Data-driven power semiconductor failure early warning method and system

Abstract

The application relates to the technical field of power electronic device monitoring, and relates to a power semiconductor fault early warning method and system based on data driving, which comprises the following steps: receiving a power semiconductor fault early warning instruction, confirming a fault early warning system based on the fault early warning instruction, acquiring an electrical evaluation node based on a power semiconductor, acquiring an electrical parameter node set based on a preset electrical monitoring time period, a preset electrical monitoring time interval and the electrical parameter monitoring unit, acquiring a frequency evaluation ratio by using a low-frequency monitoring module and a high-frequency monitoring module, confirming a first-level monitoring state of the power semiconductor based on the electrical parameter node set, the frequency evaluation ratio and the electrical evaluation node, acquiring a monitoring temperature set and a temperature change rate set by using a temperature monitoring module set, confirming a second-level monitoring state of the power semiconductor based on the comprehensive monitoring temperature value and the temperature change rate set, and realizing power semiconductor fault early warning. The application can improve the accuracy and reliability of power semiconductor fault early warning.

Description

Technical Field

[0001] This invention relates to the field of power electronic equipment monitoring technology, and in particular to a data-driven method and system for early warning of power semiconductor faults. Background Technology

[0002] Power semiconductors play a central role in modern power electronic systems and are widely used in various fields. As the complexity of power electronic equipment continues to increase, power semiconductor failures can lead to system downtime or even damage, causing significant economic losses. Therefore, timely and accurate fault warnings for power semiconductors are crucial for ensuring the safe operation of power electronic systems.

[0003] Currently, traditional power semiconductor fault early warning mainly relies on monitoring a single electrical parameter, such as voltage or current threshold judgment. However, monitoring a single parameter cannot fully reflect the operating status of power semiconductors, and it is easy to miss or falsely report faults, which reduces the accuracy of fault early warning.

[0004] While the aforementioned methods can provide early warning of power semiconductor faults, they lack monitoring of factors such as temperature changes in the power semiconductor, resulting in poor reliability and low accuracy in fault prediction. Therefore, improving the accuracy and reliability of power semiconductor fault prediction has become an urgent problem to be solved. Summary of the Invention

[0005] This invention provides a data-driven power semiconductor fault early warning method and a computer-readable storage medium, the main purpose of which is to improve the accuracy and reliability of power semiconductor fault early warning.

[0006] To achieve the above objectives, the present invention provides a data-driven power semiconductor fault early warning method, comprising:

[0007] The system receives a fault warning command from a power semiconductor and confirms a fault warning system based on the fault warning command. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module.

[0008] Electrical evaluation nodes are obtained based on power semiconductors, wherein the electrical evaluation nodes include standard voltage, standard current, voltage warning deviation range, and current warning deviation range;

[0009] Based on the preset electrical monitoring period, the preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. The electrical parameter node set includes multiple electrical parameter nodes, and each electrical parameter node includes the monitoring voltage and the monitoring current.

[0010] The frequency evaluation ratio is obtained by using a low-frequency monitoring module and a high-frequency monitoring module. Based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes, the first-level monitoring status of the power semiconductor is confirmed. The first-level monitoring status is a first-level early warning status or other first-level status.

[0011] Once the Level 1 monitoring status is confirmed as Level 1 warning status, the temperature monitoring module set is obtained based on the temperature monitoring unit. The monitoring temperature set and the temperature change rate set are obtained using the temperature monitoring module set. The temperature monitoring module set contains multiple temperature monitoring modules, and the monitoring temperature set contains multiple monitoring temperature values. The temperature monitoring module and the monitoring temperature value correspond one-to-one.

[0012] The comprehensive monitoring temperature value is obtained based on the monitoring temperature set and the pre-constructed temperature calculation formula;

[0013] The secondary monitoring status of power semiconductors is confirmed based on the comprehensive monitoring temperature value and temperature change rate set, thereby realizing fault early warning of power semiconductors. The secondary monitoring status is a secondary early warning status or other secondary status.

[0014] Optionally, the method of obtaining electrical evaluation nodes based on power semiconductors includes:

[0015] Obtain the standard voltage and standard current of power semiconductors;

[0016] Obtain the permissible voltage deviation and voltage fault deviation values ​​of the power semiconductor;

[0017] The voltage warning deviation range is obtained based on the voltage allowable deviation value and the voltage fault deviation value, wherein the minimum and maximum values ​​corresponding to the voltage warning deviation range are the voltage allowable deviation value and the voltage fault deviation value, respectively.

[0018] Obtain the allowable current deviation and current fault deviation of the power semiconductor;

[0019] The current warning deviation range is obtained based on the allowable current deviation value and the current fault deviation value. The minimum and maximum values ​​corresponding to the current warning deviation range are the allowable current deviation value and the current fault deviation value, respectively.

[0020] By associating the standard voltage, standard current, voltage warning deviation range, and current warning deviation range, electrical evaluation nodes are obtained.

[0021] Optionally, obtaining the frequency evaluation ratio using the low-frequency monitoring module and the high-frequency monitoring module includes:

[0022] Obtain the frequency monitoring period, and then obtain multiple frequency monitoring sub-periods based on the frequency monitoring period;

[0023] The frequency monitoring sub-periods are sorted in order from front to back according to the time corresponding to the frequency monitoring sub-periods to obtain the frequency monitoring sub-period sequence;

[0024] Multiple odd frequency monitoring sub-segments and multiple even frequency monitoring sub-segments are obtained based on the frequency monitoring sub-segment sequence. Among them, the odd frequency monitoring sub-segments are the frequency monitoring sub-segments with an odd position in the frequency monitoring sub-segment sequence, and the even frequency monitoring sub-segments are the frequency monitoring sub-segments with an even position in the frequency monitoring sub-segment sequence.

[0025] The first frequency value set is obtained based on multiple odd frequency monitoring sub-periods, a preset first number of monitoring times, and a low frequency monitoring module.

[0026] By summing up the first frequency value sets, multiple first frequency value sets are obtained;

[0027] The second frequency value set is obtained based on multiple even-frequency monitoring sub-periods, a preset second monitoring number, and a high-frequency monitoring module.

[0028] By summing up the second frequency value sets, multiple second frequency value sets are obtained;

[0029] The frequency evaluation ratio is obtained based on multiple first frequency value sets, multiple second frequency value sets, and a preset method.

[0030] Optionally, the confirmation of the primary monitoring status of the power semiconductor based on the electrical parameter node set, frequency evaluation ratio, and electrical evaluation nodes includes:

[0031] The electrical parameter nodes in the electrical parameter node set are sorted in chronological order according to the time corresponding to each electrical parameter node, resulting in an electrical parameter node sequence.

[0032] Electrical parameter nodes are extracted sequentially from the electrical parameter node sequence. Based on the extracted electrical parameter nodes, reference electrical parameter nodes are identified in the electrical parameter node sequence. The reference electrical parameter nodes are adjacent to and lag behind the extracted electrical parameter nodes.

[0033] Calculate the absolute difference between the monitored voltage at the electrical parameter node and the monitored voltage at the reference electrical parameter node, and use the absolute difference as the voltage fluctuation difference;

[0034] Compare the voltage fluctuation difference with the preset voltage fluctuation threshold;

[0035] If the voltage fluctuation difference is greater than the voltage fluctuation threshold, the monitoring voltage corresponding to the reference electrical parameter node is confirmed as the fluctuating voltage.

[0036] By summarizing the fluctuating voltages, a fluctuating voltage set is obtained. The number of fluctuating voltages in the fluctuating voltage set is counted to obtain the number of fluctuating voltages.

[0037] The total number of electrical parameter nodes is obtained by counting the number of electrical parameter nodes in the electrical parameter node sequence.

[0038] Weight nodes are obtained based on the number of fluctuating voltages, the total number of electrical parameter nodes, and preset initial weight nodes. The initial weight nodes include: initial first weight, initial second weight, and initial third weight, and the weight nodes include: first weight, second weight, and third weight.

[0039] The first evaluation value is obtained based on the electrical parameter node sequence, the number of fluctuating voltages, electrical evaluation nodes, weight nodes, and pre-constructed evaluation value calculation formula;

[0040] The second evaluation value is obtained based on the electrical parameter node set and electrical evaluation nodes;

[0041] Compare the first evaluation value with the preset first warning threshold, the second evaluation value with the preset second warning threshold, and the frequency evaluation ratio with the preset frequency evaluation ratio threshold, respectively;

[0042] If the first evaluation value is greater than the first warning threshold, or the second evaluation value is greater than the second warning threshold, or the frequency evaluation ratio is greater than the frequency evaluation ratio threshold, the first-level monitoring status of the power semiconductor is confirmed as a first-level warning status.

[0043] Otherwise, the primary monitoring status of the power semiconductor will be confirmed as another primary status.

[0044] Optionally, the step of obtaining weighted nodes based on the number of fluctuating voltages, the total number of electrical parameter nodes, and preset initial weighted nodes includes:

[0045] Calculate the ratio of the number of fluctuating voltages to the total number of electrical parameter nodes to obtain the voltage ratio;

[0046] Compare the voltage ratio with the preset ratio threshold;

[0047] If the voltage ratio is less than or equal to the ratio threshold, the initial weight node is used as the weight node; otherwise, the updated first weight, updated second weight, and updated third weight are obtained based on the number of fluctuating voltages, the total number of electrical parameter nodes, and the pre-constructed updated weight calculation formula.

[0048] Compare and update the first weight with the minimum weight;

[0049] If the updated first weight is greater than or equal to the minimum weight, then the updated first weight, the updated second weight, and the updated third weight are respectively used as the first weight, the second weight, and the third weight. Otherwise, the minimum weight is used as the updated first weight, and the first weight, the second weight, and the third weight are obtained by using the updated first weight, the updated second weight, the updated third weight, and the preset normalization method.

[0050] By summing the first weight, the second weight, and the third weight, we obtain the weight nodes.

