A shipboard localization IGBT module drive monitoring system

Abstract

The application discloses a kind of shipborne localization IGBT module drive monitoring systems, specifically related to power electronic drive monitoring and smart grid control field, including: drive monitoring acquisition module is used to access the IGBT module drive board in shipborne smart grid, gate loop, power loop and auxiliary power supply loop, gathers drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal and auxiliary power supply state, and output corresponding moment monitoring record;Diagnosis disturbance injection module is used to read the drive switching time in monitoring record, and under the condition that current power conversion task is not changed, diagnostic disturbance action is injected to gate loop. By injecting diagnostic disturbance action in the process of drive switching and generating drive response fingerprint, external working condition disturbance in on-load operation and drive chain intrinsic anomaly are separated and judged, so that the problem of coexistence of false protection shutdown and missed detection is relatively improved.

Description

Technical Field

[0001] This invention relates to the field of power electronic drive monitoring and smart grid control technology, and more specifically, to a shipborne domestically produced IGBT module drive monitoring system. Background Technology

[0002] In the drive monitoring of power electronic conversion equipment, the existing processing focus is mostly on determining whether the IGBT module is abnormal during turn-on, turn-off and protection processes. The common method is to collect gate voltage, gate current, desaturation detection signal, collector voltage, device temperature or auxiliary power supply status on the drive board side, and combine it with threshold comparison, waveform feature extraction or fault code reporting to identify overcurrent, short circuit, undervoltage, overtemperature and drive mismatch. While this type of method can meet basic monitoring requirements in general industrial scenarios, in shipborne smart grid environments, power conversion units often operate under conditions of isolated grid power supply, bus fluctuations, pulse load switching, sudden changes in propulsion load, and strong electromagnetic interference. They also need to meet the requirements of continuous operation without shutdown, power interruption, or disruption of power supply and distribution. At this time, bus voltage drop, common-mode rise, parasitic oscillations, and sudden current changes will be transmitted to the drive chain and power chain simultaneously, causing phenomena such as gate waveform distortion, Miller plateau dwell time changes, forward shift of desaturation trigger boundary, and transient drop of auxiliary power supply to occur simultaneously. These phenomena may originate from external operating condition disturbances during the operation of the shipborne smart grid, or from IGBT module drive chain degradation or device state drift. This makes it difficult to identify the source of anomalies in existing processing methods based on single-point measurements, fixed criteria, or static waveform classification during load operation. On-site, it can be directly observed that the same drive parameter gives opposite judgments at different power supply task stages, alarm boundaries drift significantly after replacing modules in the same batch, and there is a coexistence of false protection shutdown and missed detection risks. The technical problem this application aims to solve is: how to distinguish between the apparent changes caused by external operating condition disturbances and the intrinsic anomalies of the IGBT module drive chain under the condition of continuous power supply and distribution of shipborne smart grids without interrupting power conversion tasks. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a shipborne domestically produced IGBT module drive monitoring system. By injecting diagnostic disturbance actions during drive switching and extracting the corresponding drive response fingerprint, and then combining the difference closure comparison with the fingerprints of the same module, the same bridge arm, and the same operating condition, a corresponding drive adjustment command is generated to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a shipborne domestically produced IGBT module drive monitoring system, comprising: The drive monitoring and acquisition module is used to connect to the IGBT module driver board, gate circuit, power circuit and auxiliary power supply circuit in the shipborne smart grid, acquire drive control signals, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal and auxiliary power supply status, and output the monitoring record at the corresponding time. The diagnostic disturbance injection module is used to read the drive switching time in the monitoring record, inject diagnostic disturbance actions into the gate circuit without changing the current power conversion task, and output the response record corresponding to the diagnostic disturbance action. The response fingerprint generation module is used to read response records, extract changes in gate load establishment, Miller plateau dwell, collector voltage drop, emitter return current, and auxiliary power supply transient drop after the diagnostic disturbance action, and generate a drive response fingerprint corresponding to the current operating condition. The difference closure determination module is used to read the drive response fingerprint, monitoring records and pre-stored historical baselines, perform difference closure comparison between the drive response fingerprint and the historical response fingerprint of the same IGBT module, the response fingerprint of the same bridge arm and the response fingerprint of the same type of operating conditions, and output the operating condition disturbance mark or drive chain abnormality mark. The monitoring output control module is used to read the operating condition disturbance flag or drive chain abnormality flag, generate monitoring result records and drive adjustment commands, and output the drive adjustment commands to the IGBT module driver board to complete the online drive monitoring and control during the operation of the shipborne smart grid.

[0005] In a preferred embodiment, the execution of the drive monitoring and acquisition module includes: The drive control signals, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal and auxiliary power status in the driver board, gate circuit, power circuit and auxiliary power supply circuit are respectively connected to the corresponding acquisition channels. Sampling is performed on each acquisition channel at the same acquisition time, and the original acquisition data group corresponding to the same acquisition time is output according to the signal access order. After obtaining the original data set, the monitoring record corresponding to the original data set is generated according to the acquisition time, bridge arm identifier, IGBT module identifier, signal type identifier and signal value writing order, and the signal values ​​under the same acquisition time are written into the same monitoring record. The continuously generated monitoring records are arranged in ascending order of acquisition time. The difference between the acquisition time of the previous monitoring record and the acquisition time of the next monitoring record is used as the recording interval. Under the conditions that the recording interval is equal to the acquisition period, the bridge arm identifier is the same, and the IGBT module identifier is the same, the corresponding monitoring records are connected in sequence to output the timing monitoring sequence of the corresponding IGBT module.

[0006] In a preferred embodiment, the execution of the diagnostic disturbance injection module includes: Read the drive switching time from the monitoring record, and extract the drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal and auxiliary power supply status that are continuously arranged between the previous acquisition time and the next acquisition time of the drive switching time. Take the acquisition time when the drive control signal flips as the switching start time, and take the first acquisition time when the gate voltage reaches the level corresponding to the drive control signal after the flip as the switching end time. Take each acquisition time between the switching start time and the switching end time as the candidate injection time, and output the candidate injection time sequence.

[0007] In a preferred embodiment, the execution of the diagnostic disturbance injection module further includes: Based on the candidate injection time sequence, a diagnostic disturbance action is generated for each candidate injection time according to the acquisition time order. Each diagnostic disturbance action includes the injection start point corresponding to the candidate injection time, the injection duration corresponding to the recording interval between the candidate injection time and the next acquisition time, the injection polarity with the same flip direction as the current drive control signal, and the injection path corresponding to the gate circuit. Each diagnostic disturbance action is then superimposed on the gate current value between the switching start point and the switching end point to obtain the disturbed gate current sequence. The disturbed gate current sequence is then integrated according to the acquisition time to obtain the disturbed gate voltage sequence, and the disturbed gate voltage sequence corresponding to each diagnostic disturbance action is output.

