Visual management method and system for all-optical network

Abstract

The invention belongs to the technical field of all-optical network management, and discloses an all-optical network visual management method and system. The method comprises the following steps that: a cloud data center constructs an all-optical network digital twin model and an artificial intelligence model; the edge computing gateway is used for collecting real-time equipment operation monitoring data and extracting real-time equipment dynamic monitoring data; the edge computing gateway is used for performing equipment fault diagnosis and equipment management strategy generation by using an artificial intelligence model; the edge computing gateway encrypts and transmits the real-time equipment operation monitoring data, the real-time equipment dynamic monitoring data, the real-time equipment fault diagnosis result, the real-time equipment management strategy and the real-time alarm signal of the optical network information transmission equipment to the cloud data center; and the cloud data center is used for performing visual display by using an all-optical network digital twinborn model. According to the invention, the problems of lack of comprehensive real-time monitoring capability, insufficient fault diagnosis precision, low visualization degree and low data processing efficiency in the prior art are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of all-optical network management, and in particular relates to an all-optical network visualized management method and system. Background Art

[0002] All-Optical Network (AON) is a network based on optical fiber communication technology that can directly transmit and exchange optical signals without converting electrical signals in the network. An all-optical network is composed of several optical network information transmission devices. These devices work together to realize the functions of optical signal transmission, amplification, insertion and removal, cross-connection, wavelength conversion, etc. in the network, ensuring that data can be transmitted efficiently and over long distances through the optical fiber network without converting to electrical signals. All-optical network management technology is a necessary condition to ensure network efficiency, reliability and security, and is crucial for network operators and service providers.

[0003] The existing all-optical network management technology has the following defects:

[0004] 1) Lack of comprehensive real-time monitoring capabilities: Existing all-optical network management technologies can only provide limited equipment status monitoring and cannot fully cover the various operating conditions of optical network information transmission equipment, which makes it difficult for network managers to obtain comprehensive network status information;

[0005] 2) Insufficient fault diagnosis accuracy: Existing all-optical network management technologies are usually based on simple rules or threshold judgments when diagnosing equipment faults. The degree of intelligence leads to low fault diagnosis accuracy and makes it difficult to accurately identify the cause of faults in complex network environments.

[0006] 3) Low visualization: The visualization of existing all-optical network management technologies is usually simple and cannot intuitively display network topology, device status, and fault information, making it difficult for network managers to quickly understand and analyze network conditions.

[0007] 4) Low data processing efficiency: When faced with a large amount of data environment, the existing all-optical network management technology has weak data processing capabilities and insufficient computing resources, resulting in slow data processing and response speeds, which is not conducive to real-time network management. Summary of the invention

[0008] In order to solve the problems of lack of comprehensive real-time monitoring capability, insufficient fault diagnosis accuracy, low visualization and low data processing efficiency in the prior art, the present invention aims to provide an all-optical network visualization management method and system.

[0009] The technical solution adopted by the present invention is:

[0010] An all-optical network visualization management method comprises the following steps:

[0011] The cloud data center builds an all-optical network digital twin model, a device dynamic monitoring model, a device fault diagnosis model, and a device management strategy generation model, and deploys the device dynamic monitoring model, device fault diagnosis model, and device management strategy generation model to all edge computing gateways;

[0012] The edge computing gateway collects real-time equipment operation monitoring data of the optical network information transmission equipment, writes the real-time equipment operation monitoring data into the corresponding equipment dynamic monitoring model, and extracts the real-time equipment dynamic monitoring data of the equipment dynamic monitoring model;

[0013] The edge computing gateway uses the device fault diagnosis model to perform device fault diagnosis on the real-time device dynamic monitoring data, and uses the device management strategy generation model to generate device management strategies for the obtained real-time device fault diagnosis results, obtains the real-time device management strategy, and generates the corresponding real-time alarm signal;

[0014] The edge computing gateway uses encryption algorithms to encrypt and transmit the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies, and real-time alarm signals of optical network information transmission equipment to the cloud data center;

[0015] The cloud data center uses the all-optical network digital twin model to visualize the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies, and real-time alarm signals of optical network information transmission equipment.

[0016] Furthermore, the cloud data center constructs an all-optical network digital twin model, a device dynamic monitoring model, a device fault diagnosis model, and a device management strategy generation model, and deploys the device dynamic monitoring model, the device fault diagnosis model, and the device management strategy generation model to all edge computing gateways, including the following steps:

[0017] The cloud data center uses digital twin technology to build a digital twin model of the all-optical network based on the basic network information of the all-optical network and the basic equipment information of several optical network information transmission devices in the all-optical network;

[0018] According to the dynamic mechanism properties of optical network information transmission equipment and equipment operation monitoring indicators, a dynamic monitoring model for equipment is constructed, and the dynamic monitoring model for equipment is used to generate a number of historical dynamic monitoring data for equipment;

[0019] Extract the all-optical network topology diagram of the all-optical network digital twin model, and use the deep learning algorithm to build a device fault diagnosis model based on the all-optical network topology diagram and several historical equipment dynamic monitoring data, and generate several historical equipment fault diagnosis results;

[0020] Based on several historical equipment fault diagnosis results, an enhanced reinforcement learning algorithm is used to build an equipment management strategy generation model, and model metadata of the equipment dynamic monitoring model, equipment fault diagnosis model, and equipment management strategy generation model are extracted;

[0021] The all-optical network topology diagram and model metadata are sent to all edge computing gateways, and in the edge computing gateways, the model is reconstructed according to the model metadata to obtain a reconstructed device dynamic monitoring model, a reconstructed device fault diagnosis model, and a reconstructed device management strategy generation model.

[0022] Furthermore, the cloud data center uses the digital twin technology to construct a digital twin model of the all-optical network based on the basic network information of the all-optical network and the basic device information of several optical network information transmission devices in the all-optical network, including the following steps:

[0023] Collect basic network information of the all-optical network and basic equipment information of several optical network information transmission equipment included in the all-optical network;

[0024] According to the network topology information in the basic network information of the all-optical network, the device location and device number of each optical network information transmission device are confirmed, and a static model of the all-optical network topology is constructed;

[0025] According to the basic device information of each optical network information transmission device, a corresponding optical network information transmission device static model is constructed at the corresponding device position of the all-optical network topology static model;

[0026] Setting data flow and visualization interface for the static model of optical network information transmission equipment to obtain the dynamic model of optical network information transmission equipment;

[0027] By integrating the static model of the all-optical network topology and the dynamic models of several optical network information transmission equipment, we can obtain the digital twin model of the all-optical network.