[0051] Optionally, the step of acquiring the monitored temperature set and the temperature change rate set using the temperature monitoring module set includes:

[0052] Perform the following operations on each temperature monitoring module in the temperature monitoring module cluster:

[0053] An initial monitoring temperature set is obtained using a temperature monitoring module, a preset temperature monitoring period, and a preset number of temperature monitoring times. The initial monitoring temperature set contains multiple initial monitoring temperature values.

[0054] The initial monitoring temperature sets are aggregated to obtain multiple initial monitoring temperature sets, and each initial monitoring temperature set corresponds one-to-one with a temperature monitoring module;

[0055] For each of the multiple initial monitoring temperature sets, the following operation is performed:

[0056] A monitoring temperature value acquisition scheme is constructed based on the initial monitoring temperature set, as shown below:

[0057]

[0058] Among them, T avg Indicates the monitored temperature value, T i T represents the i-th initial monitoring temperature value in the initial monitoring temperature set. ak This represents the preset k-th initial temperature anomaly value, α represents the preset temperature threshold, n represents the total number of initial monitoring temperature values ​​in the initial monitoring temperature set, and m represents the total number of preset initial temperature anomaly values.

[0059] The monitored temperature value is obtained based on the monitored temperature value acquisition scheme;

[0060] After confirming that each temperature monitoring module has obtained the corresponding monitored temperature value, the monitored temperature values ​​are summarized to obtain the monitored temperature set;

[0061] A set of temperature change rates is obtained based on multiple initial monitoring temperature sets.

[0062] Optionally, the step of obtaining the temperature change rate set based on multiple initial monitored temperature sets includes:

[0063] For each of the multiple initial monitoring temperature sets, the following steps are performed:

[0064] The initial monitoring temperature values ​​in the initial monitoring temperature set are sorted in chronological order according to the time corresponding to the acquisition of the initial monitoring temperature values ​​to obtain the initial monitoring temperature sequence.

[0065] The temperature change rate is obtained based on the initial monitored temperature sequence and the pre-constructed temperature change rate calculation formula;

[0066] By summarizing the temperature change rates, a set of temperature change rates is obtained.

[0067] Optionally, the temperature calculation formula is as follows:

[0068]

[0069] Among them, T b This represents the total monitored temperature value, where l indicates that the monitored temperature set contains l monitored temperature values, T j-avg w represents the j-th monitored temperature value in the monitored temperature set. j R represents the weight of the j-th monitored temperature value. θjc p represents the thermal resistance from the power semiconductor chip to the casing, and p represents the power consumption of the power semiconductor.

[0070] Optionally, the confirmation of the secondary monitoring status of the power semiconductor based on the comprehensive monitoring temperature value and temperature change rate set includes:

[0071] For each temperature change rate in the temperature change rate set, perform the following operation:

[0072] The absolute temperature change rate is obtained based on the temperature change rate, wherein the absolute temperature change rate is the absolute value of the temperature change rate.

[0073] Compare the absolute temperature change rate with a preset temperature change rate threshold;

[0074] If the absolute temperature change rate is greater than the temperature change rate threshold, the absolute temperature change rate is considered an unqualified absolute temperature change rate; otherwise, the absolute temperature change rate is considered a qualified absolute temperature change rate.

[0075] The non-compliant absolute temperature change rate and the compliant absolute temperature change rate are summarized separately to obtain the non-compliant absolute temperature change rate set and the compliant absolute temperature change rate set.

[0076] The number of non-compliant absolute temperature change rates in the non-compliant absolute temperature change rate set and the number of compliant absolute temperature change rates in the compliant absolute temperature change rate set are counted separately to obtain the number of non-compliant and compliant items.

[0077] Determine whether the comprehensive monitoring temperature value, the number of non-compliant items, and the number of compliant items meet the preset third verification condition;

[0078] When the third verification condition is met, the secondary monitoring status of the power semiconductor is confirmed as the secondary warning status;

[0079] Otherwise, the secondary monitoring status of the power semiconductor will be confirmed as another secondary status.

[0080] To achieve the above objectives, the present invention also provides a data-driven power semiconductor fault early warning system, comprising:

[0081] An electrical parameter acquisition module is used to receive fault warning commands from power semiconductors and confirm the fault warning system based on the fault warning commands. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module.

[0082] Electrical evaluation nodes are obtained based on power semiconductors, wherein the electrical evaluation nodes include standard voltage, standard current, voltage warning deviation range, and current warning deviation range;

[0083] Based on the preset electrical monitoring period, the preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. The electrical parameter node set includes multiple electrical parameter nodes, and each electrical parameter node includes the monitoring voltage and the monitoring current.

[0084] A primary monitoring and early warning module is used to obtain the frequency evaluation ratio using a low-frequency monitoring module and a high-frequency monitoring module, and to confirm the primary monitoring status of the power semiconductor based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes. The primary monitoring status is a primary early warning status or other primary status.

[0085] The integrated temperature acquisition module is used to acquire a set of temperature monitoring modules based on the temperature monitoring unit after confirming that the first-level monitoring status is a first-level warning status. It then uses the set of temperature monitoring modules to acquire a set of monitored temperatures and a set of temperature change rates. The set of temperature monitoring modules contains multiple temperature monitoring modules, and the set of monitored temperatures contains multiple monitored temperature values. Each temperature monitoring module corresponds to a monitored temperature value.

[0086] The comprehensive monitoring temperature value is obtained based on the monitoring temperature set and the pre-constructed temperature calculation formula;

[0087] The secondary monitoring and early warning module is used to confirm the secondary monitoring status of the power semiconductor based on the comprehensive monitoring temperature value and temperature change rate set, so as to realize the fault early warning of the power semiconductor. The secondary monitoring status is a secondary early warning status or other secondary status.

[0088] To address the above problems, the present invention also provides an electronic device, the electronic device comprising:

[0089] A memory that stores at least one instruction; and a processor that executes the instructions stored in the memory to implement the data-driven power semiconductor fault early warning method described above.

[0090] To address the aforementioned problems, the present invention also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor in an electronic device to implement the data-driven power semiconductor fault early warning method described above.

[0091] To address the problems described in the background art, this invention receives a fault warning command from a power semiconductor and establishes a fault warning system based on the command. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module. Electrical evaluation nodes are obtained based on the power semiconductor, including standard voltage, standard current, a voltage warning deviation range, and a current warning deviation range. This invention uses the allowable deviation value and the fault deviation value of the electrical parameters as the minimum and maximum values ​​of the electrical parameter warning deviation range, respectively, and uses the range between the allowable deviation and the fault deviation as the warning range. When the detected deviation of electrical parameter values ​​exceeds the allowable range but has not yet reached the fault deviation value, the power semiconductor can be adjusted in a timely manner to ensure its stable operation. Based on a preset electrical monitoring period, a preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. This set includes multiple electrical parameter nodes, each containing monitored voltage and current. A frequency evaluation ratio is obtained using a low-frequency monitoring module and a high-frequency monitoring module. Based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes, the primary monitoring status of the power semiconductor is confirmed. This primary monitoring status can be a primary warning status or other primary status. Therefore, this invention utilizes a low-frequency monitoring module (steady-state analysis) and... The high-frequency monitoring module (transient analysis) uses alternating sampling to comprehensively consider both the steady-state frequency mean and transient frequency bursts. This overcomes the blind spot and missed detection limitations caused by the fixed duty cycle phase in traditional single-sampling modes, improving the reliability and accuracy of early warning. A dynamic weighting method is used to obtain the weights corresponding to the three indicators used to calculate the first evaluation value. When the voltage fluctuation corresponds to a large voltage proportion, the relevant weight of the voltage fluctuation frequency indicator is dynamically increased to highlight the impact of key indicators and promptly capture potential fault risks. Once the first-level monitoring state is confirmed as a first-level early warning state, a temperature monitoring module set is obtained based on the temperature monitoring unit. This temperature monitoring module set is then used to obtain the monitored temperature set and temperature change rate set. The temperature monitoring module set contains multiple temperature monitoring modules. The monitored temperature set contains multiple monitored temperature values, and each temperature monitoring module corresponds one-to-one with a monitored temperature value. This invention utilizes the monitored temperature value acquisition scheme to identify and eliminate initial abnormal temperature values, ensuring that the calculated monitored temperature values ​​are closer to the actual situation, thus improving the accuracy and reliability of temperature monitoring. The monitored temperature values ​​provide the external temperature of the power semiconductor, while thermal characteristic parameters and power consumption are used to calculate the internal junction temperature of the power semiconductor. By comprehensively considering both the external and internal junction temperatures of the power semiconductor, the accuracy of temperature monitoring is improved. A comprehensive monitored temperature value is obtained based on the monitored temperature set and a pre-constructed temperature calculation formula. The secondary monitoring status of the power semiconductor is confirmed based on the comprehensive monitored temperature value and a set of temperature change rates, achieving fault early warning for the power semiconductor.The secondary monitoring state refers to a secondary warning state or other secondary states. This invention assesses the thermal state of power semiconductors based on a comprehensive set of monitored temperature values ​​and temperature change rates. By using a second verification condition, it accurately determines whether the power semiconductor is in a secondary warning state, enabling timely detection of potential faults and improving the reliability and safety of power semiconductor operation. Therefore, this invention can improve the accuracy and reliability of power semiconductor fault early warning. Attached Figure Description

[0092] Figure 1 This is a flowchart illustrating a data-driven power semiconductor fault early warning method according to an embodiment of the present invention.

[0093] Figure 2 A functional block diagram of a data-driven power semiconductor fault early warning system provided in an embodiment of the present invention;

[0094] Figure 3 This is a schematic diagram of the structure of an electronic device that implements the data-driven power semiconductor fault early warning method according to an embodiment of the present invention.

[0095] Explanation of reference numerals in the attached figures:

[0096] 1. Electronic device; 10. Processor; 11. Storage device; 12. Bus.

[0097] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0098] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0099] This application provides a data-driven power semiconductor fault early warning method. The executing entity of the data-driven power semiconductor fault early warning method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the data-driven power semiconductor fault early warning method can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0100] Reference Figure 1 The diagram shown is a flowchart illustrating a data-driven power semiconductor fault early warning method according to an embodiment of the present invention. In this embodiment, the data-driven power semiconductor fault early warning method includes:

[0101] S1. Receive a fault warning command from the power semiconductor, and confirm the fault warning system based on the fault warning command. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module.

