Distributed device management system and method

Abstract

The invention provides a distributed equipment management system and method. Dynamically identifying a communication protocol of target equipment through a protocol analysis and dynamic mapping module; automatically completing network access and distributed identity authentication at each equipment end by using an access authentication module; the self-healing control module monitors the running state of the equipment in real time and determines a health degree score in combination with the dynamic evaluation model, and a grading fault recovery strategy is triggered according to the score; the collaborative optimization module performs localized distributed task scheduling on the target equipment in combination with the equipment knowledge graph; and the safety detection module adopts a distributed machine learning framework to analyze equipment behaviors in real time. The closed-loop cooperative operation of the process realizes automation, high efficiency and intelligence of equipment management, reduces the system cost, and improves the efficiency and safety of equipment data processing.

Description

Technical Field

[0001] This application relates to the field of Internet of Things systems, and particularly to a distributed device management system and method. Background Art

[0002] Currently, traditional device management systems usually rely on centralized control methods, which require manual configuration of devices and setting of various parameters. Before a device is connected to the network, a cumbersome configuration and authentication process is usually required, resulting in waste of time and manpower. Traditional device management systems generally do not have a high enough level of intelligence. When device failures or security incidents occur, manual intervention is required, and the response speed is slow, leading to a reduction in the reliability of the system. At the same time, there may also be a lack of flexibility. Due to protocol differences between devices, it is difficult for the system to perform flexible management and collaboration across devices and protocols. With the continuous increase in the number of Internet of Things devices, the centralized computing and storage methods can no longer meet the needs of large-scale devices.

[0003] Therefore, it is necessary to provide a distributed device management method in combination with specific scenarios, design a closed-loop processing method for identity authentication, device processing, and information feedback, so as to realize the application of the distributed processing method in the device management system, thereby reducing system costs and improving the efficiency and security of data processing. Summary of the Invention

[0004] The purpose of this application aims to solve at least one of the above technical defects, especially the technical defects of high costs of the device management system and insufficient efficiency and security of data processing in the prior art.

[0005] In a first aspect, this application provides a distributed device management system, which includes:

[0006] A protocol parsing and dynamic mapping module, which is used to dynamically identify the communication protocols of each target device accessing the system based on a preset protocol data structure;

[0007] An access authentication module, which is used to automatically connect each target device to the network at each target device end and complete the trust verification of each target device through a distributed identity authentication mechanism;

[0008] A self-healing control module, which is used to monitor the running states of each target device, determine the health score of each target device in combination with a dynamic evaluation model, and perform self-healing control on the target device with abnormal conditions according to the health score;

[0009] A collaborative optimization module, which is used to perform local distributed task scheduling on each target device in a data processing task, and optimize the data allocation and task allocation of each target device according to the device knowledge graph;

[0010] A security detection module, which is used to perform behavior analysis using a distributed machine learning framework and dynamically adjust device data allocation and operation permissions according to the risk level.

[0011] In a second aspect, the present application provides a distributed device management method. The method is executed based on the distributed device management system described in the first aspect, and the method includes:

[0012] Through a protocol parsing and dynamic mapping module, based on a preset protocol data structure, dynamically identify the communication protocols of each target device accessing the system;

[0013] Through an access authentication module, at each of the target device ends, automatically connect each of the target devices to the network and complete the trust verification of each of the target devices through a distributed identity authentication mechanism;

[0014] Through a self-healing control module, monitor the operating status of each of the target devices, determine the health score of each of the target devices in combination with a dynamic evaluation model, and trigger a hierarchical fault repair strategy according to the health score to perform self-healing control on the target devices with abnormal conditions;

[0015] Through a collaborative optimization module, perform local distributed task scheduling on each of the target devices in a data processing task, and optimize the data allocation and task allocation of each of the target devices according to the device knowledge graph;

[0016] Through a security detection module, perform behavior analysis using a distributed machine learning framework and dynamically adjust device data allocation and operation permissions according to the risk level.

[0017] As an optional implementation manner, the monitoring the operating status of each of the target devices and determining the health score of each of the target devices in combination with a dynamic evaluation model includes:

[0018] Real-time collect the device sensor data, performance parameters, running time, and historical running records of each of the target devices;

[0019] According to the device sensor data, the performance parameters, the running time, and the historical running records, determine the dynamic weighted score through the weight parameters obtained by pre-training to determine the health score of each of the target devices.

