subsequent data analysis.

[0030] Among them, performance indicators include optical power, bit error rate and signal-to-noise ratio.

[0031] Among them, the unique multi-source data fusion collection method of the embodiment of the present application can efficiently integrate the internal equipment data of the optical network, the external environment data and the network management system data under the same data collection framework, and the collection process has little impact on the normal operation of the optical network (such as CPU utilization rate less than 5%, network bandwidth occupancy less than 1%).

[0032] Step 102, clean all collected data to remove noise data, duplicate data and obviously erroneous data, and use a data completion algorithm to reasonably fill in the missing data.

[0033] Specifically, for example, if a small amount of optical power data is missing, linear interpolation can be performed based on data at adjacent time points.

[0034] Step 103 , standardize all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format.

[0035] Among them, the data is standardized and the data of different formats and magnitudes are converted into a unified standard format for subsequent analysis and processing.

[0036] In some embodiments, after the standardization of all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format, the process further includes: All data obtained after standardization processing are stored in a memory database, a distributed storage system or a cloud storage service according to preset rules, and a data index structure is established to record the storage location, data format and data source of each data. When the distributed storage system is used for storage, the data to be stored is stored in a distributed manner on multiple storage nodes of the distributed storage system.

[0037] The data in the input model (fault prediction model, fault diagnosis model) or expert system may be data that has been stored in a memory database, a distributed storage system, or a cloud storage service, or may be data that has not yet been stored in a memory database, a distributed storage system, or a cloud storage service.

[0038] Among them, Hadoop distributed file system (such as Hadoop Distributed File System, HDFS) or similar distributed storage architecture can be used to store the cleaned and pre-processed massive optical network operation and maintenance data in multiple storage nodes according to certain rules (such as data type, acquisition time, device area, etc.), so as to improve the reliability and scalability of data storage. For example, the performance data from different devices in the same time period is stored on different storage nodes. When a node fails, it will not cause the loss of the entire data. At the same time, the storage capacity can be easily expanded by adding storage nodes. This distributed storage method has the characteristics of high reliability and high scalability, and can effectively cope with the rapid growth of data volume. This step is carried out in the data storage layer. When the data storage layer is initialized, an HDFS cluster is built, the number of storage nodes, storage strategy and other parameters are set, the cleaned data is stored in HDFS according to the predetermined storage rules, and data index and metadata information are established.

[0039] In scenarios where data read and write speeds are extremely high, in addition to distributed storage systems, in-memory databases (such as Redis) can be introduced as auxiliary storage. Some frequently accessed and analyzed hot data (such as the optical network performance data in the last hour) are stored in the in-memory database. In this way, when performing data analysis, data can be quickly obtained from the in-memory database first, improving analysis efficiency. Historical data and infrequently accessed data are still stored in the distributed storage system to achieve hierarchical management of data storage.

[0040] Among them, if the construction cost of the enterprise's own distributed storage system is high or maintenance is difficult, you can choose to use cloud storage services (such as cloud storage products provided by Alibaba Cloud, Tencent Cloud, etc.) to store optical network operation and maintenance big data. Cloud storage services have the advantages of high availability, strong scalability, and no need for enterprises to manage storage infrastructure by themselves. Enterprises only need to upload data to cloud storage and access and manage data through corresponding interfaces. At the same time, cloud storage service providers usually also provide some data security and backup functions to further ensure data security. When using cloud storage to replace distributed storage systems, you first need to register a cloud storage service account, create a corresponding storage bucket, and then modify the data upload interface of the data storage layer to the API interface of the cloud storage service, and organize and manage the data according to the requirements of cloud storage.

[0041] Among them, the data index and metadata management mechanism are designed to establish an efficient data index mechanism for the data, such as multi-dimensional indexes based on device ID, timestamp, data type, etc., to facilitate the rapid location and retrieval of required data. At the same time, the metadata of the data is managed to record the source, storage location, data format and other information of the data to facilitate data maintenance and management. For example, by establishing a composite index based on device ID and collection time, all operation and maintenance data of a certain device within a specific time period can be quickly queried.

[0042] Among them, the embodiment of the present application is based on a distributed storage optical network operation and maintenance data storage structure, including a specific data sharding strategy, a redundant storage mechanism, and an efficient data index construction method, so that the data storage has high reliability (such as a data loss rate of less than 0.01%), high scalability (can be easily expanded to PB-level storage capacity) and fast data retrieval capability (average data retrieval time is within 1 second).

[0043] For data stored in memory databases and distributed storage systems, a regular data backup plan can be developed to back up the optical network operation and maintenance big data to an off-site storage center or cloud storage platform to prevent local storage system failures or data loss. The backup frequency can be adjusted according to the importance and update frequency of the data, for example, key performance data can be backed up once a day and historical data can be backed up once a week.

