A power carrier based out-of-band management and sensing monitoring recovery method

Abstract

This invention discloses an out-of-band management and sensing monitoring recovery method based on power line carrier, belonging to the field of equipment operation and maintenance management technology. It includes the following steps: S1, real-time acquisition of equipment sensing data and data preprocessing; S2, identification of abnormal states of the target equipment, control of relays to perform power-off recovery; assessment of the degree of power-off release of the target equipment based on the equipment sensing data after power-off, and identification of the power-off release completion time; S3, determination of the power-off duration and the re-power-on time, and control of the relays to switch to the closed state to re-power the target equipment; S4, assessment of the restart stability of the target equipment based on the equipment sensing data after power-on, and determination of the target equipment recovery result. This method solves the problem that a fixed power-off duration is difficult to adapt to the discharge and restart sequences of different equipment in out-of-band management scenarios using power line carrier, leading to incomplete equipment reset, high recovery failure rate, and repeated restarts.

Description

Technical Field

[0001] This invention relates to the field of equipment operation and maintenance management technology, specifically to an out-of-band management and sensing monitoring recovery method based on power line carrier. Background Technology

[0002] With the continuous expansion of data centers, edge computing nodes, and industrial intelligent equipment, the operational status of these devices is characterized by a large number of devices, wide distribution, and high requirements for operational continuity. Out-of-band management based on the power supply network is gradually becoming an important method for equipment operation and maintenance management. Out-of-band management enables independent status awareness and remote control in the event of equipment business system anomalies, interruptions, or network disconnections, playing a crucial role in ensuring continuous equipment operation and maintenance. Especially in big data management scenarios, the stability of equipment operation, remote maintainability, and efficiency of anomaly response directly affect the overall system reliability. Therefore, intelligent management technologies targeting equipment power supply status, operational status, and remote recovery processes are receiving increasing attention.

[0003] For example, Chinese patent CN115392684 discloses a data center carbon emission monitoring system and method based on out-of-band management. This system connects to the electrical equipment in the data center through out-of-band management. The monitoring system includes: a data acquisition unit for collecting data from the electrical equipment in the data center; wherein the electrical equipment includes servers, computer infrastructure, and cooling facilities; a data processing unit for receiving and processing the data collected by the data acquisition unit to obtain the carbon emission value of the data center; and a control optimization unit for optimizing the control of the computer infrastructure and cooling facilities based on the carbon emission value and the collected data. This invention can comprehensively monitor and optimize the control of carbon-emitting electrical equipment, scientifically achieving energy conservation and emission reduction without affecting the external service capabilities of the edge data center.

[0004] However, existing technologies mostly use a fixed power outage duration to perform power-off restarts on equipment. But in power line carrier out-of-band management scenarios, due to differences in equipment power supply design, the power-off release time, internal capacitor discharge time, and time from restart to stable operation of different devices often exhibit discrete distributions. When the fixed power-off window is less than the actual discharge time required, the equipment may be powered on before completing a "complete power-off reset," resulting in incomplete reset, abnormally prolonged startup process, or repeated entry into abnormal states. The technical consequences are fluctuations in recovery success rate, increased number of repeated restarts, unpredictable recovery time, and may further exacerbate carrier control congestion and operational misjudgments.

[0005] Therefore, in order to address the above problems, there is an urgent need for an out-of-band management and sensing monitoring recovery method based on power line carrier. Summary of the Invention

[0006] Technical problems to be solved To address the shortcomings of existing technologies, this invention provides an out-of-band management and sensing monitoring recovery method based on power line carrier, which solves the problem that the fixed power outage duration is difficult to adapt to the discharge and restart sequence of different devices in the out-of-band management scenario of power line carrier, resulting in incomplete device reset, high recovery failure rate and repeated restarts.

[0007] Technical solution To achieve the above objectives, the present invention provides the following technical solution: an out-of-band management and sensing monitoring recovery method based on power line carrier, comprising the following steps: S1, real-time acquisition of device sensing data of the target device, and time alignment, anomaly correction, filtering and normalization processing of the device sensing data; S2, identification of abnormal states of the target device based on the device sensing data, and control of relays to perform power-off recovery based on the abnormal states; construction of power-off release timing data based on the device sensing data after power-off, evaluation of the degree of power-off release of the target device based on the power-off release timing data, and identification of the power-off release completion time based on the degree of power-off release; S3, adaptive determination of power-off holding duration and power-on time based on the power-off release completion time and historical power-off release records, generation of power-on permission command, and control of relays to switch to closed state to achieve power-on of the target device; S4, construction of power-on recovery timing data based on the device sensing data after power-on, evaluation of the restart stability of the target device based on the power-on recovery timing data, determination of the recovery result of the target device based on the restart stability, and execution of fault self-recovery measures when recovery fails.

[0008] Furthermore, the specific process of real-time acquisition of device sensing data from the target device, and the time alignment, anomaly correction, filtering, and normalization of the device sensing data, is as follows: The device sensing data of the target device is acquired in real time through the power line carrier network power module. This device sensing data includes: output voltage, output current, output power, I2C interface response status, and UART interface response status. The local clock of the remote control module is used as a unified clock reference to align the device sensing data in time. It is then determined whether the device sensing data is within the rated output range of the power line carrier network power module. If the sampled value exceeds the corresponding rated output range, the adjacent valid sampled value is used. The sample values ​​are corrected; the device sensing data is resampled according to a fixed sampling period to unify it to the same sampling time; the output voltage, output current, and output power are subjected to sliding median filtering, and the status values ​​of the I2C interface response status and UART interface response status are normalized; for missing values ​​in the output voltage, output current, and output power, linear interpolation is used to fill in the missing values; for missing values ​​in the I2C interface response status and UART interface response status, the sampled values ​​from the previous sampling time are used to fill in the missing values; the device sensing data is normalized; and an out-of-band recovery database is established to store the original and preprocessed device sensing data.

[0009] Furthermore, the specific process of identifying the abnormal state of the target device based on the device sensing data and controlling the relay to perform power-off recovery based on the abnormal state is as follows: the device abnormality is determined based on the device sensing data, including: whether the output power is less than the lower limit or greater than the upper limit of the output power in N consecutive sampling periods, whether the I2C interface response status is always non-response, and whether the UART interface response status is always non-response; if any abnormality determination condition is met, the target device is determined to be in an abnormal state, and a power-off recovery command is sent to the power line carrier network power module to control the relay to switch to the open state, perform power-off recovery, and record the relay open timestamp as the start time of power-off release.

[0010] Furthermore, the specific process of constructing power-off release timing data based on the device sensing data after power failure is as follows: starting from the start time of power failure release, continuously acquire the device sensing data of the target device after power failure, and assemble the power failure release timing data according to the sampling time sequence; synchronously read the no-load output power reference value and no-load output current reference value of the target device, and read the rated output voltage of the power line carrier network power module.

