Motor water pump dry-running prevention control method and system

Abstract

This application discloses a method and system for preventing dry running of a motor and water pump, belonging to the field of thermal management technology. It includes steps such as scenario pre-protection strategy, multi-source signal acquisition, multi-dimensional criterion feature extraction, dynamic weighted fusion, time-series confidence accumulation, dynamic threshold determination, graded protection recovery, vehicle thermal management linkage, and fault diagnosis. This application constructs a multi-dimensional criterion feature set including current harmonic characteristics, speed-voltage duty cycle characteristics, temperature rise rate coupling characteristics, and torque-power characteristics, and establishes a dynamic weighted fusion model and time-series confidence accumulation mechanism. This solves the problem of high false positive and false negative rates in complex operating conditions caused by existing technologies relying solely on a single current threshold for dry running determination. It achieves accurate identification and reliable determination of dry running states, and solves the problem of instantaneous imbalance in vehicle thermal management and increased risk of overheating of core components caused by direct shutdown after triggering dry running in existing technologies.

Description

Technical Field

[0001] This application belongs to the field of thermal management technology, and specifically relates to a method and system for preventing dry running of motors and water pumps. Background Technology

[0002] As a core component of the thermal management system, the electric vehicle motor water pump undertakes key functions such as coolant circulation and temperature control. Its operational reliability directly affects the stability and safety of the vehicle's thermal management system.

[0003] However, existing motor and water pump anti-dry running control technologies face many problems in practical applications. Traditional solutions mostly rely on a single parameter threshold for dry running detection, which is relatively simple and prone to misjudgment or missed detection under complex operating conditions. In addition, the fault handling strategies of existing technologies are mostly fixed modes, usually shutting down directly after a fault is triggered, lacking hierarchical protection and dynamic adjustment capabilities, and making it difficult to meet the needs of water pump protection and vehicle thermal management.

[0004] Existing technologies based on simple threshold comparison or fixed logic combinations are unable to resolve the contradiction between dynamic fusion of multiple parameters under multiple operating conditions and coordinated adaptation of vehicle thermal management, and cannot achieve intelligent anti-dry running control from multi-source signal acquisition to hierarchical protection and thermal management linkage. Summary of the Invention

[0005] To address the aforementioned issues, this application provides a method and system for preventing dry running of a motor and water pump. This method solves the problem that existing technologies rely solely on a single current threshold for dry running detection, leading to high false positive and false negative rates under complex operating conditions. It also solves the problem that existing technologies cause the vehicle's thermal management to become instantly unbalanced and the risk of overheating of core components to increase when the pump stops immediately after triggering dry running. Furthermore, it addresses the problem that existing technologies lack multi-scenario dry running prevention control, resulting in the water pump starting erroneously when coolant is not added or venting is incomplete.

[0006] In a first aspect, embodiments of this application provide a method for preventing dry running of a motor-driven water pump, comprising the following steps: The current, speed, temperature, voltage, and duty cycle signals of the motor and water pump are acquired to obtain operating data; Extract current harmonic features, speed-voltage-duty cycle features, temperature rise rate coupling features, and torque power features from the operating data to obtain a multi-dimensional criterion feature set; A dynamic weighted fusion model is established, and the weights of each feature in the multi-dimensional criterion feature set are assigned according to the vehicle operating conditions to obtain dynamic weights. A time-series confidence accumulation mechanism is introduced, which calculates and accumulates the weighted integral of the confidence scores of each feature output in the multidimensional criterion feature set within a preset time window based on dynamic weights, and obtains the fusion confidence score. A dynamic threshold is set based on the vehicle's operating conditions. The fusion confidence level is compared with the dynamic threshold. When the fusion confidence level exceeds the dynamic threshold, a protection interruption is triggered to obtain the dry running judgment result. Based on the dry running determination result, a graded protection and recovery strategy is implemented, which is linked to the vehicle's thermal management and diagnosis to obtain the motor and water pump anti-dry running control result.

