Fault prediction methods for motor controllers and vehicles
By acquiring vehicle operating condition data, quantifying power cycle and electrochemical corrosion damage respectively, and combining comprehensive damage analysis, the problem of predicting the accuracy of crater corrosion faults in motor controllers was solved, and accurate prediction of motor controller faults was achieved.
Patent Information
- Application Number
- CN202611244256.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-17
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies struggle to accurately predict crater corrosion faults in IGBT or SiC power electronic semiconductor modules within the motor controllers of new energy vehicles. Traditional diagnostic methods fail to detect fault risks in the early stages of corrosion, leading to misjudgments.
By acquiring vehicle operating condition data, cumulative damage from power cycles and cumulative damage from electrochemical corrosion are determined separately. Combining the comprehensive cumulative damage, the impact of power cycle thermal stress and electrochemical effects on crater corrosion is analyzed, enabling accurate prediction of motor controller failures.
It improves the accuracy of predicting crater corrosion faults in motor controllers and can deeply integrate the failure principles of power semiconductor modules to reduce misjudgments.
Smart Images

Figure CN123113801A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of intelligent connected vehicle technology, and in particular to a fault prediction method for a motor controller and a vehicle. Background Technology
[0002] One of the main causes of failure in power electronic semiconductor modules such as IGBT (Insulated Gate Bipolar Transistor) or SiC (Silicon Carbide) in the motor controller of new energy vehicles is "crater corrosion," which is the formation of macroscopically visible crater-like corrosion pits in the aluminum metal layer, eventually leading to the detachment of bonding wires or short circuits.
[0003] Traditional diagnostic methods rely on manual after-sales inspection and post-failure repair, which cannot detect potential failures in the early stages of corrosion. The limited number of online monitoring solutions often rely on weak correlations such as the product of temperature and humidity or the rate of temperature change for prediction, which are insufficient to reflect the true causes of crater corrosion failure and are highly prone to misdiagnosis. Summary of the Invention
[0004] In view of this, the purpose of this application is to propose a fault prediction method for motor controllers and a vehicle, thereby improving the accuracy of predicting crater corrosion faults in the power semiconductor modules of motor controllers.
[0005] To achieve the above objectives, this application provides a fault prediction method for a motor controller, comprising:
[0006] Acquire vehicle operating condition data; Based on the aforementioned operating condition data, the cumulative power cycle damage and cumulative electrochemical corrosion damage of the motor controller in the vehicle are determined. Based on the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion, the overall cumulative damage to the motor controller is determined, and the failure probability of the motor controller is determined based on the overall cumulative damage.
[0007] For the same purpose, this application also provides a vehicle, the vehicle comprising: Memory, used to store executable program code; A processor is configured to call and run the executable program code from the memory, causing the vehicle to execute the fault prediction method of the motor controller provided in any embodiment of this application.
[0008] As can be seen from the above, the fault prediction method for the motor controller provided in this application first acquires the vehicle's operating condition data, determines the cumulative power cycle damage and cumulative electrochemical corrosion damage of the motor controller through the operating condition data, then determines the comprehensive cumulative damage through the cumulative power cycle damage and cumulative electrochemical corrosion damage, and determines the fault probability of the motor controller based on the comprehensive cumulative damage, thereby achieving fault prediction for crater corrosion of the motor controller. This method, by determining the cumulative power cycle damage, can analyze the impact of power cycle thermal stress on crater corrosion, and by determining the cumulative electrochemical corrosion damage, can analyze the impact of condensation film and electrochemical action under DC bias on crater corrosion. Furthermore, by determining the comprehensive cumulative damage through the cumulative power cycle damage and cumulative electrochemical corrosion damage, it measures the probability of crater corrosion failure of the motor controller. It can simultaneously consider the impact of power cycle thermal stress and electrochemical action on crater corrosion, deeply integrates the failure principle of crater corrosion of power semiconductor modules, and ensures the accuracy of crater corrosion fault prediction. Attached Figure Description
[0009] To more clearly illustrate the technical solutions in this application or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0010] Figure 1 This is a flowchart of a fault prediction method for a motor controller provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a fault prediction device for a motor controller provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation
[0011] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.
[0012] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this application should have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms "first," "second," and similar terms used in the embodiments of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are only used to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.
[0013] Before providing a detailed explanation of the fault prediction method for motor controllers provided in this application, the technical problem solved by this method will be described first. In existing online detection solutions, the product of temperature and humidity or the rate of temperature change is typically used to quantify the corrosion of the power semiconductor modules in the motor controller.
[0014] However, condensation typically occurs when the temperature is below the dew point, and the product of temperature and humidity tends to point towards high-temperature and high-humidity scenarios. Therefore, the direction of the temperature-humidity product is opposite to the actual environmental conditions under which corrosion occurs, making it difficult to reflect the true corrosion situation and easily leading to misjudgments of corrosion faults. In addition, the correlation between the rate of temperature change and the corrosion of power semiconductor modules is low, failing to accurately match the actual physical conditions of crater corrosion, and easily leading to misjudgments of corrosion faults.
[0015] Therefore, in order to solve the above problems, this application provides a fault prediction method for a motor controller and a vehicle. By determining the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion respectively, the influence of power cycling thermal stress and electrochemical action on crater corrosion can be analyzed separately. Furthermore, by obtaining the comprehensive cumulative damage from the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion, the probability of crater corrosion failure of the motor controller can be measured. This method can deeply combine the combined effect of the two, accurately match the failure principle of crater corrosion of power semiconductor modules, and improve the prediction accuracy of crater corrosion failure of power semiconductor modules.
[0016] Figure 1 This is a flowchart illustrating a fault prediction method for a motor controller according to an embodiment of this application. This method can be executed by a fault prediction device for the motor controller, which can be integrated into an electronic device, for example, within the motor controller, via hardware and / or software. Figure 1 As shown, the method provided in this application embodiment includes the following steps: S110. Obtain vehicle operating condition data.
[0017] The operating condition data can be data related to the motor controller collected in real time during vehicle operation. For example, the operating condition data can consist of multi-source data over a period of driving time, such as the internal relative humidity of the motor controller, the internal ambient temperature of the motor controller, the controller housing temperature (i.e., the controller base plate temperature), the DC bus voltage, the junction temperature of the power semiconductor module, and the RMS value of the three-phase current. In addition, the operating condition data can also include the overall vehicle status, cumulative usage time (i.e., cumulative operating time), and cumulative mileage.
[0018] It should be noted that since each signal in the operating condition data comes from a different sensor, the corresponding sampling period can be different. For example, the sampling period for internal relative humidity and internal ambient temperature is 1 second, the sampling period for controller case temperature is 100 ms, the sampling period for junction temperature of power semiconductor module is 10 ms, the sampling period for DC bus voltage is 100 ms, and the sampling period for RMS value of three-phase current is 100 ms.
