A multi-device collaborative management system based on cloud-edge collaboration

Abstract

The present invention discloses a multi-device collaborative management system based on cloud-edge collaboration. The present invention relates to the field of Internet of Things communication technology and includes a device management module for connecting edge devices and centrally managing the edge devices, and performing optimized scheduling based on the functions and distribution positions of different edge devices; a data acquisition module connected to the device management module. This multi-device collaborative management system based on cloud-edge collaboration provides a comprehensive and accurate data foundation for multi-device collaborative management, promotes information exchange and collaborative work among devices; the optimization decision-making unit can formulate an optimized device management strategy according to the operation plan generated by the big data analysis unit and the prediction results obtained by the prediction analysis unit, and continuously optimize and adjust according to the real-time operation conditions of each edge device; enabling the system to adapt to the continuously changing device operation environment and task requirements, and realizing more efficient multi-device collaborative management.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things communication, and specifically to a multi-device collaborative management system based on cloud-edge collaboration. Background Art

[0002] In the current technical environment, multi-device collaborative management faces a series of challenges. On the one hand, the communication stability between devices is insufficient and is easily affected by factors such as network fluctuations and interference, resulting in delays, interruptions, or even failures in collaborative operations. Although existing multi-device collaborative management systems can, to a certain extent, achieve the collaborative work of devices, they still lack intelligent decision-making capabilities when facing complex tasks and dynamically changing environments. The system cannot automatically adjust collaborative strategies according to real-time situations, still requires manual intervention, and the system does not understand the functional characteristics of devices accurately and in a timely manner, making it difficult to accurately match task requirements during device scheduling. In terms of network resource utilization, the optimization of data transmission paths is not efficient enough to ensure the full utilization of bandwidth. Moreover, there is a lack of accurate prediction of device operating states and effective optimization decision-making mechanisms, making it difficult to anticipate problems that may occur with devices in advance and improve collaborative work efficiency. In summary, there is an urgent need for a multi-device collaborative management system based on cloud-edge collaboration to solve these problems. Summary of the Invention

[0003] To solve the above technical problems, the present invention is realized through the following technical solutions: A multi-device collaborative management system based on cloud-edge collaboration, including a device management module for connecting to edge devices and centrally managing the edge devices, and performing optimized scheduling based on the functions and distribution locations of different edge devices;

[0004] A data acquisition module, connected to the device management module, for real-time acquisition of the operating data of each edge device, where the operating data includes operating status data, device operation data, and environmental monitoring data, and classifying and storing the acquired data;

[0005] A cloud analysis module, connected to the data acquisition module, for receiving the classified and stored data, and analyzing the data of each edge device based on big data analysis algorithms to formulate device optimization management strategies;

[0006] A collaborative control module, connected to the device management module and the cloud analysis module, for performing collaborative control on each edge device according to the optimization management strategy generated by the cloud analysis module to ensure the optimal allocation of resources and collaborative work during the collaborative operation of the devices;

[0007] A data feedback module, connected to the collaborative control module, for feeding back the real-time operating conditions of each edge device to the cloud analysis module.

[0008] Preferably, the device management module includes:

[0009] A device registration unit, which is used to perform identity registration and permission management on newly added edge devices, ensuring that new edge devices can obtain corresponding permissions in the cloud-edge collaboration system and then join the collaborative management;

[0010] A device scheduling unit, which is used to dynamically schedule edge devices based on the physical location, function, and load conditions of the devices, ensuring the workload balance of each edge device;

[0011] A device monitoring unit, which is used to monitor the running status of each edge device in real time, identify abnormal situations during the operation of edge devices, and generate device alarm signals.

[0012] Preferably, the process of the device registration unit performing identity registration and permission management on newly added edge devices is as follows:

[0013] When a newly added edge device accesses the system, the device registration unit first discovers the newly added edge device through technical means such as the MQTT communication protocol and Bluetooth Low Energy scanning; after discovering the newly added edge device, the device registration unit verifies the identity information of the newly added edge device. The identity information verification includes checking the unique identifier and digital certificate of the newly added edge device. When the identity information provided by the newly added edge device matches the legal device information preset in the collaborative management system, the identity verification passes; otherwise, the registration will be rejected; specifically, during the identity information verification of the newly added edge device, the collaborative management system assigns a unique serial number and encryption key to each legal device. The new device must provide the correct serial number and authentication information encrypted by this key when registering to ensure the authenticity of its identity; when the identity information verification of the newly added edge device passes, the device registration unit assigns corresponding permissions to it according to the type, function, and security level of the newly added edge device; the permissions include the access right to special data and the ability to perform special operations; finally, the device registration unit stores the registration information of the newly added edge device in the collaborative management system database and updates the corresponding device list and permission management data; in this way, other modules in the system can perform effective collaborative management on the newly registered edge devices based on this information.

[0014] Preferably, the process of the device scheduling unit dynamically scheduling edge devices is as follows:

[0015] The device scheduling unit first obtains the physical location information of the edge device through network communication based on the GPS positioning data reported by the edge device itself; then, through the device information reported by the edge device, it understands the functional characteristics of the edge device and determines the function of the edge device; subsequently, it continuously monitors the load conditions of the edge device, and the load conditions include CPU usage rate, memory occupancy rate, and network bandwidth usage; during this period, the sensor device can periodically send information packets containing its own location coordinates and current working status to the device scheduling unit; when a new work task appears, the device scheduling unit analyzes the work task; determines the functional type of the edge device required for the work task, the requirement for response time, and the data processing volume, and obtains the task analysis result; for example, a video surveillance task may require a device with a high-definition camera function and has a high requirement for real-time performance, and the device scheduling unit will screen suitable devices according to these requirements; according to the task analysis result and the device information collected, the device scheduling unit screens out the edge devices that meet the conditions; subsequently, it gives priority to the edge devices whose physical location is close to the task execution area to reduce data transmission delay; at the same time, it combines whether the function of the edge device matches the task requirements and whether the load condition is within the tolerable range for screening; after screening out the corresponding edge devices, the device scheduling unit makes a scheduling decision; determines that the corresponding edge device participates in the execution of the work task and allocates the corresponding task share; when multiple devices are required to cooperate to complete the task, the device scheduling unit is used to coordinate the communication and data interaction between the edge devices; the device starts to execute the work task based on the scheduling decision, and the device scheduling unit continuously monitors the status of the edge device during the execution of the work task; when it is found that the edge device has a fault, too high load, or other abnormal conditions, it makes adjustments and re-performs the screening and scheduling decision of the edge device to ensure the smooth progress of the task.