[0102] It should be explained that a power semiconductor is a semiconductor device used to control and convert electrical energy, enabling efficient transmission and conversion of electrical energy. A fault warning command is an instruction issued by personnel who wish to provide fault warnings for the power semiconductor. A fault warning system is an integrated unit that uses an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit to monitor the operating status of the power semiconductor. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit, and the frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module. For specific applications of the units and modules, please refer to the following embodiments.

[0103] S2. Obtain electrical evaluation nodes based on power semiconductors, wherein the electrical evaluation nodes include standard voltage, standard current, voltage warning deviation range, and current warning deviation range.

[0104] It should be explained that the electrical evaluation node based on power semiconductors includes:

[0105] Obtain the standard voltage and standard current of power semiconductors;

[0106] Obtain the permissible voltage deviation and voltage fault deviation values ​​of the power semiconductor;

[0107] The voltage warning deviation range is obtained based on the voltage allowable deviation value and the voltage fault deviation value, wherein the minimum and maximum values ​​corresponding to the voltage warning deviation range are the voltage allowable deviation value and the voltage fault deviation value, respectively.

[0108] Obtain the allowable current deviation and current fault deviation of the power semiconductor;

[0109] The current warning deviation range is obtained based on the allowable current deviation value and the current fault deviation value. The minimum and maximum values ​​corresponding to the current warning deviation range are the allowable current deviation value and the current fault deviation value, respectively.

[0110] By associating the standard voltage, standard current, voltage warning deviation range, and current warning deviation range, electrical evaluation nodes are obtained.

[0111] It is understood that the standard voltage and standard current refer to the rated voltage and rated current of the power semiconductor, respectively. The voltage allowable deviation value refers to the maximum permissible deviation of the actual voltage of the power semiconductor relative to the standard voltage under stable operating conditions. The current allowable deviation value refers to the maximum permissible deviation of the actual current of the power semiconductor relative to the standard current under stable operating conditions. Both the voltage allowable deviation value and the current allowable deviation value are within the allowable fluctuation range during normal operation of the power semiconductor and will not lead to performance degradation or damage. The voltage fault deviation value and current fault deviation value refer to the difference between the actual voltage and the standard voltage and the actual current and the standard current of the power semiconductor under fault conditions. When the difference between the actual voltage and the standard voltage exceeds the voltage fault deviation value or the difference between the actual current and the standard current exceeds the current fault deviation value, it indicates that the power semiconductor is in a fault state. The voltage warning deviation value range refers to the transition range between the voltage allowable deviation value and the voltage fault deviation value when the deviation between the actual voltage and the standard voltage of the power semiconductor is during operation. The current warning deviation value range refers to the transition range between the current allowable deviation value and the current fault deviation value when the deviation between the actual current and the standard current of the power semiconductor is during operation.

[0112] For example, if the standard voltage of a power semiconductor is 100V, the allowable voltage deviation is 5V, and the voltage fault deviation is 15V, it means that the actual voltage within the range of [95V, 105V] is within the safe range, the actual voltage greater than 115V or less than 85V is within the fault range, and the actual voltage within the range of [85V, 95V] and [105V, 115V] is within the warning range. At this time, the working state of the power semiconductor is close to the fault edge, and measures can be taken in advance before the fault occurs to avoid equipment damage or performance degradation. The corresponding voltage warning deviation value range at this time is [5V, 15V]. Similarly, if the standard current of a power semiconductor is 50A, the allowable current deviation is 2A, and the fault current deviation is 5A, then the actual current within the range of [48A, 52A] is within the safe range, the actual current greater than 55A or less than 45A is within the fault range, and the actual current within the ranges of [45A, 48A] and [52A, 55A] is within the warning range. The corresponding current warning deviation range is [2A, 5A]. From this example, the electrical evaluation nodes are {100V-(5V, 15V), 50A-(2A, 5A)}. This embodiment of the invention uses the allowable electrical parameter deviation and the fault electrical parameter deviation as the minimum and maximum values ​​of the electrical parameter warning deviation range, respectively, and the range between the allowable deviation and the fault deviation as the warning range. When the deviation of the electrical parameter value exceeds the allowable range but has not yet reached the fault deviation value, the power semiconductor can be adjusted in a timely manner to ensure its stable operation.

[0113] S3. Based on the preset electrical monitoring period, the preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained, wherein the electrical parameter node set includes multiple electrical parameter nodes, and each electrical parameter node includes the monitoring voltage and the monitoring current.

[0114] Understandably, the function of the electrical parameter monitoring unit is to collect voltage and current signals in real time as monitoring voltage and monitoring current. Optionally, voltage sensors and current sensors can be used as the fault detection unit to achieve the acquisition of voltage and current signals. The monitoring voltage and monitoring current represent the actual voltage and actual current of the power semiconductor during operation, respectively.

[0115] For example, assuming the electrical monitoring period is from 8:00:00 to 8:01:40 AM on a certain day, and the electrical monitoring interval is 2 seconds, then the number of electrical parameter monitoring times is 50. During the electrical monitoring period from 8:00:00 to 8:01:40 AM, the voltage and current sensors of the electrical parameter monitoring unit collect voltage and current signals every 2 seconds, respectively, resulting in 50 sets of monitored voltage and 50 sets of monitored current. Here, we only take 10 sets of monitored voltage and 10 sets of monitored current as an example: From these 10 sets of monitored voltage and 10 sets of monitored current, we obtain the electrical... The gas parameter node set is {(8:00:02, 99V, 50A), (8:00:04, 103V, 51A), (8:00:06, 103V, 51A), (8:00:08, 102V, 50A), (8:00:10, 98V, 48A), (8:00:12, 99V, 49A), (8:00:14, 101V, 51A), (8:00:16, 99V, 49A), (8:00:18, 97V, 47A), (8:00:20, 100V, 50A)}.

[0116] S4. Obtain the frequency evaluation ratio using the low-frequency monitoring module and the high-frequency monitoring module. Confirm the first-level monitoring status of the power semiconductor based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes. The first-level monitoring status is a first-level early warning status or other first-level status.

[0117] It is understood that the first-level warning state is the state in which the power semiconductor has potential risks during the first-level monitoring but has not yet reached the fault range. Other first-level states refer to the states of the power semiconductor other than the first-level warning state during the first-level monitoring.

[0118] Specifically, the method of obtaining the frequency evaluation ratio using the low-frequency monitoring module and the high-frequency monitoring module includes:

[0119] Obtain the frequency monitoring period, and then obtain multiple frequency monitoring sub-periods based on the frequency monitoring period;

[0120] The frequency monitoring sub-periods are sorted in order from front to back according to the time corresponding to the frequency monitoring sub-periods to obtain the frequency monitoring sub-period sequence;

[0121] Multiple odd frequency monitoring sub-segments and multiple even frequency monitoring sub-segments are obtained based on the frequency monitoring sub-segment sequence. Among them, the odd frequency monitoring sub-segments are the frequency monitoring sub-segments with an odd position in the frequency monitoring sub-segment sequence, and the even frequency monitoring sub-segments are the frequency monitoring sub-segments with an even position in the frequency monitoring sub-segment sequence.

[0122] The first frequency value set is obtained based on multiple odd frequency monitoring sub-periods, a preset first number of monitoring times, and a low frequency monitoring module.

[0123] By summing up the first frequency value sets, multiple first frequency value sets are obtained;

[0124] The second frequency value set is obtained based on multiple even-frequency monitoring sub-periods, a preset second monitoring number, and a high-frequency monitoring module.

[0125] By summing up the second frequency value sets, multiple second frequency value sets are obtained;

[0126] The frequency evaluation ratio is obtained based on multiple first frequency value sets, multiple second frequency value sets, and a preset method.

[0127] For example, suppose the frequency monitoring period is from 8:00:00 to 8:00:30 AM on a certain day, and each frequency monitoring sub-segment is 5 seconds long. The corresponding frequency monitoring sub-segments are 8:00:00-8:00:05, 8:00:05-8:00:10, 8:00:10-8:00:15, 8:00:15-8:00:20, 8:00:20-8:00:25, and 8:00:25-8:00:30. The segment sequence is represented as {8:00:00-8:00:05, 8:00:05-8:00:10, 8:00:10-8:00:15, 8:00:15-8:00:20, 8:00:20-8:00:25, 8:00:25-8:00:30}, and the corresponding multiple odd-frequency monitoring sub-time periods are {8:00:00-8:00:05, 8:00:10-8:00:15, 8:00:20-8:00:25}. The corresponding even-frequency monitoring sub-periods are {8:00:05-8:00:10, 8:00:15-8:00:20, 8:00:25-8:00:30}. Taking the odd-frequency monitoring sub-period 8:00:00-8:00:05 as an example: During the period 8:00:00-8:00:05, frequency signals are collected using a low-frequency monitoring module. If the first monitoring count is 500, then the first frequency value set corresponding to this odd-frequency monitoring sub-period contains 50. There are 0 first frequency values. Each odd frequency monitoring sub-period corresponds to a set of first frequency values. Similarly, taking the even frequency monitoring sub-period 8:00:05-8:00:10 as an example: During the 8:00:05-8:00:10 period, frequency signals are collected using the high-frequency monitoring module. If the second monitoring count is 5000 times, then the second frequency value set corresponding to this even frequency monitoring sub-period contains 5000 second frequency values, and each even frequency monitoring sub-period corresponds to a set of second frequency values. Generally, in actual monitoring, multiple sets of frequency monitoring sub-periods are divided according to the frequency monitoring period to facilitate the collection of more frequency signals for early warning analysis. This example only uses 6 sets of frequency monitoring sub-periods for explanation.