[0028] Furthermore, the equipment dynamic monitoring model includes a working status dynamic monitoring function, a transmission efficiency dynamic monitoring function, an optical power change dynamic monitoring function, an optical signal quality dynamic monitoring function, a transmission rate dynamic monitoring function, a network delay dynamic monitoring function, and a network stability dynamic monitoring function;

[0029] The equipment fault diagnosis model is constructed based on the GCN-LSTM-Attention-MLP algorithm, and the equipment fault diagnosis model includes a graph structure feature extraction module constructed based on the GCN algorithm, a sequence feature extraction module constructed based on the LSTM algorithm, an attention weight module constructed based on the Attention mechanism, and an equipment fault diagnosis module constructed based on the MLP algorithm. The graph structure feature extraction module and the sequence feature extraction module are both connected to the attention weight module, and the attention weight module is connected to the equipment fault diagnosis module;

[0030] The device management strategy generation model is constructed based on the MPO-DQN algorithm, and the device management strategy generation model includes a meta-strategy optimization module constructed based on the MPO algorithm and a reinforcement learning module constructed based on the DQN algorithm. The reinforcement learning module is provided with an intelligent agent, a deep Q network and an experience replay pool. The intelligent agent is respectively connected to the deep Q network, the experience replay pool and the meta-strategy optimization module, and the meta-strategy optimization module is connected to the experience replay pool.

[0031] Furthermore, the all-optical network topology diagram of the all-optical network digital twin model is extracted, and based on the all-optical network topology diagram and several historical equipment dynamic monitoring data, a deep learning algorithm is used to build an equipment fault diagnosis model, and several historical equipment fault diagnosis results are generated, including the following steps:

[0032] Extract the all-optical network topology diagram of the all-optical network digital twin model, and associate the all-optical network topology diagram with each historical device dynamic monitoring data to obtain several model training samples;

[0033] Use the GCN-LSTM-Attention-MLP algorithm to build an initial equipment fault diagnosis model, and input several model training samples into the initial equipment fault diagnosis model;

[0034] According to a number of model training samples, the initial equipment fault diagnosis model is optimized and trained to obtain the final equipment fault diagnosis model, and a number of historical equipment fault diagnosis results are generated.

[0035] Furthermore, based on several historical equipment fault diagnosis results, an enhanced reinforcement learning algorithm is used to construct an equipment management strategy generation model, and model metadata of the equipment dynamic monitoring model, the equipment fault diagnosis model, and the equipment management strategy generation model are extracted, including the following steps:

[0036] Collecting a number of preset all-optical network management targets, and setting the preset all-optical network management targets as optimization scenarios for meta-strategy optimization, to obtain a number of optimization scenarios;

[0037] Use the MPO algorithm to build an initial meta-strategy optimization module, and use several historical equipment fault diagnosis results in several optimization scenarios to train the initial meta-strategy optimization module to obtain the final meta-strategy optimization module and generate several historical meta-strategy optimization experiences;

[0038] The device management strategy generation problem is used as the simulation environment of the DQN algorithm. The state space of the DQN algorithm is defined according to several historical device fault diagnosis results, and the action space of the DQN algorithm is defined according to several preset device management actions.

[0039] According to the impact of the preset equipment management actions on the historical equipment fault diagnosis status, define the reward function of the DQN algorithm, use the final meta-strategy optimization module to generate the initial deep Q network, and build the initial reinforcement learning module based on the initial deep Q network;

[0040] Based on the state space, action space and reward function, the initial reinforcement learning module is trained using several historical equipment fault diagnosis results to obtain the final reinforcement learning module and generate several historical reinforcement learning experiences;

[0041] Store several historical meta-strategy optimization experiences and several historical reinforcement learning experiences in the experience replay pool of the final reinforcement learning module, and integrate the final meta-strategy optimization module and the final reinforcement learning module to obtain a device management strategy generation model;

[0042] Extract model metadata of the equipment dynamic monitoring model, equipment fault diagnosis model, and equipment management strategy generation model.

[0043] Furthermore, the edge computing gateway uses the device fault diagnosis model to perform device fault diagnosis on the real-time device dynamic monitoring data, and uses the device management strategy generation model to generate a device management strategy for the obtained real-time device fault diagnosis result, obtains the real-time device management strategy, and generates a corresponding real-time alarm signal, including the following steps:

[0044] The edge computing gateway inputs the all-optical network topology diagram and real-time device dynamic monitoring data into the device fault diagnosis model;

[0045] Use the graph structure feature extraction module to extract the real-time graph structure features of the all-optical network topology graph, and use the sequence feature extraction module to extract the real-time sequence features of the real-time device dynamic monitoring data;

[0046] According to the preset attention weight value, the attention weight module is used to perform weighted fusion on the real-time graph structure features and the real-time sequence features to obtain the real-time weighted fusion features;

[0047] According to the real-time weighted fusion features, the equipment fault diagnosis module is used to diagnose equipment faults, obtain real-time equipment fault diagnosis results, and input them into the equipment management strategy generation model;

[0048] Collect real-time all-optical network management goals, and extract corresponding real-time meta-strategy optimization experience and real-time reinforcement learning experience from the experience replay pool based on the real-time all-optical network management goals;

[0049] Based on the real-time equipment fault diagnosis results, the real-time meta-strategy optimization experience and the real-time reinforcement learning experience, the equipment management strategy generation model is updated to obtain an updated reinforcement learning module;

[0050] Using the updated reinforcement learning module, the device management strategy is generated for the obtained real-time device fault diagnosis results, the real-time device management strategy is obtained, and the corresponding real-time alarm signal is generated.

[0051] Furthermore, the edge computing gateway uses an encryption algorithm to encrypt and transmit the real-time device operation monitoring data, real-time device dynamic monitoring data, real-time device fault diagnosis results, real-time device management strategies, and real-time alarm signals of the optical network information transmission device to the cloud data center, including the following steps:

[0052] The edge computing gateway compresses the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies and real-time alarm signals of the optical network information transmission equipment to obtain a real-time compressed package;

[0053] According to the private key of the edge computing gateway, an asymmetric encryption algorithm is used to encrypt the real-time compressed package to obtain the encrypted real-time compressed package, and according to the signature information of the edge computing gateway, the encrypted real-time compressed package is signed to obtain the corresponding real-time signature data;

[0054] The device number of the optical network information transmission device is extracted, the device number is used as the real-time additional information, and the encrypted real-time compressed package, real-time signature data and real-time additional information are transmitted to the cloud data center.

[0055] Furthermore, the cloud data center uses the all-optical network digital twin model to visualize the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies, and real-time alarm signals of the optical network information transmission equipment, including the following steps:

[0056] The cloud data center performs digital identity authentication on the real-time signature data. After the authentication is passed, the encrypted real-time compressed package is decrypted using an asymmetric encryption algorithm based on the public key of the edge computing gateway to obtain the decrypted real-time compressed package.