[0007] In one embodiment, current harmonic features, speed-voltage-duty cycle features, temperature rise rate coupling features, and torque-power features are extracted from operating data to obtain a multi-dimensional criterion feature set, including: Frequency domain decomposition is performed on the current signal in the operating data to extract high-frequency harmonic components and calculate the total harmonic distortion rate to obtain the current harmonic characteristics. The speed, voltage, and duty cycle signals in the operating data are mapped in three dimensions to obtain the speed-voltage-duty cycle characteristics. The temperature and current signals in the operating data are coupled and analyzed. A thermal inertia compensation coefficient is introduced to correct the measured temperature rise rate to an equivalent steady-state temperature rise rate, thus obtaining the temperature rise rate coupling characteristics. Power analysis is performed on the current and speed signals in the operating data. Threshold current and threshold power are updated in real time by identifying motor parameters online to obtain torque power characteristics.

[0008] In one embodiment, a dynamic weighted fusion model is established, and weights are assigned to each feature in the multi-dimensional criterion feature set according to the vehicle's operating conditions to obtain dynamic weights, including: Calculate the feature reliability index based on the historical misclassification rate and sensitivity of each feature; The weights of each feature are updated in real time based on the feature reliability index to obtain dynamic weights.

[0009] In one embodiment, a time-series confidence accumulation mechanism is introduced. Based on dynamic weights, the confidence scores of each feature output in the multidimensional criterion feature set are weighted and integrated within a preset time window to obtain the fused confidence score, including: Within a preset time window, the confidence scores of each feature output in the multidimensional criterion feature set are weighted and summed according to dynamic weights to obtain the instantaneous fusion confidence score. The instantaneous fusion confidence is obtained by integrating and accumulating the results within a time window.

[0010] In one embodiment, a dynamic threshold is set based on the vehicle's operating conditions. The fusion confidence level is compared with the dynamic threshold. When the fusion confidence level exceeds the dynamic threshold, a protection interruption is triggered to obtain a dry-run determination result, including: Dynamic thresholds are set based on the overall vehicle operating conditions; When the fusion confidence exceeds the dynamic threshold and the number of consecutive abnormal frames reaches a preset value, a protection interrupt is triggered and the abnormal counter is incremented to obtain the dry-run judgment result.

[0011] In one embodiment, a graded protection and recovery strategy is implemented based on the dry-turn determination result, including: The severity assessment value is calculated based on the degree to which the fusion confidence exceeds the dynamic threshold and the proportion of the current deviating from the expected current range. The target safe speed is determined based on the severity assessment value, and the speed is reduced to the target safe speed to obtain the warning current limiting result. If the abnormality persists after the warning and current limiting, the machine is restarted in a stepped speed increase manner after shutdown. The machine stops at each speed step and detects the current-speed coupling characteristics. If the current is normal and the speed is stable, it proceeds to the next step until the test speed is reached, and the shutdown-restart cycle result is obtained. After the shutdown and restart cycle failed, the water pump was locked and shut down, resulting in a fault lockout.

[0012] In one embodiment, the vehicle thermal management is linked based on the fault lockout result, including: The thermal hazard level is calculated based on the degree of approximation between the real-time temperature and the extreme temperature of each heat source in the vehicle. The coolant distribution ratio and radiator fan speed distribution are dynamically adjusted according to the thermal hazard level to obtain the overall vehicle thermal management linkage result.

[0013] In one embodiment, performing fault diagnosis based on the fault locking result includes: The duty cycle signal in the running data is detected to obtain the PWM time characteristics; Collect LIN cable fault codes; The PWM time characteristics, LIN line fault codes, and current and temperature signals in the operating data are weighted and fused to output the dry run fault confidence and fault source tracing results, thus obtaining the fault diagnosis results.

[0014] In one embodiment, before acquiring the current, speed, temperature, voltage, and duty cycle signals of the motor-pump, a scenario pre-protection strategy is executed, including: The domain controller sends a factory mode configuration word to the motor and water pump controller. The motor and water pump controller replies with an acknowledgment frame. The domain controller sends an enable lock command. The motor and water pump controller cuts off the power supply path to the motor drive circuit, thus obtaining the factory mode lock result. When the domain controller detects that the engine or motor is idling and the coolant temperature is below the cold engine threshold, it limits the maximum speed of the motor and water pump to the safe exhaust speed, thus obtaining the pre-protection result for the idling condition. After the vehicle is powered off and replenished with fluid, it is powered on again. The motor and water pump enter a low-speed exhaust mode. The current-speed coupling characteristics during the exhaust process are collected. Once the exhaust is completed, the pre-protection state is released, and the exhaust self-test results for the maintenance scenario are obtained.