[0044] Among them, the embodiment of the present application also develops data recovery tools and processes, which can quickly restore data from the backup system when data loss or damage occurs. By combining incremental backup and full backup, you can choose a suitable recovery strategy according to the actual situation when recovering data, and reduce the data recovery time. For example, first perform a full backup, and then perform incremental backups every day. When you need to restore data, if only the data of the most recent day is lost, you only need to restore the incremental backup of that day; if the data is lost for a longer period of time, restore the full backup first, and then restore the incremental backups in sequence.

[0045] Among them, the innovative data backup and recovery process of the embodiment of the present application adopts a data integrity verification mechanism based on blockchain technology to ensure the authenticity and integrity of the backup data. During the data recovery process, it can automatically select the optimal recovery path and strategy according to the importance and urgency of the data, thereby shortening the data recovery time by more than 50%.

[0046] Step 104, input all the data obtained after the standardization processing into a pre-trained performance optimization model, a pre-built expert system and / or a pre-trained fault prediction model, analyze the correlation between different network parameters through the performance optimization model, generate an optical network performance optimization plan based on the correlation, and predict the potential fault type, possible fault occurrence time and fault location of the optical network through the expert system and / or the fault prediction model.

[0047] Among them, machine learning algorithms (such as support vector machines, neural networks, etc.) are used to build an optical network fault prediction model. First, a part of the historical operation and maintenance data is used as a training set to extract features from the data (such as the trend of optical power changes, the frequency of alarms, etc.) to train the model. Then, another part of the data is used as a test set to verify and evaluate the trained model, and adjust the parameters of the model to improve the accuracy of the model. For example, the optical network data of the past year is divided into a training set and a test set at a ratio of 70% and 30%. The support vector machine algorithm is selected to build a fault prediction model, extract feature vectors from the data, train the model using the training set, and evaluate and adjust the model through the test set until the model meets the predetermined accuracy requirements. Through this model, the potential fault type, occurrence time and location can be predicted based on the current optical network operation data.

[0048] For the fault prediction model, time series analysis algorithms (such as ARIMA model) and deep learning algorithms (such as long short-term memory network LSTM) are used to analyze historical operation and maintenance data to predict possible future failures. Through long-term monitoring and analysis of optical network performance data, potential failure trends are discovered and early warning information is issued in advance so that operation and maintenance personnel can take corresponding preventive measures, such as replacing aging optical modules in advance, optimizing optical path configuration, etc., to reduce the losses caused by network failures.

[0049] Among them, for the performance optimization model, the association rule mining algorithm in data mining technology (such as Apriori algorithm or FP-Growth algorithm) is used to analyze the correlation between different equipment parameters, network configuration parameters and network performance indicators (such as network bandwidth, delay, packet loss rate, etc.) in the optical network. For example, find out the relationship between the bandwidth utilization of the optical link and the wavelength allocation of the optical signal, the traffic load of the device port and other parameters. Based on these correlations, the intelligent optimization algorithm (such as genetic algorithm, particle swarm optimization algorithm, etc.) is used to optimize the network configuration to improve the overall performance of the optical network, such as reducing network delay and improving bandwidth utilization. The configuration parameters of the optical network are optimized by using intelligent optimization algorithms such as genetic algorithm and particle swarm optimization algorithm. With network performance indicators as the optimization target, such as maximizing network bandwidth utilization and minimizing network delay, under the premise of meeting the physical limitations of optical network equipment and business requirements, the network configuration parameters (such as optical link routing selection, optical signal power allocation, etc.) are continuously adjusted to find the optimal network configuration solution to improve the overall performance of the optical network.

[0050] If the existing machine learning algorithms do not work well when building fault prediction models and performance optimization models, you can try using deep learning frameworks (such as TensorFlow, PyTorch, etc.). Deep learning frameworks can more easily build complex neural network models. For example, using convolutional neural networks (CNNs) to extract and analyze the spatiotemporal data of optical networks (such as the data on the change of optical power over time and spatial position), or using recurrent neural networks (RNNs) to model operation and maintenance data with time series characteristics (such as alarm sequences), may improve the prediction accuracy and performance optimization effect of the model. In the case of using deep learning frameworks to replace traditional machine learning algorithms to build models, it is necessary to install the corresponding deep learning framework environment, design the neural network structure according to the characteristics of the optical network data, adjust training parameters such as learning rate, number of iterations, etc., and perform model training and optimization.

[0051] Among them, in the data analysis process, in addition to data-driven machine learning and data mining algorithms, expert systems can be introduced. Expert systems are reasoning systems built based on the knowledge and experience of domain experts. By encoding the knowledge of experts in the field of optical network operation and maintenance, such as fault diagnosis experience and performance optimization strategies, into a rule base, during data analysis, the expert system first conducts a preliminary analysis and judgment of the data, gives some experience-based suggestions, and then combines the analysis results of data mining and machine learning algorithms to comprehensively draw more accurate and comprehensive conclusions. For example, for some common types of optical network faults, the expert system can quickly determine the cause of the fault and provide corresponding solutions based on pre-set rules, while for some complex situations that are difficult to judge by rules, the machine learning model conducts in-depth analysis.

