Fault monitoring method and system for air-cooling and water-cooling integrated motor controller

Abstract

The invention relates to the field of motor fault monitoring, in particular to a fault monitoring method and system for an air-cooling and water-cooling integrated motor controller, and the method comprises the steps: collecting the operation data of the motor controller according to a preset frequency, the operation data comprising a load current, a motor controller temperature, a fan speed and a water pump flow duty ratio; and for the current acquisition moment, calculating an observation noise covariance correction value of Kalman filtering, executing a Kalman filtering algorithm by using the observation noise covariance correction value to obtain a temperature estimation value of the motor controller, and judging that a fault occurs when the deviation between the temperature estimation value and the temperature of the motor controller exceeds a preset threshold value. According to the method, the problems of estimation lag and deviation amplification caused by model mismatch of traditional fixed parameter Kalman filtering can be effectively solved, and the real-time performance and accuracy of temperature state estimation are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of motor fault monitoring. In particular, it relates to a fault monitoring method and system for an integrated air-cooled and water-cooled motor controller. Background Technology

[0002] With the increasing demand for high-power-density motor controllers in fields such as electric vehicles and industrial automation, their heat dissipation reliability has become a critical challenge. Hybrid cooling solutions combining air and water cooling are widely used due to their high efficiency and compactness. To ensure reliable controller operation, real-time temperature monitoring and fault early warning are necessary.

[0003] Currently, mainstream temperature monitoring methods rely on temperature sensor data and often employ the Kalman filter algorithm. This algorithm smooths and predicts temperature observations by fusing theoretical heat generation from electrical parameters such as current and voltage, thereby improving monitoring accuracy. In traditional Kalman filtering applications, the process noise covariance and observation noise covariance are typically set to fixed values ​​or vary slowly according to a preset pattern.

[0004] However, in integrated air-cooled and water-cooled heat dissipation systems, when the cooling mode is switched (e.g., from pure air cooling to hybrid air-water cooling) or the fluid flow rate is dynamically adjusted, the system's equivalent thermal resistance and thermal time constant undergo nonlinear abrupt changes. These abrupt changes in the system's dynamic characteristics caused by changes in physical structure lead to significant mismatches in the prediction models of traditional Kalman filters based on fixed parameters. In this case, if fixed noise covariance parameters are still used, the filter will be unable to accurately determine the source of the deviation: it may mistakenly attribute errors caused by model mismatch to sensor malfunctions, or react sluggishly when a real overheating trend occurs, resulting in monitoring failure, delayed warnings, or even overheating damage to devices. Summary of the Invention

[0005] To address the technical problem that existing Kalman filter prediction models cannot adjust filter parameters according to the dynamic changes of the heat dissipation system, resulting in low accuracy of fault monitoring, this invention provides solutions in the following aspects.

[0006] In the first aspect, a fault monitoring method for an integrated air-cooled and water-cooled motor controller includes: The operating data of the motor controller is collected at a preset frequency. The operating data includes load current, motor controller temperature, fan speed and water pump flow duty cycle. For the current acquisition time, calculate the observation noise covariance correction value of Kalman filter, and use the observation noise covariance correction value to execute the Kalman filter algorithm to obtain the temperature estimate of the motor controller. When the deviation between the temperature estimate and the motor controller temperature exceeds the preset threshold, a fault is determined to have occurred. The calculation method for the observation noise covariance correction value includes: calculating the theoretical heat generation power based on the on-resistance value of the motor controller and the load current at the current acquisition time; calculating the theoretical heat dissipation capacity based on the maximum fan speed, the fan speed at the current acquisition time, and the water pump flow duty cycle; calculating the transient thermal response deviation based on the theoretical heat generation power, theoretical heat dissipation capacity, motor controller temperature, and ambient temperature at the current acquisition time; calculating the degree of dual-medium coupling at the current acquisition time based on the maximum fan speed, the fan speed at the current acquisition time, the water pump flow duty cycle, and the transient thermal response deviation; and calculating the observation noise covariance correction value based on the degree of dual-medium coupling and the basic noise covariance of the temperature sensor in the motor controller.

[0007] Preferably, the method for calculating the theoretical heat generation power includes: multiplying the square of the load current at the current acquisition time by the on-resistance value to obtain the theoretical heat generation power at the current acquisition time.

