An inductance state intelligent monitoring method based on internet of things

Abstract

The application discloses an inductance state intelligent monitoring method based on Internet of Things, which comprises the following steps: collecting inductance operation data and forming an inductance conduction monitoring sequence according to a unified time index; identifying a current working condition and combining normal calibration data to generate a normal conduction skeleton template; mapping the monitoring sequence to an electromagnetic heat conduction chain and generating a conduction fitting result; identifying conduction mismatch according to the fitting result and forming a scattering event set; reducing the scattering events according to an external working condition explanation order and retaining unexplained endogenous scattering; tracing an abnormal diffusion path along the electromagnetic heat conduction chain and generating a closed-loop monitoring result. The application uses the Internet of Things electromagnetic heat conduction skeleton scattering method, realizes inductance endogenous anomaly identification and closed-loop early warning, and has the advantages of low false alarm rate, strong explanation and high monitoring stability.

Description

Technical Field

[0001] This invention relates to the field of electronic component condition monitoring technology, and in particular to an intelligent inductor condition monitoring method based on the Internet of Things. Background Technology

[0002] Current inductor condition monitoring methods largely rely on single data points such as current, voltage, ripple, or temperature for assessment. A common approach is to collect data on inductor current, inductor terminal voltage, and ripple changes during power supply or converter operation, and then compare the results with preset thresholds, historical curves, or normality tables to determine if the inductor is at risk of failure. While this method is simple to implement and suitable for relatively stable online monitoring scenarios, its judgment criteria typically focus on a limited number of electrical characteristics, failing to fully express the combined influence of load conditions, switching control, power supply input / output status, and thermal environment on the inductor response process.

[0003] In actual operation, there is a clear electromagnetic-thermal conduction relationship between inductor current, ripple morphology, and inductor casing temperature. Load changes, switching frequency adjustments, input / output voltage fluctuations, ambient temperature changes, and onboard heat dissipation conditions can all cause similar response deviations. Existing methods typically lack normal conduction framework templates for different operating conditions, and also lack layered identification of conduction relationship mismatch, response timing mismatch, response amplitude mismatch, and response morphology mismatch. This makes it easy to misjudge external operating condition disturbances as inductor-generated anomalies, making it difficult to further trace the anomaly propagation path and form stable closed-loop monitoring results.

[0004] Therefore, how to provide an intelligent monitoring method for inductor status based on the Internet of Things is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] One objective of this invention is to propose an intelligent monitoring method for inductor status based on the Internet of Things (IoT). This invention utilizes the IoT electromagnetic thermal conduction skeleton scattering method to achieve inductor endogenous anomaly identification and closed-loop early warning, and has the advantages of low false alarm rate, strong interpretability and high monitoring stability.

[0006] An intelligent inductor status monitoring method based on the Internet of Things according to an embodiment of the present invention includes the following steps: Inductor operation data is collected through an IoT acquisition terminal and packaged into an inductor conduction monitoring data sequence according to a unified time index. Based on the inductance conduction monitoring data sequence, the inductor's operating condition is identified, an equivalent operating condition unit is generated, and a normal conduction skeleton template corresponding to the equivalent operating condition unit is generated by combining the inductor's normal calibration data. Based on the normal conduction skeleton template, the inductive conduction monitoring data sequence is mapped to the electromagnetic thermal conduction chain, and conduction consistency verification is performed to generate conduction consistency results; Based on the conduction matching results, conduction relationship mismatch, response timing mismatch, response amplitude mismatch, and response morphology mismatch are identified, and a set of scattering events is generated. The scattering event set is subjected to interpretation reduction according to the external operating condition interpretation order, generating uninterpreted endogenous scattering results; Based on unexplained endogenous scattering results, conduction anomalies are traced in the electromagnetic heat conduction chain to generate inductor endogenous anomaly diffusion paths, and inductor state closed-loop monitoring results are generated based on the inductor endogenous anomaly diffusion paths.

[0007] Optionally, the generation of the inductance conduction monitoring data sequence specifically includes: A unified time index is established using the sampling clock of the IoT acquisition terminal. The electrical response sampling record, control drive sampling record, and thermal environment sampling record corresponding to the monitored inductor are collected to form the original inductor operation data. The electrical response sampling record consists of inductor terminal voltage, inductor current, ripple morphology, and inductor housing temperature; the control drive sampling record consists of load status, switch control, and power input / output status; and the thermal environment sampling record consists of ambient temperature, onboard temperature, and heat dissipation status. Perform sampling integrity checks, timestamp rearrangement, and invalid sample removal on the original inductor operating data to form valid inductor operating data; Time alignment is performed on the valid inductor operating data according to a unified time index to form aligned inductor operating data; The control drive sampling records are read from the aligned inductor operation data and combined with the inductor terminal voltage in the electrical response sampling records to form a drive relationship data group; The inductor current, ripple morphology, and inductor case temperature are read from the electrical response sampling records of the aligned inductor operating data to form a response relationship data set. Thermal environment sampling records are read from the aligned inductor operating data to form a thermal interpretation relationship data set; The drive relationship data group, response relationship data group, and thermal interpretation relationship data group under the same unified time index are encapsulated and arranged in the order of the unified time index to generate an inductance conduction monitoring data sequence.

[0008] Optionally, the generation of the normal conduction skeleton template specifically includes: The driving relationship data set, response relationship data set, and thermal interpretation relationship data set are extracted frame by frame from the inductance conduction monitoring data sequence to form the operating condition identification data set; Based on the inductance conduction monitoring data sequence, the load status, switch control status, power input and output status and thermal environment status are identified, and the current operating condition description result is generated. Based on the current operating condition description results, the inductance conduction monitoring data sequence is merged to generate an equivalent operating condition unit; Select normal calibration segments corresponding to the equivalent operating condition unit from the normal calibration data of the inductor to form the equivalent operating condition calibration set; Based on the equivalent operating condition calibration set, the normal conduction relationship of inductance is extracted to form a set of normal conduction skeleton elements; The collection of normal conduction skeleton elements is encapsulated into a normal conduction skeleton template.

[0009] Optionally, the set of normal conduction skeleton elements consists of fixed conduction relationships and corresponding conduction attributes; The fixed conduction relationships include the normal driving relationship from load state to inductor current, the normal driving relationship from switch control to ripple mode, the normal driving relationship from power input / output state to inductor terminal voltage, the normal response relationship from inductor terminal voltage to inductor current, the normal response relationship from inductor current to ripple mode, the normal hysteresis relationship from inductor current to inductor housing temperature, the normal interpretation relationship from ambient temperature to inductor housing temperature, and the normal interpretation relationship from onboard temperature and heat dissipation state to inductor housing temperature. The corresponding conduction attributes consist of the calibration conduction direction, calibration response sequence, calibration hysteresis window, calibration amplitude conversion range, calibration silence mark, calibration ripple morphology range, calibration temperature rise recovery range, and calibration interpretation priority of each fixed conduction relationship in the equivalent operating condition calibration set.