[0008] In a preferred embodiment, the execution of the diagnostic disturbance injection module further includes: After obtaining each perturbation-induced gate voltage sequence, for each diagnostic perturbation action, the voltage difference between the perturbation-induced gate voltage sequence and the original gate voltage sequence at the switching endpoint, the number of times the voltage change direction between two adjacent acquisition times of the perturbation-induced gate voltage sequence differs from that between two adjacent acquisition times of the original gate voltage sequence, and the number of times the desaturation detection signal value within the diagnostic perturbation action's effective range differs from the desaturation detection signal value before injection are calculated. The voltage difference, the number of acquisitions with different directions, and the number of acquisitions with different desaturation detection signal values ​​are sequentially combined to form the deviation results. The target diagnostic perturbation action is determined according to the ascending order of voltage difference, the ascending order of the number of acquisitions with different directions, and the ascending order of the number of acquisitions with different desaturation detection signal values. According to the injection start point, injection duration, injection polarity, and injection path of the target diagnostic disturbance action, the injection is performed into the gate circuit. Simultaneously, the drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal, and auxiliary power supply status are collected from the injection start point to the switching end point. The target diagnostic disturbance action and the corresponding collection results are written into the same response record in the order of collection time, and the response record corresponding to the target diagnostic disturbance action is output.

[0009] In a preferred embodiment, the execution of the response fingerprint generation module includes: The system reads the action identifier, injection start point, injection duration, acquisition time, drive control signal, gate voltage, gate current, collector voltage, emitter potential, and auxiliary power supply status from the response record. These are arranged in ascending order of acquisition time to form a response sequence. The acquisition time when the drive control signal flips is taken as the switching start point, the injection start point is taken as the disturbance start point, the acquisition time corresponding to the sum of the injection start point and the injection duration is taken as the disturbance end point, and the first acquisition time when the gate voltage reaches the level corresponding to the drive control signal after the flip is taken as the switching end point. The system then outputs the switching interval sequence. Based on the switching interval sequence, the gate current at two adjacent acquisition times between the disturbance start point and the switching end point is multiplied and accumulated to obtain the gate charge establishment value. The gate voltage difference at two adjacent acquisition times between the disturbance start point and the switching end point is calculated item by item. When the gate voltage difference of two consecutive recording intervals has the same sign and the collector voltage difference of two consecutive recording intervals has opposite signs, the starting acquisition time of the first recording interval is determined as the platform start point. The acquisition time when the collector voltage difference first becomes zero in the subsequent recording interval is determined as the platform end point. The Miller platform dwell value is obtained by the difference between the platform end point and the platform start point. The collector voltage drop value is obtained by the difference between the collector voltage at the disturbance start point and the value at the switching end point. The emitter return current value is obtained by accumulating the emitter potential in the same direction difference of each recording interval between the disturbance start point and the switching end point. The auxiliary power supply transient drop value is obtained by the difference between the auxiliary power supply state at the disturbance start point and the lowest value between the disturbance start point and the switching end point. The characteristic group is then output.

[0010] In a preferred embodiment, the execution of the response fingerprint generation module further includes: After obtaining the feature set, for each response record, calculate whether the platform start point is located after the disturbance start point and before the switching end point, whether the platform end point is located after the platform start point, whether the difference sign of the collector voltage between the disturbance start point and the switching end point is consistent, and whether the minimum value of the auxiliary power supply status between the disturbance start point and the switching end point is not higher than the value of the disturbance start point. For response records that do not meet the aforementioned positional and sign relationships, delete the corresponding acquisition time and re-execute the feature calculation until the remaining response records all meet the aforementioned positional and sign relationships, and output the verified feature set. The gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power transient drop value in the verified feature group are written into the same drive response fingerprint record in the order of action identifier, bridge arm identifier, IGBT module identifier, switching start point, gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power transient drop value to generate a drive response fingerprint corresponding to the current operating condition.

[0011] In a preferred embodiment, the execution of the difference closure determination module includes: Read the current response fingerprint and the pre-stored fingerprints of the same module, the same bridge arm, and the same operating condition, calculate the corresponding differences of the gate load value, plateau value, fallback value, return current value, and drop value, and output the module difference group, bridge arm difference group, and operating condition difference group. The absolute values ​​of the corresponding differences between the module difference group and the bridge arm difference group are taken and subtracted one by one to obtain the first closed group. The absolute values ​​of the corresponding differences between the module difference group and the working condition difference group are taken and subtracted one by one to obtain the second closed group. The same-named differences between the first closed group and the second closed group are recorded as abnormal items if they have the same sign, as disturbance items if they have opposite signs, and as common items if at least one of them is zero. The number of abnormal items, the number of disturbance items, and the number of common items are output.

[0012] In a preferred embodiment, the execution of the difference closure determination module further includes: The determination is made based on the number of abnormal items, the number of disturbance items, and the number of common items. When the number of abnormal items is greater than the number of disturbance items, a drive chain abnormal flag is output. When the number of disturbance items is greater than the number of abnormal items, a working condition disturbance flag is output. When the number of abnormal items is equal to the number of disturbance items, the sum of the absolute values ​​of the fallback value, the plateau value, and the drop value in the first closed group and the second closed group are compared, and the flag corresponding to the side with the larger sum of absolute values ​​is used as the output flag.

[0013] In a preferred embodiment, the execution of the monitoring output control module includes: Read the operating condition disturbance flag or drive chain abnormal flag, and write the bridge arm identifier, module identifier, switching start point, output flag and corresponding response fingerprint into the same monitoring result record in the order of the fields to output the monitoring result record; The drive adjustment command is generated based on the output flag. When the output flag is the operating condition disturbance flag, the current drive board output remains unchanged. When the output flag is the drive chain abnormal flag, the current drive board output is switched to the shutdown output, and the drive adjustment command is output to the IGBT module drive board.

[0014] The technical effects and advantages of this invention are as follows: By injecting diagnostic disturbance actions and generating drive response fingerprints during drive switching, external operating condition disturbances during load operation are separated from intrinsic anomalies of the drive chain, thus relatively improving the problem of coexistence of false protection shutdowns and missed detections. By generating multiple diagnostic perturbation actions around the candidate injection moment and selecting the target diagnostic perturbation action according to the deviation result, the diagnostic process is based on the actual response of the current switching process, thus relatively reducing the judgment bias caused by fixed criterion drift. By extracting gate charge establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value and auxiliary power supply transient drop value to form a drive response fingerprint, the state characterization of the drive chain is expanded from a single point measurement to a combination of multiple features, thus relatively improving the completeness of anomaly identification. By performing a difference-closed comparison between the current response fingerprint and fingerprints of the same module, the same bridge arm, and the same operating condition, and outputting corresponding drive adjustment commands, the monitoring results are directly applied to the drive board control, thus improving the targeting of online processing. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the system modules of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, 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.