[0019] Therefore, after acquiring the operating condition data, each signal in the operating condition data can be aligned to a unified time grid, that is, all signals can be unified to a reference sampling period (e.g., 10ms), ensuring the timing synchronization of multi-source data and providing high-quality input for subsequent damage calculations. For example, high-frequency signals (such as controller case temperature, DC bus voltage, and three-phase current RMS value) can be directly retained according to the reference sampling period, while low-frequency signals (internal relative humidity, internal ambient temperature) can be extrapolated according to zero-order hold.
[0020] In this embodiment, considering that the junction temperature of the power semiconductor module cannot be directly measured for some vehicle models, the real-time total power loss of the power semiconductor module can be determined by first measuring the module conduction loss power and the module switching loss power, as shown in the following formula: ; In the formula, For the first The real-time total power loss under each reference sampling period is expressed in W. For the first The module conduction loss power under each reference sampling period, in W, can be obtained by fitting the effective value of the three-phase current. For the first The module switching loss power under a reference sampling period, in W, can be determined by the effective value of the three-phase current, the DC bus voltage, and the switching frequency. and The calculation can be referenced using the following formula: , ; In the formula, For the first The effective value of the three-phase current under each reference sampling period, in A; The switching frequency is expressed in Hz. For the first DC bus voltage under each reference sampling period, in V; , and All of these are fitting constants.
[0021] Furthermore, the junction temperature of the power semiconductor module can be estimated using the controller case temperature and the real-time total power loss of the power semiconductor module, as shown in the following formula: ; In the formula, For the first Junction temperature of the power semiconductor module under one reference sampling period, in °C; For the first The controller housing temperature under each reference sampling period, in °C; Junction-to-shell thermal resistance, in °C / W; For the first The real-time total power loss over a reference sampling period is expressed in W.
[0022] Furthermore, for each baseline sampling period in the operating condition data, a corresponding operating condition label can be added. For example, if the vehicle status is Ready and the vehicle speed is greater than 0, it is marked as driving; if the vehicle status is DC charging or AC charging, it is marked as charging; if the vehicle status is off and the off time does not exceed the set time (e.g., 4 hours), it is marked as hot-soaking off condition (KEY_OFF_HOT_SOAK); if the vehicle status is off and the off time is greater than the set time, it is marked as cold-soaking off condition (KEY_OFF_COLD_SOAK).
[0023] In this embodiment, to further ensure the accuracy of subsequent fault prediction, anomaly filtering can be performed on the operating condition data. Specifically, for each signal in the operating condition data, a sliding window (e.g., 60s) can be set, and the median of the signal within the sliding window can be calculated, as shown in the following formula: ; In the formula, The median, It is a sliding window. Indicates the first Move forward by one reference sampling period The subsequent time point, From Time point up to the Signal within the time range of each reference sampling period The sequence formed This indicates taking the median; Furthermore, the median absolute deviation is calculated based on the median, which describes the dispersion of the signal. The robust standard deviation is then calculated using the median absolute deviation, as shown in the following formula: , ; In the formula, This represents the absolute deviation of the median. For robust standard deviation, The correction coefficient is adapted to the normal distribution. During the cold start period (within 60 seconds of system power-on), the initial robust standard deviation can be used to avoid misjudgment during cold start.
[0024] Furthermore, a signal anomaly threshold can be calculated using the robust standard deviation (e.g., taking 4 times the robust standard deviation), and the absolute value of the difference between the signal and the corresponding median can be determined. If the difference exceeds the anomaly threshold, the signal is identified as an anomaly, and the median is used to replace it. Additionally, if the anomaly rate of the signal exceeds 20% within 60 seconds, the signal can be marked as SENSOR_SUSPECT (sensor anomaly), and all calculations dependent on this signal in the current baseline sampling period can be skipped.
[0025] S120. Based on operating condition data, determine the cumulative power cycle damage and cumulative electrochemical corrosion damage of the motor controller in the vehicle.
[0026] In this embodiment of the application, considering that both power cycle thermal stress and electrochemical effects will cause crater corrosion to the power semiconductor module of the motor controller, in order to quantify the corrosion effects caused by power cycle thermal stress and electrochemical effects respectively, the cumulative damage of power cycle and the cumulative damage of electrochemical corrosion can be determined by operating condition data.
[0027] The cumulative damage from power cycling can be defined as the cumulative damage caused by the periodic thermal stress on the power semiconductor module during the data acquisition period corresponding to the operating conditions. Specifically, during power cycling, the power semiconductor module heats up or cools down rapidly. During this period, the thermal expansion coefficients of the aluminum metal layer, silicon chip, and ceramic substrate differ greatly, repeatedly generating periodic tensile and compressive thermal stress. This thermal stress can damage the passivation oxide film on the aluminum surface. The oxide film is the natural protective layer of aluminum. The continuous cyclic stress causes microcracks, and condensation water films penetrate into the aluminum substrate along the cracks, which can significantly accelerate the rate of corrosion pitting. The already formed crater corrosion pits become stress concentration points. Each power cycle tears the edge of the pit, causing the pit to continue to expand, forming a positive feedback loop of "corrosion-stress damage to the protective layer-intensified corrosion".
[0028] For example, the number of power cycles covered by the operating condition data can be determined, and then the cumulative power cycle damage can be determined based on the number of power cycles and the predicted damage under a pre-set number of single cycles.
[0029] In this embodiment of the application, considering that the actual damage generated under different thermal cycles may differ, in order to further improve the prediction accuracy of cumulative damage from power cycles, the operating condition data can be split into multiple thermal cycles, and the damage under each thermal cycle can be calculated and then accumulated.
[0030] In one specific implementation, the operating condition data includes the junction temperature sequence of the power semiconductor module in the motor controller. Determining the cumulative power cycle damage of the motor controller in the vehicle based on the operating condition data includes the following steps: Step A1: Divide the junction temperature sequence into multiple thermal cycle subsequences to obtain the junction temperature swing and average absolute junction temperature of each thermal cycle subsequence; Step A2: For each thermal cycling subsequence, determine the number of module failure cycles corresponding to the thermal cycling subsequence based on the junction temperature swing and average absolute junction temperature of the thermal cycling subsequence. The number of module failure cycles is the number of cycles in which the power semiconductor module experiences thermal fatigue failure while operating in the thermal cycling subsequence. Step A3: Determine the cumulative power cycle damage of the motor controller based on the number of module failure cycles corresponding to each thermal cycle subsequence.
[0031] Specifically, in step A1, the junction temperatures of the power semiconductor module collected at various time points in the operating condition data can constitute a junction temperature sequence of the power semiconductor module. The junction temperature sequence can be decomposed by incremental online rainflow counting method to obtain multiple independent thermal cycle subsequences, so as to break down the continuously fluctuating junction temperature sequence into standardized and calculable temperature cycle units; thereby determining the junction temperature swing and average absolute junction temperature of each thermal cycle subsequence.