[0015] Secondly, embodiments of this application provide a method system for preventing dry running of a motor-pump, comprising: The signal acquisition module is used to acquire the current, speed, temperature, voltage and duty cycle signals of the motor and water pump to obtain operating data; The feature extraction module is used to extract current harmonic features, speed-voltage-duty cycle features, temperature rise rate coupling features, and torque power features from the operating data to obtain a multi-dimensional criterion feature set; The weight configuration module is used to establish a dynamic weighted fusion model. It configures the weights of each feature in the multi-dimensional criterion feature set according to the vehicle's operating conditions to obtain dynamic weights. The confidence accumulation module is used to introduce a time-series confidence accumulation mechanism. Based on dynamic weights, it performs weighted integral accumulation on the confidence of each feature output in the multidimensional criterion feature set within a preset time window to obtain the fused confidence. The threshold determination module is used to set a dynamic threshold based on the vehicle's operating conditions, compare the fusion confidence level with the dynamic threshold, and trigger a protection interruption when the fusion confidence level exceeds the dynamic threshold to obtain the dry running determination result. The graded protection module is used to execute graded protection and recovery strategies based on the dry running determination results, linking the vehicle's thermal management and diagnosis to obtain the motor and water pump anti-dry running control results.

[0016] Compared with the prior art, this application has the following advantages: (1) This invention constructs a multi-dimensional criterion feature set that includes current harmonic features, speed-voltage duty cycle features, temperature rise rate coupling features and torque-power features, and establishes a dynamic weighted fusion model and time sequence confidence accumulation mechanism. This solves the problem that the existing technology relies only on a single current threshold for dry running judgment, which leads to a high misjudgment rate and missed judgment rate under complex working conditions. It achieves accurate identification and reliable judgment of dry running state. (2) This invention implements a graded protection and recovery strategy based on the severity assessment value, and dynamically adjusts the coolant distribution ratio and cooling fan speed distribution by starting a vehicle thermal management network reconstruction strategy after the shutdown restart cycle fails. This solves the problem of instantaneous imbalance of vehicle thermal management and increased risk of overheating of core components caused by direct shutdown after triggering dry running in the prior art. It realizes the intelligent transition from passive fault protection to active graded control and optimized allocation of vehicle thermal management resources. (3) By implementing a scenario pre-protection strategy of factory mode locking, idling condition pre-protection and maintenance scenario exhaust self-test before signal acquisition, the problem of water pump erroneous start caused by lack of multi-scenario anti-dry running control in the existing technology is solved, and the source of dry running risk of motor water pump is blocked and prevented throughout the entire life cycle of motor water pump is realized.

[0017] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating a method for preventing dry running of a motor-driven water pump is shown. Figure 2 A schematic diagram of a motor-driven water pump anti-dry running control system is shown. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] First of all, it should be noted that after analyzing the actual application of existing motor and water pump anti-dry running control methods, it was found that traditional methods generally have problems such as single detection dimensions, fixed judgment logic, lack of time-series cumulative verification, failure to dynamically adjust in conjunction with the vehicle's operating conditions, delayed fault handling, and lack of multi-scenario pre-protection mechanisms. This results in a high rate of false positives and false negatives under complex operating conditions, and the inability to proactively warn of the accumulated dry running damage before the fault occurs. The combination of these problems makes the reliability and safety of motor and water pump operation insufficient, making it difficult to meet the intelligent anti-dry running requirements of electric vehicles in multiple scenarios.

[0022] Reference Figure 1 This invention constructs a motor-water pump anti-dry running control method based on multi-dimensional dynamic fusion and hierarchical protection. The method takes domain control as the core and achieves end-to-end anti-dry running control through steps such as scenario pre-protection strategy, multi-source signal acquisition, multi-dimensional criterion feature extraction, dynamic weighted fusion, time-series confidence accumulation, dynamic threshold determination, hierarchical protection recovery, vehicle thermal management linkage and fault diagnosis. The specific implementation of each step is described in detail below with reference to the embodiments.

