Protection circuit for improving reliability of power semiconductor device

Abstract

This invention relates to the field of power electronics technology, and more particularly to a protection circuit for improving the reliability of power semiconductor devices. This protection circuit senses and analyzes key dynamic behaviors during the operation of power semiconductor devices, acquiring characteristic information reflecting changes in the actual operating state at the initial stage of device turn-on or anomaly occurrence, and determining the device's operating state based on the characteristic analysis results. When the device is determined to be in an abnormal operating condition, the circuit implements controlled adjustment or shutdown of the gate drive signal of the power semiconductor device according to the degree and type of the abnormality, thereby limiting energy accumulation and stress rise during the abnormal period. Simultaneously, the protection strategy is dynamically adjusted based on reference behavior formed under normal operating conditions, ensuring that the protection action matches the actual operating state of the device. This achieves timely and effective protection of power semiconductor devices under abnormal operating conditions, improving the device's operational reliability under complex operating conditions.

Description

Technical Field

[0001] This invention relates to the field of power electronics technology, and in particular to protection circuits for improving the reliability of power semiconductor devices. Background Technology

[0002] Power semiconductor devices are core components in power electronic systems, widely used in power conversion, motor drives, power management, and new energy equipment to achieve switching control and power regulation of electrical energy. As power semiconductor devices develop towards higher voltage, higher frequency, and higher power density, the requirements for their operational stability and reliability are constantly increasing. Therefore, protection circuits used in conjunction with power semiconductor devices have become an important component in ensuring the safe operation of these devices.

[0003] Existing protection circuits for improving the reliability of power semiconductor devices typically monitor operating parameters such as current, voltage, or temperature, and take protective measures such as current limiting or shutdown when abnormal parameters exceed preset thresholds. However, during device turn-on or in the initial stage of an anomaly, changes in the device's operating state have significant transient characteristics. Relying solely on the aforementioned parameters for protection control makes it difficult to reflect the actual changes in the device's operating state in a timely manner, which can easily lead to delayed protection actions or a mismatch between the protection strategy and the actual operating conditions, thereby affecting the operational reliability of power semiconductor devices under complex operating conditions. Summary of the Invention

[0004] To overcome the above shortcomings, the present invention provides a protection circuit to improve the reliability of power semiconductor devices, aiming to improve the problems of delayed protection response and insufficient adaptability to actual operating conditions in the protection circuits of the prior art in the early stage of anomaly.

[0005] This invention provides the following technical solution: a protection circuit for improving the reliability of power semiconductor devices, comprising: The gate dynamic behavior sensing module is used to collect electrical signals reflecting the dynamic changes of the gate drive circuit of the power semiconductor device during the turn-on process or the early stage of an abnormality. The operating condition discrimination module based on behavioral features is connected to the gate dynamic behavior sensing module and is used to perform feature processing on the dynamic changes of the gate drive circuit within a preset time range to generate behavioral feature parameters for characterizing the operating state of the power semiconductor device. The controlled gate adjustment module, connected to the operating condition discrimination module, is used to controllably adjust the gate drive signal of the power semiconductor device according to the discrimination result of the abnormal operating condition when the power semiconductor device is determined to be in an abnormal operating condition, so as to limit the energy accumulation of the device under the abnormal operating condition.

[0006] Preferably, the gate dynamic behavior acquisition step specifically includes: Using the rising or falling edge of the gate drive signal as the trigger reference, the dynamic response signal of the gate drive circuit within a preset time after the trigger reference is captured. The dynamic response signal is sampled once or multiple times, and the sampling sequence is maintained to obtain the corresponding gate dynamic behavior timing segment; The timing segment of the gate dynamic behavior is input into the behavior feature parameter generation process.

[0007] Preferably, the steps for determining the short time window specifically include: The starting point of the window is determined by the transition time of the gate drive signal; During the sampling process, the rate of change of the gate dynamic behavior is calculated in real time, and the window endpoint is determined when the rate of change continuously meets the preset stability condition. The temporal segment between the start and end points of the window is used as the input for generating behavioral feature parameters.

[0008] Preferably, the behavioral feature parameter generation step specifically includes: In the gate dynamic behavior timing segment, at least one event point characterizing the inflection point of dynamic change is detected, and the timing segment is adaptively divided into at least two time sub-intervals based on the event point; For each time sub-interval, calculate at least one of the integral, mean, or rate of change, and normalize the calculation results. The behavioral feature parameters are generated based on the proportional or differential relationships between the normalization results of different time sub-intervals.