[0032] Furthermore, in step A2, for each thermal cycling subsequence, the junction temperature swing and average absolute junction temperature of the thermal cycling subsequence can be used to calculate the corresponding module failure cycle count. This module failure cycle count can be understood as the total allowable number of cycles for the power semiconductor module to undergo thermal fatigue failure under the conditions of that thermal cycling subsequence. As shown in the following formula: ; In the formula, Let be the module failure cycle number corresponding to the i-th thermal cycle subsequence, which is a dimensionless parameter; The junction temperature swing of the i-th thermal cycle subsequence is expressed in K. , Let be the damage fitting constant. The unit is Used to offset Dimensions , It can be obtained through experimental calibration, such as Can be taken , The value can be 5.039; The activation energy, measured in J, can be obtained through experimental calibration, such as... Can be taken ; Boltzmann's constant, in units of J / K, such as Can be taken ; The average absolute junction temperature of the i-th thermal cycle subsequence is expressed in K.
[0033] As can be seen from the above formula, for each thermal cycle subsequence, the larger the junction temperature swing under that thermal cycle subsequence, the smaller the calculated number of module failure cycles. This indicates that the larger the temperature difference, the more severe the fatigue in a single cycle, and the fewer the total allowable number of cycles the power semiconductor module can withstand. In addition, the higher the average absolute junction temperature under that thermal cycle subsequence, the smaller the calculated number of module failure cycles. This indicates that high temperature environments can accelerate the thermal fatigue aging of power semiconductor modules.
[0034] After obtaining the number of module failure cycles corresponding to each thermal cycle subsequence, further, in step A3, the predicted thermal cycle loss under the thermal cycle subsequence can be determined based on the number of module failure cycles. Then, the predicted thermal cycle losses under all thermal cycle subsequences are accumulated to obtain the cumulative power cycle damage corresponding to the operating condition data. As shown in the following formula: ; In the formula, The cumulative power cycle damage corresponding to the operating condition data is represented by a value in the range [0,1]. The data collection time range corresponding to the operating conditions. This is the index of the thermal cycle subsequence. The formula can be understood as: the energy consumed in each thermal cycle. The lifespan is determined by the linear superposition of losses under all thermal cycles.
[0035] Steps A1-A3 above, by dividing the junction temperature sequence into multiple thermal cycle subsequences, determine the corresponding module failure cycle number based on the junction temperature swing and average absolute junction temperature of each thermal cycle subsequence. Then, by combining the module failure cycle number under all thermal cycle subsequences, the cumulative power cycle damage corresponding to the entire operating condition data can be determined. This can accurately quantify the cumulative mechanical fatigue damage caused by power cycle thermal stress, corresponding to the periodic thermal stress in the cause of crater corrosion. Furthermore, by predicting the module failure cycle number under each thermal cycle subsequence, the accumulation of thermal fatigue caused by different temperature differences and temperatures can be accurately quantified, improving the accuracy of cumulative power cycle damage.
[0036] In this embodiment, the cumulative damage from electrochemical corrosion can be the cumulative damage caused by the coupling electrochemical effect between the DC bias voltage of the power semiconductor module and the condensate film within the acquisition time range corresponding to the operating condition data.
[0037] For example, the number of condensation events in the operating condition data can be identified, and the cumulative electrochemical corrosion damage can be determined based on the number of condensation events and the pre-set predicted damage under a single condensation event.
[0038] In this embodiment, considering that condensation film is a necessary medium for electrochemical corrosion, in addition to condensation film, being charged is a key condition for accelerated corrosion. Under the conditions of vehicle shutdown hot immersion, driving, and charging, there is a DC bias voltage on the bus, and an electric field can be formed at both ends of the water film, which accelerates the dissolution of aluminum metal anodic craters and rapidly corrodes them. Furthermore, temperature can also affect the degree of corrosion acceleration; the higher the temperature, the faster the electrochemical corrosion. In addition, electrochemical corrosion has a cumulative effect over time. As time goes on, the depth and area of the aluminum layer corrosion pits will continue to expand, eventually leading to bond wire detachment or short circuit. Therefore, the cumulative damage from electrochemical corrosion can be predicted by combining the damage caused by charged condensation, voltage, and temperature.
[0039] In some implementations, determining the cumulative electrochemical corrosion damage to the motor controller in a vehicle based on operating condition data includes the following steps: Step B1: For each reference sampling period in the operating condition data, based on the operating data under the reference sampling period in the operating condition data, determine the charged condensation damage, temperature influence factor and voltage influence factor under the reference sampling period. Step B2: Determine the electrochemical corrosion rate under the baseline sampling period based on the temperature influence factor, voltage influence factor, and charged condensation damage. Step B3: Determine the cumulative electrochemical corrosion damage of the motor controller based on the electrochemical corrosion rate under each benchmark sampling period.
[0040] Among them, charged condensation damage can be corrosion damage to the power semiconductor module when it is under high voltage bias and a condensation film exists. For example, in step B1, it can be determined whether the power semiconductor module is under high voltage bias and a condensation film exists by using the operating data under the reference sampling period. If so, charged condensation damage is determined according to the preset damage value.
[0041] In this embodiment, considering the significant difference in damage caused by short-term and long-term condensation, the cumulative duration of condensation can be introduced to accurately quantify the damage caused by charged condensation. The cumulative duration of condensation directly characterizes the effective window for electrochemical corrosion. Anodic dissolution corrosion of aluminum metal will only occur when there is a condensation film on the power semiconductor module and the bus has a DC bias voltage. The degree of damage accumulates linearly with the duration of charged immersion. Therefore, the cumulative duration of condensation can determine the condensation damage under the quantitative benchmark sampling period, ensuring the accuracy of electrochemical corrosion prediction.
[0042] In one specific implementation, the operating data includes the internal relative humidity of the motor controller, the internal ambient temperature, the controller case temperature, and the DC bus voltage. Based on the operating data under a reference sampling period in the operating condition data, the charged condensation damage under the reference sampling period is determined, including: Step B11: Determine the dew point temperature under the reference sampling period based on the internal relative humidity and internal ambient temperature, and determine the condensation state indicator variable of the motor controller under the reference sampling period based on the dew point temperature and the controller shell temperature. Step B12: Determine the live condensation status indication variable of the motor controller under the reference sampling period based on the DC bus voltage and condensation status indication variable; Step B13: Determine the cumulative duration of charged condensation under the reference sampling period based on the charged condensation state indicator variable, and determine the charged condensation damage under the reference sampling period based on the cumulative duration of charged condensation.
[0043] In step B11, for each reference sampling period, the dew point temperature under the reference sampling period can be calculated first using the internal relative humidity and internal ambient temperature under the reference sampling period, as shown in the following formula: ; ; In the formula, These are intermediate calculation variables, and they are dimensionless parameters. This represents the internal relative humidity, with a value range of 0 to 100 (representing 0% to 100%). , For fixed coefficients, Dimensionless The unit is °C. , It can be obtained through experimental calibration, such as Pick , Pick °C; This refers to the internal ambient temperature, expressed in °C. This refers to the dew point temperature, expressed in °C. It should be noted that this calculation method, under conditions where the internal ambient temperature is between -45 °C and 60 °C, yields a dew point temperature deviation of no more than 0.4 °C, covering typical vehicle driving scenarios.