[0128] It should be understood that both the low-frequency monitoring module and the high-frequency monitoring module are components in the frequency monitoring unit used for frequency signal acquisition. The difference lies in the number of times the low-frequency monitoring module acquires frequency signals within the corresponding monitoring sub-period (e.g., a few or tens of times per second), while the high-frequency monitoring module acquires frequency signals more frequently within the corresponding monitoring sub-period (e.g., thousands or even MHz per second). Optionally, using a low-speed, high-precision ADC (e.g., ADS1115) and a high-speed ADC (e.g., ADS8881) as the low-frequency and high-frequency monitoring modules respectively can achieve frequency signal acquisition; this is existing technology and will not be elaborated upon here. The process of obtaining the frequency evaluation ratio based on multiple first frequency value sets, multiple second frequency value sets, and a preset method is as follows: Assuming that the normal switching frequency range of a certain power semiconductor is 18KHz-22KHz, three first frequency value sets and three second frequency value sets are obtained from the above example. First, taking one of the first frequency value sets as an example: the average value of all first frequency values ​​in the first frequency value set is taken, and the relationship between the average value and the normal switching frequency range of 18KHz-22KHz is determined. If the average value is not within the range of 18KHz-22KHz, the first frequency value set is regarded as the warning first frequency value set. Second, taking one of the second frequency value sets as an example: if there are second frequency values ​​in the second frequency value set that exceed the range of 18KHz-22KHz, the second frequency value set is regarded as the warning second frequency value set. After all three first frequency value sets and three second frequency value sets are confirmed, the ratio of [(total number of all warning first frequency value sets + total number of all warning second frequency value sets) / (total number of all first frequency value sets + total number of all second frequency value sets)] is calculated, and this ratio is used as the frequency evaluation ratio. This invention, through time-series alternating sampling of a low-frequency monitoring module (steady-state analysis) and a high-frequency monitoring module (transient analysis), comprehensively considers both the steady-state frequency mean and the transient frequency burst value, overcoming the blind zone and missed detection limitations caused by the fixed duty cycle phase in the traditional single sampling mode, and improving the reliability and accuracy of early warning.

[0129] It should be explained that the confirmation of the primary monitoring status of power semiconductors based on the electrical parameter node set, frequency evaluation ratio, and electrical evaluation nodes includes:

[0130] The electrical parameter nodes in the electrical parameter node set are sorted in chronological order according to the time corresponding to each electrical parameter node, resulting in an electrical parameter node sequence.

[0131] Electrical parameter nodes are extracted sequentially from the electrical parameter node sequence. Based on the extracted electrical parameter nodes, reference electrical parameter nodes are identified in the electrical parameter node sequence. The reference electrical parameter nodes are adjacent to and lag behind the extracted electrical parameter nodes.

[0132] Calculate the absolute difference between the monitored voltage at the electrical parameter node and the monitored voltage at the reference electrical parameter node, and use the absolute difference as the voltage fluctuation difference;

[0133] Compare the voltage fluctuation difference with the preset voltage fluctuation threshold;

[0134] If the voltage fluctuation difference is greater than the voltage fluctuation threshold, the monitoring voltage corresponding to the reference electrical parameter node is confirmed as the fluctuating voltage.

[0135] By summarizing the fluctuating voltages, a fluctuating voltage set is obtained. The number of fluctuating voltages in the fluctuating voltage set is counted to obtain the number of fluctuating voltages.

[0136] The total number of electrical parameter nodes is obtained by counting the number of electrical parameter nodes in the electrical parameter node sequence.

[0137] Weight nodes are obtained based on the number of fluctuating voltages, the total number of electrical parameter nodes, and preset initial weight nodes. The initial weight nodes include: initial first weight, initial second weight, and initial third weight, and the weight nodes include: first weight, second weight, and third weight.

[0138] The first evaluation value is obtained based on the electrical parameter node sequence, the number of fluctuating voltages, electrical evaluation nodes, weighted nodes, and a pre-constructed evaluation value calculation formula, wherein the evaluation value calculation formula is as follows:

[0139]

[0140] Among them, S v V represents the first evaluation value, e represents the total number of electrical parameter nodes, and V represents the total number of electrical parameter nodes. k and V k-1 d represents the monitoring voltage corresponding to the k-th and (k-1)-th electrical parameter nodes in the electrical parameter node sequence, respectively. v This indicates the number of fluctuating voltages, where a1″, a2″, and a3″ represent the first, second, and third weights in the weighting node, respectively. V std This represents the standard voltage;

[0141] The second evaluation value is obtained based on the electrical parameter node set and electrical evaluation nodes;

[0142] Compare the first evaluation value with the preset first warning threshold, the second evaluation value with the preset second warning threshold, and the frequency evaluation ratio with the preset frequency evaluation ratio threshold, respectively;

[0143] If the first evaluation value is greater than the first warning threshold, or the second evaluation value is greater than the second warning threshold, or the frequency evaluation ratio is greater than the frequency evaluation ratio threshold, the first-level monitoring status of the power semiconductor is confirmed as a first-level warning status.

[0144] Otherwise, the primary monitoring status of the power semiconductor will be confirmed as another primary status.

[0145] Understandably, based on the above set of electrical parameter nodes, the electrical parameter node sequence is {(8:00:02, 99V, 50A), (8:00:04, 103V, 51A), (8:00:06, 103V, 51A), (8:00:08, 102V, 50A), (8:00:10, 98V, 48A), (8:00:12, 99V, 49A), (8:00:14, 101V, 51A), (8:00:16, 99V, 49A), (8:00:18, 97V, 47A), (8:00:20, 100V, 50A)}. Taking the first electrical parameter node as an example: the electrical parameter node (8:00:02, 99V, 51A) is extracted from the electrical parameter node sequence, which corresponds to... The corresponding reference electrical parameter node is (8:00:04, 96V, 49A). The absolute difference between the monitored voltage 99V in the electrical parameter node and the monitored voltage 103V in the reference electrical parameter node is calculated to be 4V. Taking 4V as the voltage fluctuation difference, if the preset voltage fluctuation threshold is 2.5V, the voltage fluctuation difference of 4V is greater than the voltage fluctuation threshold of 2.5V. The monitored voltage 103V corresponding to the electrical parameter node (8:00:04, 103V, 51A) is the fluctuating voltage. Thus, the electrical parameter nodes corresponding to all fluctuating voltages are (8:00:04, 103V, 51A), (8:00:10, 98V, 48A), and (8:00:20, 100V, 50A), with a corresponding number of fluctuating voltages of 3 and a total number of electrical parameter nodes of 10. Generally speaking, the voltage fluctuation threshold can be referenced to the voltage tolerance value. Typically, setting the voltage fluctuation threshold to between 40% and 60% of the voltage tolerance value can more accurately capture anomalies caused by voltage fluctuations. In the example, the voltage tolerance value is 5V, so the voltage fluctuation threshold can be set between 2V and 3V.

[0146] It should be understood that the first evaluation value is obtained using the electrical parameter node sequence, the number of fluctuating voltages, electrical evaluation nodes, weighted nodes, and a pre-constructed evaluation value calculation formula. This first evaluation value is a quantified value used to assess whether the power semiconductor has reached a Level 1 warning state based on voltage monitoring. It is compared with a first warning threshold to determine the Level 1 warning state. The first warning threshold is a benchmark quantified value used for judging the warning state based on voltage monitoring. The process of obtaining the first warning threshold can be referred to as follows: When the power semiconductor is in normal operating condition, multiple actual voltage values ​​are collected to obtain the electrical node sequence corresponding to these values. The evaluation value under normal operating condition can be obtained using the aforementioned evaluation value calculation formula. Multiple collections yield multiple evaluation values, and the average of these values ​​is used as the first warning threshold. Similarly, the second evaluation value is a quantified value used to assess whether the power semiconductor has reached a Level 1 warning state based on current monitoring. It is compared with a second warning threshold to determine the Level 1 warning state. The second warning threshold is a benchmark quantified value used for judging the warning state based on current monitoring. The processes for obtaining the second evaluation value and the second warning threshold are the same as those for obtaining the first evaluation value and the first warning threshold, and will not be repeated here. The frequency assessment ratio is a quantitative ratio used to assess whether a power semiconductor has reached a first-level warning state based on the frequency monitoring angle. The frequency assessment ratio threshold is a benchmark ratio used for comparison with the frequency assessment ratio.

[0147] Understandably, the process of obtaining the first evaluation value is as follows: The evaluation value calculation formula obtains the first evaluation value by using three indicators: absolute voltage deviation, relative voltage fluctuation, and voltage fluctuation frequency, and assigning corresponding weights to each indicator. Among these, absolute voltage deviation... Used to evaluate the monitoring voltage and standard voltage V corresponding to all electrical parameter nodes. std Overall deviation, relative voltage fluctuation Used to characterize the instantaneous voltage change rate and voltage fluctuation frequency of the monitoring voltage corresponding to the nodes of the detected electrical parameters. This is used to reflect the intensity of abnormal monitoring voltage fluctuations, and the weights of the three indicators are obtained through dynamic weighting.

[0148] Furthermore, the step of obtaining weighted nodes based on the number of fluctuating voltages, the total number of electrical parameter nodes, and preset initial weighted nodes includes:

[0149] Calculate the ratio of the number of fluctuating voltages to the total number of electrical parameter nodes to obtain the voltage ratio;

[0150] Compare the voltage ratio with the preset ratio threshold;

[0151] If the voltage ratio is less than or equal to the ratio threshold, the initial weight node is used as the weight node; otherwise, the updated first weight, updated second weight, and updated third weight are obtained based on the number of fluctuating voltages, the total number of electrical parameter nodes, and the pre-constructed updated weight calculation formula. The updated weight calculation formula is as follows:

[0152]

[0153] a1′=1-a3′-a2′

[0154] Where a1, a2, and a3 represent the initial first weight, initial second weight, and initial third weight, respectively, and a1′, a2′, and a3′ represent the updated first weight, updated second weight, and updated third weight, respectively. min This indicates the preset minimum weight;

[0155] Compare and update the first weight with the minimum weight;

[0156] If the updated first weight is greater than or equal to the minimum weight, then the updated first weight, the updated second weight, and the updated third weight are respectively used as the first weight, the second weight, and the third weight. Otherwise, the minimum weight is used as the updated first weight, and the first weight, the second weight, and the third weight are obtained by using the updated first weight, the updated second weight, the updated third weight, and the preset normalization method.