[0009] Preferably, the steps for establishing reference behavior specifically include: Obtain the corresponding set of behavioral feature parameters from multiple switching events that are determined to be in normal working condition, and remove feature parameters that exceed the preset outlier conditions. Statistical analysis is performed on the set of behavioral characteristic parameters after removal to generate a reference characteristic range for characterizing the distribution of behavior under normal operating conditions; The reference feature range is used as a reference behavior for deviation discrimination.

[0010] Preferably, the reference behavior update steps specifically include: Reference behavior update is allowed only when the deviation of the on / off events is less than the preset threshold for a preset number of consecutive preset number of events. During the update process, the amount of correction to the reference behavior is limited or the rate is limited so that the reference behavior gradually converges. When any switch event is determined to be an abnormal condition, the update of the reference behavior is suspended until the abnormality is resolved.

[0011] Preferably, the steps for limiting device energy accumulation under abnormal operating conditions specifically include: After determining the abnormal operating condition, an accumulated amount is generated based on the gate dynamic behavior timing segment or its behavior characteristic parameters to characterize the energy accumulation trend during the abnormal period. Before the accumulated amount reaches a preset threshold, a controlled adjustment is applied to the gate drive signal to suppress the rise of device stress, and a shutdown is triggered when the accumulated amount approaches the preset threshold. The controlled adjustment is maintained during the shutdown process until the device exits the abnormal operating condition.

[0012] Preferably, the controlled adjustment steps specifically include: The degree of abnormality is divided into at least two levels based on the deviation of the behavioral characteristic parameters; Different gate drive adjustment strategies are selected for different anomaly levels. The adjustment strategies include at least one of limiting the amplitude of the gate drive signal, limiting the rate of change, or staged adjustment. When the anomaly level escalates, switch to a higher-intensity adjustment strategy or trigger shutdown in advance.

[0013] Preferably, the steps for distinguishing abnormal operating condition types and determining corresponding adjustment paths specifically include: The type of abnormal operating condition is determined based on at least two different combinations of behavioral characteristic parameters; For different abnormal operating conditions, different gate drive controlled adjustment sequences are determined. The controlled adjustment sequence includes at least two different sequences: first limiting the amplitude of the gate drive signal and then turning it off, and first limiting the rate of change of the gate drive signal and then turning it off. The gate drive signal is adjusted according to the determined controlled adjustment sequence to complete the turn-off.

[0014] Preferably, the gate dynamic behavior processing and controlled adjustment execution steps specifically include: The timing signal of gate dynamic behavior is extracted based on the gate drive signal transition trigger. The time-series signal is subjected to event point detection and sub-interval division, and behavioral feature parameters are generated based on integral calculation or rate of change calculation for each sub-interval. An adjustment control quantity is generated based on the deviation between the behavioral characteristic parameters and the reference behavior, and the gate drive signal is subjected to amplitude limitation, rate of change limitation, or staged adjustment according to the adjustment control quantity.

[0015] The present invention has the following beneficial effects: 1. In this invention, by introducing a sensing and analysis mechanism for the internal dynamic behavior of power semiconductor devices during operation, the protection circuit can obtain information reflecting changes in the actual working state at the initial stage of device turn-on or abnormal occurrence. This breaks through the limitation of traditional protection methods that rely solely on external static parameters, enabling early identification and response to abnormal conditions, and significantly improving the timeliness and pertinence of protection actions.

[0016] 2. In this invention, by adopting a controlled gate adjustment method that matches the actual operating state of the device under abnormal operating conditions, the protection circuit can dynamically adjust the adjustment strategy according to the degree and type of abnormality, avoiding the problems of over-protection or under-protection caused by a single shutdown or fixed limit protection method, thereby reducing the electrical and thermal stress borne by the device during abnormal handling.

[0017] 3. In this invention, by establishing and updating reference behavior based on the device's own operating characteristics, the protection circuit has the ability to adaptively adjust as the device's operating state changes, avoiding the problem of insufficient adaptability caused by relying on fixed empirical thresholds, thereby improving the stability and consistency of the protection strategy under different devices, different operating conditions and long-term operating conditions. Attached Figure Description

[0018] Figure 1 This is a system architecture diagram of the protection circuit proposed in this invention to improve the reliability of power semiconductor devices. Detailed Implementation

[0019] The technical solutions in 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.