[0016] Preferably, the process by which the device monitoring unit identifies abnormal conditions during device operation and generates a device alarm signal is as follows:

[0017] The device monitoring unit continuously collects the operation status data of edge devices through the established communication connections with each edge device; the operation status data includes the CPU usage rate, memory occupancy rate, network bandwidth usage, and sensor readings of the edge devices; at the same time, it also collects the working mode and connection status of the edge devices; for example, for an industrial sensor device, the monitoring unit will obtain its measurement data and the status information of whether the device is normally connected to the network in real time; then it analyzes based on the collected operation status data of the edge devices; through the set thresholds and rules, it determines whether the edge devices are in a normal operation state; at the same time, using data analysis algorithms such as trend analysis and pattern recognition, it identifies potential abnormal situations; when the analysis result shows that the device operation status does not meet the normal standard, the device monitoring unit determines it as an abnormal situation; the abnormal situations include hardware failures, software errors, communication problems, etc.; when an abnormality is identified, the device monitoring unit immediately generates an alarm signal, and the alarm signal can include the identification information of the device, the type of abnormality, the severity, etc.; at the same time, according to the severity of the abnormality, different alarm methods are adopted, and the different alarm methods are sound alarms, pop-up prompts, sending emails and text message notifications.

[0018] The device management module further includes a device resource allocation unit for dynamically allocating computing resources, storage resources, and network resources among devices, enabling the devices to flexibly respond to changes in resource requirements, and preventing the devices from stopping working due to insufficient resources or reducing system efficiency due to resource waste.

[0019] The device management module further includes a device resource allocation unit for dynamically allocating computing resources, storage resources, and network resources among devices.

[0020] Preferably, in the multi-device collaborative management system based on cloud-edge collaboration, the optimization process for optimizing the data transmission path of network resources to ensure efficient use of bandwidth is as follows:

[0021] First, through network detection tools, understand the connection methods between different edge devices, whether they are wired or wireless connections, as well as the maximum transmission speed and average latency time of each link, determine the connection relationships between devices, the bandwidth capacity, latency, and reliability of the network links, and obtain the network topology analysis results.

[0022] Continuously monitor the data traffic situation in the network, including the amount of data sent and received by each edge device and the time distribution of the traffic; at the same time, using traffic prediction algorithms, according to historical traffic data and current task requirements, predict the future network traffic change trend for a period of time, and obtain the results of traffic monitoring and prediction.

[0023] Then, based on the results of network topology analysis and traffic monitoring and prediction, the shortest path algorithm is used to determine the optimal path for data transmission. The factors considered by the shortest path algorithm include the bandwidth, latency, reliability of the path, and the current traffic load situation;

[0024] During the data transmission process, continuously monitor the changes in network conditions and task requirements; if it is found that the current data transmission path is no longer optimal, or there are network failures or traffic congestion situations, then make dynamic adjustments in a timely manner;

[0025] In order to further improve the bandwidth utilization rate, a load balancing strategy is adopted; the data traffic is evenly distributed to different network paths to avoid overloading some paths while other paths are idle.

[0026] Preferably, the data acquisition module includes:

[0027] Data acquisition terminals, which are used to be deployed on each edge device to collect the operation data of the edge device in real time;

[0028] Data preprocessing unit, which is used to preprocess the collected operation data. The preprocessing includes denoising, data format conversion, and abnormal data elimination to obtain the original data, ensuring the accuracy and reliability of the data transmitted to the cloud analysis module;

[0029] Data storage unit, which is used to classify and store the original data obtained after preprocessing. The classified and stored data includes the operation status data, historical operation data, and environmental monitoring data of the edge device, and is also used for two-way data transmission between the edge device and the cloud, enabling seamless sharing of data between the cloud and the edge device.

[0030] Preferably, the cloud analysis module includes:

[0031] Big data analysis unit, which is used to analyze the working status of each edge device based on the original data of the edge device and generate an operation plan;

[0032] Prediction analysis unit, which is used to predict the future operation trend of the device based on the historical operation data and the current working status data to obtain the prediction result and identify the possible abnormalities and failures of the device in advance;

[0033] Optimization decision-making unit, which is used to formulate an optimized device management strategy based on the operation plan generated by the big data analysis unit and the prediction result obtained by the prediction analysis unit to optimize the collaborative working efficiency between devices;

[0034] Feedback processing unit, which is used to receive the feedback reports during the real-time operation of each edge device, adjust the new management strategy generated by the optimization decision-making unit, and ensure that the management strategy can be dynamically adjusted according to the actual operation situation of the device at any time;

[0035] Among them, the process of the big data analysis unit generating the device operation optimization plan is as follows:

[0036] Based on the original data, extract the features reflecting the device working state from the classified and stored data; among them, for the performance index data, extract the mean value, variance, maximum value, and minimum value within the time period; for the sensor data, extract the change trend and periodicity of the data; for the operation log data, extract the running time of the device and the number of task executions;

[0037] Then, adopt the feature importance evaluation method of the random forest to perform feature selection on the extracted features, remove redundant and irrelevant features, so as to improve the efficiency and accuracy of the analysis;

[0038] Subsequently, use the KMeans clustering algorithm to divide the devices into different clusters according to the feature vectors of the devices. Each cluster represents a specific working state type, and group the edge devices with similar working states; through cluster analysis, the working state distribution of the devices can be better understood, providing a basis for subsequent optimization;

[0039] Then, use a non-linear regression model to predict the future resource requirements based on the historical resource utilization data of the devices, and use the clustering algorithm to group the devices with similar working states for targeted optimization; among them, for the sensor data and performance indicators with time series characteristics, adopt the time series analysis method for prediction and anomaly detection;

[0040] According to the results of the data analysis, evaluate the working state of the devices and set evaluation indicators; by comparing the actual characteristics of the devices with the standard values of the evaluation indicators, determine the current state of the devices, and evaluate the current state of the devices as devices with problems and devices without problems;

[0041] Among them, for the devices evaluated as having problems, conduct problem diagnosis to obtain the problem diagnosis results, and combine the feature data, operation logs of the devices, and the results of the data analysis model to determine the root cause of the problems; among them, when the performance of the devices is poor, analyze the resource utilization rate, network communication status, and software operation logs of the devices, and judge the inducing factors in the state when the performance is poor. The inducing factors include hardware failures, software errors, and resource shortages;

[0042] According to the working status evaluation and problem diagnosis results of the edge device, formulate corresponding management strategies; management strategies include device configuration adjustment, task scheduling optimization, resource allocation adjustment, and software upgrade; convert management strategies into operation plans for corresponding edge devices, and evaluate the operation plans; during the evaluation period, evaluate the feasibility and cost-effectiveness of the plan and the improvement effect on device performance; for a software upgrade plan, it is necessary to consider the risks of the upgrade, the impact on the normal operation of the device, and the expected performance improvement after the upgrade; generate an operation plan for a certain device, including adjusting the power management settings of the device to reduce energy consumption, optimizing the cooling system of the device to improve stability, and evaluating the cost savings and performance improvement effects after the implementation of the operation plan; implement the operation plan on the corresponding edge device, and continuously monitor the operating status of the edge device to verify and compare the effectiveness of the operation plan; when it is found that the operation plan is not effective, adjust and improve the operation plan in a timely manner; after implementing the software upgrade plan, observe whether the performance indicators of the device are improved. If the expected effect is not achieved, further analyze the reasons and make adjustments; for example, after implementing a network resource allocation optimization plan, evaluate the implementation effect of the plan by monitoring indicators such as the network delay and upload and download speed of the device;