[0008] Preferably, the method for calculating the theoretical heat dissipation capacity includes: calculating the ratio of the fan speed at the current acquisition time to the maximum fan speed as a first ratio; setting a first weighting coefficient and a second weighting coefficient, and the sum of the first weighting coefficient and the second weighting coefficient is 1; adding the product of the first weighting coefficient and the first ratio to the product of the second weighting coefficient and the pump flow duty cycle at the current acquisition time to obtain the theoretical heat dissipation capacity at the current acquisition time.

[0009] Preferably, the method for calculating the transient thermal response deviation includes: calculating the difference between the motor controller temperature and the ambient temperature at the current acquisition time as a first difference; multiplying the first difference by the theoretical heat dissipation capacity to obtain a first product; calculating the difference between the theoretical heat generation power and the first product as a second difference; calculating the derivative of the motor controller temperature with respect to time at the current acquisition time, obtaining the absolute value of the deviation between the obtained derivative value and the second difference, and normalizing the obtained absolute value to obtain the transient thermal response deviation at the current acquisition time.

[0010] Preferably, the method for calculating the degree of dual-medium coupling includes: obtaining a first ratio between the fan speed at the current acquisition time and the maximum fan speed, multiplying the first ratio by the pump flow duty cycle at the current acquisition time to obtain a second product, performing positive correlation normalization on the second product, and multiplying it by the transient thermal response deviation to obtain the degree of dual-medium coupling at the current acquisition time.

[0011] Preferably, the method for calculating the observation noise covariance correction value includes: setting a preset coupling degree threshold, calculating the difference between the dual-medium coupling degree and the coupling degree threshold as a third difference, performing positive correlation normalization on the third difference; calculating the reciprocal of the sum of 1 and the normalized third difference, and multiplying the obtained reciprocal by the basic noise covariance of the temperature sensor in the motor controller to obtain the observation noise covariance correction value.

[0012] Preferably, the method for obtaining the fundamental noise covariance includes: placing the motor controller equipped with a temperature sensor in a stable environment and operating it under no-load or light-load conditions; collecting the output data of the temperature sensor within a preset time period at a preset frequency to obtain a data sequence; and using the variance of the data sequence as the fundamental noise covariance.

[0013] Secondly, a fault monitoring system for an integrated air-cooled and water-cooled motor controller includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned fault monitoring method for the integrated air-cooled and water-cooled motor controller is implemented.

[0014] The present invention has the following effects: 1. This invention dynamically generates the observation noise covariance correction value of Kalman filter by real-time calculation of transient thermal response deviation, which reflects the degree of deviation of system thermal dynamics, and dual-medium coupling degree, which characterizes the complexity of cooling conditions. When the thermal characteristics of the system change abruptly due to the switching of air-water mode, this method can automatically reduce the confidence in the model and increase the tracking weight of the sensor measured values. This effectively overcomes the estimation lag and bias amplification problem caused by model mismatch in traditional fixed-parameter Kalman filtering, and significantly improves the real-time performance and accuracy of temperature state estimation.

[0015] 2. This invention reflects the difference between theoretical heat dissipation capacity and actual heat dissipation effect by calculating the deviation between the estimated temperature value and the actual measured temperature of the motor controller. When the deviation exceeds the preset safety threshold, it indicates that even if the cooling system (fan, water pump) works according to the instructions, its actual heat dissipation efficiency has been seriously reduced, thus playing a role in timely warning of cooling system failure. Attached Figure Description

[0016] Figure 1 This is a flowchart of steps S1-S2 in the fault monitoring method for an integrated air-cooled and water-cooled motor controller according to an embodiment of the present invention.

[0017] Figure 2 This is a schematic diagram of steps S20-S24 in the fault monitoring method for an integrated air-cooled and water-cooled motor controller according to an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0019] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0020] Reference Figure 1 The fault monitoring method for integrated air-cooled and water-cooled motor controllers includes steps S1-S2, as follows: S1: Collect the operating data of the motor controller according to the preset frequency. The operating data includes load current, motor controller temperature, fan speed and water pump flow duty cycle.

[0021] In one embodiment, a sensor system configured within the motor controller collects its operating data at a preset sampling frequency (e.g., 50Hz). The operating data includes: load current, motor controller temperature, fan speed, and water pump flow duty cycle.

[0022] The load current is measured by a current sensor (such as a Hall sensor or a sampling resistor); the motor controller temperature is measured by a temperature sensor (such as an NTC (Negative Temperature Coefficient) thermistor or thermocouple) placed on the power module heat sink or key components; the fan speed is obtained by the feedback signal of the fan driver or a speed sensor, and is used to characterize the air cooling intensity; the water pump flow duty cycle is obtained by directly reading the duty cycle value of the PWM (Pulse Width Modulation) control signal output by the controller to the water pump driver, and is used to characterize the water cooling intensity command.