[0010] Optionally, the generation of the conduction anastomosis result specifically includes: Based on the fixed conduction relationship in the normal conduction skeleton template, the inductive conduction monitoring data sequence is mapped to the electromagnetic thermal conduction chain; The changes of each fixed conduction relationship along the electromagnetic heat conduction chain in the inductive conduction monitoring data sequence are checked, and the conduction direction check results and response timing check results are generated. According to the calibration lag window and calibration amplitude conversion range in the normal conduction skeleton template, perform lag verification and amplitude verification on the electromagnetic heat conduction chain to generate lag matching verification results and amplitude matching verification results; Perform ripple shape verification on fixed conduction relationships involving ripple shape, and perform temperature rise recovery verification on fixed conduction relationships involving inductor housing temperature, generating shape matching verification results; Perform a silent check on the fixed conduction relationship of the configured silent marker, and perform an interpretation priority check on the conduction chain from ambient temperature to inductor case temperature, onboard temperature, and heat dissipation status to inductor case temperature, generating silent consistency check results and interpretation priority check results; The transmission direction verification results, response timing verification results, lag matching verification results, amplitude matching verification results, morphological matching verification results, silent matching verification results, and interpretation priority verification results are summarized to generate transmission matching results.

[0011] Optionally, the generation of the scattering event set specifically includes: Based on the conduction direction verification results and silent matching verification results in the conduction matching results, conduction relationship mismatch is identified, and conduction relationship mismatch events are generated. Based on the response timing verification results and lag matching results in the transmission matching results, response timing mismatch is identified, and response timing mismatch events are generated. Based on the amplitude matching verification results in the conduction matching results, response amplitude mismatch is identified, and a response amplitude mismatch event is generated; Based on the morphological match verification results in the conduction match results, response morphological mismatch is identified, and response morphological mismatch events are generated. The events of conduction mismatch, response timing mismatch, response amplitude mismatch, and response morphology mismatch are aggregated according to a unified time index to generate a scattering event set.

[0012] Optionally, the generation of the unexplained endogenous scattering results specifically includes: Based on the fixed conduction relationship, mismatch type, and interpretation priority verification results corresponding to each scattering event in the scattering event set, the external operating condition interpretation order is determined, and the interpretation reduction order result is generated. Perform load state interpretation reduction on the scattering event set according to the interpretation reduction order, and generate load state interpretation reduction results; According to the order of interpretation and deletion, the load state interpretation and deletion results are processed to perform switch control interpretation and deletion, and switch control interpretation and deletion results are generated. According to the order of interpretation and deletion, the power input and output status interpretation and deletion results are performed on the switch control interpretation and deletion results to generate the power input and output status interpretation and deletion results; According to the interpretation and deletion order, the thermal environment interpretation and deletion results are performed on the power input and output state interpretation and deletion results to generate thermal environment interpretation and deletion results; Scattering events retained in the thermal environment interpretation and deletion results are registered as uninterpreted endogenous scattering results, and the fixed conduction relationship, mismatch type and corresponding inductive conduction monitoring frame of the corresponding scattering event are retained.

[0013] Optionally, the generation of the inductor state closed-loop monitoring result specifically includes: According to the unified time index, the anomalous initiation scattering event is identified from the unexplained endogenous scattering results, and the fixed conduction relationship corresponding to the anomalous initiation scattering event is identified as the anomalous initiation conduction relationship. Starting with the abnormal initial conduction relationship, the adjacent fixed conduction relationships are matched along the conduction endpoint in the electromagnetic heat conduction chain to generate candidate endogenous abnormal paths; Perform continuity and direction checks on candidate endogenous anomaly paths to generate inductor endogenous anomaly propagation paths; Based on the fixed conduction relationship and mismatch type in the diffusion path of inductor endogenous anomalies, the inductor state category is determined, including core saturation approaching state, winding heating and degradation state, suspected local short circuit state, and abnormal solder joint termination state. Early warning trigger results are generated based on the inductor state category and the propagation path of endogenous anomalies in the inductor. Based on the inductor status category, early warning triggering result, and the diffusion path of inductor endogenous anomalies, a closed-loop monitoring result of the inductor status is generated.

[0014] Optionally, the inductor state closed-loop monitoring results include inductor state category, inductor endogenous anomaly diffusion path, early warning trigger result, and normal conduction skeleton template update control content. When the inductor state category is core saturation approaching state, winding heating degradation state, suspected local short circuit state, or solder joint termination abnormal state, the normal conduction skeleton template update control content prohibits the corresponding inductor conduction monitoring frame from entering the normal conduction skeleton template update. When the unexplained endogenous scattering result is empty and the scattering events in the scattering event set have been explained and deleted, the normal conduction skeleton template update control content allows the corresponding inductor conduction monitoring frame to enter the normal conduction skeleton template update.

[0015] The beneficial effects of this invention are: This invention encapsulates electrical response sampling records, control drive sampling records, and thermal environment sampling records into an inductor conduction monitoring data sequence using an IoT acquisition terminal, indexed by a unified time. Based on the inductor's operating conditions, it generates an equivalent operating condition unit and a corresponding normal conduction framework template, establishing a unified foundation for conduction analysis of inductor terminal voltage, inductor current, ripple morphology, inductor casing temperature and load status, switch control, power input / output status, and thermal environment status. Therefore, this invention no longer relies on single current, ripple, or temperature thresholds for judgment, but can match and verify the electromagnetic response, ripple response, and thermal response of the inductor under different operating conditions, improving the completeness and accuracy of inductor status identification under complex operating conditions.

[0016] This invention maps inductance conduction monitoring data sequences to electromagnetic thermal conduction chains based on a normal conduction framework template. It then identifies conduction mismatches, response timing mismatches, response amplitude mismatches, and response morphology mismatches through conduction matching verification, generating a set of scattering events. Furthermore, the scattering event set is interpreted and reduced according to the external operating condition interpretation order. Scattering events that can be explained by load status, switch control, power input / output status, ambient temperature, onboard temperature, and heat dissipation status are eliminated, retaining only uninterpreted endogenous scattering results. This effectively distinguishes normal offsets caused by external operating condition disturbances from inductor anomalies, reducing false alarms caused by load transitions, switching frequency changes, power supply fluctuations, and thermal environment changes.