[0052] Among them, this step is carried out at the data analysis layer.

[0053] Among them, the optical network fault prediction model architecture based on deep learning in the embodiment of the present application includes a specific neural network layer design (such as a convolutional layer design for optical power data, a recurrent layer design for time series data), model training algorithm optimization (such as the use of adaptive learning rate adjustment, gradient clipping and other technologies) and a model evaluation index system (such as comprehensive consideration of accuracy, recall rate, F1 value and false alarm rate and other indicators), which can achieve high-precision diagnosis (accuracy of more than 95%) and early prediction (advance warning time of more than 2 hours) of optical network faults.

[0054] Among them, the embodiments of the present application combine the optical network performance optimization method of multiple data mining and intelligent optimization algorithms, such as the optical link configuration optimization algorithm that combines association rule mining with genetic algorithm, and the optical network resource allocation optimization algorithm that combines clustering analysis with particle swarm optimization algorithm, which can significantly improve the performance indicators of the optical network (such as network bandwidth utilization increased by more than 30%, network delay reduced by more than 20%).

[0055] In some embodiments, after predicting the potential fault type, possible fault time and fault location of the optical network by the expert system and / or the fault prediction model, the method further includes: Generate corresponding fault handling suggestions, and send the potential fault type of the optical network, the time when the fault may occur, the location where the fault occurs, and the fault handling suggestions to the operation and maintenance personnel or a pre-developed intelligent operation and maintenance assistant for processing by the operation and maintenance personnel or the intelligent operation and maintenance assistant.

[0056] When the fault prediction model detects a potential fault, it will promptly send warning information to the operation and maintenance personnel, including the type of fault, possible time and location of the fault, etc. At the same time, it will provide fault handling suggestions, such as recommending corresponding troubleshooting steps and equipment parts that may need to be replaced, to help the operation and maintenance personnel quickly locate and solve the fault.

[0057] Among them, when the fault prediction model issues an early warning, the operation and maintenance personnel are notified by SMS, email or pop-up windows of the operation and maintenance platform. The operation and maintenance personnel can view detailed fault information and handling suggestions on the operation and maintenance platform, and perform fault troubleshooting and handling according to the suggestions. In addition, this information can also be sent to the mobile devices of the operation and maintenance personnel, so that the operation and maintenance personnel can obtain optical network operation and maintenance information and perform operations in mobile scenarios (such as out-of-town inspections, remote emergency handling, etc.). Among them, a specially developed mobile application is set on the mobile device, which communicates with the back-end service of the data application layer. The operation and maintenance personnel can log in to the mobile application through mobile devices such as mobile phones or tablets, receive fault warning information, view network performance data, and perform some simple operation and maintenance operations (such as remote restart of equipment, confirmation of fault handling results, etc.). This can improve the work flexibility and response speed of the operation and maintenance personnel, and is not limited by the traditional operation and maintenance platform that can only be used in fixed office locations.

[0058] Among them, in terms of fault handling and performance optimization decision support, this application develops an intelligent operation and maintenance assistant. Based on artificial intelligence technology, the intelligent operation and maintenance assistant can automatically analyze fault warning information and performance optimization suggestions, and automatically perform some simple operation and maintenance operations according to preset strategies and rules, such as automatically adjusting the parameters of the optical link, switching backup equipment, etc. Only when encountering complex situations or requiring manual intervention, the operation and maintenance personnel will be notified to handle it. This can greatly reduce the workload of the operation and maintenance personnel, improve the operation and maintenance efficiency, and achieve more automated optical network operation and maintenance.

[0059] This step is performed at the data application layer.

[0060] In some embodiments, after generating the optical network performance optimization solution based on the association relationship and predicting the potential fault type, possible fault occurrence time and fault occurrence location of the optical network by the expert system and / or the fault prediction model, the method further includes: Generate an operation and maintenance report of the optical network regularly and present the operation and maintenance report in a visual manner, wherein the content of the operation and maintenance report includes but is not limited to a summary of the operation status of the optical network, a statistical analysis of faults, and a trend of performance indicators.

[0061] Among them, a variety of visual charts are designed to display the analysis results of optical network operation and maintenance data, including but not limited to line charts to show the trend of optical power changes over time, bar charts to compare the traffic size of different equipment ports, pie charts to analyze the proportion of different fault types, and topology diagrams to intuitively present the structure of the optical network and the connection relationship of equipment. Through intuitive visualization, operation and maintenance personnel can intuitively and quickly understand the operating status, performance indicators and existing problems of the optical network, providing a basis for formulating long-term operation and maintenance strategies.