[0020] Example 1 In a first embodiment of the present invention, the present invention provides a protection circuit to improve the reliability of power semiconductor devices, such as... Figure 1 As shown, it includes the following steps: The gate dynamic behavior sensing module is used to collect electrical signals reflecting the dynamic changes of the gate drive circuit of a power semiconductor device during the turn-on process or at the initial stage of an abnormality. Furthermore, the gate dynamic behavior acquisition step specifically includes: Using the rising or falling edge of the gate drive signal as the trigger reference, the dynamic response signal of the gate drive circuit within a preset time after the trigger reference is captured. The dynamic response signal is sampled once or multiple times, and the sampling sequence is held to obtain the corresponding gate dynamic behavior timing segment; The timing segment of the gate dynamic behavior is input into the behavior feature parameter generation process.

[0021] Specifically, transient response information of the gate drive circuit is acquired through gate dynamic behavior acquisition to obtain information during the initial stage of power semiconductor device turn-on or abnormal occurrence. When the drive signal output by the gate drive circuit transitions, the transition moment is used as the acquisition trigger reference. The transition includes the rising edge of the drive signal from low to high level, or the falling edge from high to low level. Starting from the trigger reference, the dynamic response signal in the gate drive circuit is captured within a preset time interval after triggering. This time interval covers the dynamic stage from the gate drive signal transition to the device entering a stable conduction state, so as to fully include the gate charging and gate-drain coupling effect change process.

[0022] Within a preset time interval, the dynamic response signal of the gate drive circuit is discretely sampled, and the sampling frequency is set to be no less than the characteristic frequency of the gate dynamic change.

[0023] The sampled data is processed by holding, forming a stable time-series data set. After sampling and holding, the gate dynamic behavior is represented as a discrete-time series: ; in: Indicates the first The amplitude of the dynamic response signal of the gate drive circuit at each sampling time; Indicates the sampling time relative to the trigger reference; This indicates the number of sampling points within a preset time interval.

[0024] This time series preserves complete dynamic change information of the gate drive circuit during the initial turn-on or the initial abnormal phase.

[0025] By performing a hold process on the sampled data sequence, a stable timing data set is formed to avoid transient fluctuations interfering with subsequent processing. After sampling and holding, a complete timing segment of the gate dynamic behavior is obtained.

[0026] By inputting the generated gate dynamic behavior timing segment into the subsequent behavior feature parameter generation step, behavior features reflecting the transient operating state of power semiconductor devices can be extracted.

[0027] Through the above-described gate dynamic behavior acquisition process, complete and repeatable gate dynamic behavior timing data can be obtained before the power semiconductor device enters a stable state, providing a stable data foundation for subsequent behavior feature extraction and abnormal operating condition identification.

[0028] Furthermore, the steps for determining the short time window specifically include: The start point of the window is determined by the transition time of the gate drive signal; The rate of change of the gate's dynamic behavior is calculated in real time during the sampling process, and the window endpoint is determined when the rate of change continuously meets the preset stability condition. The temporal segment between the start and end points of the window is used as the input for generating behavioral feature parameters.

[0029] Specifically, a short time window is used to limit the effective time range for generating behavioral characteristic parameters in the dynamic behavior of the gate. When the drive signal output by the gate drive circuit undergoes a transition, the transition moment is taken as the starting point of the time window.

[0030] During sampling, the rate of change of the gate dynamic behavior timing data is calculated in real time to characterize the transient changes in the gate dynamic behavior. The rate of change is calculated based on the difference between adjacent sampling points: ; in: This represents the rate of change at that sampling time.

[0031] When the rate of change corresponding to multiple consecutive sampling points meets the preset stability condition, the gate dynamic behavior is determined to have entered the quasi-steady state stage, and the first sampling moment that meets the stability condition is determined as the end point of the time window. The stability condition is determined by the absolute value of the rate of change not exceeding a preset threshold. The gate dynamic behavior timing segment corresponding to the start and end points of the window is then used as the input data for the subsequent behavior feature parameter generation steps.

[0032] By using the aforementioned short time window determination method, transient segments with discriminative value in the dynamic behavior of the gate can be adaptively extracted, avoiding interference from steady-state signals on behavioral feature extraction, and providing input data with a clear time structure and good consistency for subsequent abnormal operating condition discrimination.