[0017] This invention tracks conduction anomalies in the electromagnetic heat conduction chain based on uninterpreted endogenous scattering results, generates an inductor endogenous anomaly diffusion path, and forms a closed-loop monitoring result for the inductor state. This allows the monitoring results to reflect the anomaly initiation conduction relationship, anomaly diffusion direction, inductor state category, and early warning triggering results. By controlling the updating of the normal conduction skeleton template, it prohibits inductor conduction monitoring frames corresponding to endogenous anomalies from entering the template update, while allowing monitoring frames that have been deleted due to external operating condition interpretation to participate in the update. This reduces the risk of anomaly samples contaminating the normal baseline and improves the stability and adaptability of long-term online monitoring. Attached Figure Description

[0018] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of an intelligent inductor status monitoring method based on the Internet of Things proposed in this invention; Figure 2 This is a flowchart illustrating the construction and consistency verification of the electromagnetic heat conduction chain in an IoT-based intelligent inductor state monitoring method proposed in this invention. Figure 3 This is a flowchart illustrating the scattering event reduction and endogenous anomaly tracking of an IoT-based intelligent inductor state monitoring method proposed in this invention. Detailed Implementation

[0019] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0020] refer to Figures 1-3 A smart monitoring method for inductor status based on the Internet of Things includes the following steps: Inductor operation data is collected through an IoT acquisition terminal and packaged into an inductor conduction monitoring data sequence according to a unified time index. Based on the inductance conduction monitoring data sequence, the inductor's operating condition is identified, an equivalent operating condition unit is generated, and a normal conduction skeleton template corresponding to the equivalent operating condition unit is generated by combining the inductor's normal calibration data. Based on the normal conduction skeleton template, the inductive conduction monitoring data sequence is mapped to the electromagnetic thermal conduction chain, and conduction consistency verification is performed to generate conduction consistency results; Based on the conduction matching results, conduction relationship mismatch, response timing mismatch, response amplitude mismatch, and response morphology mismatch are identified, and a set of scattering events is generated. The scattering event set is subjected to interpretation reduction according to the external operating condition interpretation order, generating uninterpreted endogenous scattering results; Based on unexplained endogenous scattering results, conduction anomalies are traced in the electromagnetic heat conduction chain to generate inductor endogenous anomaly diffusion paths, and inductor state closed-loop monitoring results are generated based on the inductor endogenous anomaly diffusion paths.

[0021] In this embodiment, the generation of the inductance conduction monitoring data sequence specifically includes: A unified time index is established using the sampling clock of the IoT acquisition terminal. The electrical response sampling record, control drive sampling record, and thermal environment sampling record corresponding to the monitored inductor are collected to form the original inductor operation data. The electrical response sampling record consists of inductor terminal voltage, inductor current, ripple morphology, and inductor case temperature; the control drive sampling record consists of load status, switch control, and power input / output status; and the thermal environment sampling record consists of ambient temperature, onboard temperature, and heat dissipation status. Perform sampling integrity checks, timestamp rearrangement, and invalid sample removal on the original inductor operating data to form valid inductor operating data; Time alignment is performed on the valid inductor operating data according to a unified time index to form aligned inductor operating data; The control drive sampling records are read from the aligned inductor operation data and combined with the inductor terminal voltage in the electrical response sampling records to form a drive relationship data group; The inductor current, ripple morphology, and inductor case temperature are read from the electrical response sampling records of the aligned inductor operating data to form a response relationship data set. Thermal environment sampling records are read from the aligned inductor operating data to form a thermal interpretation relationship data set; The drive relationship data group, response relationship data group, and thermal interpretation relationship data group under the same unified time index are encapsulated and arranged in the order of the unified time index to generate an inductance conduction monitoring data sequence.

[0022] In this embodiment, the generation of the normal conduction skeleton template specifically includes: The driving relationship data set, response relationship data set, and thermal interpretation relationship data set are extracted frame by frame from the inductance conduction monitoring data sequence to form the operating condition identification data set; Based on the inductance conduction monitoring data sequence, the load status, switch control status, power input and output status and thermal environment status are identified, and the current operating condition description result is generated. The current operating condition description results include the load state, switch control and power input / output state in the drive relationship data group, and the ambient temperature, onboard temperature and heat dissipation state in the thermal interpretation relationship data group. The load state represents the driving conditions of the external load on the inductor current, the switch control represents the control conditions of the inductor ripple shape, the power input / output state represents the change conditions of the inductor terminal voltage, and the thermal environment state represents the external thermal interpretation conditions of the inductor housing temperature. Based on the current operating condition description results, the inductance conduction monitoring data sequence is merged to generate an equivalent operating condition unit; The operating conditions are grouped using load status identifier, switch control status identifier, power input / output status identifier, and thermal environment status identifier as fixed grouping fields. The load status identifier includes load level and load change direction, which includes load increase, load decrease, and load hold. The switch control status identifier includes switch frequency level, duty cycle level, and operating mode. The power input / output status identifier includes input voltage level, output voltage level, and terminal voltage change direction, which includes terminal voltage increase, terminal voltage decrease, and terminal voltage hold. The thermal environment status identifier includes ambient temperature level, onboard temperature level, and heat dissipation status, which includes heat dissipation on, heat dissipation off, and heat dissipation adjustment. Inductance conduction monitoring frames with the same four types of status identifiers in continuous inductance conduction monitoring frames are grouped into the same equivalent operating condition unit. Select normal calibration segments corresponding to the equivalent operating condition unit from the normal calibration data of the inductor to form the equivalent operating condition calibration set; The normal calibration data of the inductor consists of the driving relationship data group, response relationship data group and thermal interpretation relationship data group obtained by the monitored inductor under normal conditions according to a unified time index. The normal calibration segment is a continuous calibration segment in the normal calibration data of the inductor that has the same load status identifier, switch control status identifier, power input and output status identifier and thermal environment status identifier as the equivalent operating condition unit. Based on the equivalent operating condition calibration set, the normal conduction relationship of inductance is extracted to form a set of normal conduction skeleton elements; Encapsulate the set of normal conduction skeleton elements into a normal conduction skeleton template; The normal conduction skeleton template includes fixed conduction relationships and corresponding conduction attributes. The fixed conduction relationships limit the conduction objects of inductor operating data in the equivalent operating condition unit, and the corresponding conduction attributes limit the verification basis for subsequent conduction consistency checks.

[0023] In this embodiment, the normal conductive skeleton element set consists of fixed conductive relationships and corresponding conductive attributes; The fixed conduction relationships include the normal drive relationship from load state to inductor current, the normal drive relationship from switch control to ripple mode, the normal drive relationship from power input / output state to inductor terminal voltage, the normal response relationship from inductor terminal voltage to inductor current, the normal response relationship from inductor current to ripple mode, the normal hysteresis relationship from inductor current to inductor housing temperature, the normal interpretation relationship from ambient temperature to inductor housing temperature, and the normal interpretation relationship from onboard temperature and heat dissipation state to inductor housing temperature. The corresponding conduction attributes consist of the calibration conduction direction, calibration response sequence, calibration hysteresis window, calibration amplitude conversion range, calibration silence mark, calibration ripple morphology range, calibration temperature rise recovery range, and calibration interpretation priority of each fixed conduction relationship in the equivalent working condition calibration set. For each fixed conduction relationship, the side that changes first is defined as the conduction start point, and the side that responds to the change at the conduction start point is defined as the conduction end point. The calibration conduction direction and calibration response sequence are determined according to the order of change from the conduction start point to the conduction end point. The calibration lag window is determined according to the time interval between the change at the conduction start point and the response change at the conduction end point. The calibration amplitude conversion range is determined according to the correspondence between the change amplitude at the conduction start point and the change amplitude at the conduction end point. When the conduction end point in the fixed conduction relationship remains stable within the equivalent operating condition calibration set, a calibration quiescent flag is configured for that fixed conduction relationship. When the fixed conduction relationship involves ripple morphology, the calibration ripple morphology range is determined based on the periodic, symmetrical, and peak variations of the ripple morphology within the equivalent operating condition calibration set. When the fixed conduction relationship involves inductor housing temperature, the calibration temperature rise recovery range is determined based on the duration of the inductor housing temperature rise and fall within the equivalent operating condition calibration set. When the fixed conduction relationship is a normal interpretation relationship from ambient temperature to inductor housing temperature or a normal interpretation relationship from onboard temperature and heat dissipation status to inductor housing temperature, the calibration interpretation priority is determined based on the order of changes in ambient temperature, onboard temperature, and heat dissipation status relative to inductor housing temperature.