[0011] Furthermore, the specific process for evaluating the degree of power-off release of the target device by combining power-off release timing data is as follows: Based on the power-off release timing data, calculate the absolute difference between the current output power and the output power at the previous sampling time to obtain the absolute value of the output power change; divide the absolute value of the output power change by the sum of the no-load output power reference value and the smallest positive number, and calculate the negative exponent; subtract the negative exponent from the result to obtain the power change disturbance component; divide the current output voltage by the sum of the rated output voltage and the smallest positive number to obtain the residual voltage ratio; divide the current output current by the no-load... The residual current ratio is obtained by summing the output current reference value and the smallest positive number; the residual load component is obtained by taking the square root of the sum of the squares of the residual voltage ratio and the residual current ratio; the low-voltage residual power is obtained by multiplying the current output voltage and the output current; the residual power ratio is obtained by dividing the current low-voltage residual power by the sum of the no-load output power reference value and the smallest positive number; the residual power release component is obtained by adding one to the residual power ratio and taking the natural logarithm; the residual load component is multiplied by the power change disturbance component and then added to the residual power release component to obtain the incomplete power-off release value.

[0012] Furthermore, the specific process for identifying the completion time of power-off release based on the degree of power-off release is as follows: The incomplete power-off release value is compared with the completion threshold in real time. When the incomplete power-off release value is greater than the completion threshold, the target device is determined to be in an incomplete power-off release state. The relay remains in the open state, and power-off release timing data is continuously collected. When the incomplete power-off release value is less than or equal to the completion threshold for M consecutive sampling periods, the target device is determined to have completed power-off release, and the sampling time that first meets the condition is recorded as the completion time of power-off release. The power-off release start time, power-off release completion time, and incomplete power-off release value sequence are written into the device's out-of-band recovery database.

[0013] Furthermore, by combining the power-off release completion time with historical power-off release records, the power-off hold duration and power-on time are adaptively determined, a power-on permission command is generated, and the relay is controlled to switch to the closed state. The specific process for powering back the target device is as follows: the power-off release completion time is obtained by subtracting the power-off release start time from the power-off release completion time; historical power-off release completion times are read from the device's out-of-band recovery database, and the median value is taken as the power-off hold reference duration; if the current power-off release completion time is greater than the power-off hold reference duration, then the current power-off release completion time is used. The duration of power outage release is used as the power outage retention duration. If the duration of power outage release is less than or equal to the power outage retention reference duration, the power outage retention reference duration is used as the power outage retention duration. Based on the power outage release start time and the power outage retention duration, the power-on time is determined, and a power-on permission command is generated. When the power-on time is reached, the relay is controlled by the power line carrier network power module to switch from the open state to the closed state. The power outage release duration, the power outage retention duration, and the power-on time are written into the device's out-of-band recovery database.

[0014] Furthermore, the specific process for constructing power-on recovery timing data based on the device perception data after power-on, and evaluating the restart stability of the target device using the power-on recovery timing data, is as follows: Starting from the moment of power-on, continuously acquire the device perception data of the target device after power-on, and assemble the power-on recovery timing data according to the sampling time sequence; synchronously read the reference value of the target device's operating output power and the rated output voltage of the power line carrier network power module; based on the power-on recovery timing data, calculate the ratio of the current output power to the reference value of the operating output power, and input the ratio into the hyperbolic tangent function to obtain the operating power recovery component; calculate the output power change. The absolute value of the output power change is calculated by dividing the absolute value of the output power change by the sum of the operating output power reference value and the smallest positive number, and then calculating the negative exponent to obtain the power fluctuation suppression component. The current I2C interface response status and the UART interface response status are added together and divided by two to obtain the interface response recovery component. The absolute difference between the current output voltage and the rated output voltage is calculated and divided by the sum of the rated output voltage and the smallest positive number to obtain the output voltage deviation ratio. The output voltage deviation ratio is added by one and the reciprocal is taken to obtain the output voltage stability component. The operating power recovery component, the power fluctuation suppression component, the interface response recovery component, and the output voltage stability component are multiplied together to obtain the restart stability confirmation value.

[0015] Furthermore, the specific process for determining the recovery result of the target device based on the degree of restart stability is as follows: the restart stability confirmation value is compared with the stability confirmation threshold in real time. When the restart stability confirmation value is greater than or equal to the stability confirmation threshold for K consecutive sampling periods, the target device is determined to have recovered successfully, and the sampling time when the condition is first met is recorded as the recovery success time. The difference between the recovery success time and the power-on time is calculated to obtain the recovery success duration. When the restart stability confirmation value is less than the stability confirmation threshold, and the time difference between the current time and the power-on time is greater than the maximum value of the historical recovery success duration, the target device is determined to have failed to recover.

[0016] Furthermore, the specific process for implementing fault self-recovery measures when recovery fails is as follows: When the target device fails to recover, the cumulative number of consecutive recovery failures of the target device is read and incremented by one; if the cumulative number of consecutive recovery failures is less than the upper limit threshold for recovery attempts, a power outage recovery command is resent to the power line carrier network power module to re-execute the power outage recovery; if the cumulative number of consecutive recovery failures is greater than or equal to the upper limit threshold for recovery attempts, the power outage recovery is stopped and a manual maintenance prompt is output; the recovery success time, recovery success duration, restart stability confirmation value sequence, and judgment result corresponding to the recovery result are written into the device out-of-band recovery database.

[0017] Beneficial effects The present invention has the following beneficial effects: (1) The present invention collects output voltage, output current, output power and I2C / UART interface response status through power line carrier networking power module, so that the target device can still be monitored without relying on the business network and the main system, which is beneficial to improve the device anomaly identification capability in scenarios without network, network outage or confidential network.

[0018] (2) The present invention evaluates the degree of power release of the target device by power release timing data, and can identify the state of internal capacitor discharge, residual load release and power drop of the device, so as to avoid the device being re-energized before it is fully reset due to fixed power outage duration, and improve the effectiveness of power outage recovery.

[0019] (3) The present invention adaptively determines the power outage duration and the power-on time based on the power outage release completion time and historical power outage release records, so that the recovery process can adapt to the power supply design differences of different devices and reduce the problems of repeated restarts and unpredictable recovery time.

[0020] (4) The present invention evaluates the restart stability by restoring timing data through power-on and performs fault self-recovery measures when recovery fails. It can realize successful recovery confirmation, failure retry, number of attempts limit and manual maintenance prompt, thereby improving the safety, traceability and operation and maintenance reliability of the out-of-band recovery process.

[0021] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0022] Figure 1 This is a flowchart of an out-of-band management and sensing monitoring recovery method based on power line carrier communication. Figure 2 A schematic diagram illustrating the timing evolution of the restarted stable confirmation value; Figure 3 This is a diagram of the out-of-band management, sensing, monitoring and recovery system architecture based on power line carrier. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. As those skilled in the art will understand, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figures 1-3 This invention provides a technical solution: a method for out-of-band management and sensing monitoring recovery based on power line carrier communication, such as... Figure 1 As shown, the process includes the following steps: S1, real-time acquisition of device perception data of the target device, and time alignment, anomaly correction, filtering and normalization processing of the device perception data; S2, identification of abnormal states of the target device based on the device perception data, and control of relays to perform power-off recovery based on the abnormal states; construction of power-off release timing data based on the device perception data after power-off, evaluation of the degree of power-off release of the target device based on the power-off release timing data, and identification of the power-off release completion time based on the degree of power-off release; S3, adaptive determination of power-off holding duration and power-on time based on the power-off release completion time and historical power-off release records, generation of power-on permission command, and control of relays to switch to closed state to achieve power-on of the target device; S4, construction of power-on recovery timing data based on the device perception data after power-on, evaluation of the restart stability of the target device based on the power-on recovery timing data, determination of the recovery result of the target device based on the restart stability, and execution of fault self-recovery measures when recovery fails.