[0033] The operating condition discrimination module based on behavioral features is connected to the gate dynamic behavior sensing module. It is used to characterize the dynamic changes of the gate drive circuit within a preset time range and generate behavioral feature parameters to characterize the operating state of power semiconductor devices. Furthermore, the behavioral feature parameter generation steps specifically include: In the gate dynamic behavior timing segment, at least one event point characterizing the inflection point of dynamic change is detected, and the timing segment is adaptively divided into at least two time sub-intervals based on the event point; Calculate at least one of the integral, mean, or rate of change for each time sub-interval, and normalize the calculation results. Behavioral feature parameters are generated based on the proportional or differential relationships between the normalized results of different time sub-intervals.

[0034] Specifically, after obtaining the timing segment of the gate's dynamic behavior, event point detection is performed on this timing segment to identify the positions where the dynamic change trend reverses. Event points are determined by detecting changes in the sign or extreme values ​​of the rate of change of the timing data. When the rate of change changes from positive to negative or from negative to positive, the corresponding sampling time is determined as the inflection point of the dynamic change.

[0035] Using the detected event points as boundaries, the timing segment of the gate dynamic behavior is adaptively divided into at least two time sub-intervals, each time sub-interval corresponding to a different dynamic change stage.

[0036] For each time sub-interval, at least one of the integral, mean, or rate of change of the gate dynamic behavior signal within that interval is calculated, and corresponding interval characteristic values ​​are formed. The integral is used to characterize the signal accumulation characteristics within the interval, the mean is used to characterize the overall amplitude level within the interval, and the rate of change is used to characterize the intensity of dynamic changes within the interval.

[0037] To eliminate the influence of differences in signal amplitude scales across different time intervals on the discrimination results, the interval feature values ​​obtained for each time sub-interval are normalized. The normalized interval feature values ​​are expressed as follows: ; in: Indicates the first Interval feature values ​​corresponding to each time sub-interval; Represents the normalized interval eigenvalues; This indicates the number of time sub-intervals.

[0038] Therefore, by using the proportional or difference relationship between the characteristic values ​​of the normalized intervals of different time sub-intervals, behavioral characteristic parameters are generated to characterize the dynamic behavior and structural properties of the gate, and these behavioral characteristic parameters are used for subsequent abnormal operating condition discrimination.

[0039] Through the above-mentioned event point detection, adaptive interval division and interval feature construction process, the gate dynamic behavior can be transformed from the original timing signal into behavioral feature parameters with structural information, so that the dynamic change morphology difference can be effectively expressed at the feature level, providing a stable and repeatable input basis for subsequent working condition judgment.

[0040] Furthermore, the steps for establishing reference behaviors specifically include: Obtain the corresponding set of behavioral feature parameters from multiple switching events that are determined to be in normal working condition, and remove feature parameters that exceed the preset outlier conditions. Statistical analysis is performed on the set of behavioral characteristic parameters after removal to generate a reference characteristic range for characterizing the distribution of behavior under normal operating conditions; The reference feature range is used as a reference behavior for deviation discrimination.

[0041] Specifically, after generating the behavioral feature parameters, reference behaviors are established based on multiple switching events under normal operating conditions for subsequent abnormal deviation judgment.

[0042] By acquiring corresponding behavioral characteristic parameters from multiple switching events determined to be under normal operating conditions, and constructing a characteristic parameter set from these multiple switching events, this characteristic parameter set is then filtered according to a preset outlier criterion to remove abnormal characteristic parameters that significantly deviate from the overall distribution, in order to avoid the influence of occasional disturbances or measurement noise on the establishment of reference behavior.

[0043] After outlier removal, statistical processing is performed on the remaining set of behavioral feature parameters to extract statistics that characterize the distribution range of behavioral features under normal operating conditions. These statistics include at least the central value and dispersion range of the behavioral feature parameters to form a reference feature range covering dynamic behavioral fluctuations under normal operating conditions.

[0044] The obtained reference feature range is used as a reference behavior for deviation discrimination to measure the degree of deviation of the behavioral feature parameters in subsequent switching events from the normal operating condition behavior distribution.