[0023] I. Scene Pre-protection Strategy It should be noted that before the motor and water pump are put into operation, there is a risk that the water pump will start without coolant or with insufficient coolant in different scenarios. Direct operation will cause the water pump to run dry. Therefore, scenario pre-protection strategies need to be implemented before signal acquisition.

[0024] In factory mode, the domain controller sends a factory mode configuration word to the motor / pump controller. The motor / pump controller responds with an acknowledgment frame within a preset response time. Upon receiving the acknowledgment frame, the domain controller sends an enable lock command. Upon receiving the lock command, the motor / pump controller cuts off the power supply to the motor drive circuit, achieving hardware-level locking. The locked state remains even after power failure until an unlock command is received. The preset response time is set to 100ms, which covers controller communication processing and bus transmission delays, ensuring reliable handshake operation.

[0025] In the scenario where the intelligent driving module is installed, when the domain controller detects that the engine or motor is idling and the coolant temperature is below the cold start threshold, it automatically limits the maximum speed of the motor and water pump to a safe exhaust speed and continuously monitors the current and speed coupling characteristics. The restriction is only lifted after the characteristics indicate that the coolant has circulated normally. The cold start threshold is set to 30°C, which is below the lower limit of the normal coolant operating temperature, clearly distinguishing between cold and hot engine states. The safe exhaust speed is set to 15% to 20% of the rated speed. This speed range ensures coolant circulation and exhaust while keeping impeller wear within an acceptable range.

[0026] In after-sales maintenance scenarios, after the vehicle is powered off and the coolant is replenished, the motor and water pump enter a low-speed exhaust mode. Current and speed coupling characteristics during the exhaust process are collected to determine when exhaust is complete and to deactivate the pre-protection state. The low-speed exhaust mode speed is set to 10% to 15% of the rated speed, and the running time does not exceed 60 seconds. This parameter combination can effectively remove air bubbles from the pipeline while avoiding prolonged idling. The criteria for determining exhaust completion are: at the exhaust speed, the current continuously exceeds 80% of the lower limit threshold of the load current and the speed fluctuation rate is less than 5%. A current rebound indicates that the impeller has contacted the coolant load, and a stable speed indicates that air bubbles have been expelled and the hydraulic load is stabilizing.

[0027] II. Multi-source signal acquisition and operational data acquisition The signal acquisition module acquires the current, speed, temperature, voltage and duty cycle signals of the motor and water pump during operation to obtain operating data.

[0028] It should be noted that the current signal is acquired through a current sensor, reflecting the motor load status; the speed signal is acquired through a Hall sensor or encoder, reflecting the water pump impeller rotation status; the temperature signal is acquired through a coolant temperature sensor, reflecting the temperature status of the thermal management system; the voltage signal is acquired through a voltage sensor, reflecting the power supply system status; and the duty cycle signal is acquired through feedback from the PWM drive circuit, reflecting the motor drive control status. These five types of signals together constitute the operating data, providing a data foundation for subsequent multi-dimensional criterion feature extraction. The sampling frequency is set to 10kHz, which can completely capture the current harmonic characteristics and PWM switching frequency information.

[0029] III. Multidimensional Criterion Feature Extraction The feature extraction module extracts current harmonic features, speed-voltage duty cycle features, temperature rise rate coupling features, and torque-power features from the operating data to obtain a multi-dimensional criterion feature set. It should be noted that a single criterion is easily affected by noise interference under complex operating conditions, and multi-dimensional coupled features need to be constructed from the operating data to improve the robustness of dry-running identification.

[0030] The current signal in the operating data is decomposed in the frequency domain to extract high-frequency harmonic components and calculate the total harmonic distortion (THD) to obtain the current harmonic characteristics. Specifically, a fast Fourier transform is performed on the current signal to extract harmonic components in the 100Hz to 500Hz frequency band, and the total harmonic distortion is calculated using the following formula: ; In the formula, THD represents the total harmonic distortion rate. Indicates the amplitude of the fundamental current. This represents the amplitude of the nth harmonic current, where N represents the highest harmonic order considered, typically set to 10, covering the main high-frequency harmonic components. The 100Hz to 500Hz frequency band covers the main mechanical vibration frequencies and electromagnetic noise frequencies of the motor and water pump during operation. Under dry running conditions, the impeller idles, reducing load nonlinearity, and the total harmonic distortion rate is usually 30% to 50% lower than under load conditions. This characteristic can effectively distinguish between dry running and under load conditions.