[0025] Specifically, the process of real-time acquisition of device perception data from the target device and subsequent time alignment, anomaly correction, filtering, and normalization of the data is as follows: The device perception data is acquired in real-time via a power line carrier network power module. This power module is an AC-DC or DC-DC power module connected to the target device's power supply and possessing power line carrier communication, output-side sampling, and interface status detection capabilities. A remote control module communicates with the power line carrier network power module via power line carrier communication to receive the device perception data and issue management and control commands. The two modules do not depend on the target device's service network, network port, or operating system; this is an out-of-band management connection. The device sensing data includes: output voltage, output current, output power, I2C interface response status, and UART interface response status. Output voltage is the output voltage supplied by the power line carrier network power module to the target device; output current is the output current supplied by the power line carrier network power module to the target device; and output power is obtained from the output voltage and output current. The I2C and UART interface response statuses indicate whether the corresponding interface of the target device has a valid response; a normal response is recorded as 1, and no response is recorded as 0. When the target device does not have a corresponding interface configured, the corresponding interface response status is null and will not be considered a valid interface in subsequent interface status determination. The local clock of the remote control module is used as a unified clock reference for time alignment of the device sensing data. Specifically, when the power line carrier network power module collects output voltage, output current, output power, I2C interface response status, and UART interface response status, it adds a sampling timestamp to each sampled data. After receiving the device sensing data, the remote control module converts the sampling timestamp to the local clock reference, so that the various device sensing data corresponding to the same sampling time can be associated as a group of sampling records. The system determines whether the device's sensed data falls within the rated output range of the power line carrier network power module. The rated output range is derived from the module's factory specifications, including the rated output voltage range, rated output current range, and rated output power range. If a sampled value exceeds the corresponding rated output range, it is corrected using an adjacent valid sampled value. Valid sampled values ​​are those within the corresponding rated output range. The device's sensed data is resampled at a fixed sampling period to unify it to the same sampling time. The fixed sampling period is a preset period determined based on the power line carrier network power module's output sampling capability, the target device's state change rate, and the remote control module's data processing frequency. The fixed sampling period can preferably be set to 100ms to 1s and can be configured according to the target device type. During resampling, the sampling time corresponding to the fixed sampling period is used as the unified sampling point, mapping the output voltage, output current, output power, I2C interface response status, and UART interface response status to the same sampling point.The output voltage, output current, and output power are subjected to sliding median filtering. The sliding median filtering uses a preset window length, which is an odd number of sampling points, preferably 3, 5, or 7, to suppress single-point spike noise. For the output voltage, output current, and output power within the window, the median of the corresponding window is taken as the filtered sample value. The I2C interface response status and UART interface response status are normalized to 0 or 1 values. For missing values ​​in the output voltage, output current, and output power, a linear interpolation method is used to fill in the missing values. This linear interpolation method is based on the valid sample values ​​adjacent to the missing sample point and their sampling time. If the number of consecutive missing sample points exceeds the preset missing window length, the corresponding time period is marked as a data unavailable period and is not used for subsequent power-off release degree assessment and restart stability assessment. For missing values ​​in the I2C interface response status and UART interface response status, the sample value from the previous sampling time is used for filling in the missing values. If there is no valid status value at the previous sampling time, the corresponding... Interface response status is marked as null; null is not equivalent to no response and is not directly used as a basis for anomaly judgment; device sensing data is normalized, with output voltage, output current, and output power being normalized using minimum-maximum normalization, and different data items being normalized using their respective minimum and maximum values ​​within their time windows; subsequent reference values ​​for rated output voltage, no-load output current, no-load output power, and operating output power are all synchronously normalized according to the minimum and maximum values ​​of their corresponding data items to ensure that the current sampled value and the reference value in subsequent formulas are on the same normalization scale. An out-of-band recovery database is established to store raw and preprocessed device sensing data for subsequent anomaly identification, power-off release degree assessment, adaptive recovery timing generation, and recovery closed-loop determination.

[0026] In this implementation plan, the out-of-band communication relationship between the power line carrier network power module and the remote control module is clearly defined. The device sensing data is obtained by using output-side sampling and interface status detection. The data is then processed by time alignment, anomaly correction, resampling, filtering, and normalization to ensure consistency and comparability of data from different devices and at different sampling times. At the same time, the rated output range, fixed sampling period, filtering window, and normalization reference source are clearly defined to ensure that the data processing process is feasible and traceable, providing a stable and reliable data foundation for subsequent anomaly identification, power outage release assessment, adaptive recovery control, and recovery closed-loop determination.

[0027] Specifically, the process of identifying abnormal states of the target device based on device sensing data and controlling the relay to perform power-off recovery based on the abnormal state is as follows: Device abnormality determination is performed based on device sensing data, including: whether the output power is consistently less than the lower limit or greater than the upper limit within N consecutive sampling periods. The lower and upper limits of output power are determined by statistical results of the output power during the historical normal operation phase of the target device. Preferably, the lower limit of output power is the first quartile of the output power during the historical normal operation phase minus the interquartile range, and the upper limit of output power is the third quartile of the output power during the historical normal operation phase plus the interquartile range, where the interquartile range is the difference between the third and first quartiles of the output power during the historical normal operation phase. When the target device lacks historical normal operation data, the lower and upper limits of output power can be determined based on the rated operating power range of the target device, which is derived from the nameplate parameters of the target device. The process also includes checking whether all I2C interface response states and UART interface response states are non-response states. Wherein, N is the preset number of continuous judgment periods, which is determined according to a fixed sampling period, preferably 3 to 10 sampling periods, to avoid misjudgment caused by fluctuations in a single sampling point; the I2C interface response status and the UART interface response status are both represented by 0 / 1 status values, with a normal response status recorded as 1 and a no-response status recorded as 0; the no-response status means that no valid ACK signal, valid return frame, or response data conforming to the interface protocol format has been received within the corresponding sampling period; when the target device is not configured with an I2C interface or a UART interface, the corresponding interface response status is null, which is not the same as the no-response status and is not used as an abnormal judgment condition. If any of the abnormal judgment conditions are met, the target device is determined to be in an abnormal state, and a power outage recovery command is issued to the power line carrier network power module. The power outage recovery command includes the target device identifier, the power line carrier network power module identifier, the relay control object, the disconnection action type, and the command timestamp. This is used to enable the power line carrier network power module to identify the target device and control object that needs to be restored by power outage recovery, control the relay to switch to the open state, execute the power outage recovery, and record the relay disconnection timestamp as the start time of power outage release. The relay disconnection timestamp is recorded by the power line carrier network power module when it detects that the relay state has switched from the closed state to the open state, and is transmitted back to the remote control module through power line carrier communication for subsequent construction of power outage release timing data.