[0044] After calculating the dew point temperature, the difference between the controller housing temperature and the dew point temperature can be calculated to obtain the housing-side condensation margin. This housing-side condensation margin is then compared with a preset condensation margin. If the housing-side condensation margin is less than or equal to the preset condensation margin, the condensation state indicator variable under the reference sampling period is determined to be a condensation indicator (e.g., 1); otherwise, it is a no-condensation indicator (e.g., 0), as shown in the following formula: ; ; In the formula, For the first The shell-side condensation margin under a reference sampling period, in °C; For the first The controller housing temperature under each reference sampling period, in °C; For the first Dew point temperature under a reference sampling period, in °C; The preset condensation margin is in °C, and the default value can be 1°C (to avoid critical fluctuations). For the first A condensation state indicator variable under a reference sampling period is used to describe whether a condensation film exists in the power semiconductor module.
[0045] After obtaining the condensation state indicator variable, further, in step B12, it can be determined whether the DC bus voltage under the reference sampling period is not less than the preset bus voltage. If the DC bus voltage is not less than the preset bus voltage, and the condensation state indicator variable is a condensation indication quantity, then the live condensation state indicator variable under the reference sampling period is determined to be a live condensation indication quantity (e.g., 1); otherwise, there is no live condensation indication quantity (e.g., 0), as shown in the following formula: ; In the formula, For the first DC bus voltage under each reference sampling period, in V; The preset bus voltage (which can be understood as the high voltage power-on judgment threshold) is in V, and the default value can be 60V. For the first A charged condensation state indicator variable under a reference sampling period is used to describe whether a condensation film exists in the power semiconductor module and whether the high voltage bias is satisfied.
[0046] Furthermore, in step B13, considering that the true driving force of electrochemical corrosion is the presence of a condensation film and the satisfaction of high-voltage bias, if the indicator variable for charged condensation state is the charged condensation indicator quantity, then the cycle duration corresponding to the benchmark sampling cycle can be determined as the cumulative duration of charged condensation. The cumulative duration of charged condensation can then be used to measure the charged condensation damage under the benchmark sampling cycle. This charged condensation damage can characterize the risk of charged condensation exposure before the superimposed temperature and voltage acceleration. As shown in the following formula: ; In the formula, For the first The charged condensation damage under each reference sampling period is a dimensionless parameter with a value range of [0,1]. For the first The cumulative duration of charged condensation under each baseline sampling period, in hours; The preset exposure life of charged condensation, in hours, can be obtained by calibration using crater corrosion failure samples. If no calibration data is available, the default value can be 500 hours.
[0047] Steps B11-B13 above take into account that the degree of electrochemical corrosion damage accumulates linearly with the duration of immersion in electricity. By determining the indicator variable of the state of condensation in electricity, the cumulative duration of condensation in electricity is obtained. Based on the cumulative duration of condensation in electricity, the damage caused by condensation in electricity is determined. This can accurately quantify the risk of exposure to condensation in electricity without considering the effects of temperature and voltage under each benchmark sampling period, and ensure the accuracy of electrochemical corrosion prediction.
[0048] Considering the coupling of condensation duration, DC bias, and temperature in the occurrence of electrochemical corrosion, in step B1, for each reference sampling period, in addition to determining the charged condensation damage, the temperature influence factor and voltage influence factor can also be determined. The temperature influence factor reflects the degree of influence of the controller case temperature on electrochemical corrosion, while the voltage influence factor reflects the degree of influence of the DC bus voltage on electrochemical corrosion.
[0049] For example, the temperature influence factor can be calculated using the following formula: ; In the formula, For the first The temperature influence factor under each reference sampling period is a dimensionless parameter. The preset low activation energy is expressed in J / mol, such as 30000 J / mol; For the first The controller housing temperature under each reference sampling period, in °C, is compared with... After adding, the units are converted to K; This is the gas constant, with units of J / ( ), such as taking 8.314 J / ( Considering that crater corrosion is essentially an electrochemical reaction involving an aluminum metallization layer or bonding lines within a water film, and that this reaction rate accelerates with increasing absolute temperature, the formula uses the Arrhenius form to describe the temperature-accelerated effect, with the denominator containing... It represents the product of the gas constant and the absolute temperature. It needs to be converted from °C to K. Located at the molecular position, it represents the equivalent activation energy of the corrosion reaction. The higher the activation energy, the lower the corrosion rate at the same temperature, and the more sensitive it is to temperature.
[0050] The voltage influence factor can be calculated using the following formula: ; In the formula, For the first The voltage influence factor under each reference sampling period is a dimensionless parameter. For the first DC bus voltage under each reference sampling period, in V; The reference bus voltage is in volts (V), e.g., 400V. The preset bias index is used to control the nonlinear effect of DC bus bias on the corrosion rate. It can be set to 1.0 by default, or obtained by fitting experimental data. Considering that the DC bus voltage provides the bias driving force required for electrochemical migration and localized corrosion, the higher the DC bus bias, the faster the corrosion rate. Therefore, in this formula, a bias index of 1.0 is used. Describing voltage acceleration terms, compared to directly using This allows the voltage influence factor to become a dimensionless parameter. A typical bus voltage of 400 V can be used as the normalization reference. and When they are equal, the voltage acceleration term is 1.
[0051] After calculating the temperature and voltage influence factors, the electrochemical corrosion rate under the baseline sampling period can be calculated by combining these factors with the charged condensation damage. Before this, the charged condensation damage can be adjusted to obtain the charged condensation damage enhancement value, which can then be used to determine the electrochemical corrosion rate by combining the temperature, voltage, and charged condensation damage enhancement values. The result is shown in the following formula: ; In the formula, For the first The enhanced value of charged condensation damage under each reference sampling period is a dimensionless parameter. This is the sensitivity coefficient for the cumulative duration of charged condensation, such as 1.0; The pre-set exposure life for charged condensation, in hours; For the first The cumulative duration of charged condensation under a reference sampling period, in hours, indicates that prolonged and repeated condensation can damage the oxide film on the surface of the aluminum metallization layer or bonding area, leading to the accumulation of ion residues and micro-corrosion defects, thereby increasing corrosion susceptibility under subsequent conditions of the same temperature, voltage, and condensation. In this formula, the charged condensation damage enhancement value is adopted as... A consistent saturation index form can make the effect of condensation change with... The rate increases with the increase of [something], and tends to saturate after long-term exposure, avoiding linear amplification that leads to an unbounded increase in the corrosion rate. When the value equals 0, the damage enhancement value of the charged condensation is 1. Far greater than At that time, the damage enhancement value of the charged condensation tends to be close to .