[0024] In this embodiment, the generation of the conduction anastomosis result specifically includes: Based on the fixed conduction relationship in the normal conduction skeleton template, the inductive conduction monitoring data sequence is mapped to the electromagnetic thermal conduction chain; The electromagnetic heat conduction chain retains the fixed conduction relationship in the normal conduction skeleton template. The inductance conduction monitoring frame corresponding to the start point of each fixed conduction relationship and the inductance conduction monitoring frame corresponding to the end point are established according to a unified time index to establish a conduction correspondence relationship. This includes the conduction chain from load state to inductor current, the conduction chain from switch control to ripple shape, the conduction chain from power input and output state to inductor terminal voltage, the conduction chain from inductor terminal voltage to inductor current, the conduction chain from inductor current to ripple shape, the conduction chain from inductor current to inductor housing temperature, the conduction chain from ambient temperature to inductor housing temperature, and the conduction chain from onboard temperature and heat dissipation state to inductor housing temperature. The changes of each fixed conduction relationship along the electromagnetic heat conduction chain in the inductive conduction monitoring data sequence are checked, and the conduction direction check results and response timing check results are generated. The change sequence verification takes the fixed conduction relationship's conduction start point and conduction end point as the verification objects. The inductor conduction monitoring frames where the conduction start point changes and the inductor conduction monitoring frames where the conduction end point responds are read according to a unified time index. When the response change of the conduction end point is after the change of the conduction start point, a conduction direction verification result consistent with the calibrated conduction direction is formed. The response timing verification result is formed based on the frame interval between the change of the conduction start point and the response change of the conduction end point. According to the calibration lag window and calibration amplitude conversion range in the normal conduction skeleton template, perform lag verification and amplitude verification on the electromagnetic heat conduction chain to generate lag matching verification results and amplitude matching verification results; Perform ripple shape verification on fixed conduction relationships involving ripple shape, and perform temperature rise recovery verification on fixed conduction relationships involving inductor housing temperature, generating shape matching verification results; The ripple morphology verification compares the ripple morphology in the inductance conduction monitoring data sequence with the calibrated ripple morphology range. The ripple morphology includes ripple period, ripple symmetry, and ripple peak state. The temperature rise recovery verification compares the duration of temperature rise and fall of the inductor casing in the inductance conduction monitoring data sequence with the calibrated temperature rise recovery range. Perform a silent check on the fixed conduction relationship of the configured silent marker, and perform an interpretation priority check on the conduction chain from ambient temperature to inductor case temperature, onboard temperature, and heat dissipation status to inductor case temperature, generating silent consistency check results and interpretation priority check results; The silent verification checks whether the conduction endpoint of the configured silent marker remains stable after the conduction starting point of the fixed conduction relationship changes. The interpretation priority verification determines the interpretation order of ambient temperature, onboard temperature and heat dissipation status on the inductor case temperature change according to the calibration interpretation priority. The transmission direction verification results, response timing verification results, lag matching verification results, amplitude matching verification results, morphological matching verification results, silent matching verification results, and interpretation priority verification results are summarized to generate transmission matching results.

[0025] In this embodiment, the generation of the scattering event set specifically includes: Based on the conduction direction verification results and silent matching verification results in the conduction matching results, conduction relationship mismatch is identified, and conduction relationship mismatch events are generated. The conduction mismatch includes the following: the fixed conduction relationship does not form a response correspondence from the conduction start point to the conduction end point in the electromagnetic thermal conduction chain; the response change at the conduction end point does not conform to the calibrated conduction direction; and the conduction end point configured with the calibrated silent mark does not remain stable after the conduction start point changes. The fixed conduction relationship that has caused the conduction mismatch, the corresponding inductance conduction monitoring frame, and the conduction mismatch type will be written into the conduction mismatch event. Based on the response timing verification results and lag matching results in the transmission matching results, response timing mismatch is identified, and response timing mismatch events are generated. The response timing mismatch includes the response change at the conduction endpoint earlier than the change at the conduction start point, the response change at the conduction endpoint not falling into the calibration lag window, and the response change at the conduction endpoint only forming after the calibration lag window has ended. The fixed conduction relationship where the response timing mismatch occurs, the corresponding inductance conduction monitoring frame, and the response timing mismatch type are written into the response timing mismatch event. Based on the amplitude matching verification results in the conduction matching results, response amplitude mismatch is identified, and a response amplitude mismatch event is generated; The response amplitude mismatch includes situations where the response amplitude at the conduction endpoint does not fall within the calibrated amplitude conversion range after the conduction start point changes, the response amplitude at the conduction endpoint deviates from the calibrated amplitude conversion range in continuous inductance conduction monitoring frames, and the response amplitude change at the conduction endpoint with a calibrated silence flag exceeds the silence requirement. The fixed conduction relationship where response amplitude mismatch occurs, the corresponding inductance conduction monitoring frame, and the response amplitude mismatch type are written into the response amplitude mismatch event. Based on the morphological match verification results in the conduction match results, response morphological mismatch is identified, and response morphological mismatch events are generated. The response pattern mismatch includes ripple pattern not falling within the calibrated ripple pattern range, inductor housing temperature rise not falling within the calibrated temperature rise recovery range, and inductor housing temperature fall not falling within the calibrated temperature rise recovery range. The fixed conduction relationship of the response pattern mismatch, the corresponding inductor conduction monitoring frame, and the response pattern mismatch type will be written into the response pattern mismatch event. The events of conduction mismatch, response timing mismatch, response amplitude mismatch, and response morphology mismatch are aggregated according to a unified time index to generate a scattering event set. Each scattering event in the scattering event set retains its corresponding fixed conduction relationship, mismatch type, corresponding inductance conduction monitoring frame, and interpretation priority verification result.