[0043] The process of the prediction analysis unit predicting the future operation trend of the equipment based on the historical operation data and the current working status data is as follows:

[0044] First, collect the historical operation data and current working status data of the equipment; the historical operation data includes the performance indicators, sensor data and operation logs of the equipment in the past time period; the current working status data is obtained through the data acquisition terminal, including the current values ​​of various performance indicators and the connection status of the equipment; sort out the collected data, remove abnormal values ​​and noise, and ensure the accuracy and reliability of the data; for example, remove abnormal values ​​in the temperature sensor data that obviously deviate from the normal range; then extract key features based on the historical operation data and the current working status data, and the key features include statistical features, trend features, and periodic features of the data; according to the functions and prediction requirements of the edge device, train it based on the historical operation data based on the time series analysis model, adjust the parameters of the model, so that the time series analysis model can accurately fit the historical data; input the extracted key features into the trained time series analysis model to predict the future operation trend of the equipment and obtain the prediction results; the prediction content includes the performance indicator change trend of the equipment, the possible failure risk, and the change of resource demand;

[0045] The prediction results are evaluated by comparing the prediction results with the actual data, calculating the prediction error, judging its accuracy and reliability. When the prediction results are not ideal, analyze the reasons and adjust the parameters of the time series analysis model, add more features and select different prediction models.

[0046] In the multi-device collaborative management system based on cloud-edge collaboration, the process of the optimization decision-making unit further optimizing the optimized management strategy into a new management strategy is as follows:

[0047] Obtain the operation plan generated by the big data analysis unit and the prediction results obtained by the prediction analysis unit, and at the same time obtain the real-time operation data of each edge device; evaluate the existing optimized management strategy, and analyze the effectiveness and adaptability of the optimized management strategy under the current operation conditions of the edge device; consider whether the strategy can meet the actual needs of the device and whether it can achieve the goal of improving the collaborative work efficiency; compare the differences between the prediction results and the current actual operation conditions, as well as the differences between the existing optimized management strategy and the actual needs; determine the direction of adjustment and optimization; formulate an adjustment plan according to the results of the difference analysis; determine the direction and specific measures of the adjustment, including the adjustment of device task allocation, the optimization of resource configuration and the improvement of the collaborative work process; generate a new management strategy based on the adjustment plan; the new management strategy should fully consider the results of big data analysis and prediction analysis, as well as the real-time operation conditions of the device; ensure that the new management strategy can better adapt to the dynamic changes of the device and improve the collaborative work efficiency; after the implementation of the new management strategy, continuously monitor the operation conditions of the device to verify the effectiveness of the new management strategy; if problems are found in the new management strategy or the expected effect is not achieved, give feedback and make adjustments in time.

[0048] Preferably, the collaborative control module includes:

[0049] The device control unit is used to issue instructions to each edge device according to the optimized management strategy generated by the cloud analysis module, ensuring that each edge device can execute work tasks according to the optimal strategy.

[0050] The load balancing unit is used to allocate loads among edge devices to ensure that the workload is balanced during the collaborative work of the devices, and to avoid overloading or idling of some devices.

[0051] The fault tolerance control unit is used to automatically start redundant edge devices or adjust the work tasks of other edge devices when an edge device fails or its performance degrades, ensuring the stability and reliability of the overall operation of the system.

[0052] The real-time adjustment unit is used to dynamically adjust the work tasks of edge devices according to the real-time operation conditions of edge devices, ensuring the flexibility and efficiency of the system operation.

[0053] Preferably, the data feedback module includes:

[0054] A feedback analysis unit for analyzing the operation data, identifying potential problems in the execution of the management policy by the edge device, and generating a feedback report;

[0055] A feedback transmission unit for transmitting the feedback report generated by the feedback analysis unit to the cloud analysis module, enabling the cloud analysis module to perform further optimization analysis and policy adjustment based on the feedback report.

[0056] The present invention provides a multi-device collaborative management system based on cloud-edge collaboration, having the following beneficial effects:

[0057] First, in the multi-device collaborative management system based on cloud-edge collaboration, the edge devices are centrally managed and optimized for scheduling through the device management module. Dynamic scheduling is performed according to the physical location, function, and load conditions of the devices, enabling the devices to execute tasks at the most appropriate time and location, greatly improving the collaborative working efficiency among multiple devices. The collaborative control module can perform collaborative control on each edge device according to the optimized management policy generated by the cloud analysis module to ensure the coordinated operation among the devices. That is to say, when facing complex tasks and a dynamically changing environment, the system has the ability of intelligent decision-making, can automatically adjust the collaborative policy according to the real-time situation without manual intervention, and realizes the collaborative management and control among multiple devices, with the intelligent decision-making ability guaranteed.

[0058] Second, in the multi-device collaborative management system based on cloud-edge collaboration, the device resource allocation unit can realize the dynamic allocation of computing resources, storage resources, and network resources among the devices; reasonably allocate resources according to the actual needs and load conditions of the devices to avoid resource waste and bottleneck problems; for network resources, optimize the data transmission path to ensure the efficient utilization of bandwidth; adopt technologies such as the shortest path algorithm to select the optimal data transmission path according to factors such as the network topology structure and traffic monitoring prediction results, improve the network transmission efficiency, and reduce latency and congestion; the device monitoring unit can monitor the operation status of each edge device in real time, identify abnormal situations during device operation, and generate device alarm signals; promptly discover device failures, excessive loads, or other abnormal situations to take corresponding measures for handling and ensure the stable operation of the system; the fault tolerance control unit in the collaborative control module can start the working tasks of redundant edge devices when an edge device fails to ensure the continuous operation of the system; the prediction analysis unit predicts the future operation trend of the device based on historical operation data and the current working state, providing forward-looking decision-making support for device management.

