Terminal offline detection and self-restart method and device

Abstract

The present invention relates to the field of power systems, and in particular to a terminal offline detection and self-restart method and device, including collecting operation data of power terminal equipment, preprocessing the collected data, and analyzing to obtain characteristic information; combining the preprocessed terminal data information, establishing a terminal status detection neural network model, inputting the real-time collected data into a trained deep learning model, and judging the offline status of the terminal equipment through model detection; monitoring the self-restart process according to the obtained terminal historical recovery time and basic cost, and combining the output results of the terminal status detection neural network model; comparing the obtained terminal self-restart success efficiency with the preset terminal self-restart success efficiency, and optimizing the set self-restart monitoring benchmark. The present invention realizes real-time monitoring, autonomous decision-making and automatic recovery, and improves the reliability, availability and fault handling capabilities of the power system.

Description

Technical Field

[0001] The present invention relates to the field of power systems, and in particular to a terminal offline detection and self-restart method and device. Background Art

[0002] The offline detection and automatic restart method of power terminals is an innovative technology applied in the power field. Through the offline detection mechanism and automatic restart steps, the offline status monitoring and automatic recovery of power terminal equipment are realized to ensure the continuous power supply of the power system. In the traditional power system, it is necessary to rely on manual judgment of the offline status, which may be affected by the operator's experience and subjective factors, resulting in inaccurate judgment or delayed response; due to the need to wait for manual intervention, the detection and recovery of the offline status may be delayed, resulting in a longer power outage; manual detection and recovery require a lot of human resources, which increases the cost and complexity of system operation and maintenance.

[0003] With the development of advanced sensor, communication and automation technologies, the reliability and stability of power equipment can be improved, but the existing technology still has the following problems: the terminal status parameters obtained through various sensors are large and complex, and the data cannot be effectively processed and applied; the terminal cannot make decisions and controls autonomously based on the processed data, including triggering self-restart operations, sending alarm notifications, etc.; and the self-restart process cannot be effectively monitored and controlled, resulting in a decrease in the user's power supply quality and experience, and an increase in system costs.

[0004] Therefore, a terminal offline detection and automatic restart method and device are needed to achieve real-time monitoring, autonomous decision-making and automatic recovery to improve the reliability and availability of the power system and reduce the need for manual intervention. Summary of the invention

[0005] The present invention provides a terminal offline detection and automatic restart method and device, which aims to utilize advanced sensors, automated detection, and tracking and monitoring of the automatic restart process to achieve real-time monitoring, autonomous decision-making, and automatic recovery, so as to improve the reliability, availability, and fault handling capabilities of the power system and reduce the need for manual intervention.

[0006] The technical solution of the present invention is specifically as follows:

[0007] A terminal offline detection and self-restart method comprises the following steps:

[0008] Step S1. Collecting the operation data of the power terminal equipment, preprocessing the collected data, and analyzing to obtain characteristic information;

[0009] Step S2. Combine the pre-processed terminal data information to establish a terminal status detection neural network model, input the real-time collected data into the trained deep learning model, and determine the offline status of the terminal device through model detection;

[0010] Step S3. Monitor the self-restart process based on the obtained terminal historical recovery time and basic cost, and combined with the output result of the terminal status detection neural network model;

[0011] Step S4: Compare the obtained terminal self-restart success efficiency with the preset terminal self-restart success efficiency to optimize the set self-restart monitoring benchmark.

[0012] Further, step S2 specifically includes:

[0013] The input of the terminal status detection neural network model is the preprocessed data features, and the output is a binary classification, indicating the need for self-restart or normal status; if the offline state is detected, the self-restart operation is triggered; the terminal status detection neural network model includes an input layer, an analysis layer, a detection layer, and an output layer.