[0026] In this embodiment, the generation of the endogenous scattering results is not explained in detail, including: Based on the fixed conduction relationship, mismatch type, and interpretation priority verification results corresponding to each scattering event in the scattering event set, the external operating condition interpretation order is determined, and the interpretation reduction order result is generated. The external operating condition interpretation order is determined by the fixed conduction relationship corresponding to the scattering event to determine the candidate interpretation objects, and the order of deletion is determined by the conduction position of the candidate interpretation objects in the electromagnetic heat conduction chain. Specifically, when the scattering event involves inductor current, the load state interpretation is listed as a candidate interpretation object; when the scattering event involves ripple pattern, the switch control interpretation is listed as a candidate interpretation object; and when the scattering event also involves the conduction chain from inductor current to ripple pattern, the load state interpretation is arranged before the switch control interpretation. When the scattering event involves inductor terminal voltage or the conduction chain from inductor terminal voltage to inductor current, the power input / output state interpretation is listed as a candidate interpretation object. When the scattering event involves inductor housing temperature, the ambient temperature interpretation, onboard temperature, and heat dissipation state interpretation are listed as candidate interpretation objects. The order of ambient temperature interpretation, onboard temperature, and heat dissipation state interpretation is determined according to the interpretation priority verification results. The candidate interpretation objects are arranged in a fixed order: load state interpretation, switch control interpretation, power input / output state interpretation, ambient temperature interpretation, onboard temperature, and heat dissipation state interpretation. Interpretations not listed as candidate interpretation objects are not included in the interpretation reduction of the scattering event, resulting in the interpretation reduction order. Perform load state interpretation reduction on the scattering event set according to the interpretation reduction order, and generate load state interpretation reduction results; The load state interpretation and elimination is based on the conduction chain from the load state to the inductor current. Scattering events involving inductor current, ripple shape, and inductor case temperature in the scattering event set are checked. When the fixed conduction relationship corresponding to the scattering event can form a conduction direction, response timing, response amplitude, and response shape consistent with the conduction result along the conduction chain from the load state to the inductor current, the conduction chain from the inductor current to the ripple shape, or the conduction chain from the inductor current to the inductor case temperature, the scattering event is marked as a load state interpreted scattering event and eliminated from the scattering event set. Scattering events not eliminated by the load state interpretation form the load state interpretation and elimination result. According to the order of interpretation and deletion, the load state interpretation and deletion results are processed to perform switch control interpretation and deletion, and switch control interpretation and deletion results are generated. The interpretation and deletion of switch control is based on the conduction chain from switch control to ripple shape. The scattering events involving ripple shape in the load state interpretation and deletion results are checked. When the ripple shape change corresponding to the scattering event can form a conduction direction, response timing, response amplitude and response shape consistent with the conduction result along the conduction chain from switch control to ripple shape, the scattering event is marked as a scattering event that has been interpreted by switch control and deleted from the load state interpretation and deletion results. The scattering events that have not been interpreted and deleted by switch control form the switch control interpretation and deletion results. According to the order of interpretation and deletion, the power input and output status interpretation and deletion results are performed on the switch control interpretation and deletion results to generate the power input and output status interpretation and deletion results; The power input / output state interpretation and deletion is based on the conduction chain from the power input / output state to the inductor terminal voltage and the conduction chain from the inductor terminal voltage to the inductor current. Scattering events involving inductor terminal voltage, inductor current, and ripple shape in the switch control interpretation and deletion results are checked. When the changes in inductor terminal voltage, inductor current, or ripple shape corresponding to the scattering event can form a conduction direction, response timing, response amplitude, and response shape consistent with the conduction results along the conduction chain from the power input / output state to the inductor terminal voltage, the conduction chain from the inductor terminal voltage to the inductor current, or the conduction chain from the inductor current to the ripple shape, the scattering event is marked as a power input / output state interpreted scattering event and deleted from the switch control interpretation and deletion results. Scattering events not deleted by the power input / output state interpretation and deletion results are formed in the power input / output state interpretation and deletion results. According to the interpretation and deletion order, the thermal environment interpretation and deletion results are performed on the power input and output state interpretation and deletion results to generate thermal environment interpretation and deletion results; The thermal environment interpretation and deletion are based on the conduction chain from ambient temperature to inductor housing temperature, onboard temperature, and heat dissipation status to inductor housing temperature. Scattering events involving inductor housing temperature in the power input / output status interpretation and deletion results are checked. When the inductor housing temperature change corresponding to the scattering event can form a conduction direction, response timing, response amplitude, and response mode consistent with the conduction results along the conduction chain from ambient temperature to inductor housing temperature, or along the conduction chain from onboard temperature and heat dissipation status to inductor housing temperature, the scattering event is marked as a thermally interpreted scattering event and deleted from the power input / output status interpretation and deletion results. Scattering events not deleted by the thermal environment interpretation and deletion results are formed in the thermal environment interpretation and deletion results. Scattering events retained in the thermal environment interpretation and deletion results are registered as uninterpreted endogenous scattering results, and the fixed conduction relationship, mismatch type and corresponding inductive conduction monitoring frame of the corresponding scattering event are retained.

[0027] In this embodiment, the generation of the inductor state closed-loop monitoring result specifically includes: According to the unified time index, the anomalous initiation scattering event is identified from the unexplained endogenous scattering results, and the fixed conduction relationship corresponding to the anomalous initiation scattering event is identified as the anomalous initiation conduction relationship. Among them, the abnormal initiation scattering event is the first scattering event to appear in the unexplained endogenous scattering results. When there are multiple scattering events under the same unified time index, the abnormal initiation scattering event is determined in sequence according to the fixed conduction relationship involving the inductor terminal voltage, the fixed conduction relationship involving the inductor current, the fixed conduction relationship involving the ripple shape, and the fixed conduction relationship involving the inductor shell temperature. Starting with the abnormal initial conduction relationship, the adjacent fixed conduction relationships are matched along the conduction endpoint in the electromagnetic heat conduction chain to generate candidate endogenous abnormal paths; The adjacent fixed conduction relationship is the conduction starting point of the previous fixed conduction relationship, and there is a fixed conduction relationship corresponding to the scattering event in the unexplained endogenous scattering results. The candidate endogenous anomaly path connects the anomaly starting conduction relationship and its adjacent fixed conduction relationship in the order of the unified time index. Perform continuity and direction checks on candidate endogenous anomaly paths to generate inductor endogenous anomaly propagation paths; The continuity check and direction check mainly check whether adjacent scattering events in the candidate endogenous anomaly path appear continuously or at intervals within the calibration lag window according to a unified time index, and check whether the fixed conduction relationship in the candidate endogenous anomaly path extends along the calibration conduction direction. The candidate endogenous anomaly path that simultaneously satisfies the continuity check and direction check is determined as the inductor endogenous anomaly diffusion path. Based on the fixed conduction relationship and mismatch type in the diffusion path of inductor endogenous anomalies, the inductor state category is determined, including core saturation approaching state, winding heating and degradation state, suspected local short circuit state, and abnormal solder joint termination state. Among them, when the inductor's endogenous abnormal diffusion path includes a conduction chain from the inductor terminal voltage to the inductor current and a conduction chain from the inductor current to the ripple morphology, and the mismatch type includes response amplitude mismatch or response morphology mismatch, it is determined to be a core saturation approaching state. Among them, when the inductor's internal abnormal diffusion path includes a conduction chain from the inductor current to the inductor's shell temperature, and the mismatch type includes response timing mismatch or response morphology mismatch, it is determined to be a winding thermal degradation state. Among them, when the inductor's internal abnormal diffusion path includes the conduction chain from the inductor terminal voltage to the inductor current, the conduction chain from the inductor current to the ripple morphology, and the conduction chain from the inductor current to the inductor casing temperature, and the power input and output status interpretation deletes or does not delete the corresponding scattering event, it is determined to be a suspected local short circuit state. Among them, when the inductor's internal abnormal diffusion path includes a conduction chain from the power input / output state to the inductor terminal voltage or a conduction chain from the inductor terminal voltage to the inductor current, and the mismatch type includes conduction relationship mismatch or response mode mismatch, it is determined to be an abnormal solder joint termination state. Early warning trigger results are generated based on the inductor state category and the propagation path of endogenous anomalies in the inductor. The warning trigger result determines the warning level according to the number of occurrences, duration frames and number of fixed conduction relationships involved in the scattering event in the inductor endogenous abnormal diffusion path. When the inductor state category is the magnetic core saturation approaching state, winding heating degradation state, suspected local short circuit state or solder joint termination abnormal state, the warning trigger result records the corresponding inductor state category and inductor endogenous abnormal diffusion path. Based on the inductor status category, early warning triggering result, and the diffusion path of inductor endogenous anomalies, a closed-loop monitoring result of the inductor status is generated.