[0017] Refer to the instruction manual appendix Figure 1 The present invention provides a shipborne domestically produced IGBT module drive monitoring system, comprising: The drive monitoring and acquisition module is used to connect to the IGBT module driver board, gate circuit, power circuit and auxiliary power supply circuit in the shipborne smart grid, acquire drive control signals, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal and auxiliary power supply status, and output the monitoring record at the corresponding time. During the operation of the shipborne smart grid, the drive monitoring and acquisition module is used to organize multiple input signals from the drive board, gate circuit, power circuit, and auxiliary power supply circuit into continuously readable monitoring records, and further form the timing monitoring sequence of the corresponding IGBT module, which can be directly called for subsequent diagnostic disturbance injection, response fingerprint generation, and difference closure determination. To ensure that subsequent calculations can be carried out continuously, the acquisition process adopts the same time base, the bridge arm identifier and IGBT module identifier adopt a fixed generation method, the monitoring records adopt a fixed field sequence, and the continuous monitoring records are organized in a fixed connection method. The implementation process includes the following steps: First, the drive control signal in the driver board, the gate voltage and gate current in the gate circuit, the collector voltage and emitter potential in the power circuit, the desaturation detection signal, and the auxiliary power supply status in the auxiliary power supply circuit are respectively connected to the corresponding acquisition channels. All acquisition channels share the same sampling clock. When the acquisition hardware is for simultaneous sampling, the time corresponding to the trigger of the same round of sampling is taken as the acquisition time. When the acquisition hardware is for sequential sampling, the start time of the sampling period after all acquisition channels have been polled is taken as the acquisition time, and the signal values ​​obtained within the sampling period are uniformly included in the acquisition time. After obtaining the signal values, they are arranged in the order of drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal, and auxiliary power supply status to form the original acquisition data group corresponding to the acquisition time. If a certain acquisition channel does not obtain a signal value at the acquisition time, a null value mark is written in the corresponding position of the original acquisition data group, and the order of the remaining fields remains unchanged. After the initial data set is generated, the acquisition time, bridge arm identifier, IGBT module identifier, signal type identifier, and signal value are written into the same monitoring record. The bridge arm identifier is generated according to the bridge arm position encoding in the power conversion topology, the IGBT module identifier is generated according to the combination of the bridge arm identifier and the connection terminal of the driver board, and the signal type identifier corresponds one-to-one with each signal value and is written sequentially according to the order in the initial data set. All signal values ​​acquired at the same acquisition time are written into the same monitoring record, so that the monitoring record contains both object information and corresponding acquisition results. If the initial data set contains null value markers, the null value markers and the corresponding signal type identifiers are written into the monitoring record together, thereby maintaining consistency in the number and position of fields in the monitoring record. After continuously generating monitoring records, all monitoring records are first sorted in ascending order by acquisition time. Then, the difference between the acquisition time of the previous monitoring record and the next monitoring record is calculated as the recording interval, and the recording interval is compared with the acquisition period. When the recording interval is equal to the acquisition period, the bridge arm identifier is the same, and the IGBT module identifier is the same, the next monitoring record is connected to the end of the recording chain to which the previous monitoring record belongs. When the recording interval is not equal to the acquisition period, or the bridge arm identifier is different, or the IGBT module identifier is different, the next monitoring record is used as the starting record of a new recording chain. After all monitoring records are connected, the timing monitoring sequence corresponding to each IGBT module is output. If only a single monitoring record exists after sorting, that single monitoring record is directly output as the timing monitoring sequence of the corresponding IGBT module. Through the above processing, the drive monitoring and acquisition module unifies multiple input signals into the same time caliber and recording structure, forming continuous monitoring results for the corresponding IGBT module. Subsequently, when reading the drive switching process, continuous records can be directly extracted according to the acquisition time and object identifier. In practical applications: before a bridge arm IGBT module in a shipborne smart grid enters the conduction switching phase, the drive monitoring and acquisition module obtains the drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal, and auxiliary power supply status according to a unified sampling clock. The sampling results for this round are written into the same monitoring record. Then, monitoring records with the same bridge arm identifier and IGBT module identifier and continuous acquisition time are sequentially connected to form the timing monitoring sequence of the IGBT module, which can be directly read for subsequent processing.