[0023] The collected four-dimensional operational data constitute a multi-dimensional feature vector characterizing the dynamic process of heat generation and dissipation of the system, providing a necessary data foundation for subsequent calculations of transient thermal response deviation, dual-medium coupling degree, and observation noise covariance correction value.

[0024] S2: For the current acquisition time, calculate the observation noise covariance correction value of the Kalman filter, and use the observation noise covariance correction value to execute the Kalman filter algorithm to obtain the temperature estimate of the motor controller. When the deviation between the temperature estimate and the motor controller temperature exceeds the preset threshold, a fault is determined to have occurred.

[0025] Reference Figure 2 Step S2 includes steps S20-S24, as detailed below: S20: Calculate the theoretical heat generation power based on the on-resistance value of the motor controller and the load current at the current acquisition time.

[0026] Obtain the on-resistance value from the motor controller's product datasheet. Calculate the theoretical heat output based on the motor controller's on-resistance value and the load current at the current sampling moment by multiplying the square of the load current at the current sampling moment by the on-resistance value. The specific formula is as follows: In the formula, Indicates the first The theoretical heat production power at any given moment; Indicates the first Load current at any given moment; This represents the on-resistance value of the motor controller; this formula is a direct application of Joule's law, providing the most fundamental heat source input for constructing a physical model for temperature prediction.

[0027] S21: Calculate the theoretical heat dissipation capacity based on the maximum fan speed, the fan speed at the current acquisition time, and the water pump flow duty cycle.

[0028] Obtain the maximum fan speed from the cooling fan's product specifications. Calculate the theoretical heat dissipation capacity based on the maximum fan speed, the fan speed at the current data acquisition moment, and the pump flow duty cycle. This includes: calculating the ratio of the current fan speed to the maximum fan speed as the first ratio; setting a first weighting coefficient and a second weighting coefficient, with the sum of the first and second weighting coefficients being 1; and adding the product of the first weighting coefficient and the first ratio to the product of the second weighting coefficient and the pump flow duty cycle at the current data acquisition moment to obtain the theoretical heat dissipation capacity at the current data acquisition moment. The specific formula is as follows: In the formula, Indicates the first Theoretical heat dissipation capacity at any given time; Indicates the first The fan speed at any given time; Indicates the maximum fan speed; Indicates the first The pump flow rate duty cycle at any given time; This represents the preset first weighting coefficient, i.e., the air-cooling heat dissipation weighting coefficient, in this embodiment. ; This represents the preset second weighting coefficient, namely the water cooling heat dissipation weighting coefficient, in this embodiment. .

[0029] By fusing the rotation speed signal, which characterizes the air cooling intensity, and the control signal, which characterizes the water cooling intensity, into a theoretical heat dissipation capacity through a preset weighting coefficient, the dynamic changes in the overall heat dissipation potential of the system are reflected in real time.

[0030] S22: Calculate the transient thermal response deviation based on the theoretical heat generation power, theoretical heat dissipation capacity, motor controller temperature, and ambient temperature at the current acquisition time.

[0031] The ambient temperature of the motor controller at the current acquisition moment is obtained, and then the transient thermal response deviation is calculated. The specific calculation method includes: calculating the difference between the motor controller temperature and the ambient temperature at the current acquisition moment, as the first difference; multiplying the first difference by the theoretical heat dissipation capacity to obtain the first product; calculating the difference between the theoretical heat generation power and the first product, as the second difference; calculating the derivative of the motor controller temperature with respect to time at the current acquisition moment, obtaining the absolute value of the deviation between the obtained derivative value and the second difference, and normalizing the obtained absolute value to obtain the transient thermal response deviation at the current acquisition moment. The specific formula is as follows: In the formula, Indicates the first The transient thermal response deviation at any given moment; Indicates the first The temperature of the motor controller at any given time; Indicates the first The theoretical heat production power at any given moment; Indicates the first Theoretical heat dissipation capacity at any given time; Indicates the first The ambient temperature at that moment; This represents the standard normalization function.