[0020] As an optional implementation manner, the triggering a hierarchical fault repair strategy according to the health score and performing self-healing control on the target devices with abnormal conditions includes:

[0021] If the health score is lower than a first threshold, determine the corresponding target device as a faulty device, and perform a preventive maintenance process or a fault handling process on each of the faulty devices to achieve self-healing control;

[0022] Among them, the way that each of the faulty devices executes a preventive maintenance process or a fault handling process to achieve self-healing control specifically includes:

[0023] Calculating a standardized sensor anomaly score based on the device sensor data and historical operation records of each of the faulty devices;

[0024] Determining the degree and type of the fault based on the standardized sensor anomaly score and the health score of each of the faulty devices, and determining the target self-healing method according to the recovery cost parameter corresponding to the preset self-healing method in the current situation, so as to achieve self-healing control.

[0025] As an optional implementation manner, the local distributed task scheduling for each of the target devices in the data processing task includes:

[0026] Determining the cost function corresponding to each subtask when it is executed by each of the target devices in the data processing task;

[0027] Determining the task allocation method that globally minimizes the cost function, and allocating each of the subtasks to the corresponding target device according to the task allocation method.

[0028] As an optional implementation manner, the optimizing the data allocation and task allocation of each of the target devices according to the device knowledge graph includes:

[0029] Determining the correlation degree parameter between each of the target devices, determining the failure probability of each of the target devices through the device knowledge graph according to the correlation degree parameter, the historical operation records of each of the target devices, and the health score, and optimizing the data allocation and task allocation of each of the target devices according to the failure probability of each of the target devices.

[0030] As an optional implementation manner, the method further includes:

[0031] Modifying the task allocation method according to the load conditions of each of the target devices, the resource occupancy conditions of each of the subtasks, and the priority conditions of each of the subtasks, and allocating each of the subtasks to the corresponding target device according to the modified task allocation method.

[0032] As an optional implementation manner, the using a distributed machine learning framework to perform behavior analysis and dynamically adjusting device data allocation and operation permissions includes:

[0033] Training a local behavior model at the device end and aggregating weights to update the global behavior model;

[0034] Identify abnormal operations that deviate from the normal behavior threshold through a timing pattern, determine the abnormal devices and their corresponding risk levels, and execute a preset hierarchical alarm mechanism according to the risk levels, and dynamically adjust the device operation permissions;

[0035] Among them, the hierarchical alarm mechanism includes: for abnormal devices with the first risk level, sending an alarm signal, generating an operation and maintenance log, and marking the devices to be observed; for abnormal devices with the second risk level, isolating the corresponding devices in real time and notifying the management terminal;

[0036] Among them, the dynamic adjustment of device operation permissions includes: restricting the execution permission of sensitive instructions for abnormal devices with the first risk level; switching abnormal devices with the second risk level to the read-only monitoring mode.

[0037] In a third aspect, the present application provides a computer device, including one or more processors and a memory. Computer-readable instructions are stored in the memory. When the computer-readable instructions are executed by the one or more processors, the steps of the method described in the second aspect are executed.

[0038] In a fourth aspect, the present application provides a storage medium in which computer-readable instructions are stored. When the computer-readable instructions are executed by one or more processors, one or more processors are caused to execute the steps of the method described in the second aspect.

[0039] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0040] Based on any of the above embodiments, the distributed device management system and method provided by the present application solve the compatibility problem caused by device protocol differences in traditional centralized systems through the protocol parsing and dynamic mapping module, and realize the flexible access of cross-protocol devices; the access authentication module automatically completes network access and trusted verification at each target device end, significantly improving the efficiency and security of device management; the self-healing control module monitors the device operation status in real time and determines the health score based on the dynamic evaluation model, and performs automatic repair on abnormal devices in combination with the hierarchical fault repair strategy, greatly reducing the manual intervention cost and response time; the collaborative optimization module combines the device knowledge graph to perform local distributed task scheduling on the target device, optimizes data distribution and task allocation, and improves the overall utilization efficiency of resources and the task execution speed; the security detection module uses a distributed machine learning framework to analyze device behavior in real time, dynamically adjusts the permission allocation, and constructs an intelligent security protection system. The collaborative effect of each module finally realizes the automation, high efficiency and intelligence of device management, solves the problems of traditional systems relying on centralized control, cumbersome manual configuration, slow response speed, lack of flexibility and security, significantly reduces the system operation and maintenance cost, and improves the overall reliability. Description of the Drawings

[0041] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0042] Figure 1 Schematic flow diagram of the distributed device management method provided for an embodiment of the present application;

[0043] Figure 2 Internal structure diagram of the computer device provided for the embodiment of the present application. Detailed implementation manners

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0045] With the wide application of Internet of Things technology, more and more devices are connected to the network. These devices usually have different hardware, protocols and functions. This makes the management and security of devices become more and more complex. The following security problems usually exist in Internet of Things devices: A large number of Internet of Things devices need to be automatically connected and interact with the network. Traditional identity authentication methods (such as user names and passwords) may not be able to meet the needs of automated access. Internet of Things devices may use different communication protocols, resulting in inability to communicate and cooperate between devices, thus affecting the integration and reliability of the system. Internet of Things devices usually do not have sufficient security protection measures and are vulnerable to security threats such as network attacks, data leakage and illegal access.