[0052] Specifically, in step B2, the electrochemical corrosion rate is determined using the temperature influence factor, voltage influence factor, and the enhanced value of charged condensation damage, as shown in the following formula: ; In the formula, For the first Electrochemical corrosion rate under a reference sampling period; This is the calibration coefficient, such as 1.0; For the first The charged condensation state indicator variable under each reference sampling period, since crater corrosion requires the simultaneous presence of a condensation film and a high-pressure bias, therefore, As a gating factor, if no condensation occurs or the high voltage is not energized during the current reference sampling period... Electrochemical corrosion damage is not accumulated under this baseline sampling period. In this formula, the temperature influence factor, voltage influence factor, charged condensation damage enhancement value, and gating factor correspond to four types of factors: reaction activation, bias driving, historical water film exposure cumulative effect, and electrolyte channel, respectively. Condensation absence directly inhibits corrosion; therefore, a multiplicative form is used to express the "common driving" relationship. The multiplicative structure used in this formula also facilitates component calibration; for example, it can be fitted using accelerated tests at different temperatures. Fitting through tests at different DC bus voltage levels Fitting was achieved through experiments with different pre-condensation exposure times. By reverse engineering through failed samples .
[0053] For example, multiple sets of isothermal and humidity bias tests can be designed to control the controller case temperature, internal relative humidity, DC bus voltage, and energizing time, and to periodically detect the corrosion area, corrosion depth, or equivalent failure criteria of the aluminum metallization layer or bonding area. Arrhenius linear relationships can be fitted under different temperature groups to obtain... Initial values; fitting electrochemical corrosion rate with different bias voltage groups The power-law relationship is obtained. Initial values were set; different pre-condensation exposure durations were established, and the saturation enhancement relationship of electrochemical corrosion rate with increasing cumulative charging condensation time was fitted to obtain... The initial value; substituting the test samples and post-sales confirmed crater corrosion samples into the above formula, using the total cumulative damage at the time of failure or the maximum likelihood of the failure probability as the objective, is used to back-calculate and correct. For non-failed negative samples, verify the total cumulative damage distribution. If the false alarm rate is too high, make joint adjustments. , , , Or subsequent fusion weights.
[0054] After obtaining the electrochemical corrosion rate for each benchmark sampling period, further, in step B3, the electrochemical corrosion damage for each benchmark sampling period can be determined based on the electrochemical corrosion rate for each benchmark sampling period. Then, the electrochemical corrosion damage for all benchmark sampling periods is accumulated to obtain the cumulative electrochemical corrosion damage corresponding to the operating condition data. As shown in the following formula: ; In the formula, For cumulative damage from electrochemical corrosion, This indicates the time range for collecting operational data. For the first time within this collection time range Electrochemical corrosion rate under a benchmark sampling period The period length corresponding to the reference sampling period.
[0055] Steps B1-B3 above determine the charged condensation damage, temperature influence factor, and voltage influence factor under each benchmark sampling period. Then, by combining the charged condensation damage, temperature influence factor, and voltage influence factor, the electrochemical corrosion rate is determined. Based on the electrochemical corrosion rate under each benchmark sampling period, the cumulative electrochemical corrosion damage is determined, realizing electrochemical corrosion prediction based on temperature, voltage, and charged condensation state. This can match the principle of electrochemical corrosion and ensure the accuracy of electrochemical corrosion prediction.
[0056] S130. Based on the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion, determine the overall cumulative damage of the motor controller, and determine the failure probability of the motor controller based on the overall cumulative damage.
[0057] Specifically, after obtaining the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion, these two factors can be weighted and fused to obtain a comprehensive cumulative damage. This comprehensive cumulative damage reflects the overall corrosion damage of the power semiconductor module within the data acquisition time range corresponding to its operating conditions. For example, the comprehensive cumulative damage can be calculated using the following formula: ; In the formula, To comprehensively address corrosion damage, This indicates the time range for collecting operational data. For cumulative damage from power cycling, For cumulative damage from electrochemical corrosion, As the weight of thermal fatigue damage, The electrochemical corrosion damage weight is the sum of the thermal fatigue damage weight and the electrochemical corrosion damage weight, which is 1 by default. 0.4 can be taken. 0.6 can be taken.
[0058] In this embodiment of the application, considering that the probability of component failure will increase when the vehicle has been used for a long time, the comprehensive cumulative damage can be adjusted according to the vehicle's usage time in order to take into account the aging acceleration effect of long-term vehicle use.
[0059] In some implementations, after determining the combined cumulative damage of the motor controller based on cumulative damage from power cycling and cumulative damage from electrochemical corrosion, the method further includes: Obtain the vehicle's cumulative usage time and determine the vehicle's age correction factor based on the cumulative usage time; update the comprehensive cumulative damage based on the vehicle age correction factor.
[0060] The cumulative usage time can be the cumulative usage time of the power semiconductor module in the vehicle. If the power semiconductor module has been replaced, the cumulative usage time can be restarted from 0.
[0061] After obtaining the cumulative usage time, a corresponding vehicle age correction factor can be calculated. This correction factor reflects the degree of aging of the power semiconductor module with the vehicle's age; the longer the cumulative usage time, the higher the correction factor. For example, the vehicle age correction factor can be calculated using the following formula: ; In the formula, This is the vehicle age correction factor, which is a dimensionless parameter; The total usage time is in years. This is a vehicle age control coefficient used to control the impact of vehicle age. The unit can be 1 / year, used to offset the accumulated usage time. For example... The value range is [0.05, 0.2], and the default value is 0.1.
[0062] Furthermore, the vehicle age correction factor can be multiplied by the total cumulative damage to update the total cumulative damage, as shown in the following formula: ; In the formula, The corrected total cumulative damage, This represents the overall corrosion damage before correction.
[0063] In the above implementation method, considering that the longer the power semiconductor module is used, the worse the material integrity becomes, and the higher the risk of failure due to relative damage, the vehicle age correction coefficient is determined by combining the cumulative usage time. Then, the comprehensive cumulative damage is corrected by the vehicle age correction coefficient, which makes the comprehensive cumulative damage more consistent with the actual failure mechanism of the module. Furthermore, by updating the comprehensive cumulative damage, this method makes the vehicle age correction performed at the damage layer. Compared with correcting the failure probability, it can avoid the overflow of failure probability and is more in line with the principle that material aging can amplify physical damage.
[0064] After obtaining the comprehensive cumulative damage, it can be mapped to describe the probability of motor controller failure. It should be noted that if the operating data includes all data from the power semiconductor module since its commissioning, the comprehensive cumulative damage reflects all physical damage to the power semiconductor module since its inception, and can be directly converted into a failure probability. However, if the operating data is from a recently collected period, the comprehensive cumulative damage reflects the physical damage generated during that period, and needs to be combined with the historical cumulative damage before probability conversion.
[0065] In one specific implementation, determining the failure probability of the motor controller based on comprehensive cumulative damage includes the following steps: Step C1: Obtain the total cumulative damage of the motor controller and update the total cumulative damage based on the comprehensive cumulative damage; Step C2: Determine the failure probability corresponding to the total cumulative damage through the probability mapping function.
[0066] The total cumulative damage describes the total physical damage to the power semiconductor module from the time it is put into use until the start of data collection on its operating conditions. It can be obtained by superimposing historically determined comprehensive cumulative damage.
[0067] Specifically, in step C1, the currently calculated comprehensive cumulative damage can be superimposed on the total cumulative damage to update the total cumulative damage.