[0018] The diagnostic disturbance injection module is used to read the drive switching time in the monitoring record, inject diagnostic disturbance actions into the gate circuit without changing the current power conversion task, and output the response record corresponding to the diagnostic disturbance action. The diagnostic disturbance injection module is used to locate the drive switching process from the monitoring records without changing the current power conversion task, construct executable diagnostic disturbance actions, and select the target diagnostic disturbance action from multiple candidate actions that has a controllable impact on the current switching process and is easy to extract response differences in the future. The injection information corresponding to the target diagnostic disturbance action and the acquisition results after injection are then written into the response record for the response fingerprint generation module to continue reading. To ensure continuous computation, the monitoring records, acquisition time, recording interval, switching start point, switching end point, injection start point, and injection duration all use the same time base. The drive switching time is determined by the acquisition time when the drive control signal flips; the injection polarity corresponds one-to-one with the flip direction of the drive control signal; the injection path is determined by the drive output terminal and return terminal in the gate circuit corresponding to the current IGBT module; and the additional current of the diagnostic disturbance action is determined by the known output capability of the diagnostic injection branch in the drive board and the injection duration. This implementation process includes the following steps: First, the drive switching process is located and a candidate injection time sequence is formed. This provides a unified time range and continuous input for subsequent diagnostic disturbance generation. The input quantities include the drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal, auxiliary power supply status, and acquisition time from the monitoring records. During processing, continuous monitoring records corresponding to the same bridge arm identifier and the same IGBT module identifier are read in ascending order of acquisition time. The acquisition time when the drive control signal flips is determined as the switching start point. Then, the gate voltage value is read line by line along the direction of the flipped drive control signal. When the gate voltage enters the stable output level range corresponding to the driving control signal after the flip, the first acquisition time to enter the stable output level range is determined as the switching endpoint. The stable output level range is taken from the calibration record of the driving board under the current bridge arm identifier and the current IGBT module identifier. After the switching start point and switching endpoint are determined, all acquisition times between the two are sequentially written into the candidate injection time sequence and output. If no acquisition time to enter the stable output level range is read after the switching start point, the driving switching process is recorded as an empty switching record, no candidate injection time sequence is generated, and the process waits for the next driving control signal flip. Subsequently, based on the candidate injection time sequence, each diagnostic disturbance action is constructed and the corresponding gate voltage sequence after disturbance is calculated. This serves to uniformly convert the candidate injection locations into comparable disturbance response results. The input quantities include the candidate injection time sequence, switching start point, switching end point, gate current value, acquisition time, recording interval, and the output capability record of the diagnostic injection branch. During processing, diagnostic disturbance actions are generated sequentially according to the acquisition time order in the candidate injection time sequence. Each diagnostic disturbance action is written with the injection start point, injection duration, injection polarity, and injection path. The injection start point is the corresponding candidate injection time, the injection duration is the recording interval between the candidate injection time and the next acquisition time, and the injection polarity is determined by the flip direction of the drive control signal as a same-direction injection polarity. The path takes the gate drive output terminal corresponding to the current IGBT module to the return terminal; then, based on the output capability record of the diagnostic injection branch and the injection duration, the additional current corresponding to the diagnostic disturbance action is calculated, and only within the effective interval between the injection start point and the next acquisition time after the injection start point, the additional current is superimposed on the original gate current value to obtain the gate current sequence after the disturbance. Then, the gate current values ​​after the disturbance of two adjacent acquisition times between the switching start point and the switching end point are multiplied by the recording interval and accumulated to obtain the gate voltage sequence after the disturbance corresponding to each acquisition time and output it; if the candidate injection time is the switching end point, no corresponding diagnostic disturbance action is generated; if there is no subsequent acquisition time after the candidate injection time, the candidate injection time is deleted and the next candidate injection time is processed. Subsequently, deviation calculations are performed on each diagnostic disturbance action to determine the target diagnostic disturbance action. Its purpose is to select the injection scheme that maintains the original switching direction and minimizes deviation from multiple candidate actions. The input quantities are each diagnostic disturbance action, the gate voltage sequence after each disturbance, the original gate voltage sequence, the desaturation detection signal value, the switching endpoint, and the acquisition time. During processing, three types of deviations are calculated for each diagnostic disturbance action: First, read the values ​​of the gate voltage sequence after the disturbance and the original gate voltage sequence at the switching endpoint, and calculate the absolute value of the difference between the two as the voltage difference. Secondly, the voltage change direction between two adjacent acquisition times of the gate voltage sequence after disturbance is calculated sequentially according to the acquisition time sequence, and compared with the voltage change direction at the corresponding position of the original gate voltage sequence item by item. The number of times the direction is inconsistent is accumulated as the number of acquisitions with different directions. Third, read the desaturation detection signal values ​​within the range of the diagnostic disturbance action, and compare them item by item with the desaturation detection signal values ​​at the previous acquisition time before the injection start point. The number of times the values ​​are different is accumulated as the number of acquisitions with different desaturation detection signal values. Based on this, the deviation results are formed by arranging the voltage difference, the number of acquisitions with different directions, and the number of acquisitions with different desaturation detection signal values ​​in ascending order. These results are then sorted by voltage difference in ascending order, number of acquisitions with different directions in ascending order, and number of acquisitions with different desaturation detection signal values ​​in ascending order. The diagnostic disturbance action corresponding to the first position in the sort is determined as the target diagnostic disturbance action. If the deviation results of two diagnostic disturbance actions are exactly the same, the diagnostic disturbance action with the earlier injection starting point is taken as the target diagnostic disturbance action. If a diagnostic disturbance action causes the drive control signal to flip and the direction to reverse after it is applied, the diagnostic disturbance action is directly deleted and is not included in the sorting. Finally, a real injection is performed on the target diagnostic disturbance action, and a response record is generated. Its purpose is to solidify the injection information required for subsequent response fingerprint generation and the actual acquisition results after injection into the same recording unit. The input quantities are the injection start point, injection duration, injection polarity, and injection path in the target diagnostic disturbance action, as well as the drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal, auxiliary power supply status, and acquisition time synchronously acquired from the injection start point to the switching end point. During processing, injection is first performed into the gate circuit according to the injection start point, injection duration, injection polarity, and injection path of the target diagnostic disturbance action, and then from the injection start point... Starting from the point, data is continuously acquired at the same sampling clock until the switching endpoint. The action identifier, bridge arm identifier, IGBT module identifier, injection start point, injection duration, injection polarity, injection path, and corresponding acquisition results of the target diagnostic disturbance action are written into the same response record in ascending order of acquisition time. The response record corresponding to the target diagnostic disturbance action is then output. The action identifier is generated by combining the IGBT module identifier and the injection start point to ensure that the response records of the same module at different injection times can be distinguished. If the desaturation detection signal undergoes a state reversal during the actual injection process, a protection trigger flag is written into the response record, while the acquired data is retained and the current injection ends. Through the above processing, the diagnostic disturbance injection module first determines the drive switching range from the continuous monitoring records, then expands the candidate injection moments into multiple calculable diagnostic disturbance actions, and then selects the target diagnostic disturbance action based on the deviation results. The collected results after the actual injection are then written into the response record, enabling the subsequent response fingerprint generation module to extract changes in gate load establishment, Miller platform dwell time, collector voltage drop, emitter return current, and auxiliary power supply transient drop based on the clearly defined injection location, injection duration, and injection polarity. In practical applications: when an IGBT module in a shipborne smart grid switches from the off state to the on state... In the current state, the diagnostic disturbance injection module first identifies the acquisition time of the drive control signal flip from the timing monitoring sequence of the IGBT module as the switching start point, and then reads the first acquisition time when the gate voltage enters the stable output level range as the switching end point, and writes the acquisition times between the two into the candidate injection time sequence in sequence; then, it generates a diagnostic disturbance action for each candidate injection time, calculates the corresponding gate voltage sequence after disturbance and the deviation result, selects the first ranked target diagnostic disturbance action and injects it into the gate circuit, and writes the target diagnostic disturbance action and the continuous acquisition results after injection into the same response record for subsequent reading.