[0046] Currently, traditional device management systems usually rely on centralized control methods, requiring manual configuration of devices and setting of various parameters. Before a device is connected to the network, a cumbersome configuration and authentication process is usually required, resulting in a waste of time and manpower. Traditional device management systems are usually not intelligent enough, and manual intervention is required when device failures or security incidents occur, with a slow response speed, leading to a reduction in the reliability of the system. Lack of flexibility: Due to protocol differences between devices, it is difficult for the system to perform flexible management and collaboration across devices and protocols. As the number of Internet of Things devices continues to increase, the centralized computing and storage method can no longer meet the needs of large-scale devices. To solve this problem, edge computing has emerged as a new distributed computing model, which can execute computing tasks close to the data source, thereby reducing latency and bandwidth consumption.

[0047] By combining a protocol abstraction layer and a blockchain identity authentication mechanism, this application can complete the access of devices without manual intervention, with a faster device access speed, and greatly improve the flexibility and efficiency of system management. Through smart contracts, it can ensure the authenticity of the identity of newly connected devices and data security. By executing some computing tasks locally through edge computing nodes, the dependence on the central server is reduced, thereby reducing the latency and bandwidth consumption of the system. Tasks such as the initial identity verification, protocol parsing, and anomaly detection of devices can be processed at the edge node, thereby optimizing data flow, enhancing the response speed of devices, and improving the processing capacity of the system.

[0048] The technical concept of this application is that the protocol parsing and dynamic mapping module solves the compatibility problem caused by device protocol differences in traditional centralized systems and realizes the flexible access of cross-protocol devices; the access authentication module automatically completes network access and trusted verification at each target device end, significantly improving the efficiency and security of device management; the self-healing control module monitors the device operation status in real time and determines the health score based on a dynamic evaluation model, and combines a hierarchical fault repair strategy to perform automatic repair on abnormal devices, greatly reducing the manual intervention cost and response time; the collaborative optimization module combines the device knowledge graph to perform local distributed task scheduling on target devices, optimizes data allocation and task allocation, and improves the overall utilization efficiency of resources and task execution speed; the security detection module uses a distributed machine learning framework to analyze device behavior in real time, dynamically adjusts permission allocation, and constructs an intelligent security protection system. The synergistic effect of each module finally realizes the automation, high efficiency, and intelligence of device management, solves problems such as the traditional system's dependence on centralized control, cumbersome manual configuration, slow response speed, lack of flexibility and security, significantly reduces the system operation and maintenance cost, and improves the overall reliability.

[0049] The following details the method provided by this application according to the corresponding implementation manners in some actual application scenarios.

[0050] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of a distributed device management system and method provided for an embodiment of this application. As Figure 1 shown, the method includes:

[0051] S101. Through the protocol parsing and dynamic mapping module, based on a preset protocol data structure, dynamically identify the communication protocols of each target device accessing the system;

[0052] The protocol data structure provides a highly general way of protocol extension and device extension. According to the original data of each dimension of the device, it can be uniformly mapped into a standard format. At the same time, the conversion methods of each protocol are stored in a dynamic protocol table. For specific details, please refer to the relevant description.

[0053] S102. Through the access authentication module, at each of the target device ends, automatically connect each of the target devices to the network, and complete the trusted verification of each of the target devices through a distributed identity authentication mechanism;

[0054] After the target device passes the authentication and accesses the management system, it will be automatically configured as a distributed edge computing node to perform corresponding tasks such as fault self-checking, data processing, and task allocation. It should be noted that the method of this application can be run based on a distributed device management system or any one of its modules as the execution subject. In an actual application scenario, it can also be executed based on any distributed target device.

[0055] S103. Through the self-healing control module, monitor the running states of each of the target devices, determine the health score of each of the target devices in combination with a dynamic evaluation model, and trigger a hierarchical fault repair strategy according to the health score to perform self-healing control on the target devices with abnormal situations;

[0056] S104. Through the collaborative optimization module, perform local distributed task scheduling on each of the target devices in a data processing task, and optimize the data allocation and task allocation of each of the target devices according to the device knowledge graph;

[0057] S105. Through the security detection module, perform behavior analysis using a distributed machine learning framework, and dynamically adjust the device data allocation and operation permissions according to the risk level.