[0031] A three-dimensional spatial mapping is performed on the speed, voltage, and duty cycle signals from the operating data to obtain the speed-voltage-duty cycle characteristics. Specifically, a three-dimensional feature space of PWM duty cycle, speed, and voltage is established, with the normalized duty cycle as the X-axis, the normalized speed as the Y-axis, and the normalized voltage as the Z-axis. The three-dimensional features are mapped onto the dry-running state identification plane. Dynamic boundary judgment is used instead of single threshold judgment, and the judgment boundary is adaptively adjusted in combination with the current voltage fluctuation range. The voltage fluctuation range is set to ±10% of the rated voltage, which covers the typical voltage fluctuation amplitude of the battery pack during charging and discharging transients and load changes.

[0032] Coupled analysis of temperature and current signals in the operational data is performed, and a thermal inertia compensation coefficient is introduced to correct the measured temperature rise rate to an equivalent steady-state temperature rise rate, thus obtaining the temperature rise rate coupling characteristics. It should be noted that the coolant heat capacity and circulation velocity cause thermal inertia in temperature changes; the measured temperature rise rate cannot directly reflect the steady-state thermal state. Therefore, thermal inertia compensation is required for correction. The formula for calculating the equivalent steady-state temperature rise rate is as follows: ; In the formula, This represents the equivalent steady-state temperature rise rate. The measured temperature rise rate is represented by t, which represents the sampling time. This represents the thermal time constant, calculated based on the coolant's heat capacity and circulation velocity. Thermal Time Constant The value is determined based on the heat capacity and circulation velocity of the coolant; the larger the heat capacity or the lower the circulation velocity, the better. The larger the value, the typical range is from 10s to 60s. The coupling determination condition is that if the equivalent steady-state temperature rise rate is lower than the lower limit of the expected temperature rise rate and the current signal is lower than the threshold, then dry running is determined.

[0033] Power analysis is performed on the current and speed signals in the operating data. Threshold current and threshold power are updated in real time through online identification of motor parameters to obtain torque power characteristics. Specifically, the online identification of motor parameters uses a recursive least squares method to update the motor resistance and inductance parameters in real time, thereby updating the threshold current. The formula is as follows: ; In the formula, The threshold current is represented by , k represents the safety factor, with a value ranging from 1.1 to 1.5, which takes into account parameter identification errors and operating condition fluctuation margins, and V represents the supply voltage. This indicates the online identification resistor. A safety factor k of 1.2 to 1.3 is preferred, which can avoid misjudgment while maintaining an appropriate safety margin.

[0034] IV. Construction of Dynamic Weighted Fusion Model A dynamic weighted fusion model is established through a weight configuration module. Weights are assigned to each feature in the multi-dimensional criterion feature set based on the vehicle's operating conditions, resulting in dynamic weights. It should be noted that the sensitivity and misclassification rate of each criterion feature differ under different operating conditions. For example, the reliability of the temperature criterion is low under low-temperature start-up conditions, while the sensitivity of the speed criterion is high under high-speed conditions. Therefore, a dynamic weight adjustment mechanism is necessary.

[0035] The feature reliability index is calculated based on the historical misclassification rate and sensitivity of each feature. The historical misclassification rate is calculated by statistically analyzing the number of misclassifications and the total number of judgments for each feature within the past preset statistical period, using the following formula: ; In the formula, This represents the historical misclassification rate of the i-th feature. This represents the number of misclassifications for the i-th feature within the statistical period. This represents the total number of judgments for the i-th feature within the statistical period. The statistical period is set to 100 judgment cycles. This period length can balance the response speed and stability of weight updates. If the period is too short, it is easily affected by occasional noise. If the period is too long, it cannot track changes in operating conditions in a timely manner.