[0019] The response fingerprint generation module is used to read response records, extract changes in gate load establishment, Miller plateau dwell, collector voltage drop, emitter return current, and auxiliary power supply transient drop after the diagnostic disturbance action, and generate a drive response fingerprint corresponding to the current operating condition. The response fingerprint generation module converts the injection information in the response record and the continuous acquisition results after injection into a driving response fingerprint that can be directly used for difference closure determination. This process first recovers the complete switching process corresponding to the target diagnostic disturbance action from the response record, determining the switching start point, disturbance start point, disturbance end point, and switching end point. Then, it calculates the gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power supply transient drop value around this interval. Subsequently, it performs positional and sign relationship verification on the calculation results, deleting and recalculating sampling positions that do not meet the verification conditions. Finally, it writes the verified feature values ​​into the driving response fingerprint record in a fixed field order. This implementation process includes the following steps: First, the response records are organized into a switching interval sequence. This provides a unified time boundary and continuous input for subsequent feature calculations. The input quantities include the action identifier, bridge arm identifier, IGBT module identifier, injection start point, injection duration, acquisition time, drive control signal, gate voltage, gate current, collector voltage, emitter potential, and auxiliary power supply status from the response records. During processing, the acquisition results are first arranged in ascending order of acquisition time to form a response sequence. Then, the acquisition time at which the drive control signal flips is determined as the switching start point, the injection start point is determined as the disturbance start point, and the sum of the injection start point and the injection duration is mapped to the first acquisition time no earlier than this sum as the disturbance endpoint. If the injection start point and... If the sum of the injection durations falls exactly at a certain acquisition time, then that acquisition time is directly taken as the disturbance endpoint. Subsequently, the gate voltage values ​​are read one by one along the direction of the flipped drive control signal. When the gate voltage enters the stable output level range corresponding to the flipped drive control signal, the first acquisition time to enter the stable output level range is determined as the switching endpoint. The stable output level range is taken from the calibration record of the drive board under the current bridge arm identifier and the current IGBT module identifier. After the above four time points are determined, a switching interval sequence containing the switching start point, disturbance start point, disturbance endpoint, and switching endpoint is output. If the switching endpoint is not found, the response record is recorded as an unclosed response record and is not included in subsequent feature calculations. Next, feature groups are calculated based on the switching interval sequence. The purpose is to compress the switching process after the disturbance into five comparable feature values. The input quantities are the switching interval sequence, gate current, gate voltage, collector voltage, emitter potential, auxiliary power supply status, acquisition time, and the recording interval corresponding to the difference between two adjacent acquisition times. During processing, the gate current of two adjacent acquisition times between the disturbance start point and the switching end point is multiplied by the corresponding recording interval and accumulated to obtain the gate charge establishment value. Then, the gate voltage difference between two adjacent acquisition times between the disturbance start point and the switching end point is calculated item by item, and the collector voltage difference at the corresponding position is calculated item by item. When the gate voltage difference of two consecutive recording intervals has the same sign and the collector voltage difference of two consecutive recording intervals has opposite signs, the starting acquisition time of the first recording interval is determined as the platform start point. After the platform start point is determined, the collector voltage difference is read backward. When the collector voltage difference first becomes zero, the acquisition time is determined as the platform end point. If no zero value appears subsequently, the platform ends. If the sign is reversed, the acquisition time before the sign reversal is determined as the platform endpoint. If no zero value or sign reversal occurs until the switching endpoint, the switching endpoint is determined as the platform endpoint. Then, the Miller platform dwell value is obtained by the difference between the platform endpoint and the platform start point, and the collector voltage drop value is obtained by the difference between the collector voltage at the disturbance start point and the switching endpoint. Then, the emitter potential change direction corresponding to the drive control signal reversal direction is taken as the same direction, and the emitter potential difference values ​​in the same direction within each recording interval between the disturbance start point and the switching endpoint are accumulated to obtain the emitter return current value. Then, the auxiliary power supply status value at the disturbance start point and the lowest value between the disturbance start point and the switching endpoint are read, and the difference between the two is determined as the auxiliary power supply transient drop value. After the above calculations are completed, a feature group containing the gate charge establishment value, Miller platform dwell value, collector voltage drop value, emitter return current value, and auxiliary power supply transient drop value is output. If the platform start point is not found, the response record is recorded as an empty platform record and is not entered into subsequent verification. Subsequently, the feature group is checked and recalculated to remove sampling positions that disrupt the switching relationship and retain feature results that satisfy the positional and sign relationships. The input quantities are the feature group, the switching interval sequence, the response sequence, and the gate voltage, collector voltage, and auxiliary power supply state at each sampling position. During processing, it is sequentially determined whether the platform start point is located after the disturbance start point and before the switching end point, whether the platform end point is located after the platform start point, whether the sign of the difference in collector voltage between each recording interval from the disturbance start point to the switching end point is consistent, and whether the lowest value of the auxiliary power supply state between the disturbance start point and the switching end point is consistent. The value should not exceed the disturbance start point. When a sampling moment that does not satisfy the above relationship occurs for the first time, delete the gate voltage, gate current, collector voltage, emitter potential, and auxiliary power supply status values ​​corresponding to that sampling moment, and re-execute the aforementioned feature calculation while keeping the order of the remaining sampling moments unchanged. If there are still sampling moments that do not satisfy the relationship after recalculation, continue to delete the first sampling moment that does not satisfy the relationship and recalculate again until all remaining response sequences satisfy the above relationship, and output the verified feature group. If there are fewer than two remaining sampling moments after deletion, the response record is recorded as an invalid response record and no drive response fingerprint is generated. Finally, the verified feature group is written into the drive response fingerprint record. The purpose is to form a standardized output that can be directly read by the difference closure judgment module. The input quantities are the action identifier, bridge arm identifier, IGBT module identifier, switching start point, and the gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power supply transient drop value in the verified feature group. During processing, the action identifier, bridge arm identifier, IGBT module identifier, switching start point, gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power supply transient drop value are written into the same drive response fingerprint record in the order of action identifier, bridge arm identifier, IGBT module identifier, switching start point, gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power supply transient drop value. This drive response fingerprint record is used as the drive response fingerprint output under the current operating condition. The current operating condition is characterized by the switching state corresponding to the bridge arm identifier, IGBT module identifier, and switching start point. If the action identifier corresponds to multiple drive response fingerprint records, the one with the later switching start point is retained, and the rest are deleted. Through the above processing, the response fingerprint generation module compresses the continuous sampling results corresponding to the target diagnostic disturbance action into five feature values. During the generation process, it uniformly processes the disturbance endpoint mapping, Miller platform start and end positions, emitter return current direction, and abnormal sampling position deletion and recalculation. This allows subsequent differential closure determination to directly calculate based on the gate charge establishment value, Miller platform dwell value, collector voltage drop value, emitter return current value, and auxiliary power supply transient drop value. In practical applications: when an IGBT module completes the target diagnostic disturbance action injection during conduction switching, the response fingerprint generation module first determines the feature value from the response record. Define the switching start point, disturbance start point, disturbance end point, and switching end point, then calculate the gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power supply transient drop value. If the plateau end point does not have a zero difference value, then take the acquisition time before the collector voltage difference reverses as the plateau end point. If a certain sampling position disrupts the collector voltage difference sign consistency relationship, then delete the sampling position and recalculate. After all relationships are satisfied, write the action identifier, bridge arm identifier, IGBT module identifier, switching start point, and five characteristic values ​​into the same drive response fingerprint record for direct reading in subsequent difference closure determination.