[0058] The method provided by this application runs based on a distributed device management system. Correspondingly, the system includes:

[0059] A protocol parsing and dynamic mapping module, which is used to dynamically identify the communication protocols of each target device accessing the system based on a preset protocol data structure;

[0060] An access authentication module is used to automatically connect each of the target devices to the network at each of the target device ends and complete the trusted verification of each of the target devices through a distributed identity authentication mechanism;

[0061] A self-healing control module is used to monitor the running states of each of the target devices, determine the health score of each of the target devices in combination with a dynamic evaluation model, and perform self-healing control on the target devices with abnormal conditions according to the health score;

[0062] A collaborative optimization module is used to perform local distributed task scheduling on each of the target devices in a data processing task, and optimize the data allocation and task allocation of each of the target devices according to the device knowledge graph;

[0063] A security detection module is used to perform behavior analysis using a distributed machine learning framework and dynamically adjust the device data allocation and operation permissions according to the risk level.

[0064] The distributed device management system provided by this application dynamically identifies the communication protocols of the access devices based on a preset protocol data structure through a protocol parsing and dynamic mapping module, solving the problem of difficult cross-device collaboration caused by device protocol differences in traditional systems; using the access authentication module to automatically connect devices to the network at each target device end and complete trusted verification through a distributed identity authentication mechanism, avoiding the traditional cumbersome manual configuration and authentication processes, and significantly improving the efficiency and security of device access; through the self-healing control module, the running states of devices are monitored in real time and the health score is determined in combination with a dynamic evaluation model, and a hierarchical fault repair strategy is triggered according to the score, realizing the automatic repair of abnormal devices and solving the problem of slow manual intervention response; the collaborative optimization module performs local distributed task scheduling on the target devices in combination with the device knowledge graph, optimizes data allocation and task allocation, and improves the resource utilization efficiency and task execution speed; the security detection module uses a distributed machine learning framework to analyze device behaviors in real time and dynamically adjusts data allocation and operation permissions according to the risk level, enhancing the security protection ability of the system. In summary, through the collaborative effect of each module, the system realizes the automatic management, efficient collaboration and intelligent operation and maintenance of devices, significantly reducing the system cost and improving the overall reliability.

[0065] As an optional implementation manner, the monitoring of the running states of each of the target devices and determining the health score of each of the target devices in combination with a dynamic evaluation model includes:

[0066] Real-time collecting the device sensor data, performance parameters, running time and historical running records of each of the target devices;

[0067] Based on the device sensor data, the performance parameters, the running time, and the historical operation records, and using the weight parameters obtained through pre-training, determine the dynamic weighted score to determine the health score of each target device.

[0068] In this embodiment, by collecting the sensor data, performance parameters, running time, and historical operation records of the target device in real time, and combining the pre-trained weight parameters to determine the dynamic weighted score, the accurate quantification of the device health is realized. This scoring mechanism comprehensively considers the real-time operation status and historical performance of the device, avoids the one-sidedness of traditional single-index evaluation, and makes the health score more comprehensive and reliable. By introducing the dynamic weighted model, the system can more accurately identify the potential fault risks of the device, provide a scientific basis for the subsequent hierarchical fault repair strategy, and thus significantly improve the accuracy of system fault prediction and self-healing control.

[0069] As an alternative embodiment, triggering a hierarchical fault repair strategy according to the health score and performing self-healing control on the target device with abnormal conditions includes:

[0070] If the health score is lower than the first threshold, determine the corresponding target device as a faulty device, and perform a preventive maintenance process or a fault handling process on each faulty device to achieve self-healing control;

[0071] Among them, the way of performing a preventive maintenance process or a fault handling process on each faulty device to achieve self-healing control specifically includes:

[0072] Calculate the standardized sensor anomaly score based on the device sensor data and historical operation records of each faulty device;

[0073] Determine the fault degree and fault type according to the standardized sensor anomaly score and the health score of each faulty device, and determine the target self-healing method according to the recovery cost parameter corresponding to the preset self-healing method in the current situation to achieve self-healing control.

[0074] In this embodiment, the hierarchical fault repair strategy is triggered by the health score, specifically including determining the fault degree and fault type according to the standardized sensor anomaly score and the health score, and selecting the target self-healing method in combination with the preset recovery cost parameter. This strategy realizes the refined classification and processing of faulty devices, avoiding the waste of resources caused by a one-size-fits-all maintenance method. By quantifying the fault degree and type, the system can accurately match the optimal repair plan, such as performing preventive maintenance on minor faults and isolation processing on serious faults, thus significantly reducing the resource overhead and maintenance cost of self-healing control while ensuring the stability of the system.