[0057] Decompress the decrypted real-time compressed package to obtain decompressed real-time equipment operation monitoring data, decompressed real-time equipment dynamic monitoring data, decompressed real-time equipment fault diagnosis results, decompressed real-time equipment management strategy and decompressed real-time alarm signal;

[0058] According to the real-time additional information, locate the dynamic model of the target optical network information transmission equipment in the all-optical network digital twin model, and write the decompressed real-time equipment operation monitoring data, decompressed real-time equipment dynamic monitoring data, decompressed real-time equipment fault diagnosis results, decompressed real-time equipment management strategy and decompressed real-time alarm signal into the corresponding target data stream;

[0059] Based on the target data flow, using the visualization interface of the dynamic model of the target optical network information transmission equipment, the decompressed real-time equipment operation monitoring data, the decompressed real-time equipment dynamic monitoring data, the decompressed real-time equipment fault diagnosis results, the decompressed real-time equipment management strategy and the decompressed real-time alarm signal are visualized.

[0060] An all-optical network visualization management system is used to implement an all-optical network visualization management method. The all-optical network includes a plurality of optical network information transmission devices. The system includes a cloud data center, a plurality of edge computing gateways and a plurality of data acquisition devices. The cloud data center is respectively communicated with the edge computing gateways of a plurality of management areas, each edge computing gateway is respectively communicated with a plurality of data acquisition devices within the communication range, and each data acquisition device is communicated with an optical network information transmission device in a one-to-one correspondence.

[0061] The beneficial effects of the present invention are:

[0062] The present invention provides an all-optical network visualization management method and system, which ensures that network management personnel can obtain comprehensive network status information in real time by comprehensively collecting real-time operation monitoring data of optical network information transmission equipment, thereby realizing comprehensive real-time monitoring of the all-optical network; through the equipment fault diagnosis model constructed by applying artificial intelligence technology, it can analyze the real-time equipment dynamic monitoring data, realize high-precision fault diagnosis, and accurately identify the cause of faults in complex network environments; through the equipment management strategy generation model constructed by applying artificial intelligence technology, it can automatically generate management strategies according to real-time network conditions, improve the automation and intelligence level of network management, and ensure network stability and efficiency; the all-optical network visualization management method and system, which are constructed by combining digital twin technology, can realize high-precision fault diagnosis and accurately identify the cause of faults in complex network environments; through the equipment management strategy generation model constructed by applying artificial intelligence technology, it can automatically generate management strategies according to real-time network conditions, improve the automation and intelligence level of network management, and ensure network stability and efficiency; The digital twin model of the optical network can map and simulate the actual network environment in real time, improve the predictive and intelligent level of network management, and can intuitively display the network topology, equipment status, fault information, etc., to help network managers quickly understand and analyze network conditions and improve visualization. Encryption and authentication measures are adopted in the process of data transmission and processing to effectively prevent data leakage and illegal attacks, and ensure the security of all-optical network data. The edge-cloud collaborative mechanism is collected, and monitoring and artificial intelligence models are deployed on the edge computing gateway to make full use of the capabilities of edge computing, improve data processing efficiency, reduce the computing resource requirements for cloud data centers, realize rapid data processing and response, and improve the efficiency of real-time management.

[0063] Other beneficial effects of the present invention will be further described in the specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is a flowchart of the all-optical network visualization management method in the present invention.

[0065] Figure 2 It is a structural block diagram of the all-optical network visualization management system in the present invention. DETAILED DESCRIPTION

[0066] The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments.

[0067] Embodiment 1:

[0068] like Figure 1 As shown, this embodiment provides an all-optical network visualization management method, including the following steps:

[0069] S1: Cloud data center, builds an all-optical network digital twin model, a device dynamic monitoring model, a device fault diagnosis model, and a device management strategy generation model, and deploys the device dynamic monitoring model, device fault diagnosis model, and device management strategy generation model to all edge computing gateways, including the following steps:

[0070] S1-1: The cloud data center uses the digital twin technology to build a digital twin model of the all-optical network based on the basic network information of the all-optical network and the basic device information of several optical network information transmission devices in the all-optical network, including the following steps:

[0071] S1-1-1: Collect basic network information of the all-optical network and basic equipment information of several optical network information transmission equipment included in the all-optical network;

[0072] S1-1-2: According to the network topology information in the basic network information of the all-optical network, confirm the device location and device number of each optical network information transmission device, and build a static model of the all-optical network topology;

[0073] S1-1-3: According to the basic device information of each optical network information transmission device, a corresponding static model of the optical network information transmission device is constructed at the corresponding device position of the all-optical network topology static model;

[0074] S1-1-4: Setting data flow and visualization interface for the static model of optical network information transmission equipment to obtain a dynamic model of optical network information transmission equipment;

[0075] S1-1-5: Integrate the static model of the all-optical network topology and the dynamic models of several optical network information transmission devices to obtain the digital twin model of the all-optical network;

[0076] S1-2: Based on the dynamic mechanism properties of optical network information transmission equipment and equipment operation monitoring indicators, a dynamic monitoring model for equipment is constructed, and the dynamic monitoring model for equipment is used to generate a number of historical dynamic monitoring data for equipment;

[0077] The equipment dynamic monitoring model includes a working status dynamic monitoring function, a transmission efficiency dynamic monitoring function, an optical power change dynamic monitoring function, an optical signal quality dynamic monitoring function, a transmission rate dynamic monitoring function, a network delay dynamic monitoring function, and a network stability dynamic monitoring function;

[0078] The formula is:

[0079] S(t)=F(T(t),I b (t),P in (t),...)

[0080] Where S(t) is the working state dynamic monitoring function; F(*) is the working state function; T(t),I b (t),P in (t) are the temperature data, bias current and device input power in the device operation monitoring data; t is the data acquisition time indicator;

[0081]

[0082] Where η(t) is the dynamic monitoring function of transmission efficiency; P out (t) is the equipment output power in the equipment operation monitoring data;

[0083]

[0084] Where, P(t) is the dynamic monitoring function of optical power change; I(t) and ε(t) are the device input current and external interference in the device operation monitoring data; α and β are the attenuation coefficient and gain coefficient respectively;

[0085] Q(t)=G(E(t),OSNR(t),CD(t),...)

[0086] Where, Q(t) is the dynamic monitoring function of optical signal quality; G(*) is the quality function; E(t), OSNR(t), CD(t) are the eye diagram, signal-to-noise ratio and dispersion of the non-return-to-zero signal in the equipment operation monitoring data respectively;

[0087] R(t)=B·log2(1+SNR(t))

[0088] Where R(t) is the dynamic monitoring function of the transmission rate; B is the channel bandwidth; SNR(t) is the transmission signal-to-noise ratio;

[0089] D(t)=T send +T propagation +T process

[0090] Where D(t) is the network delay dynamic monitoring function; T send is the sending time; T propagation is the signal propagation time; T process is the signal processing time;

[0091] The formula of the dynamic sub-model of network stability index is:

[0092]

[0093] Where PS(t) is the network stability dynamic monitoring function; N c is the number of interruptions; N s The total number of connection attempts;

[0094] S1-3: Extract the all-optical network topology diagram of the all-optical network digital twin model, and use the deep learning algorithm to build a device fault diagnosis model based on the all-optical network topology diagram and several historical equipment dynamic monitoring data, and generate several historical equipment fault diagnosis results;