[0032] This indicates that the motor controller, whose temperature is directly measured by the temperature sensor, is in the first... The actual rate of temperature change at any given moment, i.e., the measured rate of temperature rise; Characterization in the At any given moment, the system dissipates heat power to the environment through a combination of air cooling and water cooling. Characterizing the system's thermal balance prediction based on the current electrical load and cooling state, in the first... The theoretical net heat power at any given time is positive when heat generation exceeds heat dissipation, resulting in a theoretical increase in system temperature; negative when heat dissipation exceeds heat generation, resulting in a theoretical decrease in system temperature. Therefore, its value directly reflects the rate of theoretical temperature rise (or fall).

[0033] Transient thermal response deviation The magnitude of directly quantifies the degree of transient deviation between the actual thermal behavior of the system and the predictions of the theoretical model. When A large mismatch indicates a significant difference between the actual temperature trend (such as rapid rise or abnormal drop) and the trend predicted based on the current load current and cooling parameters. This suggests that the theoretical physical model used for prediction (its parameters or structure) can no longer accurately describe the actual thermal dynamics of the current system. This mismatch typically occurs in scenarios such as cooling mode switching, sudden changes in heat dissipation performance, or sensor failure. When the value is relatively small, it indicates that the measured values ​​and predictions match well, the theoretical model can accurately reflect the thermal state of the system, and the model parameters are reliable.

[0034] S23: Calculate the degree of dual-medium coupling at the current acquisition time based on the maximum fan speed, the fan speed at the current acquisition time, the pump flow duty cycle, and the transient thermal response deviation.

[0035] In complex cooling architectures where air and water cooling coexist, the two systems do not operate independently but rather exhibit significant thermal-fluid coupling effects. For example, in high-temperature environments, the ambient air drawn in by the fan is already at a high temperature. This hot air, when flowing over the fins of the water-cooled radiator, not only fails to effectively cool it but may also generate an additional heating effect, reducing the theoretical heat dissipation efficiency of the water-cooling system. These unmodeled physical interactions make the system's thermal dynamics exceptionally complex when both air and water cooling systems are operating under high loads, and any slight imbalance can significantly increase the uncertainty of model predictions.

[0036] This invention requires real-time identification of whether the system is under highly coupled and complex operating conditions, such as strong winds and water pressure. Once such conditions are identified, greater attention and risk assessment are given to the transient thermal response deviations that have occurred, thereby providing a precise basis for the subsequent adaptive adjustment of the trust weights of the filter.

[0037] To achieve this goal, this invention defines and calculates the degree of dual-medium coupling. The specific calculation method includes: obtaining a first ratio between the fan speed at the current acquisition moment and the fan's maximum speed; multiplying this first ratio by the pump flow duty cycle at the current acquisition moment to obtain a second product; performing positive correlation normalization on the second product; and then multiplying it by the transient thermal response deviation to obtain the degree of dual-medium coupling at the current acquisition moment. The specific formula is as follows: In the formula, Indicates the first The degree of dual-medium coupling at any given time; Indicates the first The transient thermal response deviation at any given moment; Indicates the first The fan speed at any given time; Indicates the maximum fan speed; Indicates the first The pump flow rate duty cycle at any given time; Represented by natural constant An exponential function with base 1.

[0038] When both the fan speed and the water pump flow rate duty cycle are at a high level A significant increase indicates that the system is operating in a dual-strong air-water cooling mode. In this mode, the fluid interaction and heat exchange effects between the two cooling media are most intense, the system's thermal dynamics are most complex, and the uncertainty of the theoretical thermal model is correspondingly greatest. The amplification of the transient thermal response deviation at the current moment indicates that the current theoretical thermal model can no longer accurately describe the real physical dynamics of the system, that is, the model is in a state of low credibility.

[0039] When the fan speed or water pump flow rate duty cycle is at least in a low-intensity operating state, for example, during pure air cooling. Or when pure water cooling is the dominant mode , A value approaching 0 indicates that the system primarily relies on a single cooling medium, the operating conditions are relatively simple, and the model's uncertainty mainly stems from errors in basic thermal parameters. At this point... Approaching 1, the algorithm does not enhance transient thermal response deviations. This means that in such highly predictable conditions, the algorithm considers deviations to be less suspicious and likely to originate from conventional sensor noise or minor model errors.

[0040] S24: Calculate the observed noise covariance correction value based on the degree of dual-medium coupling and the basic noise covariance of the temperature sensor in the motor controller.

[0041] The core of the Kalman filter algorithm lies in resolving a fundamental trust trade-off at each iteration: whether to place more trust in state predictions based on the physical model or in the actual measurements from sensors. This trade-off is achieved by adjusting the process noise covariance and the observation noise covariance.