[0059] III. For the multi-device collaborative management system based on cloud-edge collaboration, the data acquisition module can collect the operation data of each edge device in real time and store the collected data in a classified manner. At the same time, the data storage unit also supports bidirectional data transmission between the edge device and the cloud, enabling seamless sharing of data between the cloud and the edge device. This provides a comprehensive and accurate data foundation for multi-device collaborative management, promoting information exchange and collaborative work between devices. The optimization decision-making unit can formulate an optimized device management strategy based on the operation plan generated by the big data analysis unit and the prediction results obtained by the prediction analysis unit, and continuously optimize and adjust according to the real-time operation conditions of each edge device. This dynamic optimization decision-making mechanism enables the system to adapt to the constantly changing device operation environment and task requirements, achieving more efficient multi-device collaborative management. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a flowchart of a multi-device collaborative management system based on cloud-edge collaboration according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0061] The present invention will be further described in detail below in conjunction with the drawings and specific embodiments. The embodiments of the present invention are given for the purpose of illustration and description, and are not exhaustive or limit the present invention to the disclosed form. Many modifications and variations are obvious to those of ordinary skill in the art. The embodiments are selected and described to better illustrate the principles and practical applications of the present invention, and enable those of ordinary skill in the art to understand the present invention and design various embodiments with various modifications suitable for specific purposes.

[0062] As Figure 1 shown, the present invention provides a technical solution: a multi-device collaborative management system based on cloud-edge collaboration, including a device management module for connecting edge devices and centrally managing the edge devices, and performing optimized scheduling based on the functions and distribution positions of different edge devices;

[0063] a data acquisition module connected to the device management module for collecting the operation data of each edge device in real time, where the operation data includes operation status data, device operation data, and environmental monitoring data, and storing the collected data in a classified manner;

[0064] a cloud analysis module connected to the data acquisition module for receiving the classified and stored data and analyzing the data of each edge device based on big data analysis algorithms to formulate an optimized device management strategy;

[0065] a collaborative control module connected to the device management module and the cloud analysis module for collaboratively controlling each edge device according to the optimized management strategy generated by the cloud analysis module to ensure optimal allocation of resources and collaborative work during the collaborative operation of the devices;

[0066] A data feedback module, connected to the collaborative control module, is used to feedback the real-time operation status of each edge device to the cloud analysis module.

[0067] The device management module includes:

[0068] A device registration unit, used to perform identity registration and permission management for newly added edge devices, ensuring that new edge devices can obtain corresponding permissions in the cloud-edge collaborative system and then join the collaborative management;

[0069] A device scheduling unit, used to dynamically schedule edge devices based on the physical location, function, and load conditions of the devices, ensuring the workload balance of each edge device;

[0070] A device monitoring unit, used to monitor the operation status of each edge device in real time, identify abnormal situations during the operation of edge devices, and generate device alarm signals.

[0071] The process of the device registration unit performing identity registration and permission management for newly added edge devices is as follows:

[0072] When a newly added edge device accesses the system, the device registration unit first discovers the newly added edge device through technical means such as the MQTT communication protocol and Bluetooth Low Energy scanning; after discovering the newly added edge device, the device registration unit verifies the identity information of the newly added edge device. The identity information verification includes checking the unique identifier and digital certificate of the newly added edge device. When the identity information provided by the newly added edge device matches the legal device information preset in the collaborative management system, the identity verification passes; otherwise, the registration will be rejected; when the identity information verification of the newly added edge device passes, the device registration unit assigns corresponding permissions to it according to the type, function, and security level of the newly added edge device; the permissions include the access right to special data and the ability to perform special operations; individual edge devices will be given higher permissions to be able to access sensitive data and perform important control operations, while ordinary edge devices only have limited permissions to ensure the security and stability of the system; finally, the device registration unit stores the registration information of the newly added edge device in the collaborative management system database and updates the corresponding device list and permission management data; in this way, other modules in the system can perform effective collaborative management on the newly registered edge device according to this information; when other devices need to communicate or cooperate with the newly registered device, they can query its permissions and connection information from the database to ensure the security and smooth progress of the communication.

[0073] It should be further noted that during the specific implementation process, when specifically verifying the identity information of newly added edge devices, the collaborative management system assigns a unique serial number and encryption key to each legitimate device. When a new device registers, it must provide the correct serial number and authentication information encrypted with this key to ensure the authenticity of its identity.

[0074] The process of the device scheduling unit dynamically scheduling edge devices is as follows:

[0075] The device scheduling unit first obtains the physical location information of the edge device through network communication based on the GPS positioning data reported by the edge device itself; then, through the device information reported by the edge device, it understands the functional characteristics of the edge device and determines the function of the edge device. For example, if the device information reported by the edge device mentions that the edge device is a sensor, an actuator, or a computing device; subsequently, it continuously monitors the load situation of the edge device, and the load situation includes CPU usage rate, memory occupancy rate, and network bandwidth usage; during this period, the sensor device can regularly send information packets containing its own location coordinates and current working status to the device scheduling unit; when a new work task appears, the device scheduling unit analyzes the work task; determines the functional type of the edge device required for the work task, the requirement for response time, and the data processing volume, and obtains the task analysis result; for example, a video surveillance task may require a device with a high-definition camera function and has high real-time requirements, and the device scheduling unit will screen suitable devices according to these requirements.

[0076] Based on the task analysis results and the device information collected, the device scheduling unit filters out eligible edge devices; subsequently, it gives priority to edge devices physically close to the task execution area to reduce data transmission latency; at the same time, it filters in combination with whether the functions of the edge devices match the task requirements and whether the load conditions are within the acceptable range; for a task of environmental monitoring in a specific area, the device scheduling unit will select an edge device with a light load and environmental monitoring function near the area for the allocation of corresponding work tasks; after filtering out the corresponding edge devices, the device scheduling unit makes a scheduling decision; determines that the corresponding edge device participates in the execution of the work task and allocates the corresponding task share; when multiple devices are required to cooperate to complete a task, the device scheduling unit is used to coordinate the communication and data interaction between edge devices; for a complex data processing task, multiple computing devices may be scheduled to perform calculations together and the data distribution and result aggregation methods are determined; the device starts to execute the work task based on the scheduling decision, and the device scheduling unit continuously monitors the status of edge devices during the execution of the work task; when it is found that an edge device has a fault, an excessive load or other abnormal conditions, adjustments are made, and the edge device filtering and scheduling decision are re-made to ensure the smooth progress of the task; that is to say, when a device suddenly fails during the execution of a task, the device scheduling unit will quickly select other suitable edge devices to replace it and re-allocate the work task.

[0077] It should be further noted that in the specific implementation process, the device management module centrally manages and optimally schedules edge devices, and performs dynamic scheduling according to the physical location, function and load conditions of the devices, enabling the devices to execute tasks at the most appropriate time and location, greatly improving the collaborative work efficiency among multiple devices; for example, in an intelligent factory, devices with different functions can quickly respond and cooperate according to the requirements of production tasks, reducing production time and costs.