[0062] In addition, the embodiment of the present application has a dynamic visualization function, which can update the data in the visualization chart in real time to reflect the current state changes of the optical network. For example, in the network topology diagram, when a certain optical link fails, the corresponding link color will immediately turn red and display the fault alarm information, which is convenient for operation and maintenance personnel to find and locate the fault in time.

[0063] Among them, regularly generated operation and maintenance reports can be generated through automated scripts and sent to operation and maintenance managers in HTML format or PDF format.

[0064] Among them, the embodiments of the present application design the visualization layout and interaction of optical network operation and maintenance data, including specific visualization chart combination methods, interactive operation logic (such as zooming, panning and device information viewing of network topology diagrams through gesture operations) and data display animation effects (such as fault propagation animation display), which can greatly improve the operation and maintenance personnel's understanding of the optical network status and operational efficiency (such as shortening the fault location time by more than 60%).

[0065] Among them, the embodiments of the present application can also adopt innovative applications of optical network operation and maintenance data visualization and interaction based on virtual reality (VR) or augmented reality (AR) technology, such as building a virtual three-dimensional model of optical network equipment through VR technology. Operation and maintenance personnel can intuitively view the equipment status, perform troubleshooting and network optimization operations in a virtual environment, providing a new interactive experience and technical means for optical network operation and maintenance.

[0066] In some embodiments, after periodically generating an operation and maintenance report of the optical network and presenting the operation and maintenance report in a visual manner, the method further includes: In response to data query, analysis task customization and model parameter adjustment operations performed by operation and maintenance personnel through a visual interactive interface, the corresponding data is queried, customized tasks are analyzed, and corresponding model parameters are adjusted.

[0067] Among them, a user-friendly interactive operation interface is built in advance, allowing operation and maintenance personnel to easily perform data query, analysis task customization, model parameter adjustment and other operations. For example, a simple query interface is provided, and operation and maintenance personnel can enter conditions such as equipment ID and time range to quickly obtain corresponding operation and maintenance data; in the fault diagnosis and performance optimization analysis module, a parameter adjustment interface is provided, and operation and maintenance personnel can adjust the model threshold, algorithm parameters, etc. according to actual experience and needs to improve the accuracy and applicability of the analysis results.

[0068] Specifically, when the operation and maintenance personnel operate on the interactive interface, the data analysis engine can be triggered in time to perform corresponding calculations and analysis, and the results can be fed back to the visualization interface. For example, when the operation and maintenance personnel click on a device in the topology map, the interface can immediately display the detailed performance data and historical fault records of the device, and display the probability and time range of possible future failures of the device through the fault prediction information provided by the data analysis engine.

[0069] In some embodiments, after generating the optical network performance optimization solution based on the association relationship, the method further includes: Determining whether it is necessary to notify the operation and maintenance personnel to optimize the performance of the optical network according to the optical network performance optimization plan according to the content of the optical network performance optimization plan; If not, the optical network performance is optimized according to the optical network performance optimization solution through the developed intelligent operation and maintenance assistant.

[0070] Among them, the intelligent operation and maintenance assistant is based on artificial intelligence technology, which can automatically analyze performance optimization suggestions and automatically perform some simple operation and maintenance operations according to preset strategies and rules, such as automatically adjusting the parameters of the optical link and switching backup equipment. Only when encountering complex situations or requiring manual intervention, the operation and maintenance personnel will be notified to handle them. This can greatly reduce the workload of the operation and maintenance personnel, improve the operation and maintenance efficiency, and achieve more automated optical network operation and maintenance.

[0071] If it is determined that the operation and maintenance personnel need to be notified to optimize the performance of the optical network according to the optical network performance optimization plan, the operation and maintenance personnel are notified to optimize the performance of the optical network according to the optical network performance optimization plan. In terms of performance optimization, the operation and maintenance personnel adjust the configuration of the optical network according to the performance optimization plan provided by the data application layer, such as modifying the routing of the optical link, adjusting the parameters of the device port, etc.

[0072] In some embodiments, after the standardization of all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format, the process further includes: All data obtained after the standardization process are input into a pre-trained fault diagnosis model, and the type of optical network fault and the location where the fault occurs are determined by the fault diagnosis model.

[0073] Among them, the optical network fault diagnosis model is constructed based on machine learning algorithms, such as support vector machines (SVM), decision trees (DT) or neural networks (such as convolutional neural networks CNN or recurrent neural networks RNN in deep learning) and other algorithms, to train the collected optical network operation and maintenance data and identify the characteristic patterns corresponding to different fault types. By analyzing the correlation between various data features such as abnormal changes in optical power, sudden increases in bit error rate, and equipment alarm information, a fault diagnosis model is established, which can quickly and accurately determine the type of optical network fault (such as fiber breakage, optical module failure, equipment hardware failure, etc.) and fault location (such as specific optical fiber links, equipment ports, etc.).