[0028] In this implementation scheme, the abnormality of the target device is identified by the output power exceeding the limit and the response status of the I2C and UART interfaces within a continuous sampling period, which can avoid misjudgment caused by single-point fluctuations. At the same time, the upper and lower limits of the output power are determined by the statistics of historical normal operation data, and determined by the target device nameplate parameters when historical data is lacking, ensuring that the boundary of abnormality judgment has a data source and is feasible. Furthermore, by using a power-off recovery command that includes the target device identifier, module identifier, relay control object and instruction timestamp, the precise power-off control of the abnormal device and the recording of the power-off release start time are realized, providing a reliable timing basis for subsequent power-off release evaluation and recovery closed loop.

[0029] Specifically, the process of constructing power-out release timing data based on the device sensing data after power failure is as follows: starting from the power failure release start time, continuously acquire the device sensing data of the target device after power failure. The device sensing data after power failure includes the output voltage, output current, and output power collected by the power line carrier network power module output side sampling unit after the relay switches to the open state. The power failure release start time is the moment when the relay switches from the closed state to the open state and is detected by the power line carrier network power module. The power failure release timing data is composed according to the sampling time sequence. The power failure release timing data is arranged according to a fixed sampling period. Each sampling record includes at least a sampling timestamp, output voltage, output current, and output power, which are used to characterize the process of the residual voltage, residual current, and residual power on the output side of the target device decaying over time after power failure. The system synchronously reads the no-load output power reference value and the no-load output current reference value of the target device. The no-load output power reference value is the historical median value of the output power under no-load conditions, and the no-load output current reference value is the historical median value of the output current under no-load conditions. No-load conditions refer to the output side state when the relay is in the open state and the target device load is disconnected from the output side of the power line carrier network power module. Preferably, the no-load output power reference value and the no-load output current reference value are read from the device's out-of-band recovery database, representing the median values ​​of the target device's output power and output current when the relay was in the open phase. If no historical data is available, the no-load output parameter configuration of the power line carrier network power module is used. This configuration is derived from the factory test parameters of the power line carrier network power module and is pre-stored. The system also reads the rated output voltage of the power line carrier network power module, which is derived from the module's factory specifications and used as a reference for subsequently judging the residual output voltage during the power-off release phase.

[0030] In this implementation scheme, by continuously collecting output voltage, output current, and output power after the relay is disconnected, power-off release timing data arranged according to a fixed sampling period is formed. This can objectively characterize the attenuation process of residual voltage, residual current, and residual power after the target device is powered off. At the same time, the reference values ​​of no-load output power, no-load output current, and rated output voltage all have clear sources and can be determined from historical disconnection phase data or power module factory parameters, avoiding unclear sources of reference parameters and providing a repeatable, implementable, and traceable data basis for subsequent power-off release degree assessment.

[0031] Specifically, the process of evaluating the degree of power release of the target device by combining power release timing data is as follows: Based on the power release timing data, calculate the absolute difference between the current output power and the output power at the previous sampling time to obtain the absolute value of the output power change; divide the absolute value of the output power change by the sum of the no-load output power reference value and the smallest positive number, and calculate the negative exponent; subtract the negative exponent from the result to obtain the power change disturbance component; the power change disturbance component is used to characterize whether the output power is still in a dynamic release process after the power outage. The larger the absolute value of the output power change, the more the energy storage or load release inside the target device is still changing, and the larger the power change disturbance component is; the smaller the absolute value of the output power change, the more the output power gradually stabilizes, and the smaller the power change disturbance component is. The residual voltage ratio is obtained by dividing the current output voltage by the sum of the rated output voltage and the smallest positive number; the residual current ratio is obtained by dividing the current output current by the sum of the reference value of the no-load output current and the smallest positive number; the residual load component is obtained by taking the square root of the sum of the squares of the residual voltage ratio and the residual current ratio; the residual load component is used to comprehensively characterize the load intensity of the residual voltage and residual current on the output side after power failure; the method of taking the square root of the sum of squares can simultaneously retain the influence of the residual voltage ratio and the residual current ratio on the power failure release state. When the residual degree of either one is high, the residual load component will increase, thus reflecting that the target equipment may not have completed the power failure release. Multiplying the current output voltage and output current yields the low-voltage residual power. Dividing the current low-voltage residual power by the sum of the no-load output power reference value and the smallest positive number yields the residual power ratio. Adding one to the residual power ratio and taking its natural logarithm yields the residual power release component. The residual power release component characterizes the level of residual power remaining on the output side after power failure. Using natural logarithm processing can suppress excessive amplification of values ​​when the residual power is large and maintain the ability to distinguish subtle changes when the residual power is close to the no-load output power reference value. Multiplying the residual load component by the power change disturbance component and then adding it to the residual power release component yields the incomplete power failure release value. The residual load component multiplied by the power change disturbance component characterizes the dynamic release state where "residual voltage and current still exist and power is still changing." The residual power release component is added separately to retain the basic influence of the current residual power on the degree of incomplete power failure release. All components are calculated using dimensionless data and constrained to the same order of magnitude through exponential mapping, logarithmic mapping, and ratio mapping to ensure the comparability and numerical stability of the incomplete power failure release values. Therefore, the larger the incomplete power-off release value, the stronger the residual load, residual power, or dynamic release process of the target device after power failure, and the less suitable it is to immediately power on again; the smaller the incomplete power-off release value, the closer the residual voltage, residual current, and residual power on the output side are to the no-load stable state, and the closer the target device is to the complete power-off reset state.

[0032] The specific formula for the incomplete power-off release value is as follows: ; In the formula, The value indicates the incomplete power-off release, used to assess whether there is still residual voltage, residual current, residual power and dynamic release process after the target equipment is powered off. The larger the value, the more likely the equipment has not yet completed the power-off release, and the smaller the value, the closer the equipment is to the complete power-off reset state. This indicates the current output voltage and is used to reflect whether there is still residual voltage on the output side after a power outage; This indicates the current output current and is used to reflect whether there is still residual load current on the output side after a power outage; This represents the no-load output power reference value, used as a normalized benchmark for residual power and power variation; Indicates the rated output voltage, used as a reference for the current residual level of the output voltage; This represents the no-load output current reference value, which is used as a reference benchmark for the current residual output current level. This indicates the current output power, used to reflect the output power status during the current power-off release phase; This represents the output power at the previous sampling time, used to compare with the current output power and reflect the magnitude of power change. This represents a very small positive number, used to prevent the denominator from being zero and to ensure the stability of the formula calculation. The preferred value is [value missing]. to .

[0033] In this implementation scheme, the release status of the target device after power failure is comprehensively evaluated by output power changes, residual voltage, residual current, and residual power. This can identify whether the device still has internal energy storage release, load residue, or dynamic power changes. At the same time, by using processing methods such as ratioization, square root, natural logarithm, and exponential mapping, the judgment of the degree of incomplete power failure release is made more stable and quantifiable, avoiding the device being re-powered before it has been completely de-energized and reset, thereby improving the accuracy and reliability of power failure recovery control.