[0075] As an alternative implementation, the local distributed task scheduling for each of the target devices in the data processing task includes:

[0076] Determine the cost function corresponding to each subtask in the data processing task when it is executed by each of the target devices;

[0077] Determine the task allocation method that globally minimizes the cost function, and allocate each subtask to the corresponding target device according to the task allocation method.

[0078] In this implementation, by calculating the cost function of the subtasks of the data processing task and determining the task allocation method under the goal of global minimization, scientific scheduling of tasks is achieved. This allocation method comprehensively considers the balance between task execution costs and resource consumption, ensuring the optimal solution for task allocation. Combining the correlation parameters of the device knowledge graph and historical operation records, the system further optimizes data allocation and task allocation, avoiding the load imbalance problem caused by uneven resource allocation, and improving the task execution efficiency and the overall utilization rate of the devices.

[0079] As an alternative implementation, the optimization of the data allocation and task allocation for each of the target devices according to the device knowledge graph includes:

[0080] Determine the correlation parameters between the target devices, and determine the failure probability of each target device through the device knowledge graph according to the correlation parameters, the historical operation records of each target device, and the health score, and optimize the data allocation and task allocation of each target device according to the failure probability of each target device.

[0081] In this implementation, the correlation parameters between target devices are analyzed through the device knowledge graph, and the failure probability is predicted by combining historical operation records and health scores, realizing the dynamic optimization of task allocation. This process can adjust the task allocation strategy according to the potential risks of the devices. For example, critical tasks are preferentially allocated to devices with a lower failure probability, thereby reducing the risk of task interruption caused by device failures. At the same time, by continuously optimizing data allocation and task allocation, the system further improves the overall utilization efficiency of resources and ensures the stable execution of tasks.

[0082] As an alternative implementation, the method further includes:

[0083] Modify the task allocation method according to the load conditions of each target device, the resource occupancy of each subtask, and the priority of each subtask, and allocate each subtask to the corresponding target device according to the modified task allocation method.

[0084] According to the device load status, sub-task resource occupancy, and task priorities, this embodiment dynamically corrects the task allocation method in real time, achieving dynamic adaptability of task scheduling. This correction mechanism can flexibly handle load fluctuations or sudden task requirements during system operation. For example, when the device load is too high, tasks are reallocated to idle devices, or when the priority of critical tasks is increased, the allocation weights are adjusted. Through dynamic correction, the system avoids performance degradation caused by resource bottlenecks, ensuring efficient task execution and system stability.

[0085] As an alternative embodiment, the distributed machine learning framework is adopted for behavior analysis, and device data allocation and operation permissions are dynamically adjusted according to the risk level, including:

[0086] Train a local behavior model at the device end and aggregate weights to update the global behavior model;

[0087] Through the time series pattern, identify abnormal operations that deviate from the normal behavior threshold, determine the abnormal devices and their corresponding risk levels, and execute a preset hierarchical warning mechanism according to the risk level, and dynamically adjust the device operation permissions;

[0088] Among them, the hierarchical warning mechanism includes: for abnormal devices with the first risk level, send a warning signal, generate an operation and maintenance log, and mark the devices to be observed; for abnormal devices with the second risk level, isolate the corresponding devices in real time and notify the management terminal;

[0089] Among them, the dynamic adjustment of device operation permissions includes: restricting the execution permission of sensitive instructions for abnormal devices with the first risk level; switching abnormal devices with the second risk level to the read-only monitoring mode.

[0090] This embodiment uses a distributed machine learning framework to train a local behavior model at the device end and updates the global model through weight aggregation, realizing real-time monitoring and analysis of device behavior. By identifying abnormal operations through the time series pattern and triggering a hierarchical warning mechanism according to the risk level, the system can restrict the permissions of sensitive instructions for low-risk devices and switch high-risk devices to the read-only monitoring mode. This strategy dynamically adjusts device operation permissions, effectively isolates potential threats, avoids the spread of security incidents, and significantly improves the system's security protection ability and response efficiency.

[0091] The embodiments of the present application can be combined for application. Based on an actual application scenario below, the combined application of each embodiment is described to elaborate on the corresponding technical effects of the present application in detail.