[0014] Furthermore, in the detection layer, whether the terminal is in an offline state is detected, and the specific process is as follows:

[0015] ZI2=ω2PU1+b2

[0016]

[0017] Where ZI2 represents the input of the detection layer; ω2 represents the connection weight between the detection layer and the analysis layer; b2 represents the bias of the detection layer; PU2 represents the output of the detection layer; Δt represents the delay response coefficient; δ represents the judgment coefficient; W(t) represents the working power of the terminal device in normal state; W t-1 Represents the working power at time t-1; W t represents the working power at the current moment; μ represents the harmonic factor.

[0018] Furthermore, the terminal self-restart success efficiency is calculated based on the terminal basic cost BC and the terminal historical recovery time HRT. The specific process is as follows:

[0019]

[0020] Among them, pr_RM represents the terminal self-restart success efficiency; H s represents time loss; α represents tolerance, which means the maximum deviation that the terminal can bear during the self-restart process; h represents the fusion factor; BC jIt represents the planned maintenance cost; τ1 represents the regulation coefficient to ensure the smoothness of the change in the terminal maintenance cost; NN represents the output result of the terminal status detection neural network model.

[0021] Furthermore, step S4 specifically includes:

[0022] Define the monitoring benchmark as RM, and the preset self - restart success efficiency of the terminal is pr_RM'. Optimize the self - restart monitoring benchmark according to the comparison result of pr_RM and pr_RM', and pr_RM' > 0. The specific process is as follows:

[0023] If pr_RM ≥ pr_RM', then the optimized self - restart monitoring benchmark is where ρ1 represents the first optimization parameter;

[0024] If 0 ≤ pr_RM < pr_RM', then the optimized self - restart monitoring benchmark is where ρ2 represents the second optimization parameter;

[0025] If pr_RM < 0, then the optimized self - restart monitoring benchmark is where ρ3 represents the third optimization parameter.

[0026] Furthermore, step S4 specifically includes:

[0027] During the process of optimizing the self - restart monitoring benchmark, record the logs of offline detection and self - restart operations at the same time, and combine with the warning judgment value to send warning messages to the staff.

[0028] A terminal offline detection and self - restart device includes the following:

[0029] A terminal data acquisition unit, a terminal data processing unit, a terminal status detection unit, a terminal self - restart unit, a self - restart benchmark optimization unit, and a terminal warning unit;

[0030] The terminal data acquisition unit is used to collect the status data of the terminal device, obtain the running status information of the terminal device in real time through sensors or monitoring devices, and transmit the collected data to the terminal data processing unit;

[0031] The terminal data processing unit is used to receive the data transmitted by the terminal data acquisition unit, process and analyze it, and transfer the processed data to the terminal status detection unit for processing;

[0032] The terminal status detection unit uses the processed data to perform status detection and identification. By establishing a terminal status detection neural network model, the terminal device status is detected and analyzed to determine whether there is an offline state. If the terminal device is detected to be offline, the unit will trigger the corresponding operation.

[0033] The terminal self-restart unit is responsible for implementing the automatic restart operation of the terminal device. When the terminal status detection unit detects that the device is offline, the self-restart unit will automatically start the terminal device according to the preset self-restart strategy and process, realize automatic restart by controlling the power supply or communication interface of the terminal device, and monitor the status changes during the restart process;

[0034] The self-restart benchmark optimization unit is used to set and optimize the self-restart monitoring benchmark, evaluate and adjust the self-restart success efficiency according to the actual situation, and optimize the benchmark parameters;

[0035] The terminal early warning unit is responsible for generating alarm notifications and sending them to staff in a timely manner; when the offline state of the terminal device is detected or the self-restart operation encounters an abnormality, the early warning unit will issue an alarm signal.

[0036] Beneficial effects:

[0037] 1. The present invention can monitor the operating status of the equipment in real time by collecting data information such as relevant parameters of the power terminal equipment, and take maintenance or replacement measures in time to avoid power outages or other losses caused by equipment failures; through high-dimensional feature space analysis, it can better understand the dynamic behavior of the power terminal system, so as to take measures to improve the stability and reliability of the system; and analyze long-term time series data to provide a scientific basis for the long-term planning and investment decisions of the terminal system, ensuring the sustainability of power supply; based on data preprocessing, it can provide more accurate and real-time decision support to help operators respond quickly to various situations.