[0034] Specifically, the process of identifying the completion time of power-off release based on the degree of power-off release is as follows: Real-time comparison is made between the incomplete power-off release value and the completion threshold. The completion threshold is used to distinguish whether the target device has entered a stable no-load state. The completion threshold can be adaptively determined by statistically analyzing the incomplete power-off release values ​​during the historical normal power-off release completion phases of the target device. When the incomplete power-off release value is greater than the completion threshold, the target device is determined to be in an incomplete power-off release state. The relay remains open, and power-off release timing data is continuously collected. Maintaining the relay open indicates that a re-power-on action is not performed, allowing the target device to continue in the power-off release process until the output voltage, output current, and output power further decrease. When the incomplete power-off value is less than or equal to the completion threshold for M consecutive sampling periods, the target device is determined to have completed power-off release, and the sampling time at which the condition is first met is recorded as the completion time of power-off release. Here, M is a preset number of consecutive confirmation periods for completion of release, determined based on the fixed sampling period and the stability requirements of the target device's power-off release, preferably 3 to 10 sampling periods. Confirmation using M consecutive sampling periods avoids misjudging completion of power-off release due to noise, residual voltage fluctuations, or sampling errors at a single sampling point. The sampling time at which the condition is first met is the sampling time at which the first incomplete power-off value is less than or equal to the completion threshold within the M consecutive sampling periods. The power-off release start time, power-off release completion time, and incomplete power-off value sequence are written into the device's out-of-band recovery database. The incomplete power-off value sequence is stored in chronological order of sampling time and associated with the target device identifier, power line carrier network power module identifier, relay status, and sampling timestamp for subsequent adaptive determination of power-off hold duration, tracing the power-off release process, and updating the target device's historical power-off release records.

[0035] In this implementation scheme, the incomplete power-off release value is stably determined by the release completion threshold and the confirmation mechanism of multiple consecutive sampling cycles, which can avoid misjudgment caused by single-point noise, instantaneous fluctuation of residual voltage or sampling error. At the same time, the relay is kept open when the power-off release is not completed to prevent the equipment from being powered on too early, and the power-off release process data is written into the equipment's out-of-band recovery database, providing a reliable basis for subsequent adaptive determination of power-off holding time and recovery process traceability.

[0036] Specifically, by combining the power outage release completion time with historical power outage release records, the power outage hold duration and power-on time are adaptively determined, a power-on permission command is generated, and the relay is controlled to switch to the closed state. The specific process of powering up the target device is as follows: the power outage release completion time is obtained by subtracting the power outage release start time from the power outage release completion time. This power outage release completion time is used to characterize the actual time required for the target device to complete the power outage release from the relay opening during this power outage process, reflecting the actual state of the target device's current internal energy storage release, power rail fallback, and load residual decay. Historical power outage release completion times are read from the device's out-of-band recovery database, and the median value is taken as the power outage hold reference duration. The historical power outage release completion times are derived from the corresponding power outage release completion times in the historical successful recovery records of the target device. When the number of historical successful recovery records of the target device is insufficient, historical power outage release completion times of the same model of target device can be read. The median value is used as the power outage hold reference duration to reduce the impact of occasional abnormal power outage processes, sampling disturbances, and single recovery anomalies on the reference duration. If the current power outage release completion time is longer than the power outage hold reference time, then the current power outage release completion time will be used as the current power outage hold time. If the current power outage release completion time is less than or equal to the power outage hold reference time, then the power outage hold reference time will be used as the current power outage hold time. This ensures that the current power outage hold time is constrained by both the current actual release result and historical stable release experience. When the current release process is slower than the historical median level, the power outage hold time will be extended to ensure the equipment completes a full power-off reset. When the current release process is faster than or equal to the historical median level, the historical power outage hold reference time will be retained to avoid premature power-on due to a single rapid drop, thereby reducing incomplete reset and repeated recovery failures. Here, the power outage release completion time represents the time required for the equipment to release naturally, and the power outage hold time represents the final actual control time for maintaining the power outage. Based on the power outage release start time and the current power outage duration, the power-on reset time is determined, and a power-on permission command is generated. When the power-on reset time is reached, the power line carrier network power module controls the relay to switch from the open state to the closed state. The power-on reset time is the time obtained by adding the current power outage duration to the power outage release start time. The power-on permission command includes the target device identifier, the power line carrier network power module identifier, the relay control object, the closing action type, and the command timestamp, which is used to enable the power line carrier network power module to execute the relay closing action at the determined power-on reset time. The power outage release completion time, the current power outage duration, and the power-on reset time are written to the device's out-of-band recovery database to update the target device's historical power outage release records and provide a data basis for subsequent recovery success duration statistics, stability confirmation threshold updates, and recovery sequence references for the same model of target device.

[0037] In this implementation plan, the power outage duration is determined by combining the current power outage release completion time with the median value of historical power outage release completion times. This ensures that the power-on time not only matches the current actual discharge state of the target device but also references historical stable recovery experience. This avoids incomplete reset due to an excessively short fixed power outage duration, and also avoids premature power-on due to a single rapid fallback. At the same time, the recovery timing data is written into the device's out-of-band recovery database, providing a reliable basis for subsequent recovery strategy optimization and timing reference for devices of the same model.

[0038] Specifically, the process of constructing power-on recovery timing data based on device perception data after power-on, and evaluating the restart stability of the target device using this timing data, is as follows: Starting from the moment of power-on, continuously acquire device perception data of the target device after power-on. The power-on moment is the moment when the power supply module control relay switches from an open state to a closed state. The device perception data after power-on includes output voltage, output current, output power, I2C interface response status, and UART interface response status, arranged in the order of sampling time to form power-on recovery timing data. The power-on recovery timing data is arranged according to a fixed sampling period. Each sampling record includes at least a sampling timestamp, output voltage, output current, output power, I2C interface response status, and UART interface response status, used for representation. The system tracks the dynamic process of power recovery, voltage stabilization, and interface response recovery of the target device after power-on. It synchronously reads the target device's operating output power reference value and the rated output voltage of the power line carrier network power module. The operating output power reference value is the median value of the target device's output power during its historical normal operation phase, preferentially retrieved from the device's out-of-band recovery database during periods when no abnormal state judgment was triggered and no power outage recovery was performed. When historical normal operation data for the target device is lacking, the operating output power reference value can be determined from the rated operating power on the target device's nameplate. The rated output voltage is derived from the factory specifications of the power line carrier network power module. Before being used in calculations, the operating output power reference value and rated output voltage are normalized using the same normalization benchmark as the current output power and current output voltage, ensuring that subsequent components are dimensionless. Based on the power-on recovery timing data, the ratio of the current output power to the reference value of the operating output power is calculated. This ratio is then input into the hyperbolic tangent function to obtain the operating power recovery component. The operating power recovery component is dimensionless and characterizes the degree of recovery of the target device's output power after power-on relative to the reference value of normal operating output power. The hyperbolic tangent function maps the power recovery degree to a finite range, avoiding excessive fluctuations in output power at startup that could lead to abnormally amplified evaluation results. The absolute value of the output power change is calculated, divided by the sum of the reference value of operating output power and a minimum positive number, and a negative exponent is calculated to obtain the power fluctuation suppression component. The absolute value of the output power change is the absolute difference between the current output power and the output power at the previous sampling time. The power fluctuation suppression component is dimensionless and characterizes whether the output power tends to stabilize after power-on. The smaller the absolute value of the output power change, the closer the power fluctuation suppression component is to one, indicating a more stable operating load for the target device.The interface response recovery component is obtained by adding the current I2C interface response status and the UART interface response status together and then dividing by two. Both the I2C and UART interface response statuses are dimensionless state quantities; a normal response status is recorded as 1, and a no-response status as 0. The interface response recovery component characterizes whether communication at the target device's underlying interface has been restored. When the target device does not have a corresponding interface configured, the corresponding interface response status is null, and null values ​​are not included in the calculation of the interface response recovery component. The component is averaged according to the actual number of valid interfaces. When the target device has no valid interfaces, the interface response recovery component is set to 1 by default. The absolute difference between the current output voltage and the rated output voltage is calculated. The absolute difference is divided by the sum of the rated output voltage and a very small positive number to obtain the output voltage deviation ratio. The output voltage deviation ratio is added by one and the reciprocal is taken to obtain the output voltage stability component. Both the output voltage deviation ratio and the output voltage stability component are dimensionless components used to characterize the degree of deviation of the current output voltage from the rated output voltage. The closer the current output voltage is to the rated output voltage, the closer the output voltage stability component is to one, indicating a more stable power supply state after power-on. The restart stability confirmation value is obtained by multiplying the operating power recovery component, power fluctuation suppression component, interface response recovery component, and output voltage stability component. This restart stability confirmation value is a dimensionless comprehensive evaluation result, simultaneously reflecting the degree of power recovery, power fluctuation stability, interface response recovery, and output voltage stability of the target device after power-on. The product fusion method reflects the bottleneck constraint in the recovery confirmation; that is, if any component is significantly abnormal, the restart stability confirmation value will decrease, thus avoiding misjudging the target device's successful recovery based solely on a normal single indicator.