[0042] When the system model is highly accurate, the model predictions should be assigned a higher confidence level. In this case, the algorithm should reduce the process noise covariance (indicating low model uncertainty) and relatively increase the observation noise covariance (indicating that some noise from the sensor is acceptable), so that the filter relies more on the smooth model predictions to suppress the high-frequency jitter of the sensor.

[0043] When the system model temporarily fails due to external disturbances or internal mutations, the model predictions will deviate significantly from the actual state. In this case, it is necessary to quickly shift trust to the sensors, increase the process noise covariance (characterizing model unreliability), and decrease the observation noise covariance. This forces the Kalman gain to increase, allowing the filter's state updates to more closely follow the measured values, even if they contain noise.

[0044] In the specific application of integrated air-cooled and water-cooled motor controllers, a typical cause of model failure is the abrupt change in thermal resistance and thermal time constant due to air-water coupling. Traditional fixed process noise covariance and observation noise covariance cannot cope with such abrupt changes. Therefore, this invention solves the above problem by calculating the observation noise covariance correction value.

[0045] The calculation method for the observed noise covariance correction value includes: setting a preset coupling degree threshold; calculating the difference between the dual-medium coupling degree and the coupling degree threshold as a third difference; performing positive correlation normalization on the third difference; calculating the reciprocal of the sum of 1 and the normalized third difference; and multiplying the resulting reciprocal by the fundamental noise covariance of the temperature sensor in the motor controller to obtain the observed noise covariance correction value. The specific formula is as follows: In the formula, This represents the observation noise covariance correction value at the current acquisition time; This represents the fundamental noise covariance of the temperature sensor in the motor controller; Indicates the first The degree of dual-medium coupling at any given time; This represents a preset coupling threshold; in this embodiment, ; Represented by natural constant An exponential function with base 1.

[0046] when When the value is large (and the model's credibility is low), , When the fundamental noise covariance is relatively small, the underlying noise covariance is dynamically reduced. This reduction leads the filtering algorithm to assign a higher weight to the measured value of the current temperature sensor (i.e., the motor controller temperature at the current acquisition moment) when fusing information. This results in the final temperature state estimate more closely tracking the actual measurement output of the sensor and timely correcting the bias caused by model mismatch. When the size is small (the model has high credibility), The filter operates in the default mode.

[0047] The fundamental noise covariance is a preset constant that characterizes the measurement noise variance of the temperature sensor under typical steady-state operating conditions of the motor controller. The fundamental noise covariance can be obtained as follows: The motor controller equipped with the temperature sensor is placed in a temperature-stable environment with minimal interference. The motor controller is operated under extremely light load or no-load conditions, ensuring that the heat generated by its power module is far below its heat dissipation capacity. This ensures that during data acquisition, the actual temperature change at the temperature sensor measurement point is far less than the measurement noise amplitude of the temperature sensor itself. The output data of the temperature sensor is continuously acquired within a preset time period at the system's normal operating acquisition frequency (e.g., 50Hz), forming a data sequence. The variance of this data sequence is calculated and used as the fundamental noise covariance.

[0048] It is understood that, in other embodiments, the fundamental noise covariance can also be determined using the accuracy specifications provided in the product datasheet of the temperature sensor (typically expressed as...). The estimation is performed. Accuracy indicators generally correspond to the error range at a specific confidence level (e.g., 95%). If this range is approximately... 100 times the standard deviation (for a 95% confidence level) The fundamental noise covariance can then be estimated using the following formula: .

[0049] After obtaining the observed noise covariance correction value, the Kalman filter algorithm is executed to obtain the temperature estimate of the motor controller. The absolute value of the difference between the temperature estimate and the motor controller temperature at the current acquisition time is calculated. When the absolute value of the difference exceeds the preset threshold (e.g., 7℃), it is determined that the cooling system has failed and an early warning is triggered immediately. When the absolute value of the difference does not exceed the preset threshold, it indicates that the system is within the normal or acceptable deviation range and the cooling system is working effectively.

[0050] The threshold setting is designed to distinguish between normal model dynamic errors and abnormal temperature differences caused by cooling system failure. This value is determined by analyzing historical data of normal system operation to ensure high-reliability early warning.

[0051] This application also discloses a fault monitoring system for an integrated air-cooled and water-cooled motor controller. The system includes a processor and a memory. The memory stores computer program instructions. When the computer program instructions are executed by the processor, the fault monitoring method for an integrated air-cooled and water-cooled motor controller according to the above embodiments of the present invention is implemented.