[0036] Sensitivity is determined by calculating the difference in characteristic values ​​between dry-running and loaded states, as shown in the following formula: ; In the formula, This represents the sensitivity of the i-th feature. Let represent the mean of the i-th feature under the dry rotation state. Let represent the mean of the i-th feature under load. This represents the standard deviation of the i-th feature under load. The formula measures the distinguishing ability of a feature by the multiple of the difference between the mean under dry running and under load conditions relative to the fluctuation under load conditions; the larger the multiple, the higher the sensitivity.

[0037] The feature weights are updated in real time based on the feature reliability index to obtain dynamic weights. The weight update formula is as follows: ; In the formula, The formula represents the dynamic weight of the i-th feature, and M represents the total number of features. Here, M equals 4, corresponding to the four dimensions of current harmonic features, speed-voltage duty cycle features, temperature rise rate coupling features, and torque-power features. This formula measures the overall reliability of the features by the ratio of sensitivity to historical misclassification rate. The higher the reliability, the greater the weight, and the sum of all weights is normalized to 1.

[0038] V. Time Series Confidence Accumulation Mechanism A time-series confidence accumulation mechanism is introduced through a confidence accumulation module. Based on dynamic weights, the confidence scores of each feature output in the multidimensional criterion feature set are weighted and integrated within a preset time window to obtain the fused confidence score. It should be noted that instantaneous judgment is susceptible to noise interference, which may lead to occasional misjudgments. Time-series accumulation can improve the stability and reliability of the judgment.

[0039] Within a preset time window, the confidence scores of each feature output in the multidimensional criterion feature set are weighted and summed according to dynamic weights to obtain the instantaneous fusion confidence score. The confidence score of each feature output is obtained by comparing the current value of the feature with the deviation decision boundary and mapping it to the interval between 0 and 1, as shown in the following formula: ; In the formula, This represents the confidence level of the i-th feature output at time t. This represents the eigenvalue of the i-th feature at time t. This represents the load determination boundary for the i-th feature. This represents the boundary for determining the dry-to-turn boundary of the i-th feature. When the feature value is between the load boundary and the dry-to-turn boundary, the confidence level is linearly mapped to [0,1]; when the feature value is below the dry-to-turn boundary, the confidence level is 1, and when it is above the load boundary, the confidence level is 0.

[0040] The formula for calculating the instantaneous fusion confidence score is as follows: ; In the formula, This represents the instantaneous fusion confidence at time t.

[0041] The instantaneous fusion confidence score is obtained by integrating and summing the results over a time window. The formula for integration and summation is as follows: ; In the formula, This represents the fusion confidence level, where T represents the number of sampling points within the time window. This indicates the sampling period. The time window is set to 500ms, and the sampling period is set to 10ms, corresponding to 50 sampling points. This time window length can cover one complete mechanical rotation cycle of the motor and water pump while retaining appropriate redundancy, ensuring that the cumulative results reflect the true state rather than instantaneous noise.

[0042] VI. Dynamic Threshold Determination and Protection Interruption The threshold determination module sets a dynamic threshold based on the vehicle's operating conditions. The fused confidence level is compared with the dynamic threshold; if the threshold is exceeded, a protection interruption is triggered, resulting in a dry-running judgment. It should be noted that the dynamic threshold needs to be adaptively adjusted according to the current operating conditions to avoid the failure of a fixed threshold under extreme conditions.

[0043] A dynamic threshold is set based on the vehicle's overall operating conditions. The dynamic threshold is adaptively set based on the historical maximum and minimum values ​​of the fused confidence score within the time window under the current operating conditions, using the following formula: ; In the formula, Indicates a dynamic threshold. This represents the maximum fusion confidence level within the historical time window under the current operating conditions. This represents the threshold coefficient, with a value ranging from 0.6 to 0.75. This coefficient range has been calibrated through extensive operating condition experiments. 0.6 is suitable for high-noise operating conditions to avoid missed detections, while 0.75 is suitable for low-noise operating conditions to reduce false detections. A typical value is 0.68.

[0044] When the fusion confidence exceeds the dynamic threshold and the number of consecutive abnormal frames reaches a preset value, a protection interruption is triggered and the abnormal counter is incremented to obtain the dry-run judgment result. The number of consecutive abnormal frames is set to 5 frames, corresponding to a continuous judgment time of 50ms. This setting can filter out single-point noise interference while ensuring timely response.