[0078] The process by which the device monitoring unit identifies abnormal conditions during device operation and generates a device alarm signal is as follows:

[0079] The device monitoring unit continuously collects the operation status data of edge devices through the established communication connections with each edge device; the operation status data includes the CPU usage rate, memory occupancy rate, network bandwidth usage, and sensor readings of the edge devices; at the same time, it also collects the working mode and connection status of the edge devices; for example, for an industrial sensor device, the monitoring unit will obtain its measurement data in real time and the status information of whether the device is normally connected to the network; then it analyzes based on the collected operation status data of the edge devices; by setting thresholds and rules, it determines whether the edge devices are in a normal operation state; for example, if the CPU usage rate exceeds a certain percentage, or the sensor readings exceed the preset reasonable range, it may be regarded as an abnormal situation; at the same time, it uses data analysis algorithms, such as trend analysis, pattern recognition, etc., to identify potential abnormal situations; for example, by observing the change trend of sensor data, if there is a sudden large fluctuation, it may indicate that there is a problem with the device; when the analysis result shows that the device operation status does not meet the normal standard, the device monitoring unit determines it as an abnormal situation; abnormal situations include hardware failures, software errors, communication problems, etc.; for example, if the device suddenly loses connection, or continuously outputs incorrect sensor data, it is identified as abnormal; when an abnormality is identified, the device monitoring unit immediately generates an alarm signal.

[0080] Among them, the alarm signal can include the identification information of the device, the type of abnormality, the severity, etc.; at the same time, according to the severity of the abnormality, different alarm methods are adopted, and different alarm methods include sound alarms, pop-up prompts, sending emails, and text message notifications; for example, for a serious device failure, a strong sound alarm will be issued and a text message notification will be sent to the administrator at the same time so that timely measures can be taken for handling.

[0081] The device management module further includes a device resource allocation unit for dynamically allocating computing resources, storage resources, and network resources among devices, enabling the devices to flexibly respond to changes in resource requirements, and preventing the devices from stopping working due to insufficient resources or reducing system efficiency due to resource waste.

[0082] The device management module further includes a device resource allocation unit for dynamically allocating computing resources, storage resources, and network resources among devices.

[0083] In the multi-device collaborative management system based on cloud-edge collaboration, the optimization process for optimizing the data transmission path of network resources to ensure efficient use of bandwidth is as follows:

[0084] First, through network detection tools, understand the connection methods between different edge devices, whether they are wired connections or wireless connections, as well as the maximum transmission speed and average delay time of each link, determine the connection relationships between devices, the bandwidth capacity, delay conditions, and reliability of the network links, and obtain the network topology analysis results.

[0085] Continuously monitor the data traffic situation in the network, including the amount of data sent and received by each edge device and the time distribution of the traffic; at the same time, use a traffic prediction algorithm to predict the future trend of network traffic changes based on historical traffic data and current task requirements, and obtain the results of traffic monitoring and prediction; for example, by analyzing the traffic pattern of a certain device during a specific period in the past, predict the possible traffic volume of the device during a similar period.

[0086] Then, based on the results of network topology analysis and traffic monitoring and prediction, use the shortest path algorithm to determine the optimal path for data transmission. The factors considered by the shortest path algorithm include the bandwidth, delay, reliability of the path, and the current traffic load situation; for example, when using the Dijkstra algorithm, use the bandwidth of the path as the weight, and select the path with the largest bandwidth and the smallest delay as the data transmission path.

[0087] During the data transmission process, continuously monitor the changes in network conditions and task requirements; if it is found that the current data transmission path is no longer optimal, or there are network failures or traffic congestion, then make dynamic adjustments in a timely manner; for example, if the bandwidth of a certain link suddenly drops or there is a failure, recalculate and select a new optimal path for data transmission.

[0088] In order to further improve the bandwidth utilization rate, adopt a load balancing strategy; evenly distribute the data traffic to different network paths to avoid overloading some paths while other paths are idle; that is to say, when a large amount of data is sent simultaneously by multiple devices, the data can be distributed to different paths for transmission according to the bandwidth and load conditions of the paths to ensure the load balance of the entire network.

[0089] The device resource allocation unit can realize the dynamic allocation of computing resources, storage resources, and network resources among devices; according to the actual needs and load conditions of the devices, reasonably allocate resources to avoid resource waste and bottleneck problems.

[0090] The data acquisition module includes:

[0091] Data acquisition terminals, which are used to be deployed on each edge device to collect the operation data of the edge device in real time;

[0092] A data preprocessing unit, which is used to preprocess the collected operation data. The preprocessing includes denoising, data format conversion, and abnormal data elimination to obtain the original data and ensure the accuracy and reliability of the data transmitted to the cloud analysis module.

[0093] A data storage unit is used to classify and store the original data obtained after preprocessing. The classified and stored data includes the operating status data, historical operation data, and environmental monitoring data of the edge devices. It is also used for two-way data transmission between the edge devices and the cloud, enabling seamless sharing of data between the cloud and the edge devices.

[0094] It should be further noted that in the specific implementation process, the data acquisition module can collect the operating data of each edge device in real time and classify and store the collected data. At the same time, the data storage unit also supports two-way data transmission between the edge devices and the cloud, enabling seamless sharing of data between the cloud and the edge devices. This provides a comprehensive and accurate data basis for multi-device collaborative management, promoting information exchange and collaborative work between devices. For example, in a smart home system, different smart devices can achieve more intelligent scenario linkage and services through shared data.

[0095] The cloud analysis module includes:

[0096] A big data analysis unit is used to analyze the working status of each edge device based on the original data of the edge devices and generate an operation plan.

[0097] A predictive analysis unit is used to predict the future operation trend of the device based on the historical operation data and the current working status data, obtain a prediction result, and identify potential anomalies and faults of the device in advance.

[0098] An optimization decision unit is used to formulate an optimized device management strategy based on the operation plan generated by the big data analysis unit and the prediction result obtained by the predictive analysis unit, and optimize the collaborative work efficiency between devices.

[0099] A feedback processing unit is used to receive the feedback reports during the real-time operation of each edge device, adjust the new management strategy generated by the optimization decision unit, and ensure that the management strategy can be dynamically adjusted according to the actual operation situation of the device at any time.

[0100] Among them, the process of the big data analysis unit generating the device operation optimization plan is as follows:

[0101] Based on the original data, extract the features reflecting the device working status from the classified and stored data. Among them, for the performance index data, extract the mean, variance, maximum value, and minimum value within a time period; for the sensor data, extract the change trend and periodicity of the data; for the operation log data, extract the running time and task execution times of the device. For example, extract features such as the average usage rate and the standard deviation of the usage rate within the past hour from the CPU usage rate data, and extract features such as the rising trend and the fluctuation range of the temperature from the temperature sensor data.