[0045] Through the above-mentioned reference behavior establishment process, a stable behavioral benchmark can be formed based on the historical behavioral characteristics of the device under normal operating conditions, avoiding the lack of adaptability caused by fixed thresholds, and providing a representative and consistent reference basis for subsequent abnormal deviation judgment.

[0046] Furthermore, the reference behavior update steps specifically include: Reference behavior update is allowed only when the deviation of the on / off events is less than the preset threshold for a preset number of consecutive preset number of events. During the update process, the amount of correction to the reference behavior is limited or the rate is limited so that the reference behavior gradually converges. When any switch event is determined to be an abnormal condition, the reference behavior update is suspended until the abnormality is resolved.

[0047] Specifically, after the reference behavior is established, update control is performed on the reference behavior based on the behavioral characteristic parameters of subsequent switching events to maintain consistency between the reference behavior and the actual operating state of the device.

[0048] If, within a preset number of consecutive switching events, the corresponding behavioral characteristic parameters all deviate from a preset threshold, an update operation on the reference behavior is permitted; otherwise, the reference behavior update is not performed. During the reference behavior update, a limiting or rate-limiting constraint is applied to the correction amount of the reference behavior, ensuring that the reference behavior is adjusted gradually in a controlled manner to avoid sudden changes in the reference behavior due to fluctuations in a single event. When any switching event is determined to be an abnormal operating condition, the reference behavior update process is immediately paused, and the update determination resumes after the abnormal operating condition is resolved.

[0049] By using the above-mentioned reference behavior update control method, it is possible to achieve gradual adjustment with changes in operating status while maintaining the stability of reference behavior, avoiding the contamination of reference behavior by abnormal operating conditions, and providing a continuous and effective benchmark for subsequent deviation judgment.

[0050] The controlled gate adjustment module, connected to the operating condition discrimination module, is used to controllably adjust the gate drive signal of the power semiconductor device according to the discrimination result when the power semiconductor device is in an abnormal operating condition, so as to limit the energy accumulation of the device under the abnormal operating condition. Furthermore, the steps for limiting device energy accumulation under abnormal operating conditions specifically include: After determining the abnormal operating condition, an accumulated amount is generated based on the timing segment of the gate dynamic behavior or its behavioral characteristic parameters to characterize the energy accumulation trend during the abnormal period. Before the accumulated amount reaches a preset threshold, a controlled adjustment is applied to the gate drive signal to suppress the rise of device stress, and a shutdown is triggered when the accumulated amount approaches the preset threshold. During the shutdown process, maintain controlled regulation until the device exits the abnormal operating condition.

[0051] Specifically, after an abnormal operating condition is determined, an energy-limiting control process is initiated to constrain the energy accumulation of power semiconductor devices during the abnormal period. Upon triggering of the abnormal condition, a cumulative amount is generated based on a timing segment of the gate's dynamic behavior or the corresponding behavioral characteristic parameters to characterize the energy accumulation trend during the abnormal period. This cumulative amount is obtained by accumulating the stress-related quantities of the device during the abnormal period over time, reflecting the increasing trend of device energy input during the duration of the abnormality.

[0052] Before the accumulated amount reaches a preset threshold, controlled regulation is applied to the gate drive signal to suppress further increases in device current or voltage by limiting the amplitude and rate of change of the gate drive signal or by using a staged regulation method. When the accumulated amount approaches the preset threshold, a shutdown operation is triggered to prevent the device from entering the irreversible damage region.

[0053] During the shutdown process, the above controlled adjustment is maintained so that the gate drive signal changes in a controlled state until the device completely exits the abnormal operating condition.

[0054] Energy accumulation trend during abnormal periods through cumulative amount The representation, and its calculation method is as follows: ; in: Indicates the first The surrogate quantity related to device stress within each sampling period is used to characterize the transient energy input intensity during anomalies; Indicates the time interval between adjacent sampling periods; This indicates the number of sampling points during the period of abnormal operating conditions.

[0055] Agent volume Obtained by mapping gate dynamic behavior or behavioral characteristic parameters, it is used to reflect the energy accumulation trend of the device during abnormal periods, without relying on direct power measurement.

[0056] Through the above-mentioned energy limiting control steps, the energy input of the device can be dynamically constrained during the period of abnormal operation, and the device can be shut down before the accumulated amount reaches the safety boundary, thereby avoiding irreversible damage to the device caused by the concentrated injection of abnormal energy.