[0028] In this embodiment, the closed-loop monitoring results of the inductor state include the inductor state category, the diffusion path of the inductor's intrinsic anomaly, the warning trigger result, and the normal conduction skeleton template update control content. When the inductor state category is the core saturation approaching state, the winding heating degradation state, the suspected local short circuit state, or the solder joint termination abnormal state, the normal conduction skeleton template update control content prohibits the corresponding inductor conduction monitoring frame from entering the normal conduction skeleton template update. When the unexplained intrinsic scattering result is empty and the scattering events in the scattering event set have been explained and deleted, the normal conduction skeleton template update control content allows the corresponding inductor conduction monitoring frame to enter the normal conduction skeleton template update.

[0029] Example 1: To verify the feasibility of this invention in practice, it was applied to the aging test workshop of a new energy charging module manufacturing company. The charging modules in this workshop are subjected to alternating loads, frequent start-stop cycles, and temperature fluctuations for extended periods. During operation, the power inductors are prone to phenomena such as ripple drift, slowed temperature recovery of the casing, and voltage edge disturbances. Previously, on-site assessment relied mainly on current ripple thresholds, temperature thresholds, and manual verification to determine the inductor's condition. However, when the load suddenly changes, the switching control strategy switches, or the workshop ambient temperature rises, disturbances to normal operating conditions are easily mistaken for inductor anomalies. Furthermore, when early degradation occurs within the inductor, a single threshold is insufficient to promptly distinguish the source of the anomaly, requiring maintenance personnel to repeatedly check waveforms, temperature curves, and equipment logs, resulting in a lengthy assessment process.

[0030] In this scenario, an IoT data acquisition unit is installed near the charging module testing station to synchronously collect data on the inductor's terminal voltage, inductor current, ripple pattern, inductor casing temperature, load status, switch control, power input / output status, ambient temperature, onboard temperature, and heat dissipation status. This data is then encapsulated into an inductor conduction monitoring data sequence using a unified time index. During operation, the method first identifies the inductor's current operating condition (load increase, load hold, switch frequency switching, thermal environment change), and then combines this with the normal calibration data generated during the normal calibration phase to generate a corresponding normal conduction framework template. Subsequently, the current monitoring data is mapped to the electromagnetic thermal conduction chain, and the conduction relationships from load status to inductor current, switch control to ripple pattern, inductor terminal voltage to inductor current, inductor current to ripple pattern, inductor current to inductor casing temperature, and ambient temperature to inductor casing temperature are checked to ensure they conform to the normal conduction direction, response sequence, hysteresis window, amplitude conversion range, ripple pattern range, and temperature recovery range.

[0031] In continuous operation records, when a batch of charging modules entered the load step test phase, some inductors exhibited increased current ripple and rising casing temperature. Traditional threshold judgment would directly provide an anomaly warning, while this invention first categorizes this change into a scattering event set, and then interprets and eliminates events in the following order: load state interpretation, switch control interpretation, power input / output state interpretation, ambient temperature interpretation, onboard temperature interpretation, and heat dissipation state interpretation. For scattering events that can form a complete conduction chain from load state to inductor current, from inductor current to ripple shape, and from inductor current to inductor casing temperature, the method identifies them as interpretable offsets caused by external operating conditions and does not consider them as abnormal outputs of the inductor itself. For another part of the records where the ripple shape continues to deviate and cannot be explained by load, switch control, and power input / output state, the method retains them as unexplained endogenous scattering results and continues to track conduction anomalies in the electromagnetic and thermal conduction chain to form an inductor endogenous anomaly diffusion path.

[0032] In subsequent on-site verification, maintenance personnel, based on the closed-loop monitoring results of the inductor status, examined the anomaly initiation conduction relationship and anomaly propagation path. They could directly see whether the anomaly started from a mismatch in the response from the inductor terminal voltage to the inductor current, an amplitude mismatch from the inductor current to the ripple pattern, or a timing mismatch from the response from the inductor current to the inductor casing temperature. The workshop-retained acquisition logs, waveform records, temperature rise records, maintenance records, and false alarm verification records show that this invention can eliminate normal response deviations caused by load changes, switch control switching, and thermal environment changes from anomaly judgments, while retaining endogenous scattering that cannot be explained by external operating conditions for path tracing. Therefore, on-site identification of early inductor degradation, near-core saturation, winding heating degradation, suspected local short circuits, and solder joint termination anomalies is more stable. Maintenance personnel no longer need to repeatedly check through a large number of original waveforms and temperature curves; the anomaly confirmation process is more intuitive; the normal conduction framework template is not contaminated by endogenous anomaly records; and the reliability of long-term online monitoring is improved.

[0033] Table 1 Comparative verification data of the intelligent inductor condition monitoring method in the aging test scenario of charging modules.

[0034] Table 1 shows that the method of the present invention exhibits more stable monitoring performance in terms of false alarm rate, missed alarm rate, accuracy of inductor anomaly category identification, accuracy of anomaly path location, average anomaly confirmation time, average warning lead time, and normal baseline contamination rate. The current ripple fixed threshold method mainly judges whether the ripple amplitude exceeds a preset limit. In aging test scenarios with frequent load transitions, switch control switching, and ambient temperature changes, it easily identifies normal response deviations caused by external operating conditions as inductor anomalies, resulting in a false alarm rate of 18.7%, a missed alarm rate of 12.4%, and an anomaly category identification accuracy of 71.6%. The current ripple and PWM incremental correlation coefficient method introduces the influence of switch control on ripple changes, reducing the false alarm rate to 13.4% and improving the anomaly category identification accuracy to 77.8%. However, this method mainly focuses on the correlation between ripple and PWM, and its ability to explain ambient temperature, onboard temperature, heat dissipation status, and the inductor casing temperature recovery process is insufficient. Therefore, it still has significant misjudgments in scenarios with thermally related anomalies and multiple superimposed disturbances.