[0157] By summing the first weight, the second weight, and the third weight, we obtain the weight nodes.

[0158] For example, assuming the ratio threshold is 1 / 2, if the ratio of the number of fluctuating voltages to the total number of electrical parameter nodes (i.e., the voltage ratio) is 3 / 10, and the initial weight nodes are 0.25, 0.45, and 0.3 respectively, the minimum weight a min The initial weight is 0.1. At this point, the voltage ratio 3 / 10 is less than the ratio threshold 1 / 2. Using the initial weight node as the weight node, the first, second, and third weights are 0.25, 0.45, and 0.3 respectively. If the voltage ratio is 6 / 10, which is greater than the voltage ratio 3 / 10, then the updated first, second, and third weights need to be obtained using the updated weight calculation formula. The process is as follows: First, calculate the updated third weight a1′ as 0.45, which is greater than or equal to 0.1 (a min Then calculate the updated second weight as 0.15, which is greater than or equal to 0.1 (a). min The final updated first weight is 0.4, therefore the updated first weight is greater than or equal to the minimum weight of 0.1 (a). minTherefore, the first weight is updated to 0.4, the second weight to 0.15, and the third weight to 0.45, respectively. Assume the updated first weight is less than the minimum weight of 0.1 (a). min Considering that a3′+a2′ may be greater than 1, the obtained updated first weight is negative. The updated first weight, updated second weight, and updated third weight are normalized to [0.1] using the Min-Max normalization method. These normalized updated first weight, updated second weight, and updated third weight are then used as the updated first weight, updated second weight, and updated third weight, respectively. This embodiment of the invention uses a dynamic weighting method to obtain the weights corresponding to the three indicators used to calculate the first evaluation value. When the voltage fluctuation corresponds to a large voltage ratio, the relevant weight of the voltage fluctuation frequency indicator is dynamically increased to highlight the impact of key indicators and promptly capture potential fault risks.

[0159] S5. After confirming that the first-level monitoring status is the first-level warning status, the temperature monitoring module set is obtained based on the temperature monitoring unit, and the monitoring temperature set and temperature change rate set are obtained using the temperature monitoring module set. The temperature monitoring module set contains multiple temperature monitoring modules, and the monitoring temperature set contains multiple monitoring temperature values. The temperature monitoring module and the monitoring temperature value correspond one-to-one.

[0160] It should be explained that the acquisition of the monitored temperature set and temperature change rate set using the temperature monitoring module set includes:

[0161] Perform the following operations on each temperature monitoring module in the temperature monitoring module cluster:

[0162] An initial monitoring temperature set is obtained using a temperature monitoring module, a preset temperature monitoring period, and a preset number of temperature monitoring times. The initial monitoring temperature set contains multiple initial monitoring temperature values.

[0163] The initial monitoring temperature sets are aggregated to obtain multiple initial monitoring temperature sets, and each initial monitoring temperature set corresponds one-to-one with a temperature monitoring module;

[0164] For each of the multiple initial monitoring temperature sets, the following operation is performed:

[0165] A monitoring temperature value acquisition scheme is constructed based on the initial monitoring temperature set, as shown below:

[0166]

[0167] Among them, T avg Indicates the monitored temperature value, T i T represents the i-th initial monitoring temperature value in the initial monitoring temperature set. akThis represents the preset k-th initial temperature anomaly value, α represents the preset temperature threshold, n represents the total number of initial monitoring temperature values ​​in the initial monitoring temperature set, and m represents the total number of preset initial temperature anomaly values.

[0168] The monitored temperature value is obtained based on the monitored temperature value acquisition scheme;

[0169] After confirming that each temperature monitoring module has obtained the corresponding monitored temperature value, the monitored temperature values ​​are summarized to obtain the monitored temperature set;

[0170] A set of temperature change rates is obtained based on multiple initial monitoring temperature sets.

[0171] It should be understood that the temperature monitoring unit consists of multiple temperature monitoring modules used to collect temperature values ​​near the power semiconductor. This temperature value is the initial monitoring temperature value, which is obtained directly from the temperature monitoring modules and reflects the external temperature of the power semiconductor under actual operating conditions. Optionally, the DS18B20 smart temperature sensor can be used as the temperature monitoring module to achieve the acquisition of the initial monitoring temperature value.

[0172] For example, assuming the temperature monitoring unit consists of 5 temperature monitoring modules, taking one temperature monitoring module as an example: if the temperature monitoring period is 10 minutes and the temperature monitoring times are 10, the temperature detection module collects the temperature every minute during this monitoring period to obtain 10 initial monitoring temperature values. These 10 initial monitoring temperature values ​​form the initial monitoring temperature set corresponding to the temperature detection module. If the preset temperature threshold is 2℃ and the initial monitoring temperature set is {30℃, 32℃, 31℃, 32℃, 30℃, 35℃, 31℃, 32℃, 32℃, 30℃}, a monitoring temperature value acquisition scheme is constructed based on the initial monitoring temperature set. The implementation process of the monitoring temperature value acquisition scheme is as follows: calculate all initial values ​​in the initial monitoring temperature set... The average temperature of the monitored values ​​is 31.5℃. Initial monitored temperature values ​​deviating more than 2℃ from the average of 31.5℃ are considered initial temperature anomalies, i.e., initial monitored temperature values ​​not belonging to the temperature range [29.5℃, 33.5℃]. In this example, the initial monitored temperature value of 35℃ does not belong to the temperature range [29.5℃, 33.5℃], therefore it is removed from the initial monitored temperature set. The average of the remaining initial monitored temperature values ​​after removing 35℃ is taken as the monitored temperature value corresponding to the temperature monitoring module. After confirming that all five temperature monitoring modules have obtained their corresponding monitored temperature values, the five monitored temperature values ​​are summarized to obtain the monitored temperature set. This invention utilizes a monitored temperature value acquisition scheme to identify and remove initial abnormal temperature values, ensuring that the calculated monitored temperature values ​​are closer to the actual situation, thus improving the accuracy and reliability of temperature monitoring.

[0173] In detail, the acquisition of the temperature change rate set based on multiple initial monitored temperature sets includes:

[0174] For each of the multiple initial monitoring temperature sets, the following steps are performed:

[0175] The initial monitoring temperature values ​​in the initial monitoring temperature set are sorted in chronological order according to the time corresponding to the acquisition of the initial monitoring temperature values ​​to obtain the initial monitoring temperature sequence.

[0176] The temperature change rate is obtained based on the initial monitored temperature sequence and a pre-constructed formula for calculating the temperature change rate, as shown in the following formula:

[0177]

[0178] Where ΔT represents the rate of temperature change, T g+1 T represents the (g+1)th initial monitoring temperature value in the initial monitoring temperature sequence. g Δt′ represents the g-th initial monitoring temperature value in the initial monitoring temperature sequence, and Δt′ represents the time interval between two adjacent initial monitoring temperature values ​​in the initial monitoring temperature sequence, and the time interval is related to the number of temperature monitoring.

[0179] By summarizing the temperature change rates, a set of temperature change rates is obtained.

[0180] It is understood that the temperature change rate is a manually set threshold for measuring the degree of change in the temperature value of the power semiconductor accessory during the temperature monitoring period. The temperature change rate calculation formula calculates the change between adjacent initial monitoring temperature values ​​in the initial monitoring temperature sequence and combines the time interval to obtain the comprehensive temperature change rate, which comprehensively considers the absolute amount of temperature change, the rate of temperature change, and the relative magnitude of temperature change.

[0181] For example, Δt′ is the time interval between two adjacent initial monitoring temperature values ​​in the initial monitoring temperature sequence. Assume that the temperature monitoring period is from 8:10:00 to 8:20:00 on a certain morning, and the temperature is monitored 10 times, with the temperature collected once every 1 minute. That is, the initial monitoring temperature values ​​are collected at 8:11:00, 8:12:00, 8:13:00, 8:14:00, 8:15:00, 8:16:00, 8:17:00, 8:18:00, 8:19:00, and 8:20:00 respectively. Therefore, the time interval Δt′ between two adjacent initial monitoring temperature values ​​in the corresponding initial monitoring temperature sequence is 1 minute.

[0182] S6. Obtain the comprehensive monitoring temperature value based on the monitoring temperature set and the pre-constructed temperature calculation formula.

[0183] It should be explained that the temperature calculation formula is as follows:

[0184]

[0185] Among them, T b This represents the total monitored temperature value, where l indicates that the monitored temperature set contains l monitored temperature values, T j-avg w represents the j-th monitored temperature value in the monitored temperature set. j R represents the weight of the j-th monitored temperature value. θjc p represents the thermal resistance from the power semiconductor chip to the casing, and p represents the power consumption of the power semiconductor.

[0186] Understandably, the comprehensive monitoring temperature value is a temperature index calculated by comprehensively considering the monitoring temperature values ​​collected by multiple temperature monitoring modules and combining them with the thermal characteristics and power consumption of the power semiconductor. It is used to measure the temperature of the heat-generating points inside the power semiconductor chip. Thermal characteristic parameter R θjc This describes the thermal resistance of a power semiconductor from the junction (chip) to the casing. Power dissipation (p) is the portion of the electrical energy consumed by the power semiconductor during operation that is converted into heat. Power dissipation causes the internal temperature of the power semiconductor to rise. The thermal characteristic parameter R... θjc Both power consumption (p) and thermal characteristic parameter R are key parameters for calculating the temperature rise from the junction to the casing of a power semiconductor. Generally, thermal characteristic parameter R... θjc The power consumption p can be obtained from the power semiconductor's datasheet. To more accurately obtain the power consumption p, high-precision current and voltage sensors can be used to measure the power semiconductor's operating current and voltage in real time. The power consumption p is the product of the operating current and voltage. Monitoring temperature, thermal characteristics, and power consumption are all necessary parameters for calculating the comprehensive monitoring temperature value. The difference lies in that the monitoring temperature value provides the temperature of the power semiconductor's casing, while the thermal characteristics and power consumption are used to calculate the temperature rise from the power semiconductor junction to the casing, i.e., the temperature difference between the junction and the casing. The weight of each monitoring temperature value represents the importance of the corresponding temperature monitoring module among all temperature monitoring modules. This weight can be set based on the contribution of each temperature monitoring module in the monitoring process. Generally, modern power semiconductor modules (such as automotive-grade IGBTs) usually have pre-reserved casing temperature measurement points for sensor installation and temperature acquisition. This invention uses the monitoring temperature value to provide the external temperature of the power semiconductor and uses thermal characteristics and power consumption to calculate the internal junction temperature of the power semiconductor. By comprehensively considering both the external and internal junction temperatures of the power semiconductor, the accuracy of temperature monitoring is improved.