[0057] Furthermore, the controlled regulation steps specifically include: The degree of abnormality is divided into at least two levels based on the deviation of behavioral characteristic parameters; Different gate drive adjustment strategies are selected for different anomaly levels. The adjustment strategies include at least one of limiting the amplitude of the gate drive signal, limiting the rate of change, or staged adjustment. When the anomaly level escalates, switch to a higher-intensity adjustment strategy or trigger shutdown in advance.

[0058] Specifically, after an abnormal operating condition is determined, the degree of abnormality is classified according to the degree of deviation between the behavioral characteristic parameters and the reference behavior, and the corresponding gate drive adjustment strategy is selected accordingly.

[0059] First, based on the degree of deviation of the behavioral characteristic parameters, the degree of abnormality is divided into at least two levels. The abnormality level with a smaller deviation corresponds to a mild abnormality, and the abnormality level with a larger deviation corresponds to a severe abnormality, which is used to distinguish the control requirements under different risk levels.

[0060] Different gate drive adjustment strategies are selected for different levels of anomalies. For mild anomalies, amplitude or rate of change limits are applied to the gate drive signal to reduce the stress growth rate of the device. For severe anomalies, a staged adjustment method is adopted to impose stronger constraints on the gate drive signal, or to prepare for the shutdown process based on controlled adjustment.

[0061] When the abnormality level escalates from low to high, a higher-intensity gate drive regulation strategy is switched in real time, or a shutdown operation is triggered in advance during the regulation process to prevent the abnormal operating condition from deteriorating further.

[0062] Through the above controlled adjustment steps, the gate drive control strategy can be dynamically changed according to the degree of abnormality, realizing a smooth transition from mild control to forced protection, and avoiding the problem of over- or under-protection caused by a single adjustment method under different abnormal conditions.

[0063] Furthermore, the specific steps for distinguishing abnormal operating condition types and determining corresponding adjustment paths include: The type of abnormal operating condition is determined based on at least two different combinations of behavioral characteristic parameters; For different abnormal operating conditions, different gate drive controlled adjustment sequences are determined. The controlled adjustment sequence includes at least two different sequences: first limiting the amplitude of the gate drive signal and then performing the shutdown, and first limiting the rate of change of the gate drive signal and then performing the shutdown. The gate drive signal is adjusted according to the determined controlled adjustment sequence to complete the turn-off.

[0064] Specifically, after an abnormal operating condition is determined, the abnormal operating condition type is distinguished according to the combination relationship of behavioral characteristic parameters, and the corresponding gate drive adjustment path is selected according to different abnormal operating condition types.

[0065] First, a joint analysis is performed based on at least two different behavioral characteristic parameters. The type of abnormal operating condition is determined according to the relative relationships or combined characteristics between these parameters. Different types of abnormal operating conditions correspond to different dynamic behavior patterns, which are used to distinguish the differences in the mechanisms by which abnormalities occur.

[0066] Different gate drive control adjustment sequences are pre-set for different abnormal operating conditions. The adjustment sequence includes at least the following two different paths: one path is to apply an amplitude limit to the gate drive signal before performing the turn-off operation to reduce the gate drive amplitude before completing the turn-off; the other path is to apply a rate-of-change limit to the gate drive signal before performing the turn-off operation to slow down the rate of change of the gate drive signal before completing the turn-off.

[0067] After determining the type of abnormal operating condition and the corresponding adjustment path, the gate drive signal is adjusted according to the selected controlled adjustment sequence, and the turn-off operation is completed under controlled conditions.

[0068] By differentiating abnormal operating conditions and selecting adjustment paths as described above, the gate drive control can adopt a differentiated adjustment sequence according to different abnormal mechanisms when an abnormality occurs, avoiding the incompatibility problem caused by a single turn-off method under different abnormal conditions, thereby improving the stability and controllability of the abnormality handling process.

[0069] Furthermore, the gate dynamic behavior processing and controlled regulation execution steps specifically include: The timing signal of gate dynamic behavior is extracted based on the gate drive signal transition trigger. Event point detection and sub-interval division are performed on time-series signals, and integral or rate of change calculations are performed on each sub-interval to generate behavioral feature parameters; The regulation control quantity is generated based on the deviation between the behavior characteristic parameters and the reference behavior, and the gate drive signal is subjected to amplitude limitation, rate of change limitation or staged regulation according to the regulation control quantity.