[0020] The difference closure determination module is used to read the drive response fingerprint, monitoring records and pre-stored historical baselines, perform difference closure comparison between the drive response fingerprint and the historical response fingerprint of the same IGBT module, the response fingerprint of the same bridge arm and the response fingerprint of the same type of operating conditions, and output the operating condition disturbance mark or drive chain abnormality mark. The difference closure determination module is used to perform a directional comparison between the current response fingerprint and the pre-stored historical baseline, and to distinguish whether the current deviation source is internal drive chain deviation or external operating condition disturbance. This process first retrieves fingerprints from the same module, the same bridge arm, and the same operating condition based on the current response fingerprint. Then, it calculates the module difference group, bridge arm difference group, and operating condition difference group according to a unified field direction. Subsequently, it performs closure calculation and symbol classification on the three groups of differences. Finally, it outputs drive chain anomaly markers or operating condition disturbance markers based on the number of anomalies, disturbances, and commonalities. To ensure that subsequent determination processes can be directly implemented, the current response fingerprint and the pre-stored historical baseline... All feature values ​​in the line use the same field caliber, and the difference direction is uniformly calculated by subtracting the corresponding baseline field value from the current response fingerprint field value. Fingerprints from pre-stored historical baselines are selected using a fixed retrieval order. Specifically, for fingerprints of the same module, the most recent historical fingerprint with the same module identifier and switching direction is selected; for fingerprints of the same bridge arm, the most recent historical fingerprint with the same bridge arm identifier and closest switching start point is selected; for fingerprints of the same operating condition, the most recent historical fingerprint with consistent switching direction, bridge arm identifier, and auxiliary power supply status is selected. When multiple candidate fingerprints exist under the same retrieval result, the one with the later switching start point is selected. This implementation process includes the following steps: First, the current response fingerprint and the pre-stored historical baseline are matched to form a difference input. This is to provide a one-to-one comparison object for subsequent closure calculations. The input includes the bridge arm identifier, IGBT module identifier, switching start point, gate load value, plateau value, fallback value, return current value, and drop value from the current response fingerprint, as well as the drive control signal, bridge arm identifier, IGBT module identifier, and auxiliary power supply status from the monitoring records, and all drive response fingerprint records from the pre-stored historical baseline. During processing, the fingerprint of the same module is first retrieved based on the IGBT module identifier and switching direction in the current response fingerprint; then, the fingerprint of the same bridge arm is retrieved based on the bridge arm identifier and switching start point; and finally, the fingerprint of the same operating condition is retrieved based on the switching direction, bridge arm identifier, and auxiliary power supply status. The switching direction and auxiliary power status are taken from the monitoring record of the switching starting point corresponding to the current response fingerprint. After the three types of fingerprints are determined, the differences in grid load value, platform value, fall value, return current value, and drop value are calculated item by item according to the direction of subtracting the corresponding baseline field value from the current response fingerprint field value. The corresponding differences between the current response fingerprint and the fingerprint of the same module are written into the module difference group, the corresponding differences between the current response fingerprint and the fingerprint of the same bridge arm are written into the bridge arm difference group, and the corresponding differences between the current response fingerprint and the fingerprint of the same working condition are written into the working condition difference group for subsequent closure calculation reading. If no corresponding record is found for any of the three types of fingerprints, the current response fingerprint is recorded as a record to be supplemented and the subsequent judgment is paused until the corresponding baseline is written before continuing execution. Subsequently, closed-loop calculations are performed on the module difference group, bridge arm difference group, and operating condition difference group to generate classification results. The purpose is to separate internal and external deviations through two rounds of difference closure. The input quantities are the differences in grid load value, platform value, fall value, return flow value, and drop value in the module difference group, bridge arm difference group, and operating condition difference group. During processing, the absolute values ​​of the corresponding differences in the module difference group and bridge arm difference group are taken separately and then subtracted item by item to obtain the first closed group. Then, the absolute values ​​of the corresponding differences in the module difference group and operating condition difference group are taken separately and then subtracted item by item to obtain the second closed group. After obtaining the first and second closed groups, the signs of the same-named differences are compared item by item according to the field name. When the same-named differences in the first and second closed groups are both positive or both are negative, the result is... This field is recorded as an anomaly. When the differences between the same names in the first and second closed groups are positive and negative respectively, this field is recorded as a disturbance. When at least one of the differences between the same names in the first and second closed groups is zero, this field is recorded as a commonality. Subsequently, the number of anomalies, disturbances, and commonality are accumulated and output. An anomaly indicates that the deviation direction of the current response fingerprint relative to the fingerprint of the same module is in the same direction as the deviation direction relative to the fingerprint of the same bridge arm and the fingerprint of the same working condition. A disturbance indicates that the deviation direction of the current response fingerprint relative to different control objects is inconsistent. A commonality indicates that the current response fingerprint does not form a corresponding deviation in at least one type of control. If a field in the first or second closed group does not have a corresponding difference, this field is directly recorded as a commonality and the remaining fields are processed. Subsequently, output flags are generated based on the number of anomalies, disturbances, and commonalities. The purpose is to convert the closed-loop calculation results into judgment results that can be directly used to control the output. The input quantities are the number of anomalies, disturbances, commonalities, the difference in fallback values, the difference in plateau values, and the difference in drop values ​​in the first and second closed-loop groups. During processing, when the number of anomalies is greater than the number of disturbances, a drive chain anomaly flag is output; when the number of disturbances is greater than the number of anomalies, a working condition disturbance flag is output. When the number of anomalies equals the number of disturbances, the sum of the absolute values ​​of the fallback value difference in the first and second closed-loop groups, the sum of the absolute values ​​of the plateau value difference in the first and second closed-loop groups, and the sum of the absolute values ​​of the drop value difference in the first and second closed-loop groups are calculated respectively. The sum of the absolute values ​​in the combined group and the second closed group is then added together to form a comparison between the sum of the absolute values ​​on the abnormal side and the sum of the absolute values ​​on the disturbance side. The side with the larger sum of absolute values ​​is designated as the output label. If the sums of absolute values ​​on both sides are the same, the label corresponding to the side with the larger absolute difference in fall value is taken as the output label. If the absolute difference in fall value is still the same, the label corresponding to the side with the larger absolute difference in platform value is taken as the output label. If the absolute difference in platform value is still the same, the label corresponding to the side with the larger absolute difference in drop value is taken as the output label. After the output label is determined, it is written into the judgment result field corresponding to the current response fingerprint for direct reading by the monitoring output control module. Through the above processing, the difference closure determination module compares the current response fingerprint with the fingerprints of the same module, the same bridge arm, and the same operating condition under a unified standard. It then uses two rounds of closure calculations to categorize deviation sources into abnormal items, disturbance items, and common items. Finally, it determines the output flag based on the number of items and the strength of the difference, enabling the subsequent monitoring output control module to directly execute drive adjustments based on the output flag. In practical applications: after the current response fingerprint of a certain IGBT module is written, the difference closure determination module first retrieves the corresponding fingerprint of the same module, the fingerprint of the same bridge arm, and the current auxiliary fingerprint from the pre-stored historical baseline. Fingerprints with the same power state and switching direction are used to form module difference groups, bridge arm difference groups, and operating condition difference groups. Then, closed-loop calculations are performed on the three groups of differences. If the fields corresponding to the gate load value, plateau value, and fallback value are classified as abnormal items, while the fields corresponding to the return current value and drop value are classified as common items, then the number of abnormal items is greater than the number of disturbance items, and the drive chain abnormality flag is directly output. If the number of abnormal items is the same as the number of disturbance items, then the sum of the absolute values ​​of the fallback value difference, plateau value difference, and drop value difference in the first closed group and the second closed group is compared. The flag corresponding to the side with the larger sum of absolute values ​​is written into the judgment result field for subsequent reading.