[0074] Among them, the optical network fault prediction model architecture based on deep learning in the embodiment of the present application includes a specific neural network layer design (such as a convolutional layer design for optical power data, a recurrent layer design for time series data), model training algorithm optimization (such as the use of adaptive learning rate adjustment, gradient clipping and other technologies) and a model evaluation index system (such as comprehensive consideration of accuracy, recall rate, F1 value and false alarm rate and other indicators), which can achieve high-precision diagnosis (accuracy of more than 95%) and early prediction (advance warning time of more than 2 hours) of optical network faults.

[0075] This application has the following advantages: (1) Powerful data processing capabilities Fast data processing speed: When processing massive optical network operation and maintenance data, the data processing platform of this technology adopts a distributed computing architecture, and the data processing speed is more than 10 times faster than the traditional single-machine processing method. For example, to analyze 10TB of optical network operation and maintenance data generated in a month, the traditional method takes more than 50 hours, while this technology only takes about 5 hours to complete a series of operations such as data cleaning, analysis and storage, meeting the requirements of optical network operation and maintenance for data real-time and timeliness, and can provide timely support for operation and maintenance decisions.

[0076] Strong data integration capability: This technology can seamlessly integrate multi-source heterogeneous data from different manufacturers, different types of optical network equipment, and network management systems, integrating more than 20 types of data, including equipment logs, performance indicators, alarm information, topology, etc. In a complex hybrid optical network environment, it successfully integrated data from optical transmission equipment from 5 different manufacturers, 3 different types of optical switching equipment, and 2 sets of network management systems, and built a unified optical network operation and maintenance data view, laying the foundation for comprehensive and in-depth data analysis, avoiding analysis difficulties and information loss caused by scattered data and inconsistent formats.

[0077] (2) Advanced algorithm model High accuracy of fault prediction model: The fault prediction model built with deep learning algorithm has an accuracy of 95% and over 90% on the training set and test set respectively. Through feature learning of a large amount of optical network operation and maintenance data, the model can accurately capture complex feature relationships related to faults, such as abnormal fluctuations in optical power, bit error rate change trends, and alarm sequence patterns. In the fault prediction test of 1,000 optical network nodes, the model accurately predicted potential faults of more than 900 nodes, and the false alarm rate was controlled within 5%. Compared with traditional rule-based fault prediction models, it has significant advantages in accuracy and reliability.

[0078] High efficiency of performance optimization algorithm: The performance optimization algorithm based on data mining and intelligent optimization algorithm can find a near-optimal solution in a short time during the network configuration optimization process. For example, when optimizing the routing of a network with 200 optical links, the traditional trial and error method requires thousands of attempts and calculations, which takes several days, while the optimization algorithm of this technology only needs hundreds of calculations and can obtain the optimized routing solution within 2 hours, which significantly improves network performance, such as increasing bandwidth utilization by more than 30% and reducing network latency by about 25%.

[0079] (3) Good visualization and interactivity Visualization is intuitive and easy to understand: The visualization interface of this technology provides more than 10 different types of visualization charts, including real-time dynamic optical network topology, performance indicator trend charts, fault distribution heat maps, etc. Through intuitive graphical display, operation and maintenance personnel can quickly understand the overall operation status and key information of the optical network in a short time. During a network performance monitoring process, by viewing the bandwidth utilization trend chart on the visualization interface, operation and maintenance personnel can immediately find abnormal fluctuations in bandwidth utilization in a certain area during a specific period of time, and quickly locate the relevant optical links for further investigation. Compared with viewing traditional data reports, the efficiency of information acquisition has increased by more than 5 times.

[0080] Convenient and efficient interactive operations: Users can easily perform operations such as data query, analysis task customization, and model parameter adjustment through a visual interactive interface. For example, operation and maintenance personnel can customize data analysis tasks for a specific time period, specific area, or specific equipment by simply clicking and dragging the mouse on the interface, and the system can generate corresponding analysis reports within a few minutes. In terms of model parameter adjustment, users can flexibly adjust the threshold of the fault prediction model or the parameters of the performance optimization algorithm based on actual operation and maintenance experience and needs. After the adjustment, the system can immediately re-analyze and calculate according to the new parameters and provide real-time feedback on the results, which greatly improves the operation and maintenance personnel's control over the system and the flexibility of operation and maintenance work.

[0081] (4) Strong system scalability and compatibility Good system scalability: This technology adopts microservice architecture and modular design. The addition of new functional modules and the upgrade of existing modules have little impact on the overall performance of the system. Within two years after the system was launched, five new data analysis modules and three data acquisition interface expansion modules were added according to business needs and technological development. The system response time increased by only about 10%, and there was no need for large-scale re-architecting and code reconstruction of the entire system, which effectively reduced the cost and difficulty of system maintenance and upgrades, ensuring that the system can continue to adapt to the changing needs of optical network operation and maintenance.