[0052] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0053] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for fault monitoring of an air- and water-cooled integrated motor controller, characterized by, include: The operating data of the motor controller is collected at a preset frequency. The operating data includes load current, motor controller temperature, fan speed and water pump flow duty cycle. For the current acquisition time, calculate the observation noise covariance correction value of Kalman filter, and use the observation noise covariance correction value to execute the Kalman filter algorithm to obtain the temperature estimate of the motor controller. When the deviation between the temperature estimate and the motor controller temperature exceeds the preset threshold, a fault is determined to have occurred. The calculation method for the observation noise covariance correction value includes: calculating the theoretical heat generation power based on the on-resistance value of the motor controller and the load current at the current acquisition time; calculating the theoretical heat dissipation capacity based on the maximum fan speed, the fan speed at the current acquisition time, and the water pump flow duty cycle; calculating the transient thermal response deviation based on the theoretical heat generation power, theoretical heat dissipation capacity, motor controller temperature, and ambient temperature at the current acquisition time; calculating the degree of dual-medium coupling at the current acquisition time based on the maximum fan speed, the fan speed at the current acquisition time, the water pump flow duty cycle, and the transient thermal response deviation; and calculating the observation noise covariance correction value based on the degree of dual-medium coupling and the basic noise covariance of the temperature sensor in the motor controller.

2. The fault monitoring method for an integrated air-cooled and water-cooled motor controller according to claim 1, characterized in that, The method for calculating the theoretical heat generation power includes multiplying the square of the load current at the current acquisition moment by the on-resistance value to obtain the theoretical heat generation power at the current acquisition moment.

3. The fault monitoring method for an integrated air-cooled and water-cooled motor controller according to claim 1, characterized in that, The method for calculating the theoretical heat dissipation capacity includes: calculating the ratio of the fan speed at the current acquisition time to the maximum fan speed, as the first ratio; setting a first weighting coefficient and a second weighting coefficient, and the sum of the first weighting coefficient and the second weighting coefficient is 1; adding the product of the first weighting coefficient and the first ratio to the product of the second weighting coefficient and the pump flow duty cycle at the current acquisition time to obtain the theoretical heat dissipation capacity at the current acquisition time.

4. The fault monitoring method for an integrated air-cooled and water-cooled motor controller according to claim 1, characterized in that, The calculation method for transient thermal response deviation includes: calculating the difference between the motor controller temperature and the ambient temperature at the current acquisition time as the first difference; multiplying the first difference by the theoretical heat dissipation capacity to obtain the first product; calculating the difference between the theoretical heat generation power and the first product as the second difference; calculating the derivative of the motor controller temperature with respect to time at the current acquisition time, obtaining the absolute value of the deviation between the obtained derivative value and the second difference, and normalizing the obtained absolute value to obtain the transient thermal response deviation at the current acquisition time.

5. The fault monitoring method for an integrated air-cooled and water-cooled motor controller according to claim 3, characterized in that, The method for calculating the degree of dual-medium coupling includes: obtaining a first ratio between the fan speed at the current acquisition time and the maximum fan speed, multiplying the first ratio by the pump flow duty cycle at the current acquisition time to obtain a second product, performing positive correlation normalization on the second product, and multiplying it by the transient thermal response deviation to obtain the degree of dual-medium coupling at the current acquisition time.

6. The fault monitoring method for an integrated air-cooled and water-cooled motor controller according to claim 1, characterized in that, The method for calculating the observation noise covariance correction value includes: setting a coupling degree threshold, calculating the difference between the dual-medium coupling degree and the coupling degree threshold as the third difference, performing positive correlation normalization on the third difference; calculating the reciprocal of the sum of 1 and the normalized third difference, and multiplying the obtained reciprocal by the basic noise covariance of the temperature sensor in the motor controller to obtain the observation noise covariance correction value.

7. The fault monitoring method for an integrated air-cooled and water-cooled motor controller according to claim 6, characterized in that, The method for obtaining the fundamental noise covariance includes: placing a motor controller equipped with a temperature sensor in a stable environment and operating it under no-load or light-load conditions; collecting the output data of the temperature sensor within a preset time period at a preset frequency to obtain a data sequence; and using the variance of the data sequence as the fundamental noise covariance.

8. A fault monitoring system for an integrated air-cooled and water-cooled motor controller, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the fault monitoring method for an integrated air-cooled and water-cooled motor controller according to any one of claims 1-7.

Images