[0039] The specific formula for restarting the stable confirmation value is as follows: ; In the formula, This represents the restart stability confirmation value, which is used to comprehensively evaluate the target device's operational recovery status, power fluctuation stability status, interface communication recovery status, and output voltage stability status after power-on. The larger the value, the closer the target device is to a stable recovery operational status. This indicates the current output power, used to reflect the operating load status after power-on; This indicates the output power at the previous sampling time, used to reflect output power fluctuations; This represents the reference value for normal operating output power, used as a benchmark for the current degree of output power recovery and power fluctuation. This indicates the current output voltage, used to reflect the stable state of the output voltage after power-on. This indicates the rated output voltage, which is used as a reference for the stability of the output voltage. This indicates the current I2C interface response status, reflecting whether the I2C interface has resumed normal communication. A normal response is represented by 1, and a no response is represented by 0. This indicates the current UART interface response status, reflecting whether the UART interface has resumed normal communication. A normal response is 1, and a no response is 0. This represents a very small positive number, used to prevent the denominator from being zero and to ensure the stability of the formula calculation. The preferred value is [value missing]. to .

[0040] In this embodiment, Table 1 is a restart stability confirmation value data table. The normalized normal operating output power reference value is 0.72, and the normalized rated output voltage is 0.78. The table details the normalized output power, output power at the previous sampling time, output voltage, I2C interface response status, UART interface response status, and restart stability confirmation value for five specific moments. Specifically, moment 1 corresponds to a normalized output power of 0.32, a previous sampling output power of 0.00, an output voltage of 0.54, an I2C interface response status of 0, a UART interface response status of 0, and a restart stability confirmation value of 0.0000; moment 2 corresponds to a normalized output power of 0.55, a previous sampling output power of 0.32, an output voltage of 0.68, an I2C interface response status of 1, a UART interface response status of 0, and a restart stability confirmation value of 0.2072; moment 3 corresponds to a normalized output power of 0.70, a previous sampling output power of 0.55, and an output voltage of 0. 0.75, I2C interface response status is 1, UART interface response status is 1, restart stability confirmation value is 0.5861; the normalized output power at time 4 is 0.73, the output power at the previous sampling time is 0.70, the output voltage is 0.77, I2C interface response status is 1, UART interface response status is 1, restart stability confirmation value is 0.7267; the normalized output power at time 5 is 0.72, the output power at the previous sampling time is 0.73, the output voltage is 0.78, I2C interface response status is 1, UART interface response status is 1, restart stability confirmation value is 0.7511.

[0041] Table 1 Restart Stability Confirmation Values ​​Data Table

[0042] like Figure 2 The figure shows a schematic diagram of the timing evolution of the restart stability confirmation value. The figure uses the sampling time as the horizontal axis and the restart stability confirmation value as the vertical axis, displaying the stable recovery process of the target device after power-on in a broken line format; the dashed line in the figure represents the stability confirmation threshold, used to determine whether the target device has entered the stable recovery state. (Refer to Table 1 and...) Figure 2It can be seen that during the period from time 1 to time 2, the restart stability confirmation value was low, indicating that the target device had just completed a power-on, and the output power, output voltage, and interface response status had not yet fully recovered, and the device was still in an unstable state. During the period from time 3, the restart stability confirmation value rose to 0.5861 and exceeded the stability confirmation threshold, indicating that the target device had basically completed interface recovery and operation recovery. During the period from time 4 to time 5, the restart stability confirmation value further increased and tended to stabilize, indicating that the output power fluctuation gradually decreased, the output voltage approached the rated output voltage, and both the I2C interface and UART interface resumed normal response, and the target device had entered a stable operating state.

[0043] In this implementation plan, the restart stability of the target device is comprehensively evaluated by measuring the output power, power changes, output voltage, and I2C / UART interface response status after power-on. This assessment can simultaneously reflect the recovery of operating load, convergence of power fluctuations, stability of power supply voltage, and recovery of underlying interface communication. Each component is calculated using normalized dimensionless data to ensure comparability between different data items. Furthermore, the product fusion method reflects the bottleneck constraints in recovery confirmation, avoiding misjudgments caused by a single normal indicator, thereby improving the accuracy and reliability of the target device recovery success determination.

[0044] Specifically, the process for determining the recovery result of the target device based on the degree of restart stability is as follows: The restart stability confirmation value is compared with the stability confirmation threshold in real time. The stability confirmation threshold is used to characterize the lower limit requirement that the restart stability confirmation value should meet when the target device reaches a stable recovery state. The stability confirmation threshold is preferably determined statistically from the restart stability confirmation values ​​recorded in the historical recovery success records of the target device. When the restart stability confirmation value is greater than or equal to the stability confirmation threshold for K consecutive sampling periods, the target device is determined to have recovered successfully, and the sampling time at which the condition is first met is recorded as the recovery success time. Here, K is the preset number of consecutive confirmation periods for stable recovery, which is determined based on the fixed sampling period and the restart stability confirmation requirements of the target device, preferably 3 to 10 sampling periods. Using K consecutive sampling periods for confirmation is used to avoid misjudging recovery success due to short-term power recovery, instantaneous interface response, and short-term output voltage stability at a single sampling point. The sampling time at which the condition is first met is the sampling time at which the restart stability confirmation value is greater than or equal to the stability confirmation threshold in the K consecutive sampling periods. The difference between the successful recovery time and the power-on time is calculated to obtain the successful recovery duration. This duration characterizes the actual time required for the target device to reach a stable recovery state after power-on, and is written to the device's out-of-band recovery database as a data source for subsequent historical successful recovery duration statistics. If the restart stability confirmation value is less than the stability confirmation threshold, and the time difference between the current time and the power-on time is greater than the maximum historical successful recovery duration, the target device is considered to have failed to recover. The maximum historical successful recovery duration is derived from the target device's historical successful recovery records in the out-of-band recovery database. When the number of historical successful recovery records for the target device is less than a preset minimum number of records, the target device's factory startup time is used as a substitute benchmark for the historical successful recovery duration.