[0095] The equipment fault diagnosis model is constructed based on the graph convolutional network (GCN)-long short-term memory network (LSTM)-attention-multilayer perceptron (MLP) algorithm, and the equipment fault diagnosis model includes a graph structure feature extraction module constructed based on the GCN algorithm, a sequence feature extraction module constructed based on the LSTM algorithm, an attention weight module constructed based on the Attention mechanism, and an equipment fault diagnosis module constructed based on the MLP algorithm. The graph structure feature extraction module and the sequence feature extraction module are both connected to the attention weight module, and the attention weight module is connected to the equipment fault diagnosis module;

[0096] The graph structure feature extraction module propagates features on the all-optical network topology graph through operations similar to convolution, extracts node features of several nodes in the all-optical network topology graph, edge features of the positional relationship between nodes, and forms complex, high-dimensional global graph structure features, where nodes are device information of the static model of optical network information transmission equipment, and edges are topological connection relationships between optical network information transmission equipment in the all-optical network; the sequence feature extraction module is used to extract sequence features of dynamic monitoring data of equipment; the attention weight module performs weighted fusion of multimodal features, enhances the model's attention to important features, and improves the accuracy and efficiency of model prediction; the equipment fault diagnosis module is used to perform equipment fault diagnosis and prediction based on weighted fusion features, and analyze the fault conditions of optical network information transmission equipment;

[0097] By combining the all-optical network topology diagram, it is possible to accurately analyze the deep impact of optical network information transmission equipment at different locations and in different connection relationships in the all-optical network, thus improving the accuracy of equipment fault diagnosis.

[0098] The all-optical network topology diagram of the all-optical network digital twin model is extracted, and based on the all-optical network topology diagram and several historical equipment dynamic monitoring data, a deep learning algorithm is used to build an equipment fault diagnosis model and generate several historical equipment fault diagnosis results, including the following steps:

[0099] S1-3-1: extract the all-optical network topology diagram of the all-optical network digital twin model, and associate the all-optical network topology diagram with each historical device dynamic monitoring data to obtain several model training samples;

[0100] S1-3-2: Use the GCN-LSTM-Attention-MLP algorithm to build an initial equipment fault diagnosis model, and input several model training samples into the initial equipment fault diagnosis model;

[0101] S1-3-3: Optimize and train the initial equipment fault diagnosis model based on several model training samples to obtain the final equipment fault diagnosis model and generate several historical equipment fault diagnosis results;

[0102] Equipment fault diagnosis results include abnormal situation diagnosis results (transmission anomaly, data anomaly, equipment anomaly, etc.), fault type diagnosis results (hardware fault: such as line damage, module damage, power supply problem, cooling system failure, etc.; software fault: such as system crash, application error, configuration error, software compatibility problem, etc.; network fault: such as connection interruption, network configuration error, routing problem, bandwidth bottleneck, etc.), fault level diagnosis results (serious fault: affects the operation of the entire system and needs to be handled immediately; general fault: affects some functions and can be handled within a certain period of time; minor fault: has little impact on system operation and can be handled in routine maintenance), fault location diagnosis results (point out the specific location where the fault occurs, such as a certain optical network signal The equipment number and location of the information transmission equipment or optical fiber segment), fault cause diagnosis results (direct cause: direct factors that lead to the fault, such as hardware damage, software bugs, human operation errors, etc.; indirect cause: deeper factors that may lead to the fault, such as design defects, environmental factors, improper maintenance, etc.), fault impact diagnosis results (impact scope: business scope and service objects affected by the fault; duration: the length of time from the fault to recovery; economic loss: direct and indirect economic losses that may be caused by the fault) and diagnosis confidence results (high confidence: the diagnosis result is very reliable; medium confidence: the diagnosis result has a certain degree of reliability, but may need further verification; low confidence: the diagnosis result is uncertain and requires more information and testing to confirm), etc.;

[0103] S1-4: Based on several historical equipment fault diagnosis results, an enhanced reinforcement learning algorithm is used to build an equipment management strategy generation model, and model metadata of the equipment dynamic monitoring model, equipment fault diagnosis model, and equipment management strategy generation model are extracted;

[0104] The device management policy generation model is constructed based on the Meta-Policy Optimization (MPO)-Deep Q Network (DQN) algorithm, and the device management policy generation model includes a meta-policy optimization module constructed based on the MPO algorithm and a reinforcement learning module constructed based on the DQN algorithm. The reinforcement learning module is provided with an intelligent agent, a deep Q network and an experience replay pool. The intelligent agent is respectively connected to the deep Q network, the experience replay pool and the meta-policy optimization module, and the meta-policy optimization module is connected to the experience replay pool.

[0105] The meta-strategy optimization module is used to optimize the deep Q network parameters in the reinforcement learning module so that these parameters can quickly adapt to new all-optical network management goals and unseen all-optical network management goals, improve the generalization ability of the model, and update the deep Q network based on previous learning experience even under unseen equipment fault diagnosis results; through meta-strategy optimization, the DQN algorithm can converge faster, reduce the amount of data and time required for training, and significantly improve learning efficiency; the reinforcement learning module uses previous learning experience to output and select appropriate equipment management strategies based on real-time equipment fault diagnosis results;

[0106] Based on several historical equipment fault diagnosis results, an enhanced reinforcement learning algorithm is used to build an equipment management strategy generation model, and the model metadata of the equipment dynamic monitoring model, equipment fault diagnosis model, and equipment management strategy generation model are extracted, including the following steps:

[0107] S1-4-1: Collecting a number of preset all-optical network management targets, and setting the preset all-optical network management targets as optimization scenarios for meta-strategy optimization, to obtain a number of optimization scenarios;

[0108] The goals of all-optical network management are to optimize network performance, optimize network reliability, maximize energy efficiency, and minimize network costs;

[0109] S1-4-2: Use the MPO algorithm to build an initial meta-strategy optimization module, and use several historical equipment fault diagnosis results in several optimization scenarios to train the initial meta-strategy optimization module to obtain the final meta-strategy optimization module and generate several historical meta-strategy optimization experiences;

[0110] S1-4-3: Use the device management strategy generation problem as the simulation environment of the DQN algorithm, define the state space of the DQN algorithm based on several historical device fault diagnosis results, and define the action space of the DQN algorithm based on several preset device management actions;

[0111] Device management actions include automatic repair actions (restart device: for some transient faults that can be resolved by restarting; reset configuration: restore to the default or known good configuration state; update firmware / software: install the latest firmware or software updates to fix known defects), resource allocation actions (load balancing: redistribute network traffic to avoid overloaded devices; route adjustment: change the packet routing path to bypass the faulty node; bandwidth allocation: adjust network bandwidth allocation according to demand), fault isolation actions (switch to backup device: use redundant device to take over the work of the faulty device; disconnect faulty node: remove the faulty device from the network to protect other parts of the network), early warning and notification actions (send alarm: send real-time alarm to network administrators , including fault information and recommended repair measures; update status indication: update the device status on the management interface so that administrators can quickly identify problems), maintenance and optimization actions (arrange maintenance: arrange on-site or remote maintenance according to the severity of the fault and the importance of the device; performance optimization: adjust device parameters to optimize performance, such as adjusting cache size, queue length, etc.), data collection and analysis actions (increase monitoring frequency: increase data collection frequency for suspected or confirmed faulty devices; in-depth analysis: conduct in-depth analysis of detailed logs and performance data of faulty devices) and security-related actions (start security protocol: start security protocol when potential security threats are detected; block attack: take measures to block the attack source for confirmed security attacks);