[0038] 2. The present invention can realize automated power terminal detection by using a deep learning model for terminal status detection, reduce the need for manual intervention, improve detection efficiency and reduce manual burden, especially in large-scale power systems; automatically identify offline status and accurately classify it as a situation requiring self-restart, thereby improving operational efficiency; by automatically triggering the self-restart operation, the offline status can be quickly restored, shortening the fault handling time, helping to reduce the downtime of terminal equipment, reducing the impact on users, and improving user satisfaction; at the same time, it can be used to evaluate the effect of the self-restart operation, improve the self-restart strategy, and better understand the offline mode and fault cause of the terminal equipment, thereby optimizing the operation and maintenance of the power system.

[0039] 3. By monitoring the self-restart process and calculating the self-restart success efficiency, the present invention can evaluate the effect and performance of the terminal self-restart, understand the reliability and stability of the self-restart operation, and help decision makers and operation and maintenance personnel evaluate the effectiveness of the self-restart strategy and determine whether improvement or optimization is needed; at the same time, combined with the terminal's historical recovery time and basic cost, the self-restart success efficiency can be calculated to conduct a cost-benefit analysis; by collecting data and feedback information during the self-restart process, more in-depth analysis and adjustments can be made to improve the self-restart strategy and increase the self-restart success rate, so as to establish a continuous improvement cycle and continuously improve the reliability and automated operation and maintenance capabilities of the power system.

[0040] 4. The present invention can evaluate the rationality of the set self-restart monitoring benchmark by comparing the actual self-restart success efficiency with the preset self-restart success efficiency; by comparing and analyzing and optimizing the self-restart monitoring benchmark, it can identify the problems and bottlenecks in the self-restart process, and take corresponding measures to improve them, thereby improving the self-restart efficiency, reducing power outage time and maintenance costs, and improving the reliability and availability of the power system; by recording key events and operation information, it is possible to understand the status of the terminal equipment and the changes in the self-restart process, and when necessary, send alarm notifications to relevant personnel to attract their attention in a timely manner, and take appropriate measures to deal with abnormal situations, quickly solve problems and reduce potential losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A flowchart of a terminal offline detection and self-restart method of the present invention;

[0042] Figure 2 The present invention is a module diagram of a terminal offline detection and self-restart device. DETAILED DESCRIPTION

[0043] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. It should also be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0044] Refer to the attached Figure 1 This embodiment provides a terminal offline detection and self-restart method, including the following steps:

[0045] The power terminal equipment sends heartbeat signals periodically. Specifically, the equipment needs to establish a reliable communication connection with the server. Various communication protocols, such as TCP / IP, HTTP, etc., can be used to send and receive heartbeat signals through communication between the equipment and the server.

[0046] The device periodically sends a heartbeat signal to the server to indicate the device's online status; the heartbeat signal can be a specific data packet or a simple request to inform the server of the active status of the device; the sending frequency can be configured according to specific needs, usually at a fixed time interval.

[0047] The server receives the heartbeat signal sent by the device and performs corresponding processing; the server can check the arrival time of the heartbeat signal. If the heartbeat signal is not received for a long time, it can be determined that the device is offline; the server can take corresponding measures according to specific needs, including recording offline events, sending alarm notifications, and trying to reconnect the device.

[0048] If the server determines that the device is offline, it can send an instruction to the device to restart; after receiving the restart instruction, the device executes the restart operation to restore the normal operation of the device; and the self-restart process is monitored during this process.

[0049] S1. Collect the operating data of power terminal equipment, including equipment status, power parameters, network connection information, etc.; pre-process the collected data, including data cleaning, feature extraction, feature correlation research and other steps.