[0082] Strong system compatibility: This technology has good compatibility with existing mainstream optical network operation and maintenance management systems and related standard protocols, and can achieve seamless docking with more than 90% of common optical network equipment and management systems. For example, in the process of integrating with a well-known manufacturer's network management system (NMS), only a small amount of configuration adjustment and interface adaptation is required to achieve two-way data transmission and sharing. The data analysis function and visualization interface of this technology can be directly called on the NMS platform, providing users with a unified and convenient operation and maintenance management experience, avoiding integration difficulties and functional limitations caused by system compatibility issues.

[0083] This application has the following effects: (1) Fault prediction and rapid processing Improved fault prediction accuracy: In a one-year actual optical network operation and maintenance test, after applying this technology, the fault prediction accuracy rate was greatly improved from 60% of the traditional method to more than 90%. Through retrospective analysis of more than 500 fault cases, it was found that this technology can accurately predict the occurrence of more than 450 faults, with an average advance warning time of 3 hours, while the traditional method can only warn about 30 minutes in advance. This allows the operation and maintenance team to have sufficient time for preventive maintenance and fault preparation, significantly reducing the number and duration of business interruptions caused by faults. For example, in a regional network with 1,000 optical network nodes, the number of business interruptions has been reduced from an average of 10 times per month to less than 3 times, and the duration of each interruption has been shortened from an average of 60 minutes to less than 20 minutes.

[0084] Accelerated fault location and diagnosis: Comparative tests show that the average time for fault location and diagnosis using this technology has been reduced from the original 2 hours to less than 30 minutes. Based on machine learning training of a large amount of fault data, the system can quickly analyze fault characteristics and match them with known fault patterns. In a typical optical link failure, the traditional method requires operation and maintenance personnel to check multiple devices and link links one by one, which takes about 150 minutes. However, this technology uses intelligent diagnosis to accurately determine the fault point as excessive loss at the connection point of a certain section of optical fiber within 20 minutes, and provides detailed fault cause analysis and repair suggestions, greatly improving the efficiency of fault repair.

[0085] (2) Network performance optimization Improved bandwidth utilization: After performance monitoring and optimization practices of multiple optical networks, this technology has increased network bandwidth utilization by an average of 35%. Through in-depth analysis of traffic data at different time periods and for different business types, the idle bandwidth resources in the network can be accurately identified and re-allocated reasonably. For example, in the optical network of a large enterprise, the bandwidth utilization in some areas was only about 40% during peak business hours, resulting in a large amount of bandwidth waste. After optimization by this technology, the peak bandwidth utilization rate was increased to more than 75%, effectively meeting the company's growing business data transmission needs. For example, the smooth operation of services such as high-definition video conferencing and large-scale data backup has been significantly improved.

[0086] Reduced network latency: In multiple metropolitan optical network application cases, the average network latency was reduced by 40%. Through comprehensive optimization of network topology, routing strategies, and equipment parameters, the path length and processing links of data transmission were reduced. Taking the optical network of an online game service provider as an example, the average latency of players before optimization was about 80 milliseconds, resulting in a poor gaming experience. After optimization using this technology, the average latency was reduced to less than 48 milliseconds, and the game operation response of players was faster, and the jamming phenomenon was significantly reduced, which greatly improved user satisfaction and business competitiveness.

[0087] (3) Reduced operation and maintenance costs Reduction of labor costs: After applying this technology, the daily operation and maintenance workload of a medium-sized optical network operation and maintenance team (30 people) was reduced by about 60% due to the realization of automatic fault handling and performance optimization functions. Most of the troubleshooting and network optimization tasks that originally required the collaboration of multiple people can now be completed automatically by the system or only require a small number of people to perform simple monitoring and intervention. This enables the operation and maintenance team to streamline staffing and reduce labor costs by about 40% without affecting the quality of operation and maintenance. The saved human resources can be invested in more innovative and strategic work, such as research and planning of new network technologies.

[0088] Equipment maintenance cost savings: Based on accurate fault prediction and equipment performance monitoring, preventive maintenance of equipment can be effectively implemented, and the number of emergency repairs and replacements caused by sudden equipment failures has been greatly reduced. In a network with 500 optical network devices, the annual average failure rate of equipment has been reduced from 15% to less than 5%, and the cost of equipment repair and replacement has been reduced by more than 50%. At the same time, by optimizing equipment operating parameters, the service life of the equipment has been extended, and the overall equipment replacement cycle is expected to be extended by 30%, further reducing long-term equipment procurement and maintenance costs.