[0112] S1-4-4: Based on the impact of preset equipment management actions on the historical equipment fault diagnosis status, define the reward function of the DQN algorithm, use the final meta-strategy optimization module to generate the initial deep Q network, and build the initial reinforcement learning module based on the initial deep Q network;

[0113] S1-4-5: Based on the state space, action space and reward function, use several historical equipment fault diagnosis results to train the initial reinforcement learning module, obtain the final reinforcement learning module, and generate several historical reinforcement learning experiences;

[0114] S1-4-6: storing several historical meta-strategy optimization experiences and several historical reinforcement learning experiences in the experience replay pool of the final reinforcement learning module, and integrating the final meta-strategy optimization module and the final reinforcement learning module to obtain a device management strategy generation model;

[0115] S1-4-7: Extract model metadata of equipment dynamic monitoring model, equipment fault diagnosis model and equipment management strategy generation model;

[0116] S1-5: Send the all-optical network topology diagram and model metadata to all edge computing gateways, and reconstruct the model in the edge computing gateway according to the model metadata to obtain a reconstructed device dynamic monitoring model, a reconstructed device fault diagnosis model, and a reconstructed device management strategy generation model;

[0117] S2: The edge computing gateway collects the real-time device operation monitoring data of the optical network information transmission device, writes the real-time device operation monitoring data into the corresponding device dynamic monitoring model, and extracts the real-time device dynamic monitoring data of the device dynamic monitoring model, including the following steps:

[0118] S2-1: Edge computing gateway, receiving initial real-time equipment operation monitoring data of optical network information transmission equipment sent by data acquisition devices within the communication range;

[0119] S2-2: preprocessing the initial real-time equipment operation monitoring data to obtain preprocessed real-time equipment operation monitoring data, and writing the preprocessed real-time equipment operation monitoring data into a corresponding equipment dynamic monitoring model;

[0120] S2-2: when the data upload cycle arrives, extract the real-time device dynamic monitoring data of the device dynamic monitoring model in the current state; the real-time device dynamic monitoring data includes the real-time working state dynamic monitoring data of the optical network information transmission equipment, the real-time transmission efficiency dynamic monitoring data, the real-time optical power change dynamic monitoring data, the real-time optical signal quality dynamic monitoring data, the real-time transmission rate dynamic monitoring data, the real-time network delay dynamic monitoring data and the real-time network stability dynamic monitoring data;

[0121] S3: The edge computing gateway uses the device fault diagnosis model to perform device fault diagnosis on the real-time device dynamic monitoring data, and uses the device management strategy generation model to generate a device management strategy for the obtained real-time device fault diagnosis results, obtains the real-time device management strategy, and generates a corresponding real-time alarm signal, including the following steps:

[0122] S3-1: Edge computing gateway, which inputs the all-optical network topology diagram and real-time device dynamic monitoring data into the device fault diagnosis model;

[0123] S3-2: Use the graph structure feature extraction module to extract the real-time graph structure features of the all-optical network topology graph, and use the sequence feature extraction module to extract the real-time sequence features of the real-time device dynamic monitoring data;

[0124] S3-3: According to the preset attention weight value, use the attention weight module to perform weighted fusion on the real-time graph structure features and the real-time sequence features to obtain the real-time weighted fusion features;

[0125] S3-4: Based on the real-time weighted fusion features, use the equipment fault diagnosis module to perform equipment fault diagnosis, obtain real-time equipment fault diagnosis results, and input them into the equipment management strategy generation model;

[0126] S3-5: Collect real-time all-optical network management goals, and extract corresponding real-time meta-strategy optimization experience and real-time reinforcement learning experience from the experience playback pool according to the real-time all-optical network management goals;

[0127] S3-6: Based on the real-time equipment fault diagnosis results, the real-time meta-strategy optimization experience and the real-time reinforcement learning experience, the equipment management strategy generation model is updated to obtain an updated reinforcement learning module, including the following steps:

[0128] S3-6-1: Based on the real-time meta-policy optimization experience, the meta-policy optimization module is used to update the deep Q network of the reinforcement learning module in the device management policy generation model to obtain an updated deep Q network;

[0129] S3-6-2: Generate a number of real-time device management actions based on the real-time reinforcement learning experience, and update the action space of the reinforcement learning module based on the real-time device management actions to obtain an updated action space A' = [a'1, ..., a' j" ,...,a' I ], where a' j" is the updated j-th "action value", j" is the action indicator, and I is the total number of action space dimensions;

[0130] S3-6-3: Perform state analysis on the real-time equipment fault diagnosis results to obtain several real-time equipment fault diagnosis states, and update the state space of the reinforcement learning module according to the several real-time equipment fault diagnosis states to obtain the updated state space S'=[s'1,...,s' i" ,...,s' I' ], where s' i" is the updated i'th state value, i' is the state indicator, and I' is the total number of state space dimensions;

[0131] S3-6-4: Integrate the updated deep Q network, the updated action space, and the updated state space to obtain an updated reinforcement learning module;

[0132] S3-7: Using the updated reinforcement learning module, generating a device management strategy for the obtained real-time device fault diagnosis result, obtaining a real-time device management strategy, and generating a corresponding real-time alarm signal, including the following steps:

[0133] S3-7-1: Using the agent of the updated reinforcement learning module to control the updated deep Q network, generating a Q value for each possible real-time device management action in the updated action space for each real-time device fault diagnosis state in the updated state space;

[0134] S3-7-2: according to the reward function, obtain the reward value of each possible real-time device management action for each real-time device fault diagnosis state, and update the corresponding Q value according to the reward value to obtain an updated Q value;

[0135] The formula is:

[0136] Q(s' p' ,a' p' )=(1-α")·Q(s p' ,a p' )+α"·(R(s p' ,a p' ,s' p' )+γ·Q max (s p' ,a p' ))

[0137] In the formula, Q(s' p' ,a' p' ) is the updated state value s' p' and the updated action value a' p' The corresponding updated Q value; Q(s p' ,a p' ) is the state value s p' and action value a p' The corresponding Q value; α" is the learning rate; Q max (s p' ,a p' ) is the state value s p' and action value a p' The highest Q value corresponding to the value; p' is the comprehensive indicator; γ is the update parameter; R(s p' ,a' p' ,s' p' ) is the updated action value a' p' Set the state value s p' Transform to the updated state value s' p' The reward value of