[0187] S7. Based on the comprehensive monitoring temperature value and temperature change rate set, confirm the secondary monitoring status of the power semiconductor to realize fault early warning of the power semiconductor, wherein the secondary monitoring status is a secondary early warning status or other secondary status.

[0188] It should be explained that the secondary monitoring status confirmation of power semiconductors based on comprehensive monitoring of temperature values ​​and temperature change rate sets includes:

[0189] For each temperature change rate in the temperature change rate set, perform the following operation:

[0190] The absolute temperature change rate is obtained based on the temperature change rate, wherein the absolute temperature change rate is the absolute value of the temperature change rate.

[0191] Compare the absolute temperature change rate with a preset temperature change rate threshold;

[0192] If the absolute temperature change rate is greater than the temperature change rate threshold, the absolute temperature change rate is considered an unqualified absolute temperature change rate; otherwise, the absolute temperature change rate is considered a qualified absolute temperature change rate.

[0193] The non-compliant absolute temperature change rate and the compliant absolute temperature change rate are summarized separately to obtain the non-compliant absolute temperature change rate set and the compliant absolute temperature change rate set.

[0194] The number of non-compliant absolute temperature change rates in the non-compliant absolute temperature change rate set and the number of compliant absolute temperature change rates in the compliant absolute temperature change rate set are counted separately to obtain the number of non-compliant and compliant items.

[0195] Determine whether the comprehensive monitoring temperature value, the number of non-compliant items, and the number of compliant items meet the preset third verification condition, wherein the third verification condition is as follows:

[0196] |T b -T′|>∈,

[0197] Where T′ represents the preset comprehensive temperature benchmark value, ∈ represents the preset comprehensive temperature difference threshold, Q1 represents the number of non-compliant items, Q2 represents the number of compliant items, and σ represents the preset monitoring ratio;

[0198] When the third verification condition is met, the secondary monitoring status of the power semiconductor is confirmed as the secondary warning status;

[0199] Otherwise, the secondary monitoring status of the power semiconductor will be confirmed as another secondary status.

[0200] Understandably, the secondary monitoring state refers to the monitoring state determined after confirming the primary monitoring state as a primary warning state, through further evaluation of the power semiconductor's state based on the comprehensive monitoring temperature value and temperature change rate set. The difference between the secondary and primary monitoring states is that the secondary monitoring state confirms the power semiconductor's state by evaluating its thermal condition, while the primary monitoring state confirms its state by judging whether the power semiconductor's electrical and frequency parameters are abnormal. The secondary warning state refers to the state in which the power semiconductor, during secondary monitoring, has both its comprehensive monitoring temperature value and temperature change rate exceeding the safe range but not yet reaching the fault range. Other secondary states refer to the power semiconductor's states during secondary monitoring other than the secondary warning state.

[0201] It should be understood that the temperature change rate threshold is a manually set benchmark value used to evaluate the temperature change rate. For example, assuming the temperature monitoring module collects the temperature every minute during the monitoring period, the corresponding temperature change rate set obtained from the five temperature monitoring modules is {-0.45, -0.51, -0.49, -0.46, -0.53}. Here, we only take -0.45 as an example: where -0.45 represents an average temperature decrease of 0.45℃ per minute. -0.51, -0.49, -0.46, and -0.53 achieve the same effect as -0.45, and will not be elaborated further here. The absolute temperature change rate represents the absolute value of the temperature change rate. Based on the temperature change rate set {-0.45, -0.51, -0.49, -0.46, -0.53}, the absolute temperature change rates corresponding to each temperature change rate are 0.45, 0.51, 0.49, 0.46, and 0.53, respectively. Assuming the temperature change rate threshold is 0.50, temperature change rates with an absolute temperature change rate greater than the threshold of 0.50 are considered unqualified absolute temperature change rates, while those with a lower threshold are considered qualified absolute temperature change rates. From the example, the unqualified absolute temperature change rates are 0.51 and 0.53, and the qualified absolute temperature change rates are 0.45, 0.49, and 0.46. Therefore, the number of unqualified values ​​Q1 is 2, and the number of qualified values ​​Q2 is 3. The comprehensive temperature reference value and the comprehensive temperature difference threshold are used to evaluate the operating temperature of the power semiconductor. These can be set according to the power semiconductor datasheet. When the difference between the comprehensive temperature monitoring value and the comprehensive temperature reference value exceeds the comprehensive temperature difference threshold, When the monitoring ratio is greater than the threshold, the second verification condition is met, indicating that the secondary monitoring status of the power semiconductor is in a secondary warning state, requiring confirmation of the power semiconductor. This invention assesses the thermal state of the power semiconductor based on a comprehensive set of monitored temperature values ​​and temperature change rates. By accurately determining whether the power semiconductor is in a secondary warning state through the second verification condition, potential faults can be detected in a timely manner, improving the reliability and safety of power semiconductor operation.

[0202] To address the problems described in the background art, this invention receives a fault warning command from a power semiconductor and establishes a fault warning system based on the command. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module. Electrical evaluation nodes are obtained based on the power semiconductor, including standard voltage, standard current, a voltage warning deviation range, and a current warning deviation range. This invention uses the allowable deviation value and the fault deviation value of the electrical parameters as the minimum and maximum values ​​of the electrical parameter warning deviation range, respectively, and uses the range between the allowable deviation and the fault deviation as the warning range. When the detected deviation of electrical parameter values ​​exceeds the allowable range but has not yet reached the fault deviation value, the power semiconductor can be adjusted in a timely manner to ensure its stable operation. Based on a preset electrical monitoring period, a preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. This set includes multiple electrical parameter nodes, each containing monitored voltage and current. A frequency evaluation ratio is obtained using a low-frequency monitoring module and a high-frequency monitoring module. Based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes, the primary monitoring status of the power semiconductor is confirmed. This primary monitoring status can be a primary warning status or other primary status. Therefore, this invention utilizes a low-frequency monitoring module (steady-state analysis) and... The high-frequency monitoring module (transient analysis) uses alternating sampling to comprehensively consider both the steady-state frequency mean and transient frequency bursts. This overcomes the blind spot and missed detection limitations caused by the fixed duty cycle phase in traditional single-sampling modes, improving the reliability and accuracy of early warning. A dynamic weighting method is used to obtain the weights corresponding to the three indicators used to calculate the first evaluation value. When the voltage fluctuation corresponds to a large voltage proportion, the relevant weight of the voltage fluctuation frequency indicator is dynamically increased to highlight the impact of key indicators and promptly capture potential fault risks. Once the first-level monitoring state is confirmed as a first-level early warning state, a temperature monitoring module set is obtained based on the temperature monitoring unit. This temperature monitoring module set is then used to obtain the monitored temperature set and temperature change rate set. The temperature monitoring module set contains multiple temperature monitoring modules. The monitored temperature set contains multiple monitored temperature values, and each temperature monitoring module corresponds one-to-one with a monitored temperature value. This invention utilizes the monitored temperature value acquisition scheme to identify and eliminate initial abnormal temperature values, ensuring that the calculated monitored temperature values ​​are closer to the actual situation, thus improving the accuracy and reliability of temperature monitoring. The monitored temperature values ​​provide the external temperature of the power semiconductor, while thermal characteristic parameters and power consumption are used to calculate the internal junction temperature of the power semiconductor. By comprehensively considering both the external and internal junction temperatures of the power semiconductor, the accuracy of temperature monitoring is improved. A comprehensive monitored temperature value is obtained based on the monitored temperature set and a pre-constructed temperature calculation formula. The secondary monitoring status of the power semiconductor is confirmed based on the comprehensive monitored temperature value and a set of temperature change rates, achieving fault early warning for the power semiconductor.The secondary monitoring state refers to a secondary warning state or other secondary states. This invention assesses the thermal state of power semiconductors based on a comprehensive set of monitored temperature values ​​and temperature change rates. By using a second verification condition, it accurately determines whether the power semiconductor is in a secondary warning state, enabling timely detection of potential faults and improving the reliability and safety of power semiconductor operation. Therefore, this invention can improve the accuracy and reliability of power semiconductor fault early warning.

[0203] like Figure 2 The diagram shown is a functional block diagram of a data-driven power semiconductor fault early warning system provided in an embodiment of the present invention.

[0204] The data-driven power semiconductor fault early warning system 100 described in this invention can be installed in electronic devices. Depending on the functions implemented, the data-driven power semiconductor fault early warning system 100 may include an electrical parameter acquisition module 101, a primary monitoring and early warning module 102, a comprehensive temperature acquisition module 103, and a secondary monitoring and early warning module 104. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device.

[0205] The electrical parameter acquisition module 101 is used to receive fault warning commands from power semiconductors and confirm the fault warning system based on the fault warning commands. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module.

[0206] Electrical evaluation nodes are obtained based on power semiconductors, wherein the electrical evaluation nodes include standard voltage, standard current, voltage warning deviation range, and current warning deviation range;

[0207] Based on the preset electrical monitoring period, the preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. The electrical parameter node set includes multiple electrical parameter nodes, and each electrical parameter node includes the monitoring voltage and the monitoring current.

[0208] The first-level monitoring and early warning module 102 is used to obtain the frequency evaluation ratio using the low-frequency monitoring module and the high-frequency monitoring module, and to confirm the first-level monitoring status of the power semiconductor based on the electrical parameter node set, the frequency evaluation ratio and the electrical evaluation node, wherein the first-level monitoring status is a first-level early warning status or other first-level status.