[0045] In this implementation plan, a mechanism for confirming stable restart values ​​through a stable confirmation threshold and multiple consecutive sampling cycles is used to continuously determine the stable restart confirmation value. This avoids false judgments of successful recovery caused by short-term power rebound, instantaneous interface response, or short-term voltage stabilization. At the same time, the recovery failure is determined by using the recovery success duration record and the maximum historical recovery success duration. This ensures that the recovery results have both real-time status and historical time sequence basis. When historical data is insufficient, the factory start-up duration is used as a substitute benchmark, thereby improving the accuracy, feasibility, and traceability of recovery success / failure determination.

[0046] Specifically, the process of implementing fault self-recovery measures when recovery fails is as follows: when the target device fails to recover, the cumulative number of consecutive recovery failures of the target device is read and incremented by one; the cumulative number of consecutive recovery failures is used to record the number of times the target device fails to meet the recovery success conditions in the continuous recovery process; when the target device recovers successfully, the cumulative number of consecutive recovery failures is cleared to zero to avoid historical failure records affecting subsequent independent recovery processes. If the cumulative number of consecutive recovery failures is less than the upper limit threshold, a power outage recovery command is reissued to the power line carrier network power module to re-execute the power outage recovery. The upper limit threshold is pre-configured by the target device's operation and maintenance strategy and equipment safety protection requirements to limit the impact caused by frequent relay switching and repeated power outages and power-ups of the target device. When re-executing the power outage recovery, the process returns to the power outage release status identification process, records the new power outage release start time, and re-evaluates the power outage release degree. If the cumulative number of consecutive recovery failures is greater than or equal to the upper limit threshold, the power outage recovery is stopped, and a manual maintenance prompt is output. This manual maintenance prompt can be output through the alarm interface of the remote out-of-band management platform. The prompt includes at least the target device identifier, the power line carrier network power module identifier, the cumulative number of consecutive recovery failures, the time of the last recovery failure, and the corresponding judgment result, prompting maintenance personnel to manually check the target device, power module, relays, and interface connection status. The recovery success time, recovery success duration, restart stability confirmation value sequence, and judgment result corresponding to the recovery result are written to the device's out-of-band recovery database. Specifically, when the determination result is successful recovery, the successful recovery time and duration are written; when the determination result is unsuccessful recovery, the unsuccessful recovery time, the cumulative number of consecutive unsuccessful recovery attempts, and the manual inspection prompt result are written; the restart stable confirmation value sequence is stored in the order of sampling time for subsequent recovery process traceability, recovery count statistics, and recovery strategy optimization.

[0047] like Figure 3The diagram shows the architecture of an out-of-band management and sensing monitoring recovery system based on power line carrier. It illustrates the overall architecture of this system, which constructs a private power line carrier network via AC power supply lines. This network simultaneously provides power and carries out-of-band communication, enabling remote status sensing and recovery control of target device nodes. The remote out-of-band management platform connects to the private power line carrier network via the Internet or a dedicated network to perform anomaly monitoring, status analysis, remote recovery, and log management. The AC power supply lines and power line carrier communication links are used to enable power line carrier data interaction with the target devices. Each target device node includes a relay control and monitoring unit, an AC-DC power module, an output-side sampling unit, an MCU control unit, and an out-of-band management agent module. The system comprises several components: a relay control and monitoring unit for power outage and power-on control and monitoring relay status; an AC-DC power supply module for providing stable output power to the target device; an output-side sampling unit for real-time acquisition of output voltage and current and calculation of output power to generate device sensing data; an MCU control unit for device status acquisition, recovery logic calculation, and GPIO, I2C, and UART interface status detection; and an out-of-band management agent module for power line carrier communication and out-of-band management protocol interaction. Overall, it demonstrates a complete out-of-band recovery closed-loop process that utilizes power supply-side sampling data to identify abnormal device states, assess the degree of power outage release, control the power-on process, and confirm a stable restart state. Furthermore, it can still achieve remote recovery control via a private power line carrier network even in the event of a main system malfunction or network failure in the target device.

[0048] In this implementation plan, the power outage recovery process is constrained by accumulating the number of consecutive recovery failures. After a recovery failure, the power outage recovery can be automatically re-executed, and the repeated operation will stop when the upper limit of the number of recovery attempts is reached, so as to avoid the impact caused by frequent switching of relays and repeated power outages and power-on of equipment. At the same time, manual maintenance prompts are output through the remote out-of-band management platform, and the recovery process data is written into the equipment's out-of-band recovery database, thereby improving the safety, traceability and operation and maintenance efficiency of the fault self-recovery process.

[0049] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0050] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. As those skilled in the art will understand, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for out-of-band management and sensing monitoring recovery based on power line carrier, characterized in that, Includes the following steps: S1 collects the device perception data of the target device in real time, and performs time alignment, anomaly correction, filtering and normalization on the device perception data. S2, based on the equipment sensing data, identify the abnormal state of the target equipment, and control the relay to perform power outage recovery based on the abnormal state; construct power outage release timing data based on the equipment sensing data after power outage, evaluate the degree of power outage release of the target equipment in combination with the power outage release timing data, and identify the power outage release completion time according to the degree of power outage release. S3, combining the power-off release completion time with historical power-off release records, adaptively determines the power-off holding time and the power-on time, generates a power-on permission command, and controls the relay to switch to the closed state to realize the target device's power-on; S4. Based on the device perception data after power-on, construct power-on recovery timing data, combine the power-on recovery timing data to evaluate the restart stability of the target device, determine the recovery result of the target device based on the restart stability, and execute fault self-recovery measures when recovery fails.

2. The method for out-of-band management and sensing monitoring recovery based on power line carrier as described in claim 1, characterized in that, The specific process of acquiring the device sensing data of the target device in real time and performing time alignment, anomaly correction, filtering and normalization on the device sensing data is as follows: The power line carrier network power module collects device perception data of the target device in real time. The device perception data includes: output voltage, output current, output power, I2C interface response status and UART interface response status. Using the local clock of the remote control module as a unified clock reference, time alignment is performed on the device sensing data; it is determined whether the device sensing data is within the rated output range of the power line carrier network power module. If the sampled value exceeds the corresponding rated output range, it is corrected using adjacent valid sampled values; the device sensing data is resampled according to a fixed sampling period to unify it to the same sampling time; sliding median filtering is applied to the output voltage, output current, and output power, and the status values ​​of the I2C interface response status and UART interface response status are normalized; missing values ​​in the output voltage, output current, and output power are filled in using a linear interpolation method; missing values ​​in the I2C interface response status and UART interface response status are filled in using the sampled value from the previous sampling time; the device sensing data is normalized; and an out-of-band recovery database is established to store the original and preprocessed device sensing data.