[0138] S3-7-3: Repeat the above Q value update step until the number of iterations reaches the iteration number threshold, and obtain the final Q value of each possible real-time device management action for each real-time device fault diagnosis state;

[0139] S3-7-4: using a greedy strategy, selecting the possible real-time device management action corresponding to the highest final Q value as the execution real-time device management action corresponding to the real-time device fault diagnosis state;

[0140] S3-7-5: Integrate all real-time equipment fault diagnosis states to execute real-time equipment management actions, obtain real-time equipment management strategies, and generate corresponding real-time alarm signals;

[0141] S4: The edge computing gateway uses an encryption algorithm to encrypt and transmit the real-time device operation monitoring data, real-time device dynamic monitoring data, real-time device fault diagnosis results, real-time device management strategies, and real-time alarm signals of the optical network information transmission equipment to the cloud data center, including the following steps:

[0142] S4-1: The edge computing gateway compresses the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies and real-time alarm signals of the optical network information transmission equipment to obtain a real-time compressed package;

[0143] S4-2: According to the private key of the edge computing gateway, use an asymmetric encryption algorithm to encrypt the real-time compressed package to obtain the encrypted real-time compressed package, and sign the encrypted real-time compressed package according to the signature information of the edge computing gateway to obtain the corresponding real-time signature data;

[0144] The formula is:

[0145] M u0 =E(SK u0 ,m u0 )

[0146] Where M u0 is the real-time compressed package after encryption; E(*) is the asymmetric encryption function; SK u0 is the private key of edge computing gateway u0; m u0 For real-time compression packages;

[0147] R u0 =rP

[0148]

[0149] In the formula, {ID u0 ,M u0 ,,γ"={K u0 ,R u0 ,B u0}} is the real-time signature data of edge computing gateway u0; R u0 ,B u0 ,γ" are signature parameters; r is a random number; is the prime order; P is the base point of the prime domain; H2 is a one-way mapping hash function; {K u0 ,KID u0} is the signature information of edge computing gateway u0; K u0 Registration parameters for edge computing gateway u0; KID uo is the registration ID of the edge computing gateway u0; ID uo is the gateway ID of edge computing gateway u0; M u0 It is a real-time compressed package after encryption;

[0150] The private key of the edge computing gateway belongs to the public-private key pair of the edge computing gateway. The public-private key pair includes a private key and a public key. The public-private key pair is generated by a trusted third-party organization based on the authentication information of the edge computing gateway using an asymmetric key generation algorithm. The public key is published to the cloud data center, and the private key is only stored in the corresponding edge computing gateway to ensure the confidentiality and security of the public-private key pair and improve the security of data transmission. The signature information of the edge computing gateway is registered and generated by a trusted third-party organization based on the authentication information of the edge computing gateway using digital identity technology, and is only stored in the corresponding edge computing gateway and the trusted third-party organization.

[0151] S4-3: extracting the device number of the optical network information transmission device, using the device number as the real-time additional information, and transmitting the encrypted real-time compressed package, the real-time signature data, and the real-time additional information to the cloud data center;

[0152] S5: The cloud data center uses the all-optical network digital twin model to visualize the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies, and real-time alarm signals of the optical network information transmission equipment, including the following steps:

[0153] S5-1: The cloud data center performs digital identity authentication on the real-time signature data. After the authentication is passed, the encrypted real-time compressed package is decrypted using an asymmetric encryption algorithm based on the public key of the edge computing gateway to obtain a decrypted real-time compressed package.

[0154] Call a trusted third-party organization to perform digital identity authentication on real-time signature data. The formula is:

[0155] βB u0 P=β u0 H2(R u0 ,M u0 ,ID u0 ,K u0 )R u0 +β u0 K u0 +β u0 H1(ID u0 ,Ku0 )

[0156] In the formula, β u0 is the signature authentication parameter. If the left formula is equal to the right formula, the digital identity authentication is successful. H1 and H2 are both one-way mapping hash functions.

[0157] m' u0 =E - (PK u0 ,M u0 )

[0158] In the formula, m' u0 It is a real-time compressed package after decryption; E - (*) is an asymmetric decryption function; PK u0 is the public key of edge computing gateway u0;

[0159] S5-2: decompress the decrypted real-time compressed package to obtain the decompressed real-time equipment operation monitoring data, the decompressed real-time equipment dynamic monitoring data, the decompressed real-time equipment fault diagnosis results, the decompressed real-time equipment management strategy and the decompressed real-time alarm signal;

[0160] S5-3: According to the real-time additional information, locate the dynamic model of the target optical network information transmission equipment in the all-optical network digital twin model, and write the decompressed real-time equipment operation monitoring data, decompressed real-time equipment dynamic monitoring data, decompressed real-time equipment fault diagnosis results, decompressed real-time equipment management strategy and decompressed real-time alarm signal into the corresponding target data stream;

[0161] S5-4: Based on the target data stream, use the visualization interface of the target optical network information transmission equipment dynamic model to visualize the decompressed real-time equipment operation monitoring data, the decompressed real-time equipment dynamic monitoring data, the decompressed real-time equipment fault diagnosis results, the decompressed real-time equipment management strategy, and the decompressed real-time alarm signal.

[0162] Embodiment 2:

[0163] like Figure 2 As shown, this embodiment provides an all-optical network visualization management system for implementing an all-optical network visualization management method. The all-optical network includes a number of optical network information transmission devices. The system includes a cloud data center, a number of edge computing gateways and a number of data acquisition devices. The cloud data center is communicated with the edge computing gateways of a number of management areas respectively, and each edge computing gateway is communicated with a number of data acquisition devices within the communication range respectively, and each data acquisition device is communicated with an optical network information transmission device one-to-one.

[0164] A data collection device, used to collect real-time equipment operation monitoring data of optical network information transmission equipment at corresponding positions in the all-optical network, and send it to the edge computing gateway within the communication range;

[0165] The edge computing gateway is used to receive the real-time equipment operation monitoring data of the optical network information transmission equipment sent by the data acquisition device, write the real-time equipment operation monitoring data into the corresponding equipment dynamic monitoring model, and extract the real-time equipment dynamic monitoring data of the equipment dynamic monitoring model; use the equipment fault diagnosis model to perform equipment fault diagnosis on the real-time equipment dynamic monitoring data, and use the equipment management policy generation model to generate equipment management policies for the obtained real-time equipment fault diagnosis results, obtain real-time equipment management policies, and generate corresponding real-time alarm signals; use the encryption algorithm to encrypt and transmit the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management policies, and real-time alarm signals of the optical network information transmission equipment to the cloud data center;

[0166] The cloud data center is used to build an all-optical network digital twin model, a device dynamic monitoring model, a device fault diagnosis model, and a device management policy generation model, and deploy the device dynamic monitoring model, the device fault diagnosis model, and the device management policy generation model to all edge computing gateways; the all-optical network digital twin model is used to visualize the real-time device operation monitoring data, real-time device dynamic monitoring data, real-time device fault diagnosis results, real-time device management strategies, and real-time alarm signals of optical network information transmission equipment.