[0050] After basic preprocessing of the data information related to the power terminal equipment, the preprocessed power terminal data information is converted into a high-dimensional feature space. The high-dimensional feature space can provide more dimensions and flexibility, so that the complex relationship between the power terminal data information can be better captured and understood. The specific process is as follows:

[0051]

[0052] Among them, Wh represents the result of conversion to high-dimensional feature space; P r Represents the rth one in the basic preprocessed power terminal data information, r∈[1,R]; represents the standardized weight of the rth data; α1 represents the balance coefficient; α2 represents the sensitivity coefficient; Represents the transformation index of the rth data.

[0053] The data information converted to the high-dimensional feature space is associated with the time series, and the feature information is further analyzed. The specific process is as follows:

[0054]

[0055] Where Gf represents the analysis result; Q y represents the weight when extracting different information; A represents the feature adjustment parameter; Wh lIt represents the l-th dimension data information located in the high-dimensional feature set; Jd represents the dynamic feature weight grid, which is used to adjust the weights between features according to the dynamic changes of different data; B represents the bias when extracting different information.

[0056] By collecting data information such as relevant parameters of power terminal equipment, the present invention can monitor the operating status of the equipment in real time, and take maintenance or replacement measures in time to avoid power outages or other losses caused by equipment failures; through high-dimensional feature space analysis, it can better understand the dynamic behavior of the power terminal system, so as to take measures to improve the stability and reliability of the system; and analyze long-term time series data to provide a scientific basis for the long-term planning and investment decisions of the terminal system, ensuring the sustainability of power supply; based on data preprocessing, it can provide more accurate and real-time decision support to help operators respond quickly to various situations.

[0057] S2. According to the preprocessed terminal data information, which is used as the input of the deep learning model, a terminal status detection neural network model is established, and the real-time collected data is input into the trained deep learning model, and the offline status of the terminal device is judged by model detection. The input is the preprocessed data feature, and the output is a binary classification, indicating whether self-restart is required or in normal state. If the deep learning model detects an offline state, a self-restart operation is triggered. The terminal status detection neural network model includes an input layer, an analysis layer, a detection layer, and an output layer.

[0058] Define the preprocessed terminal data information as D x , as training sample data input to the input layer, D x ={D x1 ,D x2 ,...,D xm}, m represents the number of elements in each group of training samples; the input layer is fully connected to the analysis layer, and the preprocessed terminal sample data is subjected to feature analysis in the analysis layer. The specific process is as follows:

[0059] ZI1=ω1D x +b1

[0060]

[0061] Among them, ZI1 represents the input in the analysis layer; PU1 represents the output of the analysis layer; ω1 represents the connection weight between the input layer and the analysis layer; b1 represents the bias of the analysis layer; β represents the adjustment coefficient; θ represents the stable constant; δ1 and δ2 represent the corresponding dynamic response parameters respectively; represents the average critical value of the terminal data obtained at time t; represents the normalized processed value of the analysis layer input obtained at time t; μ0 represents the analysis parameter; Represents the weight corresponding to time t-1.

[0062] In the detection layer, whether the terminal is in an offline state is further detected. The specific process is as follows:

[0063] ZI2=ω2PU1+b2

[0064]

[0065] Where ZI2 represents the input of the detection layer; ω2 represents the connection weight between the detection layer and the analysis layer; b2 represents the bias of the detection layer; PU2 represents the output of the detection layer; Δt represents the delay response coefficient; δ represents the judgment coefficient; W(t) represents the working power of the terminal device in normal state; W t-1 Represents the working power at time t-1; W t represents the working power at the current moment; μ represents the harmonic factor.

[0066] if The output result is Indicates that the terminal device is in normal state and can continue to work normally; if The output result is -1, indicating that the terminal device is offline and needs to be restarted to solve possible faults, generate an alarm notification, and send it to the staff in time to restore the normal working state of the distribution terminal.