[0070] Specifically, after an abnormal operating condition is determined, gate drive control under abnormal conditions is achieved by processing the dynamic behavior of the gate and synchronously performing controlled adjustments.

[0071] When the gate drive signal transitions, the transition time is used as the trigger reference to extract the corresponding gate dynamic behavior timing signal. The extraction range covers the dynamic change stage in the early stage of the abnormality to reflect the transient gate response characteristics of the device under abnormal conditions.

[0072] Event point detection is performed on the intercepted gate dynamic behavior timing signal. By analyzing the dynamic change trend, the inflection point of change is identified, and the detected event point is used as the boundary to complete the sub-interval division, so that different sub-intervals correspond to different dynamic change stages.

[0073] For each time sub-interval obtained by division, integral calculation or rate of change calculation is performed to generate behavioral feature parameters that characterize the properties of each dynamic stage.

[0074] After obtaining the behavioral characteristic parameters, these parameters are compared with established reference behavior to obtain a deviation degree reflecting the extent of deviation from the current dynamic behavior. A corresponding adjustment control quantity is then generated based on this deviation degree. According to the adjustment control quantity, at least one of amplitude limiting, rate of change limiting, or staged adjustment is applied to the gate drive signal to ensure the gate drive process operates under controlled conditions and matches the degree of abnormality.

[0075] Through the above-mentioned gate dynamic behavior processing and controlled adjustment execution process, after an abnormal operating condition occurs, the gate dynamic response can be transformed into a structured behavioral characteristic, and an adjustment control quantity adapted to the degree of abnormality can be generated accordingly. This enables continuous and controllable adjustment of the gate drive signal under abnormal conditions, avoiding abrupt control behavior under abnormal operating conditions.

[0076] Example 2: When power semiconductor devices are powered by a high-voltage DC bus, gate drive anomalies can easily occur during turn-on, leading to a rapid rise in device current and a concentrated accumulation of energy within the device in a short period. This, in turn, causes a sharp increase in electrical and thermal stress, affecting the long-term operational reliability of the device. To address these issues, this invention provides a protection circuit to improve the reliability of power semiconductor devices, the structure of which is as follows: Figure 1 As shown. The specific implementation process of this system is as follows: S1. During the power semiconductor device turn-on process, when the gate drive signal transitions, it triggers the acquisition of gate dynamic behavior. The dynamic response signal of the gate drive circuit in the initial stage of turn-on is captured and a corresponding gate dynamic behavior timing segment is formed. Then, event point detection and sub-interval division are performed on the timing segment, and behavioral feature parameters reflecting dynamic changes are generated based on each sub-interval.

[0077] S2. Compare the behavioral characteristic parameters corresponding to the current switching event with the established reference behavior, calculate the deviation degree reflecting the degree of deviation from the dynamic behavior, and determine whether the current switching process has entered an abnormal operating condition based on the deviation degree. When an abnormal operating condition is determined, the abnormal operating condition type is distinguished based on the combination relationship of multiple behavioral characteristic parameters.

[0078] S3. After an abnormal operating condition is determined, a cumulative amount is generated based on the gate dynamic behavior or behavioral characteristic parameters to characterize the energy accumulation trend during the abnormal period. Depending on the degree of the abnormality and the type of abnormal operating condition, a corresponding gate drive controlled adjustment strategy is determined, including at least one of applying amplitude limiting, rate-of-change limiting, or staged adjustment to the gate drive signal.

[0079] S4. Adjust the gate drive signal according to the selected controlled adjustment strategy, and trigger shutdown under controlled conditions when the abnormality worsens or the energy accumulation approaches the preset threshold, so that the power semiconductor device can safely exit the abnormal operating condition.

[0080] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. 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 protection circuit for improving the reliability of power semiconductor devices, characterized in that, include: The gate dynamic behavior sensing module is used to collect electrical signals reflecting the dynamic changes of the gate drive circuit of the power semiconductor device during the turn-on process or the early stage of an abnormality. The operating condition discrimination module based on behavioral features is connected to the gate dynamic behavior sensing module and is used to perform feature processing on the dynamic changes of the gate drive circuit within a preset time range to generate behavioral feature parameters for characterizing the operating state of the power semiconductor device. The controlled gate adjustment module, connected to the operating condition discrimination module, is used to controllably adjust the gate drive signal of the power semiconductor device according to the discrimination result of the abnormal operating condition when the power semiconductor device is determined to be in an abnormal operating condition, so as to limit the energy accumulation of the device under the abnormal operating condition.

2. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The specific steps for acquiring gate dynamic behavior include: Using the rising or falling edge of the gate drive signal as the trigger reference, the dynamic response signal of the gate drive circuit within a preset time after the trigger reference is captured. The dynamic response signal is sampled once or multiple times, and the sampling sequence is maintained to obtain the corresponding gate dynamic behavior timing segment; The timing segment of the gate dynamic behavior is input into the behavior feature parameter generation process.

3. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The steps for determining the short time window specifically include: The starting point of the window is determined by the transition time of the gate drive signal; During the sampling process, the rate of change of the gate dynamic behavior is calculated in real time, and the window endpoint is determined when the rate of change continuously meets the preset stability condition. The temporal segment between the start and end points of the window is used as the input for generating behavioral feature parameters.

4. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The specific steps for generating behavioral feature parameters include: In the gate dynamic behavior timing segment, at least one event point characterizing the inflection point of dynamic change is detected, and the timing segment is adaptively divided into at least two time sub-intervals based on the event point; For each time sub-interval, calculate at least one of the integral, mean, or rate of change, and normalize the calculation results. The behavioral feature parameters are generated based on the proportional or differential relationships between the normalization results of different time sub-intervals.

5. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The steps for establishing reference behaviors specifically include: Obtain the corresponding set of behavioral feature parameters from multiple switching events that are determined to be in normal working condition, and remove feature parameters that exceed the preset outlier conditions. Statistical analysis is performed on the set of behavioral characteristic parameters after removal to generate a reference characteristic range for characterizing the distribution of behavior under normal operating conditions; The reference feature range is used as a reference behavior for deviation discrimination.

6. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The specific steps for updating reference behavior include: Reference behavior update is allowed only when the deviation of the on / off events is less than the preset threshold for a preset number of consecutive preset number of events. During the update process, the amount of correction to the reference behavior is limited or the rate is limited so that the reference behavior gradually converges. When any switch event is determined to be an abnormal condition, the update of the reference behavior is suspended until the abnormality is resolved.

7. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The specific steps for limiting device energy accumulation under abnormal operating conditions include: After determining the abnormal operating condition, an accumulated amount is generated based on the gate dynamic behavior timing segment or its behavior characteristic parameters to characterize the energy accumulation trend during the abnormal period. Before the accumulated amount reaches a preset threshold, a controlled adjustment is applied to the gate drive signal to suppress the rise of device stress, and a shutdown is triggered when the accumulated amount approaches the preset threshold. The controlled adjustment is maintained during the shutdown process until the device exits the abnormal operating condition.

8. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The controlled adjustment steps specifically include: The degree of abnormality is divided into at least two levels based on the deviation of the behavioral characteristic parameters; Different gate drive adjustment strategies are selected for different anomaly levels. The adjustment strategies include at least one of limiting the amplitude of the gate drive signal, limiting the rate of change, or staged adjustment. When the anomaly level escalates, switch to a higher-intensity adjustment strategy or trigger shutdown in advance.

9. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The specific steps for distinguishing abnormal operating condition types and determining corresponding adjustment paths include: The type of abnormal operating condition is determined based on at least two different combinations of behavioral characteristic parameters; For different abnormal operating conditions, different gate drive controlled adjustment sequences are determined. The controlled adjustment sequence includes at least two different sequences: first limiting the amplitude of the gate drive signal and then turning it off, and first limiting the rate of change of the gate drive signal and then turning it off. The gate drive signal is adjusted according to the determined controlled adjustment sequence to complete the turn-off.

10. The protection circuit for improving the reliability of power semiconductor devices according to claim 1, characterized in that, The specific steps for gate dynamic behavior processing and controlled regulation execution include: The timing signal of gate dynamic behavior is extracted based on the gate drive signal transition trigger. The time-series signal is subjected to event point detection and sub-interval division, and behavioral feature parameters are generated based on integral calculation or rate of change calculation for each sub-interval. An adjustment control quantity is generated based on the deviation between the behavioral characteristic parameters and the reference behavior, and the gate drive signal is subjected to amplitude limitation, rate of change limitation, or staged adjustment according to the adjustment control quantity.

Images