[0167] The present invention provides an all-optical network visualization management method and system, which ensures that network management personnel can obtain comprehensive network status information in real time by comprehensively collecting real-time operation monitoring data of optical network information transmission equipment, thereby realizing comprehensive real-time monitoring of the all-optical network; through the equipment fault diagnosis model constructed by applying artificial intelligence technology, it can analyze the real-time equipment dynamic monitoring data, realize high-precision fault diagnosis, and accurately identify the cause of faults in complex network environments; through the equipment management strategy generation model constructed by applying artificial intelligence technology, it can automatically generate management strategies according to real-time network conditions, improve the automation and intelligence level of network management, and ensure network stability and efficiency; the all-optical network visualization management method and system, which are constructed by combining digital twin technology, can realize high-precision fault diagnosis and accurately identify the cause of faults in complex network environments; through the equipment management strategy generation model constructed by applying artificial intelligence technology, it can automatically generate management strategies according to real-time network conditions, improve the automation and intelligence level of network management, and ensure network stability and efficiency; The digital twin model of the optical network can map and simulate the actual network environment in real time, improve the predictive and intelligent level of network management, and can intuitively display the network topology, equipment status, fault information, etc., to help network managers quickly understand and analyze network conditions and improve visualization. Encryption and authentication measures are adopted in the process of data transmission and processing to effectively prevent data leakage and illegal attacks, and ensure the security of all-optical network data. The edge-cloud collaborative mechanism is collected, and monitoring and artificial intelligence models are deployed on the edge computing gateway to make full use of the capabilities of edge computing, improve data processing efficiency, reduce the computing resource requirements for cloud data centers, realize rapid data processing and response, and improve the efficiency of real-time management.

[0168] The present invention is not limited to the above optional implementations, and anyone can derive other various forms of products under the enlightenment of the present invention. The above specific implementations should not be understood as limiting the scope of protection of the present invention. The scope of protection of the present invention should be based on the definition in the claims, and the description can be used to interpret the claims.

Claims

1. A method for visual management of an all-optical network, characterized in that: The steps include: The cloud data center builds an all-optical network digital twin model, a device dynamic monitoring model, a device fault diagnosis model, and a device management strategy generation model, and deploys the device dynamic monitoring model, device fault diagnosis model, and device management strategy generation model to all edge computing gateways; The edge computing gateway collects real-time equipment operation monitoring data of the optical network information transmission equipment, writes the real-time equipment operation monitoring data into the corresponding equipment dynamic monitoring model, and extracts the real-time equipment dynamic monitoring data of the equipment dynamic monitoring model; The edge computing gateway uses the device fault diagnosis model to perform device fault diagnosis on the real-time device dynamic monitoring data, and uses the device management strategy generation model to generate device management strategies for the obtained real-time device fault diagnosis results, obtains the real-time device management strategy, and generates the corresponding real-time alarm signal; The edge computing gateway uses encryption algorithms to encrypt and transmit the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies, and real-time alarm signals of optical network information transmission equipment to the cloud data center; The cloud data center uses the all-optical network digital twin model to visualize the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies, and real-time alarm signals of optical network information transmission equipment.

2. The all-optical network visualization management method according to claim 1, characterized in that: The cloud data center builds an all-optical network digital twin model, a device dynamic monitoring model, a device fault diagnosis model, and a device management strategy generation model, and deploys the device dynamic monitoring model, device fault diagnosis model, and device management strategy generation model to all edge computing gateways, including the following steps: The cloud data center uses digital twin technology to build a digital twin model of the all-optical network based on the basic network information of the all-optical network and the basic equipment information of several optical network information transmission devices in the all-optical network; According to the dynamic mechanism properties of optical network information transmission equipment and equipment operation monitoring indicators, a dynamic monitoring model for equipment is constructed, and the dynamic monitoring model for equipment is used to generate a number of historical dynamic monitoring data for equipment; Extract the all-optical network topology diagram of the all-optical network digital twin model, and use the deep learning algorithm to build a device fault diagnosis model based on the all-optical network topology diagram and several historical equipment dynamic monitoring data, and generate several historical equipment fault diagnosis results; Based on several historical equipment fault diagnosis results, an enhanced reinforcement learning algorithm is used to build an equipment management strategy generation model, and model metadata of the equipment dynamic monitoring model, equipment fault diagnosis model, and equipment management strategy generation model are extracted; The all-optical network topology diagram and model metadata are sent to all edge computing gateways, and in the edge computing gateways, the model is reconstructed according to the model metadata to obtain a reconstructed device dynamic monitoring model, a reconstructed device fault diagnosis model, and a reconstructed device management strategy generation model.

3. The method for visual management of an all-optical network according to claim 2, characterized in that: The cloud data center uses digital twin technology to build a digital twin model of the all-optical network based on the basic network information of the all-optical network and the basic device information of several optical network information transmission devices in the all-optical network, including the following steps: Collect basic network information of the all-optical network and basic equipment information of several optical network information transmission equipment included in the all-optical network; According to the network topology information in the basic network information of the all-optical network, the device location and device number of each optical network information transmission device are confirmed, and a static model of the all-optical network topology is constructed; According to the basic device information of each optical network information transmission device, a corresponding optical network information transmission device static model is constructed at the corresponding device position of the all-optical network topology static model; Setting data flow and visualization interface for the static model of optical network information transmission equipment to obtain the dynamic model of optical network information transmission equipment; By integrating the static model of the all-optical network topology and the dynamic models of several optical network information transmission equipment, we can obtain the digital twin model of the all-optical network.

4. The method for visual management of an all-optical network according to claim 3, characterized in that: The equipment dynamic monitoring model includes a working status dynamic monitoring function, a transmission efficiency dynamic monitoring function, an optical power change dynamic monitoring function, an optical signal quality dynamic monitoring function, a transmission rate dynamic monitoring function, a network delay dynamic monitoring function, and a network stability dynamic monitoring function; The equipment fault diagnosis model is constructed based on the GCN-LSTM-Attention-MLP algorithm, and the equipment fault diagnosis model includes a graph structure feature extraction module constructed based on the GCN algorithm, a sequence feature extraction module constructed based on the LSTM algorithm, an attention weight module constructed based on the Attention mechanism, and an equipment fault diagnosis module constructed based on the MLP algorithm. The graph structure feature extraction module and the sequence feature extraction module are both connected to the attention weight module, and the attention weight module is connected to the equipment fault diagnosis module; The device management strategy generation model is constructed based on the MPO-DQN algorithm, and the device management strategy generation model includes a meta-strategy optimization module constructed based on the MPO algorithm and a reinforcement learning module constructed based on the DQN algorithm. The reinforcement learning module is provided with an intelligent agent, a deep Q network and an experience replay pool. The intelligent agent is respectively connected to the deep Q network, the experience replay pool and the meta-strategy optimization module, and the meta-strategy optimization module is connected to the experience replay pool.