[0067] Finally, the detection layer passes the status detection results of the device terminal to the output layer for output, and monitors the self-restart process of the terminal device based on the output results.

[0068] The present invention can realize automated power terminal detection by using a deep learning model for terminal status detection, reduce the need for manual intervention, improve detection efficiency and reduce manual burden, especially in large-scale power systems; automatically identify offline status and accurately classify it as a situation requiring self-restart, thereby improving operational efficiency; by automatically triggering the self-restart operation, the offline status can be quickly restored, the fault handling time can be shortened, which helps to reduce the downtime of terminal equipment, reduce the impact on users, and improve user satisfaction; at the same time, it can be used to evaluate the effect of the self-restart operation, improve the self-restart strategy, and better understand the offline mode and fault cause of the terminal equipment, thereby optimizing the operation and maintenance of the power system.

[0069] S3. Monitor the self-restart process based on the obtained terminal historical recovery time and basic cost, and combined with the output results of the terminal status detection neural network model.

[0070] During the self-restart process, the power supply to the power terminal equipment is cut off through power control devices or relays to completely cut off power; a period of time is waited to ensure that the equipment is completely powered off and fully discharged; then power is supplied to the power terminal equipment again to restart it; it is also necessary to monitor the startup process of the equipment to ensure that the equipment restarts successfully.

[0071] In an embodiment of the present invention, the monitoring benchmark is set as RM, the terminal historical recovery time is set as HRT, the basic cost is set as BC, and the output result of the terminal status detection neural network model is expressed as NN. Specifically, RM is set according to historical data and expert experience, HRT represents the terminal historical recovery time, which is the average time required for the terminal device to complete self-restart in the past, and BC represents the basic cost, which is the fixed cost of the self-restart process, including energy consumption, equipment maintenance and other aspects.

[0072] Now, based on the terminal basic cost BC and the terminal historical recovery time HRT, the terminal self-restart success efficiency is calculated. The specific process is as follows:

[0073]

[0074] Among them, pr_RM represents the terminal self-restart success efficiency; H s represents time loss; α represents tolerance, which means the maximum deviation that the terminal can bear during the self-restart process; h represents the fusion factor; BC j represents the planned maintenance cost; τ1 represents the control coefficient, which ensures the smoothness of changes in terminal maintenance costs.

[0075] By monitoring the self-restart process and calculating the self-restart success efficiency, the present invention can evaluate the effect and performance of the terminal self-restart, understand the reliability and stability of the self-restart operation, and help decision makers and operation and maintenance personnel evaluate the effectiveness of the self-restart strategy and determine whether improvement or optimization is needed; at the same time, combined with the terminal's historical recovery time and basic cost, the self-restart success efficiency can be calculated to conduct a cost-benefit analysis; by collecting data and feedback information during the self-restart process, more in-depth analysis and adjustments can be made to improve the self-restart strategy and increase the self-restart success rate, so as to establish a continuous improvement cycle and continuously improve the reliability and automated operation and maintenance capabilities of the power system.

[0076] S4. Compare the obtained terminal self-restart success efficiency with the preset terminal self-restart success efficiency, and optimize the set self-restart monitoring benchmark. At the same time, record the logs of offline detection and self-restart operations, and send alarm notifications to staff when necessary to handle abnormal situations in a timely manner. The specific process is as follows:

[0077] Now, define the preset terminal self - restart success efficiency as pr_RM‘, and optimize the self - restart monitoring benchmark according to the comparison result between pr_RM and pr_RM‘, and pr_RM‘>0. The specific process is as follows:

[0078] If pr_RM≥pr_RM‘, the optimized self - restart monitoring benchmark is where ρ1 represents the first optimization parameter;

[0079] If 0≤pr_RM<pr_RM‘, the optimized self - restart monitoring benchmark is where ρ2 represents the second optimization parameter;

[0080] If pr_RM<0, the optimized self - restart monitoring benchmark is where ρ3 represents the third optimization parameter.