[0021] The monitoring output control module is used to read the operating condition disturbance flag or drive chain abnormality flag, generate monitoring result records and drive adjustment commands, and output the drive adjustment commands to the IGBT module driver board to complete the online drive monitoring and control during the operation of the shipborne smart grid. The monitoring output control module converts the output flags given by the difference closure judgment module into storable monitoring result records and executable drive adjustment commands, and sends the drive adjustment commands to the IGBT module driver board to complete the online drive monitoring and control during the operation of the shipborne smart grid. This process first reads the operating condition disturbance flags or drive chain anomaly flags, and writes the bridge arm identifier, module identifier, switching start point, output flag, and corresponding response fingerprint into the same monitoring result record. Then, based on the output flag, it determines whether the driver board maintains the current output or switches to a shutdown output, thus ensuring that the monitoring judgment result and the drive execution result are closed within the same processing chain. This implementation process includes the following steps: First, a monitoring result record is generated. The purpose of this process is to fix the current judgment result, the corresponding object, the corresponding time, and the corresponding response fingerprint into the same record. The input quantities are the operating condition disturbance flag or drive chain abnormal flag output by the difference closure judgment module, the bridge arm identifier and module identifier in the current response fingerprint, the switching start point corresponding to the current response fingerprint, and the current response fingerprint itself. During processing, the output flag is read first, and then written into the same monitoring result record in the order of the fields of bridge arm identifier, module identifier, switching start point, output flag, and corresponding response fingerprint. Among them, the corresponding response fingerprint is written with action identifier, grid load value, plateau value, fall value, return current value, and drop value. After the monitoring result record is written, it is output to the monitoring result storage area for subsequent operation traceability and control result verification. If the output flag is missing, the current response fingerprint is recorded as a record to be judged, no monitoring result record is generated, and the output flag is waited for the difference closure judgment module to rewrite it. Subsequently, a drive adjustment command is generated and executed, the purpose of which is to directly map the judgment result to the drive board control action; the input quantities are the output mark, bridge arm identifier, module identifier, switching start point, and the drive control signal state of the current drive board at the time the output mark is generated; during processing, when the output mark is a working condition disturbance mark, a hold command is generated, and the drive control signal state at the time the output mark is generated is written into the drive adjustment command as the hold target, and then the drive adjustment command is output to the corresponding IGBT module drive board, so that the IGBT module drive board keeps the current output unchanged until the next monitoring result record is generated; when the output mark is a drive chain abnormality mark, a shutdown command is generated, and the shutdown level control information is written into the drive adjustment command, and then the drive adjustment command is output to the corresponding IGBT module drive board, so that the IGBT module drive board switches from the current output to the shutdown output and holds it until the next drive switching time; if no execution acknowledgment is received from the corresponding IGBT module drive board after the drive adjustment command is output, an unexecuted mark is added to the monitoring result record, and the same drive adjustment command is output again; Through the above processing, the monitoring output control module connects the output flag, monitoring result record, and drive adjustment command into the same closed loop, so that the judgment result can not only be retained, but also directly act on the IGBT module driver board, thereby completing the online monitoring and control of the shipborne smart grid during operation. In practical applications: when an IGBT module under a certain bridge arm obtains a condition disturbance flag after differential closure judgment, the monitoring output control module first writes the bridge arm identifier, module identifier, switching start point, condition disturbance flag, and corresponding response fingerprint into the monitoring result record, and then keeps the drive control signal state of the IGBT module driver board unchanged at the time the output flag is generated; when the IGBT module obtains a drive chain abnormality flag after differential closure judgment, the monitoring output control module writes the corresponding information into the monitoring result record and outputs a shutdown command to the IGBT module driver board, causing the IGBT module driver board to switch to shutdown output.

[0022] Working Principle: This scheme first uses a drive monitoring and acquisition module to acquire signals from the drive board, gate circuit, power circuit, and auxiliary power supply circuit under the same time reference. Multiple signals at the same acquisition time are written into a monitoring record and then connected into a timing monitoring sequence according to bridge arm identifiers, IGBT module identifiers, and acquisition times. Subsequently, the diagnostic disturbance injection module locates the drive switching process from the timing monitoring sequence, selects candidate injection times between the switching start and end points, constructs diagnostic disturbance actions for each, calculates their corresponding post-disturbance responses, and selects the target diagnostic disturbance action that has a controlled impact on the original switching process for injection, thus forming a... The response record; the response fingerprint generation module then extracts the gate load establishment value, Miller platform dwell value, collector voltage drop value, emitter return current value, and auxiliary power supply transient drop value from the response record to generate the drive response fingerprint; the difference closure judgment module compares the current drive response fingerprint with the fingerprints of the same module, the same bridge arm, and the same operating condition to determine whether the current deviation is closer to an internal abnormality in the drive chain or an external operating condition disturbance. Finally, the monitoring output control module writes the judgment result into the monitoring result record and outputs a drive adjustment command to the IGBT module driver board to maintain the current output or switch to the off output, thereby completing online monitoring and control; Taking a bridge arm IGBT module in a shipborne smart grid as an example, when the propulsion load switches or bus fluctuations occur, the system first continuously collects the drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal, and auxiliary power supply status during the IGBT module's turn-on and turn-off processes, and organizes them into a continuous monitoring record. When the system detects that the IGBT module has undergone drive switching, it will select an appropriate injection moment during the switching process to apply a diagnostic disturbance action, and then extract the drive response fingerprint based on the actual response after injection. If the deviation between the current drive response fingerprint and the historical fingerprint of the same module is more concentrated in the module's own changes, the system will output a drive chain abnormality flag and control the IGBT module's drive board to switch to turn-off output. If the deviation between the current drive response fingerprint and the fingerprint of the same operating condition and the fingerprint of the same bridge arm is closer to the external operating fluctuations, the system will output an operating condition disturbance flag and keep the current drive board output unchanged. In this way, it can identify the real drive chain abnormality under load operation conditions and avoid misjudging external operating condition changes as device failures.

[0023] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A shipborne domestically produced IGBT module drive monitoring system, characterized in that, include: The drive monitoring and acquisition module is used to connect to the IGBT module driver board, gate circuit, power circuit and auxiliary power supply circuit in the shipborne smart grid, acquire drive control signals, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal and auxiliary power supply status, and output the monitoring record at the corresponding time. The diagnostic disturbance injection module is used to read the drive switching time in the monitoring record, inject diagnostic disturbance actions into the gate circuit without changing the current power conversion task, and output the response record corresponding to the diagnostic disturbance action. The response fingerprint generation module is used to read response records, extract changes in gate load establishment, Miller plateau dwell, collector voltage drop, emitter return current, and auxiliary power supply transient drop after the diagnostic disturbance action, and generate a drive response fingerprint corresponding to the current operating condition. The difference closure determination module is used to read the drive response fingerprint, monitoring records and pre-stored historical baselines, perform difference closure comparison between the drive response fingerprint and the historical response fingerprint of the same IGBT module, the response fingerprint of the same bridge arm and the response fingerprint of the same type of operating conditions, and output the operating condition disturbance mark or drive chain abnormality mark. The monitoring output control module is used to read the operating condition disturbance flag or drive chain abnormality flag, generate monitoring result records and drive adjustment commands, and output the drive adjustment commands to the IGBT module driver board to complete the online drive monitoring and control during the operation of the shipborne smart grid.

2. The shipborne domestically produced IGBT module drive monitoring system according to claim 1, characterized in that: The execution of the drive monitoring and acquisition module includes: The drive control signals, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal and auxiliary power status in the driver board, gate circuit, power circuit and auxiliary power supply circuit are respectively connected to the corresponding acquisition channels. Sampling is performed on each acquisition channel at the same acquisition time, and the original acquisition data group corresponding to the same acquisition time is output according to the signal access order. After obtaining the original data set, the monitoring record corresponding to the original data set is generated according to the acquisition time, bridge arm identifier, IGBT module identifier, signal type identifier and signal value writing order, and the signal values ​​under the same acquisition time are written into the same monitoring record. The continuously generated monitoring records are arranged in ascending order of acquisition time. The difference between the acquisition time of the previous monitoring record and the acquisition time of the next monitoring record is used as the recording interval. Under the conditions that the recording interval is equal to the acquisition period, the bridge arm identifier is the same, and the IGBT module identifier is the same, the corresponding monitoring records are connected in sequence to output the timing monitoring sequence of the corresponding IGBT module.

3. The shipborne domestically produced IGBT module drive monitoring system according to claim 2, characterized in that: The execution of the diagnostic disturbance injection module includes: Read the drive switching time from the monitoring record, and extract the drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal and auxiliary power supply status that are continuously arranged between the previous acquisition time and the next acquisition time of the drive switching time. Take the acquisition time when the drive control signal flips as the switching start time, and take the first acquisition time when the gate voltage reaches the level corresponding to the drive control signal after the flip as the switching end time. Take each acquisition time between the switching start time and the switching end time as the candidate injection time, and output the candidate injection time sequence.