[0209] The integrated temperature acquisition module 103 is used to acquire a set of temperature monitoring modules based on the temperature monitoring unit after confirming that the first-level monitoring state is a first-level warning state, and to acquire a set of monitoring temperatures and a set of temperature change rates using the set of temperature monitoring modules. The set of temperature monitoring modules contains multiple temperature monitoring modules, and the set of monitoring temperatures contains multiple monitoring temperature values. The temperature monitoring modules and the monitoring temperature values ​​correspond one-to-one.

[0210] The comprehensive monitoring temperature value is obtained based on the monitoring temperature set and the pre-constructed temperature calculation formula;

[0211] The secondary monitoring and early warning module 104 is used to confirm the secondary monitoring status of the power semiconductor based on the comprehensive monitoring temperature value and temperature change rate set, so as to realize the fault early warning of the power semiconductor. The secondary monitoring status is a secondary early warning status or other secondary status.

[0212] In detail, the modules in the data-driven power semiconductor fault early warning system 100 described in this embodiment of the invention employ the same methods as described above. Figure 1 The method uses the same technical means as the data-driven power semiconductor fault early warning method described in the article and can produce the same technical effect, so it will not be repeated here.

[0213] like Figure 3 The diagram shown is a schematic representation of an electronic device that implements a data-driven power semiconductor fault early warning method according to an embodiment of the present invention.

[0214] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a data-driven power semiconductor fault early warning method program.

[0215] The memory 11 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as a portable hard drive. In other embodiments, the memory 11 can be an external storage device of the electronic device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device 1. Furthermore, the memory 11 includes both internal storage units and external storage devices of the electronic device 1. The memory 11 can be used not only to store application software and various types of data installed on the electronic device 1, such as the code of a data-driven power semiconductor fault warning method program, but also to temporarily store data that has been output or will be output.

[0216] In some embodiments, the processor 10 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control unit of the electronic device, connecting various components of the entire electronic device through various interfaces and lines. It executes programs or modules stored in the memory 11 (e.g., a data-driven power semiconductor fault warning method program) and calls data stored in the memory 11 to perform various functions of the electronic device 1 and process data.

[0217] The bus 12 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 can be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to realize the connection and communication between the memory 11 and at least one processor 10, etc.

[0218] Figure 3 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 3The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0219] For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) to power various components. Preferably, the power supply can be logically connected to the at least one processor 10 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0220] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the electronic device 1 and other electronic devices.

[0221] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired or wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device 1 and to display a visual user interface.

[0222] The data-driven power semiconductor fault early warning method program stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When run in the processor 10, it can achieve the following:

[0223] The system receives a fault warning command from a power semiconductor and confirms a fault warning system based on the fault warning command. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module.

[0224] Electrical evaluation nodes are obtained based on power semiconductors, wherein the electrical evaluation nodes include standard voltage, standard current, voltage warning deviation range, and current warning deviation range;

[0225] Based on the preset electrical monitoring period, the preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. The electrical parameter node set includes multiple electrical parameter nodes, and each electrical parameter node includes the monitoring voltage and the monitoring current.

[0226] The frequency evaluation ratio is obtained by using a low-frequency monitoring module and a high-frequency monitoring module. Based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes, the first-level monitoring status of the power semiconductor is confirmed. The first-level monitoring status is a first-level early warning status or other first-level status.

[0227] Once the Level 1 monitoring status is confirmed as Level 1 warning status, the temperature monitoring module set is obtained based on the temperature monitoring unit. The monitoring temperature set and the temperature change rate set are obtained using the temperature monitoring module set. The temperature monitoring module set contains multiple temperature monitoring modules, and the monitoring temperature set contains multiple monitoring temperature values. The temperature monitoring module and the monitoring temperature value correspond one-to-one.

[0228] The comprehensive monitoring temperature value is obtained based on the monitoring temperature set and the pre-constructed temperature calculation formula;

[0229] The secondary monitoring status of power semiconductors is confirmed based on the comprehensive monitoring temperature value and temperature change rate set, thereby realizing fault early warning of power semiconductors. The secondary monitoring status is a secondary early warning status or other secondary status.

[0230] Specifically, the processor 10's implementation method for the above instructions can be found in [reference needed]. Figures 1 to 3 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0231] Furthermore, if the modules / units integrated in the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0232] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following:

[0233] The system receives a fault warning command from a power semiconductor and confirms a fault warning system based on the fault warning command. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module.

[0234] Electrical evaluation nodes are obtained based on power semiconductors, wherein the electrical evaluation nodes include standard voltage, standard current, voltage warning deviation range, and current warning deviation range;

[0235] Based on the preset electrical monitoring period, the preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. The electrical parameter node set includes multiple electrical parameter nodes, and each electrical parameter node includes the monitoring voltage and the monitoring current.

[0236] The frequency evaluation ratio is obtained by using a low-frequency monitoring module and a high-frequency monitoring module. Based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes, the first-level monitoring status of the power semiconductor is confirmed. The first-level monitoring status is a first-level early warning status or other first-level status.

[0237] Once the Level 1 monitoring status is confirmed as Level 1 warning status, the temperature monitoring module set is obtained based on the temperature monitoring unit. The monitoring temperature set and the temperature change rate set are obtained using the temperature monitoring module set. The temperature monitoring module set contains multiple temperature monitoring modules, and the monitoring temperature set contains multiple monitoring temperature values. The temperature monitoring module and the monitoring temperature value correspond one-to-one.

[0238] The comprehensive monitoring temperature value is obtained based on the monitoring temperature set and the pre-constructed temperature calculation formula;

[0239] The secondary monitoring status of power semiconductors is confirmed based on the comprehensive monitoring temperature value and temperature change rate set, thereby realizing fault early warning of power semiconductors. The secondary monitoring status is a secondary early warning status or other secondary status.

[0240] In the embodiments provided by this invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and actual implementations may have other classification methods.

[0241] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0242] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0243] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0244] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A data-driven power semiconductor fault early warning method, characterized in that, The method includes: The system receives a fault warning command from a power semiconductor and confirms a fault warning system based on the fault warning command. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module. Electrical evaluation nodes are obtained based on power semiconductors, wherein the electrical evaluation nodes include standard voltage, standard current, voltage warning deviation range, and current warning deviation range; Based on the preset electrical monitoring period, the preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. The electrical parameter node set includes multiple electrical parameter nodes, and each electrical parameter node includes the monitoring voltage and the monitoring current. The frequency evaluation ratio is obtained by using a low-frequency monitoring module and a high-frequency monitoring module. Based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes, the first-level monitoring status of the power semiconductor is confirmed. The first-level monitoring status is a first-level early warning status or other first-level status. The method of obtaining the frequency evaluation ratio using the low-frequency monitoring module and the high-frequency monitoring module includes: Obtain the frequency monitoring period, and then obtain multiple frequency monitoring sub-periods based on the frequency monitoring period; The frequency monitoring sub-periods are sorted in order from front to back according to the time corresponding to the frequency monitoring sub-periods to obtain the frequency monitoring sub-period sequence; Multiple odd frequency monitoring sub-segments and multiple even frequency monitoring sub-segments are obtained based on the frequency monitoring sub-segment sequence. Among them, the odd frequency monitoring sub-segments are the frequency monitoring sub-segments with an odd position in the frequency monitoring sub-segment sequence, and the even frequency monitoring sub-segments are the frequency monitoring sub-segments with an even position in the frequency monitoring sub-segment sequence. The first frequency value set is obtained based on multiple odd frequency monitoring sub-periods, a preset first number of monitoring times, and a low frequency monitoring module. By summing up the first frequency value sets, multiple first frequency value sets are obtained; The second frequency value set is obtained based on multiple even-frequency monitoring sub-periods, a preset second monitoring number, and a high-frequency monitoring module. By summing up the second frequency value sets, multiple second frequency value sets are obtained; The frequency evaluation ratio is obtained based on multiple first frequency value sets, multiple second frequency value sets, and a preset method. Once the Level 1 monitoring status is confirmed as Level 1 warning status, the temperature monitoring module set is obtained based on the temperature monitoring unit. The monitoring temperature set and the temperature change rate set are obtained using the temperature monitoring module set. The temperature monitoring module set contains multiple temperature monitoring modules, and the monitoring temperature set contains multiple monitoring temperature values. The temperature monitoring module and the monitoring temperature value correspond one-to-one. The comprehensive monitoring temperature value is obtained based on the monitoring temperature set and the pre-constructed temperature calculation formula; The secondary monitoring status of power semiconductors is confirmed based on the comprehensive monitoring temperature value and temperature change rate set, thereby realizing fault early warning of power semiconductors. The secondary monitoring status is a secondary early warning status or other secondary status.

2. The data-driven power semiconductor fault early warning method as described in claim 1, characterized in that, The electrical evaluation node based on power semiconductors includes: Obtain the standard voltage and standard current of power semiconductors; Obtain the permissible voltage deviation and voltage fault deviation values ​​of the power semiconductor; The voltage warning deviation range is obtained based on the voltage allowable deviation value and the voltage fault deviation value, wherein the minimum and maximum values ​​corresponding to the voltage warning deviation range are the voltage allowable deviation value and the voltage fault deviation value, respectively. Obtain the allowable current deviation and current fault deviation of the power semiconductor; The current warning deviation range is obtained based on the allowable current deviation value and the current fault deviation value. The minimum and maximum values ​​corresponding to the current warning deviation range are the allowable current deviation value and the current fault deviation value, respectively. By associating the standard voltage, standard current, voltage warning deviation range, and current warning deviation range, electrical evaluation nodes are obtained.