[0102] Then, the feature importance evaluation method of random forest is adopted to perform feature selection on the extracted features, removing redundant and irrelevant features to improve the efficiency and accuracy of analysis; that is to say, the correlation between each feature and the device working state can be calculated, and features with higher correlation can be selected for subsequent analysis;

[0103] Subsequently, the KMeans clustering algorithm is used to divide the devices into different clusters according to the feature vectors of the devices. Each cluster represents a specific type of working state, grouping edge devices with similar working states; through clustering analysis, the distribution of the device working states can be better understood, providing a basis for subsequent optimization; for example, a group of industrial devices are clustered according to their features such as temperature and CPU usage rate, divided into three clusters: high load, medium load, and low load;

[0104] Then, a non - linear regression model is used to predict future resource requirements based on the historical resource utilization data of the devices, and the clustering algorithm is used to group devices with similar working states for targeted optimization; for example, a non - linear regression model is established to predict the relationship between the response time of the device and CPU usage rate, network latency; for the relationship between the performance indicators (such as response time, throughput, etc.) of the device and influencing factors (such as device load, network conditions, etc.); predicting the performance of the device under different conditions, providing a reference for resource allocation and task scheduling; among them, for sensor data and performance indicators with time - series characteristics, time - series analysis methods are used for prediction and anomaly detection; similarly, the autoregressive integrated moving average model can be used to predict the temperature data of the device to detect potential overheating problems in advance; the isolation forest algorithm in the anomaly detection algorithm is used to detect abnormal changes in the device performance indicators to timely detect device failures; through time - series analysis of the network traffic data of the device, predicting the network traffic trend in the next period of time, providing a basis for network resource allocation;

[0105] According to the results of data analysis, the working state of the device is evaluated, and evaluation indicators are set, such as device load level (high, medium, low), performance performance (excellent, good, average, poor), health status (normal, warning, failure); by comparing the actual features of the device with the standard values of the evaluation indicators, the current state of the device is determined, and the current state of the device is evaluated as devices with problems and devices without problems; for example, according to indicators such as the CPU usage rate and memory occupancy rate of the device, the device load level is divided into three levels: high, medium, and low; if a device's CPU usage rate exceeds 80% and memory occupancy rate exceeds 70%, then the device is considered to be in a high - load state;

[0106] Among them, for the devices evaluated as having problems, problem diagnosis is carried out to obtain the problem diagnosis results, and the root cause of the problem is determined by combining the characteristic data of the devices, operation logs, and the results of the data analysis model. Among them, when the performance of the device is poor, the resource utilization rate, network communication status, and software operation logs of the device are analyzed to judge the inducing factors in the state of poor performance. The inducing factors include hardware failures, software errors, and resource shortages. For example, if the response time of a device increases significantly, by checking indicators such as network latency, CPU usage, and disk read / write speed, it is determined whether the problem is caused by network problems or excessive device load.

[0107] According to the working state evaluation and problem diagnosis results of the edge devices, corresponding management strategies are formulated. The management strategies include device configuration adjustment, task scheduling optimization, resource allocation adjustment, and software upgrade. For example, for high-load devices, task priorities can be adjusted, or more computing resources or storage resources can be allocated. For devices with network problems, network configurations can be optimized, data transmission paths can be adjusted, etc. For example, if a device has an over-high temperature, possible management strategies include adding heat dissipation devices and adjusting the working frequency of the device to reduce heat generation. The management strategies are transformed into operation plans for the corresponding edge devices, and the operation plans are evaluated. During the evaluation, the feasibility, cost-effectiveness, and the improvement effect on device performance of the plans are considered. For example, for a software upgrade plan, the risks of the upgrade, the impact on the normal operation of the device, and the expected performance improvement after the upgrade need to be considered. For example, an operation plan for a certain device is generated, including adjusting the power management settings of the device to reduce energy consumption and optimizing the heat dissipation system of the device to improve stability, and the cost savings and performance improvement effects after the implementation of the operation plan are evaluated. The operation plan is implemented on the corresponding edge devices, and the running states of the edge devices are continuously monitored to verify and compare the effectiveness of the operation plans. When it is found that the effect of the operation plan is not good, the operation plan is adjusted and improved in a timely manner. For example, after implementing a software upgrade plan, observe whether the performance indicators of the device are improved. If the expected effect is not achieved, further analyze the reasons and make adjustments. For example, after implementing a network resource allocation optimization plan, evaluate the implementation effect of the plan by monitoring indicators such as the network latency and upload / download speed of the device.

[0108] Among them, the process by which the prediction analysis unit predicts the future operation trend of the device based on historical operation data and current working state data is as follows:

[0109] First, collect the historical operation data and current working status data of the equipment; the historical operation data includes the performance indicators of the equipment in the past time period (such as CPU usage, memory occupancy, network bandwidth usage, etc.), sensor data (such as temperature, humidity, pressure, etc.) and the operation log of the equipment; the current working status data is obtained through the data acquisition terminal, including the current values ​​of various performance indicators and the connection status of the equipment; sort out the collected data, remove abnormal values ​​and noise, and ensure the accuracy and reliability of the data; for example, remove abnormal values ​​in the temperature sensor data that obviously deviate from the normal range; then extract key features based on the historical operation data and the current working status data, the key features include statistical features of the data (such as mean, variance, maximum value, minimum value, etc.), trend features (such as rising or falling trends of the data), periodic features (if the data has periodic waves For example, extract the average usage rate and usage rate change trend in the past period of time from the CPU usage rate data; according to the function and prediction requirements of the edge device, train it with historical operation data based on the time series analysis model, adjust the parameters of the model, so that the time series analysis model can accurately fit the historical data; for example, use the time series analysis model to train the temperature data of the device, and by continuously adjusting the parameters of the model, the model can better capture the time series characteristics of the temperature data; input the extracted key features into the trained time series analysis model to predict the future operation trend of the device and obtain the prediction results; the prediction content includes the performance indicator change trend of the device, the possible failure risk, and the change of resource demand; for example, predict whether the CPU usage of the device will continue to rise in the future, or whether the device is at risk of failure due to overheating;

[0110] Evaluate the prediction results by comparing them with the actual data and calculating the prediction error; judge their accuracy and reliability; when the prediction results are not ideal, analyze the reasons and adjust the parameters of the time series analysis model, add more features and choose different prediction models; for example, if it is found that the predicted CPU usage rate deviates greatly from the actual value, further analyze the data features, adjust the model parameters or try other more suitable models;

[0111] In the multi-device collaborative management system based on cloud-edge collaboration, the simplified process of the optimization decision unit further optimizing the optimization management strategy into a new management strategy is as follows:

[0112] Obtain the operation plan generated by the big data analysis unit and the prediction results obtained by the prediction analysis unit, and at the same time obtain the real-time operation data of each edge device; Evaluate the existing optimization management strategy and analyze the effectiveness and adaptability of the optimization management strategy under the current operation conditions of the edge device; Consider whether the strategy can meet the actual needs of the device and whether it can achieve the goal of improving the collaborative work efficiency; For example, check whether the regulations on device resource allocation in the optimization management strategy match the current load conditions of the device; Compare the differences between the prediction results and the current actual operation conditions, as well as the differences between the existing optimization management strategy and the actual needs; Determine the direction of adjustment and optimization; For example, if the prediction results show that the future device load will increase and the existing strategy does not have corresponding resource allocation measures, then this is a difference point that needs to be optimized; According to the results of the difference analysis, formulate an adjustment plan; Determine the direction and specific measures of the adjustment; including the adjustment of device task allocation, the optimization of resource configuration, and the improvement of the collaborative work process; For example, if it is found that some devices may experience resource shortages in the future, the plan is to transfer some tasks from these devices to devices with abundant resources; Based on the adjustment plan, generate a new management strategy; The new management strategy should fully consider the results of big data analysis and prediction analysis, as well as the real-time operation conditions of the device; Ensure that the new management strategy can better adapt to the dynamic changes of the device and improve the collaborative work efficiency; For example, formulate a new resource allocation plan and dynamically adjust the allocation of computing resources, storage resources, and network resources according to the device load prediction and real-time status; After the implementation of the new management strategy, continuously monitor the operation conditions of the device to verify the effectiveness of the new management strategy; If problems are found in the new management strategy or the expected effect is not achieved, give feedback and make adjustments in a timely manner; For example, evaluate the effectiveness of the new management strategy by comparing the collaborative work efficiency indicators of the device before and after the implementation of the new management strategy, and further optimize according to the actual situation.

[0113] Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art and related fields without creative efforts shall fall within the protection scope of the present invention. The structures, devices, and operation methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art without special instructions and limitations.

Claims

1. A multi-device collaborative management system based on cloud-edge collaboration, characterized in that: include: The device management module is used to connect edge devices and centrally manage them, optimizing the scheduling based on the functions of different edge devices and their distribution locations; The device management module includes: a device scheduling unit for dynamically scheduling edge devices based on the physical location, function and load of the device; The data acquisition module is connected to the device management module and is used to collect the operating data of each edge device in real time, including operating status data, device operation data and environmental monitoring data, and classify and store the collected data; The cloud analysis module is connected to the data acquisition module to receive the classified and stored data, analyze the data of each edge device based on the big data analysis algorithm, and formulate device optimization management strategies; A collaborative control module, connected to the device management module and the cloud analysis module, is used to collaboratively control each edge device according to the optimization management strategy generated by the cloud analysis module; The data feedback module is connected to the collaborative control module and is used to feed back the real-time operation status of each edge device to the cloud analysis module; The process of the device scheduling unit dynamically scheduling the edge device is as follows: The device scheduling unit first obtains the physical location information of the edge device through network communication based on the GPS positioning data reported by the edge device itself; then, through the device information reported by the edge device, it understands the functional characteristics of the edge device, determines the function of the edge device, and then continuously monitors the load of the edge device, which includes CPU usage, memory occupancy, and network bandwidth usage; when a new work task appears, the device scheduling unit analyzes the work task; determines the functional type of the edge device required for the work task, the response time requirement, and the data processing volume, and obtains the task analysis result; based on the task analysis result and the collected device information, the device scheduling unit selects the edge devices that meet the conditions; then optimizes First consider edge devices that are physically close to the task execution area; at the same time, screen the edge devices based on whether their functions match the task requirements and whether the load conditions are within an acceptable range; after screening the corresponding edge devices, the device scheduling unit makes a scheduling decision; determines that the corresponding edge devices participate in the execution of the work task, and allocates the corresponding task share; when multiple devices are required to collaborate to complete a task, the device scheduling unit is used to coordinate the communication and data interaction between edge devices; the device starts to execute the work task based on the scheduling decision, and the device scheduling unit continuously monitors the status of the edge devices during the execution of the work task; when it is found that the edge device fails or the load is too high, adjustments are made, and the edge device screening and scheduling decisions are made again.

2. According to claim 1, a multi-device collaborative management system based on cloud-edge collaboration is characterized in that: The device management module includes: Device registration unit, used to register the identity and manage the permissions of newly added edge devices; The device monitoring unit is used to monitor the operating status of each edge device in real time, identify abnormal conditions in the operation of the edge device and generate device alarm signals.

3. According to claim 2, a multi-device collaborative management system based on cloud-edge collaboration is characterized in that: The process of the device registration unit performing identity registration and authority management on the newly added edge device is as follows: When a new edge device is connected to the system, the device registration unit first discovers the new edge device through the MQTT communication protocol and Bluetooth low energy scanning technology; after discovering the new edge device, the device registration unit verifies the identity information of the new edge device. The identity information verification includes checking the unique identifier and digital certificate of the new edge device. When the identity information provided by the new edge device matches the legitimate device information pre-set by the collaborative management system, the identity authentication is passed; otherwise, the registration will be rejected; when the identity information of the new edge device is verified, the device registration unit assigns corresponding permissions to it according to the type, function and security level of the new edge device; permissions include access to special data and the ability to perform special operations; finally, the device registration unit stores the registration information of the new edge device in the collaborative management system database, and updates the corresponding device list and permission management data.

4. According to claim 3, a multi-device collaborative management system based on cloud-edge collaboration is characterized in that: The process of the equipment monitoring unit identifying abnormal conditions in equipment operation and generating equipment alarm signals is as follows: The device monitoring unit continuously collects the operating status data of the edge devices through the communication connection established with each edge device; the operating status data includes the CPU usage rate, memory occupancy rate, network bandwidth usage and sensor readings of the edge devices; at the same time, the working mode and connection status of the edge devices are also collected; then, analysis is performed based on the collected edge device operating status data; through the set thresholds and rules, it is determined whether the edge device is in a normal operating state; when the analysis results show that the operating status of the device does not meet the normal standards, the device monitoring unit determines it as an abnormal situation; when an abnormality is identified, the device monitoring unit immediately generates an alarm signal; The device management module also includes a device resource allocation unit for implementing dynamic allocation of computing resources, storage resources and network resources among devices.

5. According to claim 4, a multi-device collaborative management system based on cloud-edge collaboration is characterized in that: In the multi-device collaborative management system based on cloud-edge collaboration, the optimization process for optimizing the data transmission path for network resources to ensure efficient use of bandwidth is as follows: First, use network detection tools to understand whether the connection between different edge devices is wired or wireless, as well as the maximum transmission speed and average delay time of each link, determine the connection relationship between each device, the bandwidth capacity, delay and reliability of the network link, and obtain the network topology analysis results; Continuously monitor the data traffic in the network, including the amount of data sent and received by each edge device and the time distribution of the traffic. At the same time, use the traffic prediction algorithm to predict the trend of network traffic changes in the future based on historical traffic data and current task requirements, and obtain the results of traffic monitoring prediction. Then, based on the results of network topology analysis and traffic monitoring prediction, the shortest path algorithm is used to determine the optimal path for data transmission. The shortest path algorithm takes into account factors such as the path's bandwidth, delay, reliability, and current traffic load. During the data transmission process, continuously monitor changes in network conditions and task requirements; if it is found that the current data transmission path is no longer optimal, or if network failures or traffic congestion occur, make dynamic adjustments in a timely manner.