5. The method for visual management of an all-optical network according to claim 4, characterized in that: The all-optical network topology diagram of the all-optical network digital twin model is extracted, and based on the all-optical network topology diagram and several historical equipment dynamic monitoring data, a deep learning algorithm is used to build an equipment fault diagnosis model and generate several historical equipment fault diagnosis results, including the following steps: Extract the all-optical network topology diagram of the all-optical network digital twin model, and associate the all-optical network topology diagram with each historical device dynamic monitoring data to obtain several model training samples; Use the GCN-LSTM-Attention-MLP algorithm to build an initial equipment fault diagnosis model, and input several model training samples into the initial equipment fault diagnosis model; According to a number of model training samples, the initial equipment fault diagnosis model is optimized and trained to obtain the final equipment fault diagnosis model, and a number of historical equipment fault diagnosis results are generated.

6. The method for visual management of an all-optical network according to claim 5, characterized in that: Based on several historical equipment fault diagnosis results, an enhanced reinforcement learning algorithm is used to build an equipment management strategy generation model, and the model metadata of the equipment dynamic monitoring model, equipment fault diagnosis model, and equipment management strategy generation model are extracted, including the following steps: Collecting a number of preset all-optical network management targets, and setting the preset all-optical network management targets as optimization scenarios for meta-strategy optimization, to obtain a number of optimization scenarios; Use the MPO algorithm to build an initial meta-strategy optimization module, and use several historical equipment fault diagnosis results in several optimization scenarios to train the initial meta-strategy optimization module to obtain the final meta-strategy optimization module and generate several historical meta-strategy optimization experiences; The device management strategy generation problem is used as the simulation environment of the DQN algorithm. The state space of the DQN algorithm is defined based on several historical device fault diagnosis results, and the action space of the DQN algorithm is defined based on several preset device management actions. According to the impact of preset equipment management actions on the historical equipment fault diagnosis status, define the reward function of the DQN algorithm, use the final meta-strategy optimization module to generate the initial deep Q network, and build the initial reinforcement learning module based on the initial deep Q network; Based on the state space, action space and reward function, the initial reinforcement learning module is trained using several historical equipment fault diagnosis results to obtain the final reinforcement learning module and generate several historical reinforcement learning experiences; Store several historical meta-strategy optimization experiences and several historical reinforcement learning experiences in the experience replay pool of the final reinforcement learning module, and integrate the final meta-strategy optimization module and the final reinforcement learning module to obtain a device management strategy generation model; Extract model metadata of the equipment dynamic monitoring model, equipment fault diagnosis model, and equipment management strategy generation model.

7. The method for visual management of an all-optical network according to claim 6, characterized in that: The edge computing gateway uses the device fault diagnosis model to perform device fault diagnosis on the real-time device dynamic monitoring data, and uses the device management strategy generation model to generate device management strategy for the obtained real-time device fault diagnosis results, obtains the real-time device management strategy, and generates the corresponding real-time alarm signal, including the following steps: The edge computing gateway inputs the all-optical network topology diagram and real-time device dynamic monitoring data into the device fault diagnosis model; Use the graph structure feature extraction module to extract the real-time graph structure features of the all-optical network topology graph, and use the sequence feature extraction module to extract the real-time sequence features of the real-time device dynamic monitoring data; According to the preset attention weight value, the attention weight module is used to perform weighted fusion on the real-time graph structure features and the real-time sequence features to obtain the real-time weighted fusion features; According to the real-time weighted fusion features, the equipment fault diagnosis module is used to diagnose equipment faults, obtain real-time equipment fault diagnosis results, and input them into the equipment management strategy generation model; Collect real-time all-optical network management goals, and extract corresponding real-time meta-strategy optimization experience and real-time reinforcement learning experience from the experience replay pool based on the real-time all-optical network management goals; Based on the real-time equipment fault diagnosis results, the real-time meta-strategy optimization experience and the real-time reinforcement learning experience, the equipment management strategy generation model is updated to obtain an updated reinforcement learning module; Using the updated reinforcement learning module, the device management strategy is generated for the obtained real-time device fault diagnosis results, the real-time device management strategy is obtained, and the corresponding real-time alarm signal is generated.

8. The method for visual management of an all-optical network according to claim 7, characterized in that: The edge computing gateway uses an encryption algorithm to encrypt and transmit the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies, and real-time alarm signals of the optical network information transmission equipment to the cloud data center, including the following steps: The edge computing gateway compresses the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies and real-time alarm signals of the optical network information transmission equipment to obtain a real-time compressed package; According to the private key of the edge computing gateway, an asymmetric encryption algorithm is used to encrypt the real-time compressed package to obtain the encrypted real-time compressed package, and according to the signature information of the edge computing gateway, the encrypted real-time compressed package is signed to obtain the corresponding real-time signature data; The device number of the optical network information transmission device is extracted, the device number is used as the real-time additional information, and the encrypted real-time compressed package, real-time signature data and real-time additional information are transmitted to the cloud data center.

9. The method for visual management of an all-optical network according to claim 8, characterized in that: The cloud data center uses the all-optical network digital twin model to visualize the real-time equipment operation monitoring data, real-time equipment dynamic monitoring data, real-time equipment fault diagnosis results, real-time equipment management strategies, and real-time alarm signals of optical network information transmission equipment, including the following steps: The cloud data center performs digital identity authentication on the real-time signature data. After the authentication is passed, the encrypted real-time compressed package is decrypted using an asymmetric encryption algorithm based on the public key of the edge computing gateway to obtain the decrypted real-time compressed package. Decompress the decrypted real-time compressed package to obtain decompressed real-time equipment operation monitoring data, decompressed real-time equipment dynamic monitoring data, decompressed real-time equipment fault diagnosis results, decompressed real-time equipment management strategy and decompressed real-time alarm signal; According to the real-time additional information, locate the dynamic model of the target optical network information transmission equipment in the all-optical network digital twin model, and write the decompressed real-time equipment operation monitoring data, decompressed real-time equipment dynamic monitoring data, decompressed real-time equipment fault diagnosis results, decompressed real-time equipment management strategy and decompressed real-time alarm signal into the corresponding target data stream; Based on the target data flow, using the visualization interface of the dynamic model of the target optical network information transmission equipment, the decompressed real-time equipment operation monitoring data, the decompressed real-time equipment dynamic monitoring data, the decompressed real-time equipment fault diagnosis results, the decompressed real-time equipment management strategy and the decompressed real-time alarm signal are visualized.

10. An all-optical network visualization management system, used to implement the all-optical network visualization management method according to any one of claims 1 to 9, wherein the all-optical network comprises a plurality of optical network information transmission devices, characterized in that: The system includes a cloud data center, several edge computing gateways and several data acquisition devices. The cloud data center is communicated with the edge computing gateways of several management areas respectively, and each of the edge computing gateways is communicated with several data acquisition devices within the communication range respectively. Each of the data acquisition devices is communicated with an optical network information transmission device one by one.

Images