[0092] The protocol parsing and dynamic mapping module is used to construct a unified data format using the protocol abstraction layer and support mainstream protocols. Through the protocol adapter, identify the communication protocol of the device and establish a dynamic protocol mapping table;

[0093] Build a unified data format using the Protocol Abstraction Layer (PAL) and support mainstream protocols. Through the protocol adapter, identify the communication protocol of the device and establish a dynamic protocol mapping table, including:

[0094] Set the original data format of the device as:

[0095]

[0096] where P represents the communication protocol type of the device, V represents the measured value or status of the device, T represents the timestamp, and M represents the metadata of the device;

[0097] The PAL maps the original data D of the device through the conversion function f raw uniformly to the standardized PAL format:

[0098]

[0099] where the PAL format uses the JSON structure;

[0100] The dynamic protocol mapping table stores the conversion rules for different protocols. Suppose a device supports multiple protocols (P1, P2, P3), and the data conversion relationship between different protocols is represented by a conversion matrix:

[0101]

[0102] where f ij represents the conversion function from protocol P i to P j and f ii = 1 indicates that the protocol itself does not require conversion.

[0103] The access authentication module is used to achieve device access without manual configuration based on zero-configuration technology. It adopts a blockchain identity authentication mechanism, uses smart contracts to perform trusted verification on newly connected devices, and combines edge computing to perform preliminary identity verification and protocol parsing locally;

[0104] Adopt a blockchain identity authentication mechanism and use smart contracts to perform trusted verification on newly connected devices, including:

[0105] When a device first accesses the network, it automatically selects an IP address. The device selects a random IP within the address range through the IPv4 Automatic Private IP Address mechanism. For IPv6, the device automatically generates a globally unique IP based on the network prefix. The device broadcasts its recognizable name in the local network through multicast DNS and resolves the correspondence between the device name and IP address in the local network.

[0106] Each device generates a unique device identity identifier when it first accesses:

[0107]

[0108] Among them, the MAC is the physical address of the device, the public key is used for encrypted communication and authentication, the hash function H() ensures the uniqueness of the identity, and the device identity information is stored in the blockchain ledger;

[0109] The device generates an identity identifier IDdevice and sends a registration request. The blockchain network verifies the uniqueness of the device identity and records its identity information. When the device connects to the network, it submits an authentication request to the blockchain. The blockchain queries the stored device identity and performs a matching verification: if the device is already registered, the authentication passes and it is allowed to join the network; if the device is not registered, access is denied.

[0110] The device submits its identity information, including the device ID and signature, to the blockchain, which is decrypted using the device's public key stored in the blockchain. If the match is successful, it indicates that the device identity is authentic and trustworthy, and access to the network is allowed.

[0111] The self-healing control module is used to monitor the working state of the device, establish a health score model, and adopt a dynamic self-healing algorithm to automatically repair when a fault is detected;

[0112] Adopting a dynamic self-healing algorithm to automatically repair when a fault is detected, including:

[0113] Set the health score H i of device D i to be calculated from multiple parameters:

[0114]

[0115] Among them, S i is the score for abnormal sensor data of the device, P i is the performance parameter of the device, A i is the historical fault record of the device, T i is the running time of the device, and α1, α2, α3, α4 are weight parameters obtained by training with historical data;

[0116] For example, in a specific scenario, the judgment threshold for the device state can be set as follows: if H i > 80, the device is healthy and running normally; if 50 ≤ H i ≤ 80, the device has a minor anomaly and preventive maintenance is recommended; if H i < 50, the device has a poor health state and immediate repair or replacement is required;

[0117] Actually, when H i is lower than the preset threshold H min , the dynamic self-healing mechanism can be triggered.

[0118] Furthermore, the method can also be implemented in combination with sensor standardization anomaly scoring:

[0119] Set device D i Collected sensor data set:

[0120]

[0121] where x j represents the j-th sensor data of the device;

[0122] Calculate the standardized anomaly score of each sensor data:

[0123]

[0124] where μ j is the historical mean, σ j is the standard deviation, if ∣Z j ∣>δ, then this data point is an anomaly;

[0125] When device anomaly is detected or the health score is lower than the threshold H min , the dynamic self-healing algorithm is used for automatic repair, and the self-healing strategy selects the recovery plan according to the device anomaly type:

[0126]

[0127] where R(D i ) is the self-healing strategy selected by device D i , and C(r k ) is the recovery cost of strategy r k ;

[0128] If the device has a minor anomaly, software-level repair is performed; if the device has a resource allocation problem, parameter adjustment is performed; if the device has a serious fault, hardware-level repair is performed.