[0068] Furthermore, in step C2, the total cumulative damage can be mapped to a failure probability reflecting the failure risk of the power semiconductor module. Specifically, a probability mapping function for implementing the probability mapping can be pre-constructed. This probability mapping function can adopt the Weibull mapping form and be obtained by fitting a large number of samples labeled as crater corrosion, as shown in the following equation: ; In the formula, This represents the probability of failure. Total cumulative damage; For scale parameters; These are shape parameters used to control the steepness of the curve. The larger the value, the more pronounced the inflection point from a slow increase to a rapid increase in the failure probability. In the early stages, the failure probability rises slowly as the total cumulative damage increases; however, once the total cumulative damage reaches a certain level, the failure probability accelerates. This corresponds to the characteristic of slow incubation in the early stages of crater corrosion, followed by a rapid increase in failure risk after the corrosion pit has formed. The scale parameter and shape parameter can be obtained by fitting failure data; the scale parameter can be set to 1.0 by default, and the shape parameter can be set to 2.5 by default.
[0069] Based on the above probability mapping function, the probability threshold for triggering a fault warning can also be substituted into the above probability mapping function to calculate the damage threshold for triggering a fault warning. Taking a probability threshold of 0.5 as an example, it is shown in the following formula: ; In the formula, The damage threshold that triggers a fault warning.
[0070] Steps C1-C2 above update the total cumulative damage by comprehensively accumulating the cumulative damage, and then convert the total cumulative damage into the failure probability through the probability mapping function. This yields a failure probability that increases with the increase of the total cumulative damage, and conforms to the degradation pattern of slow development in the early stage and accelerated risk increase in the later stage.
[0071] In this embodiment of the application, in order to further improve the accuracy of the failure probability, the failure probability can also be corrected by combining the model prediction results.
[0072] In some implementations, after determining the failure probability of the motor controller based on the comprehensive cumulative damage, the following steps are also included: Step D1: Input the operating condition data into the pre-trained corrosion prediction model to obtain the predicted corrosion probability output by the corrosion prediction model; Step D2: Update the failure probability based on the predicted corrosion probability.
[0073] The corrosion prediction model can be trained using a sample set, which can include positive and negative sample data. Positive sample data can be data collected from vehicles identified by the after-sales system as having crater corrosion, collected at a set time (e.g., 60 days) before the failure occurred. Negative sample data can be data collected from vehicles of the same model series, in the same region, within the same mileage range, and without any related failures.
[0074] Specifically, the sample set can be imbalanced, for example, by collecting positive and negative sample data from the sample set in a 1:5 ratio, and applying a higher loss weight to the positive sample data, in order to avoid the corrosion prediction model from being biased towards predicting fault-free data.
[0075] Furthermore, the corrosion prediction model can be trained using a sample set. This corrosion prediction model can adopt a neural network model architecture. The corrosion prediction model can predict the corresponding corrosion probability based on positive or negative sample data. Then, based on the corrosion probability and preset labels (1 represents confirmed crater corrosion, and 0 represents no crater corrosion), the loss value is calculated. The parameters in the corrosion prediction model are adjusted in reverse by the loss value until the iterative convergence condition is met, and the trained corrosion prediction model is obtained.
[0076] Specifically, in step D1, the current vehicle operating condition data can be input into the trained corrosion prediction model so that the corrosion prediction model can output the predicted corrosion probability.
[0077] Furthermore, in step D2, the predicted corrosion probability output by the corrosion prediction model can be weighted and fused with the failure probability derived from power cycling and electrochemical corrosion to obtain a new failure probability, as shown in the following formula: ; In the formula, The updated failure probability. The probability of failure before the update. The predicted corrosion probability is the output of the corrosion prediction model. This is the fault probability weight, which can be set to 0.6 by default. This is an interval phase function used to constrain the weighted result of the fault probability and the predicted corrosion probability within the range of [0,1] to avoid probability out of bounds.
[0078] In steps D1-D2 above, considering that there may be a large number of interfering factors in actual driving that affect the fault probability determined based on the physical analysis formula, such as environmental differences, user driving habits, and sensor measurement drift, in order to avoid such hidden effects, the fault probability is compensated by a pre-trained corrosion prediction model, which can further improve the accuracy of the fault probability.
[0079] In this embodiment, after determining the probability of a fault, a fault warning can be issued based on the probability. For example, if the probability of a fault exceeds a set threshold, a prompt message is generated and displayed through the vehicle's infotainment system, voice playback, dashboard display, or head-up display.
[0080] In addition to determining the probability of failure, the remaining usable mileage of the power semiconductor module can be estimated by comprehensively accumulating damage, and users can be given early warnings based on this remaining usable mileage.
[0081] In some implementations, after determining the combined cumulative damage to the motor controller based on cumulative damage from power cycling and cumulative damage from electrochemical corrosion, the following steps are also included: Step E1: Determine the damage growth rate of the motor controller based on the comprehensive cumulative damage; Step E2: Determine the estimated remaining mileage of the vehicle based on the rate of damage growth.
[0082] Specifically, in step E1, a linear regression can be performed on all the comprehensive cumulative damage calculated within a preset regression time (e.g., 30 days) to obtain the damage growth rate, which can reflect the growth rate of the crater corrosion damage of the power semiconductor module.
[0083] Furthermore, in step E2, the difference between the total cumulative damage and the damage threshold that triggers the fault warning can be calculated. Then, based on this difference, the damage growth rate, and the average daily mileage within the preset regression time, the estimated remaining mileage is determined, as shown in the following formula: ; In the formula, The remaining distance is estimated in km; The damage threshold that triggers a fault warning. To comprehensively assess cumulative damage, The rate of damage growth; The average daily mileage during the preset regression period is expressed in km.
[0084] In this embodiment of the application, the calculated estimated remaining mileage can also be used to trigger a fault warning. For example, if the estimated remaining mileage is less than a preset mileage threshold, a prompt message can be generated and displayed through the vehicle interface, voice playback, instrument panel display, or head-up display.
[0085] Steps E1-E2 above estimate the damage growth rate of the motor controller by comprehensively accumulating damage, and then determine the estimated remaining mileage of the vehicle based on the damage growth rate. This can predict the remaining mileage that will trigger a fault warning, making it easier to provide timely warnings to users and prevent faults from affecting their travel.
[0086] In this embodiment of the application, in order to ensure the reliability of fault warning, multi-level hierarchical warning can be achieved by combining the estimated remaining mileage, the rate of damage growth, and the probability of failure.
[0087] In one example, the fault prediction method for a motor controller provided in this application embodiment further includes the following steps: Step F1: Determine the warning level based on the estimated remaining mileage, damage growth rate, and failure probability; Step F2: Control the operation of the vehicle according to the warning level.
[0088] Specifically, in step F1, the estimated mileage can be compared with a preset mileage threshold, the damage growth rate can be compared with a preset rate threshold, and the failure probability can be compared with a preset probability threshold. There can be multiple probability thresholds, and different probability thresholds correspond to different warning levels.