3. The method for out-of-band management and sensing monitoring recovery based on power line carrier as described in claim 1, characterized in that, The specific process of identifying the abnormal state of the target device based on the device sensing data, and controlling the relay to perform power-off recovery based on the abnormal state is as follows: Based on the device sensing data, the device anomaly is determined, including: whether the output power is less than the lower limit or greater than the upper limit of the output power in N consecutive sampling periods, whether the I2C interface response status is always unresponsive, and whether the UART interface response status is always unresponsive. If any of the abnormal judgment conditions are met, the target device is determined to be in an abnormal state, and a power outage recovery command is sent to the power line carrier network power module to control the relay to switch to the open state, perform power outage recovery, and record the relay open timestamp as the start time of power outage release.

4. The method for out-of-band management and sensing monitoring recovery based on power line carrier as described in claim 3, characterized in that, The specific process for constructing power outage release timing data based on device sensing data after a power outage is as follows: Starting from the moment of power failure release, continuously acquire device sensing data after power failure of the target device, and assemble power failure release timing data according to the sampling time sequence; synchronously read the no-load output power reference value and no-load output current reference value of the target device, and read the rated output voltage of the power line carrier network power module.

5. The method for out-of-band management and sensing monitoring recovery based on power line carrier as described in claim 4, characterized in that, The specific process for evaluating the power-off release degree of the target device by combining power-off release timing data is as follows: Based on the power-off release timing data, the absolute difference between the current output power and the output power at the previous sampling time is calculated to obtain the absolute value of the output power change. The absolute value of the output power change is divided by the sum of the no-load output power reference value and the smallest positive number, and the negative exponent is calculated. The power change disturbance component is obtained by subtracting the negative exponent from the result. Divide the current output voltage by the sum of the rated output voltage and the smallest positive number to obtain the residual voltage ratio; divide the current output current by the sum of the no-load output current reference value and the smallest positive number to obtain the residual current ratio. The residual load component is obtained by taking the square root of the sum of the squares of the residual voltage ratio and the residual current ratio. Multiply the current output voltage by the output current to obtain the low-voltage residual power; divide the current low-voltage residual power by the sum of the no-load output power reference value and the smallest positive number to obtain the residual power ratio; add one to the residual power ratio and take the natural logarithm to obtain the residual power release component. Multiply the residual load component by the power change disturbance component, and then add it to the residual power release component to obtain the incomplete power release value.

6. The method for out-of-band management and sensing monitoring recovery based on power line carrier according to claim 5, characterized in that, The specific process for identifying the completion time of power outage release based on the degree of power outage release is as follows: The system compares the incomplete power-off release value with the complete power-off release threshold in real time. When the incomplete power-off release value is greater than the complete power-off release threshold, the target device is determined to be in an incomplete power-off release state. The relay is kept in the open state, and the power-off release timing data is continuously collected. When the power failure release value is less than or equal to the release completion threshold for M consecutive sampling periods, the target device is determined to have completed the power failure release, and the sampling time when the condition is first met is recorded as the power failure release completion time. Write the sequence of power-off release start time, power-off release completion time, and power-off release incomplete value into the device's out-of-band recovery database.

7. The method for out-of-band management and sensing monitoring recovery based on power line carrier as described in claim 1, characterized in that, The specific process of combining the power-off release completion time with historical power-off release records to adaptively determine the power-off holding time and the power-on time, generating a power-on permission command, and controlling the relay to switch to the closed state to achieve the target device's power-on is as follows: The power outage release completion time is obtained by subtracting the power outage release start time from the power outage release completion time; the historical power outage release completion time is read from the out-of-band recovery database of the device, and the median value is taken as the power outage hold reference time. If the duration of this power outage release is longer than the power outage retention reference duration, then the duration of this power outage release will be used as the power outage retention duration; if the duration of this power outage release is less than or equal to the power outage retention reference duration, then the power outage retention reference duration will be used as the power outage retention duration. Based on the power outage release start time and the power outage duration, the power restoration time is determined, and a power restoration permission command is generated. When the power restoration time is reached, the power line carrier network power module controls the relay to switch from the open state to the closed state. The duration of this power outage release, the duration of this power outage, and the time of power restoration are written into the device's out-of-band recovery database.

8. The method for out-of-band management and sensing monitoring recovery based on power line carrier as described in claim 1, characterized in that, The specific process of constructing power-on recovery timing data based on device perception data after power-on, and then evaluating the restart stability of the target device using the power-on recovery timing data, is as follows: From the moment of power-on, continuously acquire device sensing data of the target device after power-on, and assemble power-on recovery timing data according to the sampling time sequence; synchronously read the reference value of the target device's operating output power and the rated output voltage of the power line carrier networking power module; Based on the power-on recovery timing data, the ratio of the current output power to the reference value of the operating output power is calculated, and the ratio is input into the hyperbolic tangent function to obtain the operating power recovery component; Calculate the absolute value of the output power change, divide the absolute value of the output power change by the sum of the operating output power reference value and the smallest positive number, and calculate the negative exponent to obtain the power fluctuation suppression component. Add the current I2C interface response status to the UART interface response status and divide by two to obtain the interface response recovery component; Calculate the absolute difference between the current output voltage and the rated output voltage, and divide the absolute difference by the sum of the rated output voltage and the smallest positive number to obtain the output voltage deviation ratio. Add one to the output voltage deviation ratio and take the reciprocal to obtain the stable component of the output voltage; Multiply the operating power recovery component, power fluctuation suppression component, interface response recovery component, and output voltage stability component to obtain the restart stability confirmation value.

9. The method for out-of-band management and sensing monitoring recovery based on power line carrier as described in claim 8, characterized in that, The specific process for determining the recovery result of the target device based on the degree of restart stability is as follows: The restart stability confirmation value is compared with the stability confirmation threshold in real time. When the restart stability confirmation value is greater than or equal to the stability confirmation threshold for K consecutive sampling periods, the target device is determined to have recovered successfully. The sampling time when the condition is first met is recorded as the recovery success time. The difference between the recovery success time and the power-on time is calculated to obtain the recovery success time. If the restart stability confirmation value is less than the stability confirmation threshold, and the time difference between the current time and the power-on time is greater than the maximum value of the historical successful recovery time, the target device is determined to have failed to recover.

10. The method for out-of-band management and sensing monitoring recovery based on power line carrier according to claim 9, characterized in that, The specific process for implementing fault self-recovery measures when recovery fails is as follows: When the target device fails to recover, the cumulative number of consecutive recovery failures of the target device is read and incremented by one; if the cumulative number of consecutive recovery failures is less than the upper limit threshold of the recovery count, the power outage recovery command is resent to the power line carrier network power module and the power outage recovery is re-executed. If the cumulative number of consecutive recovery failures is greater than or equal to the upper limit threshold for recovery attempts, the power outage recovery process will be stopped and a manual inspection prompt will be output. Write the recovery success time, recovery success duration, restart stability confirmation value sequence, and judgment result corresponding to the recovery result into the device out-of-band recovery database.

Images