[0129] The collaborative optimization module is used to execute part of the computing tasks locally through the edge computing nodes, adopt distributed task scheduling, optimize the data flow between devices, and analyze the associations between devices in combination with the knowledge graph;

[0130] Adopt distributed task scheduling, optimize the data flow between devices, and analyze the associations between devices, including:

[0131] Set each task T j to be executed by the edge node E i , and the task allocation is modeled through the following optimization function:

[0132]

[0133] Among them, f(E i , T j ) is the cost function for device E i to execute task T j . n is the number of tasks, and m is the number of edge nodes;

[0134] The load of each node needs to be balanced to avoid overloading some nodes. Assuming the load of the node is L i , the goal is to minimize the load difference:

[0135]

[0136] Among them, is the average load of all nodes;

[0137] The knowledge graph combines the historical data and health score of the device to predict the potential faults of the device. The association information between devices enhances the accuracy of fault prediction. Assuming the fault probability of device Di is Pfail(Di), the fault prediction is represented by the following model:

[0138]

[0139] Among them, α j is the weight of the influence of device D j on device D i , R ij is the association strength between devices, and H j is the health score of device D j ;

[0140] The security detection module is used to perform security analysis using federated learning, identify abnormal device behaviors by combining behavioral analysis, set a hierarchical alarm mechanism, and automatically adjust device permissions when security anomalies occur.

[0141] Identify abnormal device behaviors by combining behavioral analysis, set a hierarchical alarm mechanism, and automatically adjust device permissions when security anomalies occur, including:

[0142] Use time series analysis methods to model the normal behaviors of devices. Assuming the state of device D i at time t is described by S i (t), the normal behavior pattern of the device is modeled through historical state data S i (t1), S i (t2), …, S i (tn), and detect anomalies through model errors. Set a threshold . When the difference between the actual behavior and the predicted behavior of the device exceeds this threshold, it is determined as an abnormal behavior;

[0143] The LSTM will be trained locally on each device. Each device trains its own model based on local historical data and shares the updated model weights with other devices.

[0144] For example, in an actual scenario, through a set threshold function, an alarm is triggered when it is detected that the behavior of the device exceeds the normal range. The alarm threshold function is defined as:

[0145]

[0146] Where 1, 2, 3 are the thresholds for different alarm levels respectively;

[0147] According to the anomaly detection results of the device, normal devices access and execute tasks normally, slightly abnormal devices restrict sensitive operations of the device, and severely abnormal devices automatically switch the device to a restricted mode, only allowing basic communication and monitoring tasks.

[0148] The embodiments of the present application also provide a distributed device management virtual device to implement the corresponding management method through computer program modular design. The acquisition module executes the original data acquisition process that may be involved in the present application, and the processing module executes the data processing process in the present application. For specific reference, see the implementation manner on the method side, which will not be elaborated here.

[0149] It should be noted that it should be understood that the division of each module of the above device is only a logical function division. In actual implementation, it can be fully or partially integrated into a physical entity, or physically separated. And these modules can all be implemented in the form of software called by processing elements; they can also all be implemented in the form of hardware; or some modules can be implemented in the form of software called by processing elements, and some modules can be implemented in the form of hardware. For example, the processing module can be a separately established processing element, or can be integrated in a certain chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a certain processing element of the above device to implement the functions of the above determined modules. The implementation of other modules is similar. In addition, these modules can be fully or partially integrated together, or can be independently implemented. Here, the processing element can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed through the integrated logic circuit in the processor element or the instructions in the form of software.

[0150] Schematically, as Figure 2 shown, Figure 2 is a schematic internal structure diagram of a computer device provided by the embodiments of the present application. The computer device 300 can be provided as a server. Refer toFigure 2 , the computer device 300 includes a processing component 302, which further includes one or more processors, and memory resources represented by a memory 301 for storing instructions executable by the processing component 302, such as application programs. The application programs stored in the memory 301 may include one or more modules each corresponding to a set of instructions. In addition, the processing component 302 is configured to execute instructions to perform the methods of any of the above embodiments.

[0151] The computer device 300 may further include a power component 303 configured to perform power management of the computer device 300, a wired or wireless network interface 304 configured to connect the computer device 300 to a network, and an input / output (I / O) interface 305. The computer device 300 may operate based on an operating system stored in the memory 301, such as Windows Server TM, Mac OS XTM, Unix TM, Linux TM, Free BSDTM or the like.

[0152] Those skilled in the art can understand that Figure 2 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0153] The embodiments of this application provide a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to execute the methods provided in any of the embodiments.

[0154] Finally, it should also be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including the element.

[0155] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0156] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A distributed device management system, characterized in that, The system includes: A protocol parsing and dynamic mapping module, which is used to dynamically identify the communication protocols of each target device accessing the system based on a preset protocol data structure; An access authentication module, which is used to automatically connect each target device to the network at each target device end and complete the trusted verification of each target device through a distributed identity authentication mechanism; A self-healing control module, which is used to monitor the running states of each target device, determine the health score of each target device in combination with a dynamic evaluation model, and perform self-healing control on the target devices with abnormal conditions according to the health score; A collaborative optimization module, which is used to perform local distributed task scheduling on each target device in a data processing task and optimize the data distribution and task distribution of each target device according to a device knowledge graph; A security detection module, which is used to perform behavior analysis using a distributed machine learning framework and dynamically adjust the device data distribution and operation permissions according to the risk level.