[0089] For example, the warning level can be divided into four levels: L0 Normal, L1 Attention, L2 Warning, and L3. If the probability of failure is less than the first probability threshold (e.g., 0.3), and the estimated mileage is less than the preset mileage threshold, and the damage growth rate is less than the preset rate threshold, then the warning level is determined to be L0 Normal; if the probability of failure is not less than the first probability threshold and less than the second probability threshold (e.g., 0.5), or the damage growth rate is greater than the preset rate threshold, then the warning level is determined to be L1 Attention; if the probability of failure is not less than the second probability threshold and less than the third probability threshold (e.g., 0.8), then the warning level is determined to be L2 Warning; if the probability of failure is not less than the third probability threshold, or the estimated mileage is greater than the preset mileage threshold, then the warning level is determined to be L3 Emergency.
[0090] After determining the warning level, in step F2, vehicle control can be performed according to the warning level. The control strategies under different warning levels can be different. The higher the warning level, the greater the promptness of the control strategy.
[0091] Using the previous example, if the warning level is L0 (normal), the operating condition data will be recorded locally on the vehicle; if the warning level is L1 (concern), the operating condition data will be uploaded to the cloud, and a maintenance and inspection reminder will be generated to remind the user to check the power semiconductor module during the next maintenance; if the warning level is L2 (warning), a fault risk warning will be generated, displayed on the dashboard, and a fault notification will be sent to a remote terminal (such as a 4S store); if the warning level is L3 (emergency), the vehicle's driving will be controlled according to the preset power, and a recall reminder will be generated, displayed on the dashboard, and simultaneously sent to the cloud.
[0092] In this embodiment, to avoid false alarms triggered by a single vibration, an anti-vibration mechanism can be set. For example, if the number of times the judgment conditions corresponding to the above-mentioned warning level are met exceeds a preset number within multiple set evaluation periods, the vehicle's operation is controlled according to the warning level. This method can avoid false alarms caused by instantaneous operating condition fluctuations. Furthermore, the downgrading of the warning level can also employ a symmetrical hysteresis mechanism. If, within multiple consecutive set evaluation periods, the difference between the fault probability and the probability threshold corresponding to the next lower warning level is less than a set value (e.g., -0.05), the warning level can be downgraded.
[0093] In this embodiment of the application, while performing fault warning, a report text can also be generated when the warning level is detected to be rising. The report text includes a fault probability curve, the most recent charged condensation event, the rain flow spectrum, and recommended measures.
[0094] Steps F1-F2 above determine the warning level by estimating the remaining mileage, the rate of damage growth, and the probability of failure. Then, the vehicle's driving is controlled according to the warning level, realizing a layered fault warning system. Compared with setting only a single parameter as the trigger condition for the fault warning, this system can ensure the rationality of the warning. Furthermore, through layered warnings, interference to users can be reduced in low-risk scenarios, while ensuring timely activation of safety control in high-risk scenarios, thus balancing the driving experience and the overall vehicle safety.
[0095] The fault prediction method for motor controllers provided in this application first acquires the vehicle's operating condition data, then determines the cumulative power cycle damage and cumulative electrochemical corrosion damage of the motor controller based on the operating condition data, and then determines the comprehensive cumulative damage based on the cumulative power cycle damage and cumulative electrochemical corrosion damage. Based on the comprehensive cumulative damage, the failure probability of the motor controller is determined, thereby achieving fault prediction for crater corrosion of the motor controller. This method, by determining the cumulative power cycle damage, can analyze the impact of power cycle thermal stress on crater corrosion, and by determining the cumulative electrochemical corrosion damage, can analyze the impact of condensation film and electrochemical action under DC bias on crater corrosion. Furthermore, by determining the comprehensive cumulative damage based on the cumulative power cycle damage and cumulative electrochemical corrosion damage, the probability of crater corrosion failure of the motor controller is measured. This method can simultaneously consider the impact of power cycle thermal stress and electrochemical action on crater corrosion, deeply integrates the failure principle of crater corrosion of power semiconductor modules, and ensures the accuracy of crater corrosion failure prediction.
[0096] It should be noted that the method in this embodiment can be executed by a single device, such as a computer or server. The method can also be applied in a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method in this embodiment, and the multiple devices will interact with each other to complete the method described.
[0097] It should be noted that the above description describes some embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0098] Based on the same inventive concept, corresponding to any of the above embodiments, this application also provides a fault prediction device for a motor controller. Figure 2 This is a schematic diagram of the structure of a fault prediction device for a motor controller provided in an embodiment of this application, with reference to... Figure 2 The fault prediction device for the motor controller includes a data acquisition module 210, a damage prediction module 220, and a fault prediction module 230, wherein: The data acquisition module 210 is used to acquire the vehicle's operating condition data; Damage prediction module 220 is used to determine the cumulative power cycle damage and cumulative electrochemical corrosion damage of the motor controller in the vehicle based on operating condition data. The fault prediction module 230 is used to determine the comprehensive cumulative damage of the motor controller based on the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion, and to determine the failure probability of the motor controller based on the comprehensive cumulative damage.
[0099] Based on the above implementation, the operating condition data includes the junction temperature sequence of the power semiconductor module in the motor controller, and the damage prediction module 220 includes a mechanical damage prediction unit, which is used for: The junction temperature sequence is divided into multiple thermal cycle subsequences to obtain the junction temperature swing and average absolute junction temperature of each thermal cycle subsequence. For each thermal cycling subsequence, the number of module failure cycles corresponding to the thermal cycling subsequence is determined based on the junction temperature swing and the average absolute junction temperature of the thermal cycling subsequence. The number of module failure cycles is the number of cycles in which the power semiconductor module experiences thermal fatigue failure while operating in the thermal cycling subsequence. The cumulative power cycle damage of the motor controller is determined based on the number of module failure cycles corresponding to each thermal cycle subsequence.
[0100] Based on the above embodiments, the damage prediction module 220 includes an electrochemical damage prediction unit, which is used for: For each reference sampling period in the operating condition data, based on the operating data under the reference sampling period, determine the charged condensation damage, temperature influence factor, and voltage influence factor under the reference sampling period. The electrochemical corrosion rate under the reference sampling period was determined based on the temperature influence factor, voltage influence factor, and charged condensation damage. The cumulative electrochemical corrosion damage of the motor controller is determined based on the electrochemical corrosion rate under each benchmark sampling period.
[0101] Based on the above implementation method, the operating data includes the internal relative humidity, internal ambient temperature, controller case temperature, and DC bus voltage of the motor controller. The electrochemical damage prediction unit is also used for: The dew point temperature under the reference sampling period is determined based on the internal relative humidity and internal ambient temperature, and the condensation state indicator variable of the motor controller under the reference sampling period is determined based on the dew point temperature and the controller shell temperature. Based on the DC bus voltage and condensation status indication variables, determine the live condensation status indication variables of the motor controller under the reference sampling period; The cumulative duration of charged condensation under the reference sampling period is determined based on the charged condensation state indicator variable, and the charged condensation damage under the reference sampling period is determined based on the cumulative duration of charged condensation.