[0081] In the embodiment of the present invention, according to the optimized self - restart monitoring benchmark RM’, set the warning threshold ε, Q represents the warning judgment value, k0 represents the parameter for adjusting the warning trigger degree; if Q≤ - 0.05, trigger a warning, indicating that there is a problem during the self - restart process and staff intervention for maintenance and inspection is required.

[0082] By comparing the actual self - restart success efficiency with the preset self - restart success efficiency, the present invention can evaluate the rationality of the set self - restart monitoring benchmark; by comparing, analyzing and optimizing the self - restart monitoring benchmark, the problems and bottlenecks existing in the self - restart process can be identified, and corresponding measures can be taken for improvement to improve the self - restart efficiency, reduce the power outage time and maintenance cost, and improve the reliability and availability of the power system; by recording key event and operation information, the status of the terminal device and the changes during the self - restart process can be understood. When necessary, sending an alarm notification to relevant personnel can attract their attention in time, and appropriate measures can be taken to handle abnormal situations, quickly solve problems and reduce potential losses.

[0083] Referring to the appendix Figure 2 , this embodiment provides a terminal offline detection and self - restart device, including the following:

[0084] Terminal data acquisition unit, terminal data processing unit, terminal status detection unit, terminal self - restart unit, self - restart benchmark optimization unit, terminal warning unit;

[0085] The terminal data acquisition unit is used to collect the status data of the terminal device, including but not limited to parameters such as current, voltage, temperature, etc., obtain the operating status information of the terminal device in real time through sensors or monitoring equipment, and transmit the collected data to the subsequent processing unit;

[0086] The terminal data processing unit is used to receive the data transmitted by the terminal data acquisition unit, process and analyze it, perform operations such as filtering, correction, and feature extraction on the collected data to obtain more accurate and reliable terminal device status information, and pass the processed data to other related units for further processing;

[0087] The terminal status detection unit uses the processed data to perform status detection and identification. By establishing a terminal status detection neural network model, the terminal device status is detected and analyzed to determine whether there is an offline state. If the terminal device is detected to be offline, the unit will trigger corresponding operations, such as starting the self-restart unit.

[0088] The terminal self-restart unit is responsible for implementing the automatic restart operation of the terminal device. When the terminal status detection unit detects that the device is offline, the self-restart unit will automatically start the terminal device according to the preset self-restart strategy and process, realize automatic restart by controlling the power supply or communication interface of the terminal device, and monitor the status changes during the restart process;

[0089] The self-restart benchmark optimization unit is used to set and optimize the self-restart monitoring benchmark, evaluate and adjust the self-restart success efficiency according to the actual situation, and optimize the benchmark parameters to improve the efficiency and reliability of the self-restart;

[0090] The terminal early warning unit is responsible for generating alarm notifications and sending them to relevant personnel in a timely manner; when the offline status of the terminal device is detected or the self-restart operation encounters an abnormality, the early warning unit will send an alarm signal to notify the operation and maintenance personnel or relevant departments for timely processing.