[0089] (4) Decision support and strategic planning Reduced decision-making error rate: In the decision-making process of optical network expansion and upgrade projects, after applying the comprehensive data analysis and forecast reports provided by this technology, the decision-making error rate has been significantly reduced from the previous 20% to less than 5%. Through accurate analysis of historical operation and maintenance data, business growth trends, and network performance bottlenecks, a scientific basis is provided for decision-making. For example, in a network expansion decision, the expansion scale estimated by the traditional method based on experience deviated greatly from the actual demand, resulting in some resources being idle and wasted. However, this technology accurately predicted the business traffic growth and network performance requirements in the next two years, making the expansion scale decision more accurate and reasonable, avoiding unnecessary investment and waste of resources.

[0090] Improved accuracy of strategic planning: Through the mining and analysis of long-term optical network operation and maintenance big data, the accuracy of enterprises in optical network strategic planning has been greatly improved. For example, in the planning of new technology introduction and network architecture upgrade, it is possible to accurately predict the changes in business development demand for the network 1-2 years in advance, so that enterprises can adjust their strategic layout in a timely manner and reserve technology and resources in advance. In the past three years, due to the precise implementation of strategic planning, the company's market share in the optical network field has increased by 25%, and business revenue has increased by more than 30%, laying a solid foundation for the company's long-term development.

[0091] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present application, which will not be described one by one here.

[0092] In specific implementation, the present application is not limited by the execution order of the various steps described. If no conflict occurs, some steps can be performed in other orders or simultaneously.

[0093] As can be seen from the above, the optical network performance optimization solution acquisition and fault diagnosis method provided in the embodiment of the present application first collects the operation logs, performance indicators and alarm information of various devices in the optical network according to the pre-set data collection frequency, and collects the optical network topology data, network-level performance data, equipment configuration information and computer room environment data, and then cleans all the collected data to remove noise data, duplicate data and obviously erroneous data, and uses a data completion algorithm to reasonably fill in the missing data, and then standardizes all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format, and finally performs standardization processing. All the data obtained after processing are input into a pre-trained performance optimization model, a pre-built expert system and / or a pre-trained fault prediction model. The performance optimization model is used to analyze the correlation between different network parameters, and an optical network performance optimization plan is generated based on the correlation. The expert system and / or the fault prediction model is used to predict the potential fault type, possible fault occurrence time and fault location of the optical network, so as to not only obtain a better optical network performance optimization plan, but also improve the fault locating speed and positioning accuracy, thereby improving the operation and maintenance efficiency and reliability of the optical network, reducing the operation and maintenance difficulty and cost, and providing a strong guarantee for the stable operation of the optical network.

[0094] An embodiment of the present application also provides an optical network performance optimization solution acquisition and fault diagnosis device, which is integrated in an optical network performance optimization solution acquisition and fault diagnosis system. The device includes a processor and a memory, and a computer program is stored in the memory. The processor calls the computer program stored in the memory to execute the optical network performance optimization solution acquisition and fault diagnosis method described in any of the above embodiments.

[0095] See also Figure 2 , Figure 2A structural diagram of an optical network performance optimization solution acquisition and fault diagnosis device provided in an embodiment of the present application. The device 30 of the optical network performance optimization solution acquisition and fault diagnosis method includes a memory 120, one or more processors 180, and one or more applications, wherein the one or more applications are stored in the memory 120 and configured to be executed by the processor 180; the memory 120 can be used to store applications and data. The application stored in the memory 120 contains executable code. The application can be composed of various functional modules. The processor 180 executes various functional applications and data processing by running the application stored in the memory 120. In addition, the memory 120 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage device. Accordingly, the memory 120 may also include a memory controller to provide the processor 180 with access to the memory 120.

[0096] The processor 180 is the control center of the device, and uses various interfaces and lines to connect various parts of the entire terminal, and executes various functions of the device and processes data by running or executing applications stored in the memory 120, and calling data stored in the memory 120, so as to monitor the device as a whole. Optionally, the processor 180 may include one or more processing cores; preferably, the processor 180 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs.

[0097] Specifically in this embodiment, a computer program is stored in the memory 120 , and the processor 180 executes the optical network performance optimization solution acquisition and fault diagnosis method described in any of the above embodiments by calling the computer program stored in the memory 120 .

[0098] An embodiment of the present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program runs on a computer, the computer executes the optical network performance optimization solution acquisition and fault diagnosis method described in any of the above embodiments.

[0099] It should be noted that, for the optical network performance optimization solution acquisition and fault diagnosis method described in this application, ordinary testers in this field can understand that the entire or partial process of the optical network performance optimization solution acquisition and fault diagnosis method described in the embodiment of this application can be completed by controlling the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, such as stored in the memory of the optical network performance optimization solution acquisition and fault diagnosis system, and executed by at least one processor in the optical network performance optimization solution acquisition and fault diagnosis system, and the execution process may include the process of the embodiment of the optical network performance optimization solution acquisition and fault diagnosis method. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM, Read Only Memory), a random access memory (RAM, Random Access Memory), etc.