3. The data-driven power semiconductor fault early warning method as described in claim 2, characterized in that, The first-level monitoring status of power semiconductors, confirmed based on electrical parameter node set, frequency evaluation ratio, and electrical evaluation nodes, includes: The electrical parameter nodes in the electrical parameter node set are sorted in chronological order according to the time corresponding to each electrical parameter node, resulting in an electrical parameter node sequence. Electrical parameter nodes are extracted sequentially from the electrical parameter node sequence. Based on the extracted electrical parameter nodes, reference electrical parameter nodes are identified in the electrical parameter node sequence. The reference electrical parameter nodes are adjacent to and lag behind the extracted electrical parameter nodes. Calculate the absolute difference between the monitored voltage at the electrical parameter node and the monitored voltage at the reference electrical parameter node, and use the absolute difference as the voltage fluctuation difference; Compare the voltage fluctuation difference with the preset voltage fluctuation threshold; If the voltage fluctuation difference is greater than the voltage fluctuation threshold, the monitoring voltage corresponding to the reference electrical parameter node is confirmed as the fluctuating voltage. By summarizing the fluctuating voltages, a fluctuating voltage set is obtained. The number of fluctuating voltages in the fluctuating voltage set is counted to obtain the number of fluctuating voltages. The total number of electrical parameter nodes is obtained by counting the number of electrical parameter nodes in the electrical parameter node sequence. Weight nodes are obtained based on the number of fluctuating voltages, the total number of electrical parameter nodes, and preset initial weight nodes. The initial weight nodes include: initial first weight, initial second weight, and initial third weight, and the weight nodes include: first weight, second weight, and third weight. The first evaluation value is obtained based on the electrical parameter node sequence, the number of fluctuating voltages, electrical evaluation nodes, weight nodes, and pre-constructed evaluation value calculation formula; The second evaluation value is obtained based on the electrical parameter node set and electrical evaluation nodes; Compare the first evaluation value with the preset first warning threshold, the second evaluation value with the preset second warning threshold, and the frequency evaluation ratio with the preset frequency evaluation ratio threshold, respectively; If the first evaluation value is greater than the first warning threshold, or the second evaluation value is greater than the second warning threshold, or the frequency evaluation ratio is greater than the frequency evaluation ratio threshold, the first-level monitoring status of the power semiconductor is confirmed as a first-level warning status. Otherwise, the primary monitoring status of the power semiconductor will be confirmed as another primary status.

4. The data-driven power semiconductor fault early warning method as described in claim 3, characterized in that, The process of obtaining weighted nodes based on the number of fluctuating voltages, the total number of electrical parameter nodes, and preset initial weighted nodes includes: Calculate the ratio of the number of fluctuating voltages to the total number of electrical parameter nodes to obtain the voltage ratio; Compare the voltage ratio with the preset ratio threshold; If the voltage ratio is less than or equal to the ratio threshold, the initial weight node is used as the weight node; otherwise, the updated first weight, updated second weight, and updated third weight are obtained based on the number of fluctuating voltages, the total number of electrical parameter nodes, and the pre-constructed updated weight calculation formula. Compare and update the first weight with the minimum weight; If the updated first weight is greater than or equal to the minimum weight, then the updated first weight, the updated second weight, and the updated third weight are respectively used as the first weight, the second weight, and the third weight. Otherwise, the minimum weight is used as the updated first weight, and the first weight, the second weight, and the third weight are obtained by using the updated first weight, the updated second weight, the updated third weight, and the preset normalization method. By summing the first weight, the second weight, and the third weight, we obtain the weight nodes.

5. The data-driven power semiconductor fault early warning method as described in claim 4, characterized in that, The method of acquiring the monitored temperature set and temperature change rate set using the temperature monitoring module set includes: Perform the following operations on each temperature monitoring module in the temperature monitoring module cluster: An initial monitoring temperature set is obtained using a temperature monitoring module, a preset temperature monitoring period, and a preset number of temperature monitoring times. The initial monitoring temperature set contains multiple initial monitoring temperature values. The initial monitoring temperature sets are aggregated to obtain multiple initial monitoring temperature sets, and each initial monitoring temperature set corresponds one-to-one with a temperature monitoring module; For each of the multiple initial monitoring temperature sets, the following operation is performed: A monitoring temperature value acquisition scheme is constructed based on the initial monitoring temperature set, as shown below: ； ； in, Indicates the monitored temperature value. Indicates the initial monitoring temperature set. An initial monitored temperature value, Indicates the preset first An initial temperature anomaly value, This indicates the preset temperature threshold. Indicates that the initial monitored temperature is collectively shared. An initial monitored temperature value, Indicates shared ownership A preset initial temperature anomaly value; The monitored temperature value is obtained based on the monitored temperature value acquisition scheme; After confirming that each temperature monitoring module has obtained the corresponding monitored temperature value, the monitored temperature values ​​are summarized to obtain the monitored temperature set; A set of temperature change rates is obtained based on multiple initial monitoring temperature sets.

6. The data-driven power semiconductor fault early warning method as described in claim 5, characterized in that, The method of obtaining a set of temperature change rates based on multiple initial monitoring temperature sets includes: For each of the multiple initial monitoring temperature sets, the following steps are performed: The initial monitoring temperature values ​​in the initial monitoring temperature set are sorted in chronological order according to the time corresponding to the acquisition of the initial monitoring temperature values ​​to obtain the initial monitoring temperature sequence. The temperature change rate is obtained based on the initial monitored temperature sequence and the pre-constructed temperature change rate calculation formula; By summarizing the temperature change rates, a set of temperature change rates is obtained.

7. The data-driven power semiconductor fault early warning method as described in claim 6, characterized in that, The temperature calculation formula is as follows: ； in, This indicates the comprehensive monitoring temperature value. Indicates that the temperature monitoring set has a total Each monitored temperature value, Indicates the number of temperature monitoring centers Each monitored temperature value, Indicates the first The weight of each monitored temperature value This indicates the thermal resistance from the power semiconductor chip to the casing. This indicates the power consumption of the power semiconductor.

8. The data-driven power semiconductor fault early warning method as described in claim 7, characterized in that, The secondary monitoring status of the power semiconductor, based on the comprehensive monitoring of temperature values ​​and temperature change rate sets, includes: For each temperature change rate in the temperature change rate set, perform the following operation: The absolute temperature change rate is obtained based on the temperature change rate, wherein the absolute temperature change rate is the absolute value of the temperature change rate. Compare the absolute temperature change rate with a preset temperature change rate threshold; If the absolute temperature change rate is greater than the temperature change rate threshold, the absolute temperature change rate is considered an unqualified absolute temperature change rate; otherwise, the absolute temperature change rate is considered a qualified absolute temperature change rate. The non-compliant absolute temperature change rate and the compliant absolute temperature change rate are summarized separately to obtain the non-compliant absolute temperature change rate set and the compliant absolute temperature change rate set. The number of non-compliant absolute temperature change rates in the non-compliant absolute temperature change rate set and the number of compliant absolute temperature change rates in the compliant absolute temperature change rate set are counted separately to obtain the number of non-compliant and compliant items. Determine whether the comprehensive monitoring temperature value, the number of non-compliant items, and the number of compliant items meet the preset third verification condition; When the third verification condition is met, the secondary monitoring status of the power semiconductor is confirmed as the secondary warning status; Otherwise, the secondary monitoring status of the power semiconductor will be confirmed as another secondary status.

9. A data-driven power semiconductor fault early warning system, characterized in that, The system includes: An electrical parameter acquisition module is used to receive fault warning commands from power semiconductors and confirm the fault warning system based on the fault warning commands. The fault warning system includes an electrical parameter monitoring unit, a frequency monitoring unit, and a temperature monitoring unit. The frequency monitoring unit includes a low-frequency monitoring module and a high-frequency monitoring module. Electrical evaluation nodes are obtained based on power semiconductors, wherein the electrical evaluation nodes include standard voltage, standard current, voltage warning deviation range, and current warning deviation range; Based on the preset electrical monitoring period, the preset electrical monitoring time interval, and the electrical parameter monitoring unit, an electrical parameter node set is obtained. The electrical parameter node set includes multiple electrical parameter nodes, and each electrical parameter node includes the monitoring voltage and the monitoring current. A primary monitoring and early warning module is used to obtain the frequency evaluation ratio using a low-frequency monitoring module and a high-frequency monitoring module, and to confirm the primary monitoring status of the power semiconductor based on the electrical parameter node set, the frequency evaluation ratio, and the electrical evaluation nodes. The primary monitoring status is a primary early warning status or other primary status. The method of obtaining the frequency evaluation ratio using the low-frequency monitoring module and the high-frequency monitoring module includes: Obtain the frequency monitoring period, and then obtain multiple frequency monitoring sub-periods based on the frequency monitoring period; The frequency monitoring sub-periods are sorted in order from front to back according to the time corresponding to the frequency monitoring sub-periods to obtain the frequency monitoring sub-period sequence; Multiple odd frequency monitoring sub-segments and multiple even frequency monitoring sub-segments are obtained based on the frequency monitoring sub-segment sequence. Among them, the odd frequency monitoring sub-segments are the frequency monitoring sub-segments with an odd position in the frequency monitoring sub-segment sequence, and the even frequency monitoring sub-segments are the frequency monitoring sub-segments with an even position in the frequency monitoring sub-segment sequence. The first frequency value set is obtained based on multiple odd frequency monitoring sub-periods, a preset first number of monitoring times, and a low frequency monitoring module. By summing up the first frequency value sets, multiple first frequency value sets are obtained; The second frequency value set is obtained based on multiple even-frequency monitoring sub-periods, a preset second monitoring number, and a high-frequency monitoring module. By summing up the second frequency value sets, multiple second frequency value sets are obtained; The frequency evaluation ratio is obtained based on multiple first frequency value sets, multiple second frequency value sets, and a preset method. The integrated temperature acquisition module is used to acquire a set of temperature monitoring modules based on the temperature monitoring unit after confirming that the first-level monitoring status is a first-level warning status. It then uses the set of temperature monitoring modules to acquire a set of monitored temperatures and a set of temperature change rates. The set of temperature monitoring modules contains multiple temperature monitoring modules, and the set of monitored temperatures contains multiple monitored temperature values. Each temperature monitoring module corresponds to a monitored temperature value. The comprehensive monitoring temperature value is obtained based on the monitoring temperature set and the pre-constructed temperature calculation formula; The secondary monitoring and early warning module is used to confirm the secondary monitoring status of the power semiconductor based on the comprehensive monitoring temperature value and temperature change rate set, so as to realize the fault early warning of the power semiconductor. The secondary monitoring status is a secondary early warning status or other secondary status.

Images