[0102] Based on the above implementation method, the fault prediction module 230 is further configured to: Obtain the vehicle's cumulative usage time and determine the vehicle's age correction factor based on the cumulative usage time; The overall cumulative damage is updated based on the vehicle age correction factor.
[0103] Based on the above implementation method, the fault prediction module 230 is further configured to: Obtain the total cumulative damage of the motor controller and update the total cumulative damage based on the comprehensive cumulative damage; The failure probability corresponding to the total cumulative damage is determined by a probability mapping function.
[0104] Based on the above implementation method, the fault prediction module 230 is further configured to: The operating condition data is input into the pre-trained corrosion prediction model to obtain the predicted corrosion probability output by the corrosion prediction model. The failure probability is updated based on the predicted corrosion probability.
[0105] Based on the above implementation method, the fault prediction device for the motor controller further includes a mileage prediction module, which is used for: The damage growth rate of the motor controller is determined based on the comprehensive cumulative damage. The estimated remaining mileage of the vehicle is determined based on the rate of damage growth.
[0106] Based on the above embodiments, the fault prediction device for the motor controller further includes an early warning module, which is used for: The warning level is determined based on the estimated remaining mileage, the rate of damage increase, and the probability of failure. Vehicle operation is controlled according to the warning level.
[0107] For ease of description, the above devices are described in terms of function, divided into various modules. Of course, in implementing this application, the functions of each module can be implemented in one or more software and / or hardware.
[0108] The apparatus of the above embodiments is used to implement the fault prediction method of the corresponding motor controller in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0109] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the fault prediction method of the motor controller described in any of the above embodiments.
[0110] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 3 A specific hardware structure of an electronic device is shown, which may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.
[0111] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.
[0112] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.
[0113] The input / output interface 1030 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components within the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touchscreens, microphones, various sensors, etc., while output devices may include displays, speakers, vibrators, indicator lights, etc.
[0114] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).
[0115] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.
[0116] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.
[0117] The electronic devices described above are used to implement the fault prediction method of the corresponding motor controller in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0118] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the fault prediction method of the motor controller as described in any of the above embodiments.
[0119] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.
[0120] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the fault prediction method of the motor controller as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0121] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides a vehicle, the vehicle comprising: Memory, used to store executable program code; A processor is configured to call and run executable program code from memory, causing the vehicle to perform the fault prediction method of the motor controller provided in any embodiment of this application.
[0122] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this application (including the claims) is limited to these examples; within the framework of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this application as described above, which are not provided in the details for the sake of brevity.
[0123] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this application, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this application, and this also takes into account the fact that the details of the implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this application will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this application, it will be apparent to those skilled in the art that the embodiments of this application can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.
[0124] Although this application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.
[0125] The embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of this application.
Claims
1. A method of failure prediction for a motor controller, the method comprising: include: Acquire vehicle operating condition data; Based on the aforementioned operating condition data, the cumulative power cycle damage and cumulative electrochemical corrosion damage of the motor controller in the vehicle are determined. Based on the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion, the overall cumulative damage to the motor controller is determined, and the failure probability of the motor controller is determined based on the overall cumulative damage.
2. The method of claim 1, wherein, The operating condition data includes the junction temperature sequence of the power semiconductor module in the motor controller. Based on the operating condition data, the cumulative power cycle damage of the motor controller in the vehicle is determined, including: The junction temperature sequence is divided into multiple thermal cycle subsequences to obtain the junction temperature swing and average absolute junction temperature of each thermal cycle subsequence. For each of the thermal cycling subsequences, the number of module failure cycles corresponding to the thermal cycling subsequence is determined based on the junction temperature swing and the average absolute junction temperature of the thermal cycling subsequence. The number of module failure cycles is the number of cycles in which the power semiconductor module experiences thermal fatigue failure while operating under the thermal cycling subsequence. The cumulative power cycle damage of the motor controller is determined based on the number of module failure cycles corresponding to each of the aforementioned thermal cycle subsequences.
3. The method of claim 1, wherein, Based on the aforementioned operating condition data, the cumulative electrochemical corrosion damage to the motor controller in the vehicle is determined, including: For each reference sampling period in the operating condition data, based on the operating data under the reference sampling period in the operating condition data, determine the charged condensation damage, temperature influence factor, and voltage influence factor under the reference sampling period. The electrochemical corrosion rate under the reference sampling period is determined based on the temperature influence factor, the voltage influence factor, and the charged condensation damage. The cumulative electrochemical corrosion damage of the motor controller is determined based on the electrochemical corrosion rate under each of the aforementioned benchmark sampling periods.
4. The method of claim 3, wherein, The operating data includes the internal relative humidity, internal ambient temperature, controller case temperature, and DC bus voltage of the motor controller. Based on the operating data under the reference sampling period in the operating condition data, the charged condensation damage under the reference sampling period is determined, including: The dew point temperature under the reference sampling period is determined based on the internal relative humidity and the internal ambient temperature, and the condensation state indication variable of the motor controller under the reference sampling period is determined based on the dew point temperature and the controller housing temperature. Based on the DC bus voltage and the condensation state indication variable, the live condensation state indication variable of the motor controller under the reference sampling period is determined; The cumulative duration of charged condensation under the reference sampling period is determined based on the charged condensation state indicator variable, and the charged condensation damage under the reference sampling period is determined based on the cumulative duration of charged condensation.
5. The method of claim 1, wherein, After determining the overall cumulative damage of the motor controller based on the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion, the method further includes: Obtain the cumulative usage time of the vehicle, and determine the vehicle age correction factor based on the cumulative usage time; The overall cumulative damage is updated based on the vehicle age correction factor.
6. The method of claim 1, wherein, The failure probability of the motor controller is determined based on the comprehensive cumulative damage, including: Obtain the total cumulative damage of the motor controller, and update the total cumulative damage based on the comprehensive cumulative damage; The failure probability corresponding to the total cumulative damage is determined by a probability mapping function.
7. The method of claim 1, wherein, After determining the failure probability of the motor controller based on the comprehensive cumulative damage, the method further includes: The operating condition data is input into a pre-trained corrosion prediction model to obtain the predicted corrosion probability output by the corrosion prediction model. The failure probability is updated based on the predicted corrosion probability.
8. The method of claim 1, wherein, After determining the overall cumulative damage of the motor controller based on the cumulative damage from power cycling and the cumulative damage from electrochemical corrosion, the method further includes: The damage growth rate of the motor controller is determined based on the comprehensive cumulative damage. The estimated remaining mileage of the vehicle is determined based on the rate of damage growth.
9. The method of claim 8, wherein, The method further includes: The warning level is determined based on the estimated remaining mileage, the damage growth rate, and the failure probability. The operation of the vehicle is controlled according to the warning level.
10. A vehicle characterized by comprising: The vehicles include: Memory, used to store executable program code; A processor is configured to call and run the executable program code from the memory, causing the vehicle to perform the fault prediction method of the motor controller as described in any one of claims 1 to 9.