2. A distributed device management method, characterized in that The method is executed based on the distributed device management system as described in claim 1, and the method includes: Through the protocol parsing and dynamic mapping module, dynamically identify the communication protocols of each target device accessing the system based on a preset protocol data structure; Through the access authentication module, automatically connect each target device to the network at each target device end and complete the trusted verification of each target device through a distributed identity authentication mechanism; Through the self-healing control module, monitor the running states of each target device, determine the health score of each target device in combination with a dynamic evaluation model, and trigger a hierarchical fault repair strategy according to the health score to perform self-healing control on the target devices with abnormal conditions; Through the collaborative optimization module, perform local distributed task scheduling on each target device in a data processing task and optimize the data distribution and task distribution of each target device according to a device knowledge graph; Through the security detection module, perform behavior analysis using a distributed machine learning framework and dynamically adjust the device data distribution and operation permissions according to the risk level.

3. The method according to claim 2, wherein The monitoring of the running states of each target device and the determination of the health score of each target device in combination with a dynamic evaluation model include: Real-time collect the device sensor data, performance parameters, running time, and historical running records of each target device; According to the device sensor data, the performance parameters, the running time, and the historical running records, determine the dynamic weighted score through the weight parameters obtained by pre-training to determine the health score of each target device.

4. The method according to claim 3, characterized in that, The triggering of a hierarchical fault repair strategy according to the health score and the performance of self-healing control on the target devices with abnormal conditions include: If the health score is lower than a first threshold, determine the corresponding target device as a faulty device, and perform a preventive maintenance process or a fault handling process on each faulty device to achieve self-healing control; Among them, the manner of performing a preventive maintenance process or a fault handling process on each faulty device to achieve self-healing control specifically includes: Calculate the standardized sensor anomaly score based on the device sensor data and historical operation records of each faulty device; Determine the fault degree and fault type based on the standardized sensor anomaly score and the health score of each faulty device, and determine the target self-healing method according to the recovery cost parameter corresponding to the preset self-healing method in the current situation, so as to achieve self-healing control.

5. The method according to claim 1, wherein The local distributed task scheduling for each target device in the data processing task includes: Determine the cost function corresponding to each subtask in the data processing task when executed by each target device; Determine the task allocation method that minimizes the global cost function, and allocate each subtask to the corresponding target device according to the task allocation method.

6. The method according to claim 5, characterized in that, The optimization of the data allocation and task allocation of each target device according to the device knowledge graph includes: Determine the correlation degree parameter between each target device. According to the correlation degree parameter, the historical operation records of each target device, and the health score, determine the fault probability of each target device through the device knowledge graph, and optimize the data allocation and task allocation of each target device according to the fault probability of each target device.

7. The method according to any one of claims 5-6, characterized in that, The method further includes: Modify the task allocation method according to the load conditions of each target device, the resource occupancy of each subtask, and the priority of each subtask, and allocate each subtask to the corresponding target device according to the modified task allocation method.

8. The method according to claim 1, wherein The behavior analysis is performed using a distributed machine learning framework, and the device data allocation and operation permissions are dynamically adjusted, including: Train a local behavior model at the device end and aggregate the weights to update the global behavior model; Identify abnormal operations that deviate from the normal behavior threshold through a time series pattern, determine the abnormal device and the corresponding risk level, and execute a preset hierarchical alarm mechanism according to the risk level, and dynamically adjust the device operation permissions; Among them, the hierarchical alarm mechanism includes: for abnormal devices with the first risk level, send an alarm signal, generate an operation and maintenance log, and mark the devices to be observed; for abnormal devices with the second risk level, isolate the corresponding device in real time and notify the management terminal; Among them, the dynamic adjustment of the device operation permissions includes: restricting the execution permission of sensitive instructions for abnormal devices with the first risk level; switching abnormal devices with the second risk level to the read-only monitoring mode.

9. A computer device, characterized in that, It includes one or more processors and a memory. Computer-readable instructions are stored in the memory. When the computer-readable instructions are executed by the one or more processors, the steps of the method according to any one of claims 2-8 are executed.

10. A storage medium, characterized in that, Computer-readable instructions are stored in the storage medium. When the computer-readable instructions are executed by one or more processors, one or more processors are caused to execute the steps of the method according to any one of claims 2-8.

Images