[0035] The LSTM autoencoder reconstruction error method can identify abnormal fluctuations in the inductor operation process by utilizing timing reconstruction deviations, reducing the false alarm rate and false negative rate to 11.2% and 8.5%, respectively. Compared with the fixed threshold and ripple correlation methods, it has better timing adaptability. However, the anomaly basis of this method mainly comes from the reconstruction error, which makes it difficult to directly explain whether the anomaly comes from the conduction mismatch between the inductor terminal voltage, inductor current, ripple shape, or inductor shell temperature. Therefore, the anomaly path localization accuracy is 78.4%, which still cannot fully meet the field needs for anomaly source tracing. The GAT-LSTM multivariate graph anomaly detection method further utilizes multivariate relationships and temporal changes for joint modeling, achieving an anomaly category identification accuracy of 84.5% and an anomaly path localization accuracy of 81.6%. However, its graph relationships rely more on data learning and do not incorporate the fixed conduction relationships from load state to inductor current, switch control to ripple shape, power input / output state to inductor terminal voltage, inductor terminal voltage to inductor current, inductor current to ripple shape, inductor current to inductor housing temperature, and thermal environment to inductor housing temperature into the monitoring logic. Therefore, it still has shortcomings in interpreting the boundary between external operating condition disturbances and inductor-borne anomalies.

[0036] The method of this invention reduces the false alarm rate to 6.8%, the false negative rate to 5.9%, the accuracy rate of inductor anomaly category identification to 88.7%, and the accuracy rate of anomaly path location to 86.9%. This effect mainly stems from three aspects: First, the inductor conduction monitoring data sequence places electrical response sampling records, control drive sampling records, and thermal environment sampling records under a unified time index, providing a consistent data time reference for inductor state judgment. Second, the normal conduction skeleton template defines fixed conduction relationships and corresponding conduction attributes, allowing for item-by-item verification of conduction direction, response timing, hysteresis window, amplitude conversion range, ripple morphology range, and temperature rise recovery range, thereby expanding anomaly identification from single amplitude judgment to electromagnetic thermal conduction process judgment. Third, after the scattering event set is reduced through load state interpretation, switch control interpretation, power input / output state interpretation, and thermal environment interpretation, only unexplained endogenous scattering results are retained, which can eliminate explainable offsets caused by external operating conditions from anomaly judgment, reducing false alarms and improving the ability to locate anomaly sources.

[0037] In terms of average anomaly confirmation time and average warning lead time, the method of this invention shortens the average anomaly confirmation time to 13.8 minutes and increases the average warning lead time to 33.8 minutes. This is because the method does not simply check the original waveform or temperature rise curve after an anomaly occurs, but directly establishes an inductor-borne anomaly diffusion path, allowing maintenance personnel to check the anomaly's starting location, conduction direction, and mismatch type along a fixed conduction relationship. Compared to the GAT-LSTM multivariate graph anomaly detection method, the average anomaly confirmation time is reduced by 5.8 minutes, and the average warning lead time is increased by 6.4 minutes, indicating that a fixed electromagnetic thermal conduction chain and unexplained endogenous scattering results can improve anomaly interpretation efficiency. Regarding the normal baseline contamination rate, this method is 3.9%, significantly lower than other comparative methods. This is because the inductor state closed-loop monitoring results include a normal conduction skeleton template update control. When the inductor state category is near core saturation, winding heating degradation, suspected local short circuit, or solder joint termination anomaly, the corresponding inductor conduction monitoring frame will not enter the normal conduction skeleton template update, thereby reducing the risk of anomaly samples being absorbed into the normal baseline.

[0038] In summary, the method of this invention can transform inductor operating data from a single threshold judgment to a consistency check under a fixed conduction relationship in complex operating scenarios such as charging module aging tests. It can locate the source of anomalies by identifying scattering events, interpreting and reducing external operating conditions, and tracing the diffusion path of inductor endogenous anomalies. This can reduce false alarms caused by load changes, switching control changes, power supply fluctuations, and thermal environment changes, and improve the identification stability of near-core saturation, winding heating degradation, suspected local short circuits, and abnormal solder joint terminations. At the same time, it reduces the contamination of the normal baseline by normal conduction skeleton template update control, making the long-term online monitoring process more reliable, interpretable, and valuable for maintenance guidance.

[0039] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for intelligent monitoring of inductor status based on the Internet of Things, characterized in that, Includes the following steps: Inductor operation data is collected through an IoT acquisition terminal and packaged into an inductor conduction monitoring data sequence according to a unified time index. Based on the inductance conduction monitoring data sequence, the inductor's operating condition is identified, an equivalent operating condition unit is generated, and a normal conduction skeleton template corresponding to the equivalent operating condition unit is generated by combining the inductor's normal calibration data. Based on the normal conduction skeleton template, the inductive conduction monitoring data sequence is mapped to the electromagnetic thermal conduction chain, and conduction consistency verification is performed to generate conduction consistency results; Based on the conduction matching results, conduction relationship mismatch, response timing mismatch, response amplitude mismatch, and response shape mismatch are identified, and a set of scattering events is generated. The scattering event set is subjected to interpretation reduction according to the external operating condition interpretation order, generating uninterpreted endogenous scattering results; Based on unexplained endogenous scattering results, conduction anomalies are traced in the electromagnetic heat conduction chain to generate an inductor endogenous anomaly diffusion path, and inductor state closed-loop monitoring results are generated based on the inductor endogenous anomaly diffusion path.

2. The method for intelligent monitoring of inductor status based on the Internet of Things according to claim 1, characterized in that, The generation of the inductance conduction monitoring data sequence specifically includes: A unified time index is established using the sampling clock of the IoT acquisition terminal. The electrical response sampling record, control drive sampling record, and thermal environment sampling record corresponding to the monitored inductor are collected to form the original inductor operation data. The electrical response sampling record consists of inductor terminal voltage, inductor current, ripple morphology, and inductor housing temperature; the control drive sampling record consists of load status, switch control, and power input / output status; and the thermal environment sampling record consists of ambient temperature, onboard temperature, and heat dissipation status. Perform sampling integrity checks, timestamp rearrangement, and invalid sample removal on the original inductor operating data to form valid inductor operating data; Time alignment is performed on the valid inductor operating data according to a unified time index to form aligned inductor operating data; The control drive sampling records are read from the aligned inductor operation data and combined with the inductor terminal voltage in the electrical response sampling records to form a drive relationship data group; The inductor current, ripple morphology, and inductor case temperature are read from the electrical response sampling records of the aligned inductor operating data to form a response relationship data set. Thermal environment sampling records are read from the aligned inductor operating data to form a thermal interpretation relationship data set; The drive relationship data group, response relationship data group, and thermal interpretation relationship data group under the same unified time index are encapsulated and arranged in the order of the unified time index to generate an inductance conduction monitoring data sequence.