[0100] The above is a detailed introduction to the optical network performance optimization solution acquisition and fault diagnosis method, device and system provided by the embodiments of the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application; at the same time, for those skilled in the art, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for obtaining an optical network performance optimization solution and diagnosing faults, characterized in that: The method comprises: Collect the operation logs, performance indicators and alarm information of various devices in the optical network according to the pre-set data collection frequency, as well as collect the optical network topology data, network-level performance data, device configuration information and computer room environment data; Clean all collected data to remove noise data, duplicate data, and obviously erroneous data, and use data completion algorithms to reasonably fill in missing data; Standardize all the data obtained after cleaning to convert data of different formats and magnitudes into a unified standard format; All data obtained after the standardization process are input into a pre-trained performance optimization model, a pre-built expert system and / or a pre-trained fault prediction model, the correlation between different network parameters is analyzed by the performance optimization model, an optical network performance optimization plan is generated based on the correlation, and the potential fault type, possible fault occurrence time and fault occurrence location of the optical network are predicted by the expert system and / or the fault prediction model.

2. The optical network performance optimization solution acquisition and fault diagnosis method according to claim 1, characterized in that: After predicting the potential fault type, possible fault time and fault location of the optical network by the expert system and / or the fault prediction model, the method further includes: Generate corresponding fault handling suggestions, and send the potential fault type of the optical network, the time when the fault may occur, the location where the fault occurs, and the fault handling suggestions to the operation and maintenance personnel or a pre-developed intelligent operation and maintenance assistant for processing by the operation and maintenance personnel or the intelligent operation and maintenance assistant.

3. The optical network performance optimization solution acquisition and fault diagnosis method according to claim 1, characterized in that: After generating the optical network performance optimization scheme based on the association relationship, and predicting the potential fault type, possible fault occurrence time and fault occurrence location of the optical network by the expert system and / or the fault prediction model, the method further includes: Generate an operation and maintenance report of the optical network regularly and present the operation and maintenance report in a visual manner, wherein the content of the operation and maintenance report includes but is not limited to a summary of the operation status of the optical network, a statistical analysis of faults, and a trend of performance indicators.

4. The optical network performance optimization solution acquisition and fault diagnosis method according to claim 3, characterized in that: After the operation and maintenance report of the optical network is generated regularly and the operation and maintenance report is presented in a visual manner, the method further includes: In response to data query, analysis task customization and model parameter adjustment operations performed by operation and maintenance personnel through a visual interactive interface, the corresponding data is queried, customized tasks are analyzed, and corresponding model parameters are adjusted.

5. The optical network performance optimization solution acquisition and fault diagnosis method according to claim 1, characterized in that: After generating the optical network performance optimization scheme based on the association relationship, the method further includes: Determining whether it is necessary to notify the operation and maintenance personnel to optimize the performance of the optical network according to the optical network performance optimization plan according to the content of the optical network performance optimization plan; If not, the optical network performance is optimized according to the optical network performance optimization solution through the developed intelligent operation and maintenance assistant.

6. The optical network performance optimization solution acquisition and fault diagnosis method according to claim 1, characterized in that: After the above-mentioned standardization processing of all the data obtained after cleaning is performed to convert the data of different formats and magnitudes into a unified standard format, it also includes: All data obtained after standardization processing are stored in a memory database, a distributed storage system or a cloud storage service according to preset rules, and a data index structure is established to record the storage location, data format and data source of each data. When the distributed storage system is used for storage, the data to be stored is stored in a distributed manner on multiple storage nodes of the distributed storage system.

7. The optical network performance optimization solution acquisition and fault diagnosis method according to claim 1, characterized in that: After the above-mentioned standardization processing of all the data obtained after cleaning is performed to convert the data of different formats and magnitudes into a unified standard format, it also includes: All data obtained after the standardization process are input into a pre-trained fault diagnosis model, and the type of optical network fault and the location where the fault occurs are determined by the fault diagnosis model.

8. An optical network performance optimization solution acquisition and fault diagnosis device, characterized in that: The device includes a processor and a memory, wherein a computer program is stored in the memory, and the processor is used to execute the optical network performance optimization solution acquisition and fault diagnosis method as described in any one of claims 1 to 7 by calling the computer program stored in the memory.

9. An optical network performance optimization solution acquisition and fault diagnosis system, characterized in that: The optical network performance optimization solution acquisition and fault diagnosis system includes the optical network performance optimization solution acquisition and fault diagnosis device as described in claim 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program runs on a computer, the computer executes the optical network performance optimization solution acquisition and fault diagnosis method according to any one of claims 1 to 7.

Images