[0091] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0092] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0093] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0094] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0095] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A terminal offline detection and self-restart method, characterized in that: The following steps are involved: Step S1. Collecting the operation data of the power terminal equipment, preprocessing the collected data, and analyzing to obtain characteristic information; Step S2. Combine the preprocessed terminal data information to establish a terminal status detection neural network model, input the real-time collected data into the trained model, and judge the offline status of the terminal device through model detection; the terminal status detection neural network model includes an input layer, an analysis layer, a detection layer, and an output layer; The detection layer detects whether the terminal is in an offline state. The specific process is as follows: ZI2=ω2PU1+b2 Where ZI2 represents the input of the detection layer; ω2 represents the connection weight between the detection layer and the analysis layer; b2 represents the bias of the detection layer; PU2 represents the output of the detection layer; Δt represents the delay response coefficient; δ represents the judgment coefficient; W(t) represents the working power of the terminal device in normal state; W t-1 Represents the working power at time t-1; W t represents the working power at the current moment; μ represents the harmonic factor; PU1 represents the output of the analysis layer; Step S3. Monitor the self-restart process based on the obtained terminal historical recovery time and basic cost, and combined with the output result of the terminal status detection neural network model; According to the terminal basic cost BC and the terminal historical recovery time HRT, the terminal self-restart success efficiency is calculated. The specific process is as follows: Among them, pr_RM represents the terminal self-restart success efficiency; H s represents time loss; α represents tolerance, which means the maximum deviation that the terminal can bear during the self-restart process; h represents the fusion factor; BC j represents the planned maintenance cost; τ1 represents the control coefficient, which ensures the stability of the terminal maintenance cost change; NN represents the output result of the terminal status detection neural network model; Step S4: Compare the obtained terminal self-restart success efficiency with the preset terminal self-restart success efficiency to optimize the set self-restart monitoring benchmark.

2. A terminal offline detection and self-restart method according to claim 1, characterized in that: The step S2 specifically includes: The input of the terminal status detection neural network model is the preprocessed data features, and the output is a binary classification, indicating the need for self-restart or normal status; if the offline state is detected, the self-restart operation is triggered.

3. A terminal offline detection and self-restart method according to claim 1, characterized in that: The step S4 specifically includes: Define the monitoring benchmark as RM, and the preset terminal self-restart success efficiency as pr_RM'. According to the comparison results of pr_RM and pr_RM', optimize the self-restart monitoring benchmark, and pr_RM'>0. The specific process is as follows: If pr_RM ≥ pr_RM', the optimized self-restart monitoring benchmark is Wherein, ρ1 represents the first optimization parameter; If 0 ≤ pr_RM < pr_RM‘, the optimized self-restart monitoring baseline is where ρ2 represents the second optimization parameter; If pr_RM<0, the optimized self-restart monitoring benchmark is Wherein, ρ3 represents the third optimization parameter.

4. A terminal offline detection and self-restart method according to claim 1, characterized in that: The step S4 specifically includes: During the optimization of the self-restart monitoring benchmark, the logs of offline detection and self-restart operations are recorded at the same time. Combined with the warning judgment value, the warning message is sent to the staff.

5. A terminal offline detection and self-restart device, applied to the terminal offline detection and self-restart method according to claim 1, characterized in that: Includes the following: Terminal data acquisition unit, terminal data processing unit, terminal status detection unit, terminal self-restart unit, self-restart benchmark optimization unit, terminal early warning unit; The terminal data acquisition unit is used to collect the status data of the terminal device, obtain the operating status information of the terminal device in real time through the sensor or monitoring device, and transmit the collected data to the terminal data processing unit; The terminal data processing unit is used to receive the data transmitted by the terminal data acquisition unit, process and analyze it, and pass the processed data to the terminal status detection unit for processing; The terminal status detection unit uses the processed data to perform status detection and identification, and detects and analyzes the status of the terminal device by establishing a terminal status detection neural network model to determine whether there is an offline state. If the terminal device is detected to be offline, the unit will trigger a corresponding operation; The terminal self-restart unit is responsible for implementing the automatic restart operation of the terminal device. When the terminal status detection unit detects that the device is offline, the self-restart unit will automatically start the terminal device according to the preset self-restart strategy and process, realize automatic restart by controlling the power supply or communication interface of the terminal device, and monitor the state change during the restart process; The self-restart benchmark optimization unit is used to set and optimize the self-restart monitoring benchmark, evaluate and adjust the self-restart success efficiency according to actual conditions, and optimize the benchmark parameters; The terminal early warning unit is responsible for generating an alarm notification and sending it to the staff in a timely manner; when the offline state of the terminal device is detected or the self-restart operation encounters an abnormality, the early warning unit will send out an alarm signal.

Images