3. The method for intelligent monitoring of inductor status based on the Internet of Things according to claim 1, characterized in that, The generation of the normal conduction skeleton template specifically includes: The driving relationship data set, response relationship data set, and thermal interpretation relationship data set are extracted frame by frame from the inductance conduction monitoring data sequence to form the operating condition identification data set; Based on the inductance conduction monitoring data sequence, the load status, switch control status, power input and output status and thermal environment status are identified, and the current operating condition description result is generated. Based on the current operating condition description results, the inductance conduction monitoring data sequence is merged to generate an equivalent operating condition unit; Select normal calibration segments corresponding to the equivalent operating condition unit from the normal calibration data of the inductor to form the equivalent operating condition calibration set; Based on the equivalent operating condition calibration set, the normal conduction relationship of inductance is extracted to form a set of normal conduction skeleton elements; The collection of normal conduction skeleton elements is encapsulated into a normal conduction skeleton template.

4. The method for intelligent monitoring of inductor status based on the Internet of Things according to claim 3, characterized in that, The set of normal conduction skeleton elements consists of fixed conduction relationships and corresponding conduction attributes; The fixed conduction relationships include the normal driving relationship from load state to inductor current, the normal driving relationship from switch control to ripple mode, the normal driving relationship from power input / output state to inductor terminal voltage, the normal response relationship from inductor terminal voltage to inductor current, the normal response relationship from inductor current to ripple mode, the normal hysteresis relationship from inductor current to inductor housing temperature, the normal interpretation relationship from ambient temperature to inductor housing temperature, and the normal interpretation relationship from onboard temperature and heat dissipation state to inductor housing temperature. The corresponding conduction attributes consist of the calibration conduction direction, calibration response sequence, calibration hysteresis window, calibration amplitude conversion range, calibration silence mark, calibration ripple morphology range, calibration temperature rise recovery range, and calibration interpretation priority of each fixed conduction relationship in the equivalent operating condition calibration set.

5. The method for intelligent monitoring of inductor status based on the Internet of Things according to claim 1, characterized in that, The generation of the conduction anastomosis result specifically includes: Based on the fixed conduction relationship in the normal conduction skeleton template, the inductive conduction monitoring data sequence is mapped to the electromagnetic thermal conduction chain; The changes of each fixed conduction relationship along the electromagnetic heat conduction chain in the inductive conduction monitoring data sequence are checked, and the conduction direction check results and response timing check results are generated. According to the calibration lag window and calibration amplitude conversion range in the normal conduction skeleton template, perform lag verification and amplitude verification on the electromagnetic heat conduction chain to generate lag matching verification results and amplitude matching verification results; Perform ripple shape verification on fixed conduction relationships involving ripple shape, and perform temperature rise recovery verification on fixed conduction relationships involving inductor housing temperature, generating shape matching verification results; Perform a silent check on the fixed conduction relationship of the configured silent marker, and perform an interpretation priority check on the conduction chain from ambient temperature to inductor case temperature, onboard temperature, and heat dissipation status to inductor case temperature, generating silent consistency check results and interpretation priority check results; The transmission direction verification results, response timing verification results, lag matching verification results, amplitude matching verification results, morphological matching verification results, silent matching verification results, and interpretation priority verification results are summarized to generate transmission matching results.

6. The method for intelligent monitoring of inductor status based on the Internet of Things according to claim 1, characterized in that, The generation of the scattering event set specifically includes: Based on the conduction direction verification results and silent matching verification results in the conduction matching results, conduction relationship mismatch is identified, and conduction relationship mismatch events are generated. Based on the response timing verification results and lag matching results in the transmission matching results, response timing mismatch is identified, and response timing mismatch events are generated. Based on the amplitude matching verification results in the conduction matching results, response amplitude mismatch is identified, and a response amplitude mismatch event is generated; Based on the morphological match verification results in the conduction match results, response morphological mismatch is identified, and response morphological mismatch events are generated. The events of conduction mismatch, response timing mismatch, response amplitude mismatch, and response morphology mismatch are aggregated according to a unified time index to generate a scattering event set.

7. The method for intelligent monitoring of inductor status based on the Internet of Things according to claim 1, characterized in that, The generation of the unexplained endogenous scattering results specifically includes: Based on the fixed conduction relationship, mismatch type, and interpretation priority verification results corresponding to each scattering event in the scattering event set, the external operating condition interpretation order is determined, and the interpretation reduction order result is generated. Perform load state interpretation reduction on the scattering event set according to the interpretation reduction order, and generate load state interpretation reduction results; According to the order of interpretation and deletion, the load state interpretation and deletion results are processed to perform switch control interpretation and deletion, and switch control interpretation and deletion results are generated. According to the order of interpretation and deletion, the power input and output status interpretation and deletion results are performed on the switch control interpretation and deletion results to generate the power input and output status interpretation and deletion results; According to the interpretation and deletion order, the thermal environment interpretation and deletion results are performed on the power input and output state interpretation and deletion results to generate thermal environment interpretation and deletion results; Scattering events retained in the thermal environment interpretation and deletion results are registered as uninterpreted endogenous scattering results, and the fixed conduction relationship, mismatch type and corresponding inductive conduction monitoring frame of the corresponding scattering event are retained.

8. The method for intelligent monitoring of inductor status based on the Internet of Things according to claim 1, characterized in that, The generation of the inductor state closed-loop monitoring results specifically includes: According to the unified time index, the anomalous initiation scattering event is identified from the unexplained endogenous scattering results, and the fixed conduction relationship corresponding to the anomalous initiation scattering event is identified as the anomalous initiation conduction relationship. Starting with the abnormal initial conduction relationship, the adjacent fixed conduction relationships are matched along the conduction endpoint in the electromagnetic heat conduction chain to generate candidate endogenous abnormal paths; Perform continuity and direction checks on candidate endogenous anomaly paths to generate inductor endogenous anomaly propagation paths; Based on the fixed conduction relationship and mismatch type in the diffusion path of inductor endogenous anomalies, the inductor state category is determined, including core saturation approaching state, winding heating and degradation state, suspected local short circuit state, and abnormal solder joint termination state. Early warning trigger results are generated based on the inductor state category and the propagation path of endogenous anomalies in the inductor. Based on the inductor status category, early warning triggering result, and the diffusion path of inductor endogenous anomalies, a closed-loop monitoring result of the inductor status is generated.

9. The method for intelligent monitoring of inductor status based on the Internet of Things according to claim 8, characterized in that, The closed-loop monitoring results of the inductor status include the inductor status category, the diffusion path of the inductor's intrinsic anomaly, the warning trigger result, and the normal conduction skeleton template update control content. When the inductor status category is the core saturation approaching state, the winding heating degradation state, the suspected local short circuit state, or the solder joint termination abnormal state, the normal conduction skeleton template update control content prohibits the corresponding inductor conduction monitoring frame from entering the normal conduction skeleton template update. When the unexplained intrinsic scattering result is empty and the scattering events in the scattering event set have been explained and deleted, the normal conduction skeleton template update control content allows the corresponding inductor conduction monitoring frame to enter the normal conduction skeleton template update.

Images