6. A multi-device collaborative management system based on cloud-edge collaboration according to claim 5, characterized in that: The data acquisition module comprises: Data collection terminal, which is deployed on each edge device to collect the operating data of the edge device in real time; A data preprocessing unit is used to preprocess the collected operation data, including denoising, data format conversion and abnormal data elimination to obtain original data; The data storage unit is used to classify and store the raw data obtained after preprocessing. The classified stored data includes the operating status data, historical operation data and environmental monitoring data of the edge device. It is also used for two-way transmission of data between the edge device and the cloud, so that data can be seamlessly shared between the cloud and edge devices.

7. The multi-device collaborative management system based on cloud-edge collaboration according to claim 6, characterized in that: The cloud analysis module includes: A big data analysis unit, used to analyze the working status of each edge device based on the original data of the edge device and generate an operation plan; A prediction analysis unit is used to predict the future operation trend of the equipment based on historical operation data and current working status data to obtain prediction results; An optimization decision-making unit is used to formulate an equipment optimization management strategy based on the operation plan generated by the big data analysis unit and the prediction results obtained by the prediction analysis unit; A feedback processing unit is used to receive feedback reports from each edge device in real time and adjust the new management strategy generated by the optimization decision unit; The process of the big data analysis unit generating the equipment operation optimization plan is as follows: Based on the original data, extract the features reflecting the working status of the equipment from the classified and stored data; for the performance indicator data, extract the mean, variance, maximum and minimum values ​​within the time period; for the sensor data, extract the change trend and periodicity of the data; for the operation log data, extract the running time of the equipment and the number of task executions; then use the random forest feature importance evaluation method to select the extracted features and remove redundant and irrelevant features. Then use the KMeans clustering algorithm to divide the equipment into different clusters according to the feature vector of the equipment. Each cluster represents a specific working status type, and the edge devices with similar working status are grouped; Then, a nonlinear regression model is used to predict future resource requirements based on the historical resource utilization data of the equipment, and a clustering algorithm is used to group equipment with similar working status. For sensor data and performance indicators with time series characteristics, time series analysis methods are used for prediction and anomaly detection. Based on the results of data analysis, the working status of the equipment is evaluated and evaluation indicators are set; the current status of the equipment is determined by comparing the actual characteristics of the equipment with the standard values ​​of the evaluation indicators, and the current status of the equipment is evaluated and divided into equipment with problems and equipment without problems; For devices that are evaluated to have problems, problem diagnosis is performed to obtain problem diagnosis results, and the root cause of the problem is determined by combining the device's characteristic data, operation logs, and the results of the data analysis model; when the performance of the device is poor, the resource utilization rate, network communication status, and software operation logs of the device are analyzed to determine the inducing factors of the poor performance, wherein the inducing factors include hardware failure, software error, and insufficient resources; Formulate management strategies based on the working status evaluation and problem diagnosis results of edge devices; management strategies include device configuration adjustment, task scheduling optimization, resource allocation adjustment, and software upgrade; convert management strategies into operation plans for edge devices and evaluate the operation plans; during the evaluation period, evaluate the feasibility and cost-effectiveness of the plans as well as the effect of improving device performance, implement the operation plans on edge devices, continuously monitor the operating status of edge devices, and verify and compare the effectiveness of the operation plans; when the operation plan is found to be ineffective, adjust and improve the operation plan; The process of the prediction analysis unit predicting the future operation trend of the equipment based on the historical operation data and the current working status data is as follows: First, collect the historical operation data and current working status data of the equipment; the historical operation data includes the performance indicators, sensor data and operation logs of the equipment in the past time period; the current working status data is obtained through the data acquisition terminal, including the current performance indicator values ​​and the connection status of the equipment; sort the collected data to remove outliers and noise; then extract key features from the historical operation data and the current working status data, the key features include the statistical features, trend features and periodic features of the data; according to the functions and prediction requirements of the edge device, train it based on the historical operation data based on the time series analysis model, adjust the parameters of the model, and make the time series analysis model fit the historical data; input the extracted key features into the trained time series analysis model to predict the future operation trend of the equipment and obtain the prediction results; the prediction content includes the performance indicator change trend of the equipment, the risk of failure, and the change of resource demand; Evaluate the forecast results by comparing them with the actual data and calculating the forecast error; when the forecast results are not ideal, analyze the reasons and adjust the parameters of the time series analysis model; In the multi-device collaborative management system based on cloud-edge collaboration, the simplified process of the optimization decision unit further optimizing the optimization management strategy into a new management strategy is as follows: Obtain the operation plan generated by the big data analysis unit and the prediction results obtained by the prediction analysis unit, and obtain the real-time operation data of each edge device; evaluate the existing optimization management strategy, and analyze the effectiveness and adaptability of the optimization management strategy under the current edge device operation conditions; compare the difference between the prediction results and the current actual operation conditions, as well as the difference between the existing optimization management strategy and the actual needs; determine the direction that needs to be adjusted and optimized; formulate an adjustment plan based on the results of the difference analysis; determine the direction and specific measures of the adjustment; including the adjustment of equipment task allocation, optimization of resource allocation and improvement of collaborative workflow; generate a new management strategy based on the adjustment plan; after the implementation of the new management strategy, continuously monitor the operation of the equipment to verify the effectiveness of the new management strategy.

8. The multi-device collaborative management system based on cloud-edge collaboration according to claim 7, characterized in that: The collaborative control module includes: The device control unit is used to issue instructions to each edge device according to the optimization management strategy generated by the cloud analysis module; A load balancing unit, used to distribute the load among edge devices; A fault-tolerant control unit, used to start the working tasks of the redundant edge devices when an edge device fails; The real-time adjustment unit is used to dynamically adjust the work tasks of the edge device according to the real-time operation status of the edge device.

9. The multi-device collaborative management system based on cloud-edge collaboration according to claim 8, characterized in that: The data feedback module comprises: A feedback analysis unit, which is used to analyze the operation data, identify potential problems in the edge device execution management strategy, and generate feedback reports; The feedback transmission unit is used to transmit the feedback report generated by the feedback analysis unit to the cloud analysis module, so that the cloud analysis module can perform further optimization analysis and strategy adjustment based on the feedback report.

Images