4. The shipborne domestically produced IGBT module drive monitoring system according to claim 3, characterized in that: The execution of the diagnostic disturbance injection module also includes: Based on the candidate injection time sequence, a diagnostic disturbance action is generated for each candidate injection time according to the acquisition time order. Each diagnostic disturbance action includes the injection start point corresponding to the candidate injection time, the injection duration corresponding to the recording interval between the candidate injection time and the next acquisition time, the injection polarity with the same flip direction as the current drive control signal, and the injection path corresponding to the gate circuit. Each diagnostic disturbance action is then superimposed on the gate current value between the switching start point and the switching end point to obtain the disturbed gate current sequence. The disturbed gate current sequence is then integrated according to the acquisition time to obtain the disturbed gate voltage sequence, and the disturbed gate voltage sequence corresponding to each diagnostic disturbance action is output.

5. The shipborne domestically produced IGBT module drive monitoring system according to claim 4, characterized in that: The execution of the diagnostic disturbance injection module also includes: After obtaining each perturbation-induced gate voltage sequence, for each diagnostic perturbation action, the voltage difference between the perturbation-induced gate voltage sequence and the original gate voltage sequence at the switching endpoint, the number of times the voltage change direction between two adjacent acquisition times of the perturbation-induced gate voltage sequence differs from that between two adjacent acquisition times of the original gate voltage sequence, and the number of times the desaturation detection signal value within the diagnostic perturbation action's effective range differs from the desaturation detection signal value before injection are calculated. The voltage difference, the number of acquisitions with different directions, and the number of acquisitions with different desaturation detection signal values ​​are sequentially combined to form the deviation results. The target diagnostic perturbation action is determined according to the ascending order of voltage difference, the ascending order of the number of acquisitions with different directions, and the ascending order of the number of acquisitions with different desaturation detection signal values. According to the injection start point, injection duration, injection polarity, and injection path of the target diagnostic disturbance action, the injection is performed into the gate circuit. Simultaneously, the drive control signal, gate voltage, gate current, collector voltage, emitter potential, desaturation detection signal, and auxiliary power supply status are collected from the injection start point to the switching end point. The target diagnostic disturbance action and the corresponding collection results are written into the same response record in the order of collection time, and the response record corresponding to the target diagnostic disturbance action is output.

6. The shipborne domestically produced IGBT module drive monitoring system according to claim 5, characterized in that: The execution of the response fingerprint generation module includes: The system reads the action identifier, injection start point, injection duration, acquisition time, drive control signal, gate voltage, gate current, collector voltage, emitter potential, and auxiliary power supply status from the response record. These are arranged in ascending order of acquisition time to form a response sequence. The acquisition time when the drive control signal flips is taken as the switching start point, the injection start point is taken as the disturbance start point, the acquisition time corresponding to the sum of the injection start point and the injection duration is taken as the disturbance end point, and the first acquisition time when the gate voltage reaches the level corresponding to the drive control signal after the flip is taken as the switching end point. The system then outputs the switching interval sequence. Based on the switching interval sequence, the gate current at two adjacent acquisition times between the disturbance start point and the switching end point is multiplied and accumulated to obtain the gate charge establishment value. The gate voltage difference at two adjacent acquisition times between the disturbance start point and the switching end point is calculated item by item. When the gate voltage difference of two consecutive recording intervals has the same sign and the collector voltage difference of two consecutive recording intervals has opposite signs, the starting acquisition time of the first recording interval is determined as the platform start point. The acquisition time when the collector voltage difference first becomes zero in the subsequent recording interval is determined as the platform end point. The Miller platform dwell value is obtained by the difference between the platform end point and the platform start point. The collector voltage drop value is obtained by the difference between the collector voltage at the disturbance start point and the value at the switching end point. The emitter return current value is obtained by accumulating the emitter potential in the same direction difference of each recording interval between the disturbance start point and the switching end point. The auxiliary power supply transient drop value is obtained by the difference between the auxiliary power supply state at the disturbance start point and the lowest value between the disturbance start point and the switching end point. The characteristic group is then output.

7. The shipborne domestically produced IGBT module drive monitoring system according to claim 6, characterized in that: The execution of the response fingerprint generation module also includes: After obtaining the feature set, for each response record, calculate whether the platform start point is located after the disturbance start point and before the switching end point, whether the platform end point is located after the platform start point, whether the difference sign of the collector voltage between the disturbance start point and the switching end point is consistent, and whether the minimum value of the auxiliary power supply status between the disturbance start point and the switching end point is not higher than the value of the disturbance start point. For response records that do not meet the aforementioned positional and sign relationships, delete the corresponding acquisition time and re-execute the feature calculation until the remaining response records all meet the aforementioned positional and sign relationships, and output the verified feature set. The gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power transient drop value in the verified feature group are written into the same drive response fingerprint record in the order of action identifier, bridge arm identifier, IGBT module identifier, switching start point, gate load establishment value, Miller plateau dwell value, collector voltage drop value, emitter return current value, and auxiliary power transient drop value to generate a drive response fingerprint corresponding to the current operating condition.

8. A shipborne domestically produced IGBT module drive monitoring system according to claim 7, characterized in that: The execution of the difference closure determination module includes: Read the current response fingerprint and the pre-stored fingerprints of the same module, the same bridge arm, and the same operating condition, calculate the corresponding differences of the gate load value, plateau value, fallback value, return current value, and drop value, and output the module difference group, bridge arm difference group, and operating condition difference group. The absolute values ​​of the corresponding differences between the module difference group and the bridge arm difference group are taken and subtracted one by one to obtain the first closed group. The absolute values ​​of the corresponding differences between the module difference group and the working condition difference group are taken and subtracted one by one to obtain the second closed group. The same-named differences between the first closed group and the second closed group are recorded as abnormal items if they have the same sign, as disturbance items if they have opposite signs, and as common items if at least one of them is zero. The number of abnormal items, the number of disturbance items, and the number of common items are output.

9. A shipborne domestically produced IGBT module drive monitoring system according to claim 8, characterized in that: The execution of the difference closure determination module also includes: The determination is made based on the number of abnormal items, the number of disturbance items, and the number of common items. When the number of abnormal items is greater than the number of disturbance items, a drive chain abnormal flag is output. When the number of disturbance items is greater than the number of abnormal items, a working condition disturbance flag is output. When the number of abnormal items is equal to the number of disturbance items, the sum of the absolute values ​​of the fallback value, the plateau value, and the drop value in the first closed group and the second closed group are compared, and the flag corresponding to the side with the larger sum of absolute values ​​is used as the output flag.

10. A shipborne domestically produced IGBT module drive monitoring system according to claim 9, characterized in that: The execution of the monitoring output control module includes: Read the operating condition disturbance flag or drive chain abnormal flag, and write the bridge arm identifier, module identifier, switching start point, output flag and corresponding response fingerprint into the same monitoring result record in the order of the fields to output the monitoring result record; The drive adjustment command is generated based on the output flag. When the output flag is the operating condition disturbance flag, the current drive board output remains unchanged. When the output flag is the drive chain abnormal flag, the current drive board output is switched to the shutdown output, and the drive adjustment command is output to the IGBT module drive board.

Images