[0045] VII. Tiered Protection and Recovery Strategy The graded protection module executes graded protection and recovery strategies based on the dry running determination results. It should be noted that the severity of dry running varies, and direct shutdown may cause an instantaneous imbalance in the vehicle's thermal management; therefore, differentiated protection must be implemented according to the severity.

[0046] The severity assessment value is calculated based on the degree to which the fusion confidence exceeds the dynamic threshold and the degree of change in the current signal amplitude, using the following formula: ; In the formula, D represents the severity assessment value. and The weighting coefficients are 0.5 and 0.5 respectively, which reflect the fusion confidence bias and the degree of current deviation in a balanced manner. Indicates the rated operating current. This represents the actual current. Based on the severity assessment value, the target safe speed is determined and then reduced to obtain the warning current limiting result. The formula for calculating the target safe speed is as follows: ; In the formula, Indicates the target safe rotation speed. Indicates the rated speed. This represents the speed reduction factor, which ranges from 0.3 to 0.7, with a typical value of 0.5. This value strikes a balance between protecting the mechanical components of the water pump and maintaining the basic cooling cycle.

[0047] If the abnormality persists after the warning and current limiting measures are implemented, the pump is restarted using a stepped speed increase method after shutdown. At each speed step, the pump pauses and monitors the current and speed signals. If normal, it proceeds to the next speed step until the test speed is reached, thus obtaining the shutdown-restart cycle result. The stepped speed increase is set with four speed steps: 25%, 50%, 75%, and 100% of the test speed. Each speed step pauses for 2 seconds, and the current and speed coupling characteristics are monitored. This number of steps and pause time fully verifies the hydraulic load recovery at each speed range. If the shutdown-restart cycle fails, the pump is locked down, resulting in a fault lockout.

[0048] 8. Integrated Vehicle Thermal Management Based on the fault location results, a vehicle thermal management network reconfiguration strategy was initiated. It should be noted that after the water pump stops, the vehicle's cooling capacity drops sharply, requiring a reallocation of thermal management resources to prevent overheating of core components.

[0049] The thermal hazard level is calculated based on the degree of approximation between the real-time temperature and the extreme temperature of each heat source in the vehicle, using the following formula: ; In the formula, This indicates the thermal hazard level of the i-th heat source. This represents the real-time temperature of the i-th heat source. Indicates ambient temperature. This represents the limiting temperature of the i-th heat source. The formula normalizes the temperature of each heat source to [0,1], where 0 represents the ambient temperature and 1 represents the limiting temperature. The larger the value, the higher the thermal risk.

[0050] The coolant distribution ratio and radiator fan speed are dynamically adjusted based on the thermal hazard level to obtain the overall vehicle thermal management linkage result. The coolant distribution ratio is allocated according to the thermal hazard level of each heat source, as shown in the following formula: ; In the formula, This represents the coolant distribution ratio for the i-th heat source. This indicates the thermal hazard level of the i-th heat source. Let represent the thermal hazard level of the j-th heat source, and K represent the total number of heat sources, including motors, batteries, and electronic control systems. The cooling fan speed is determined by mapping the highest thermal hazard level among all heat sources, as shown in the following formula: ; In the formula, Indicates the speed of the cooling fan. This indicates the maximum fan speed.

[0051] IX. Fault Diagnosis Fault diagnosis is performed based on the fault location results. It should be noted that dry run faults need to be distinguished from controller faults and sensor faults to avoid false alarms.

[0052] The duty cycle signal in the operating data is detected to obtain the PWM time characteristics, namely the duration of the PWM high and low levels and the duty cycle change rate; LIN line fault codes are collected; the PWM time characteristics, LIN line fault codes, and current and temperature signals in the operating data are weighted and fused to output the dry-running fault confidence and fault tracing results, thus obtaining the fault diagnosis results. The weighted fusion adopts the DS evidence theory, and the basic probability allocation is determined based on the historical reliability of each information source, as shown in the following formula: ; In the formula, m(A) represents the basic probability distribution of proposition A. This indicates the historical reliability of information source A. The historical reliability of information source B is represented by its physical meaning and calculation method, which are related to... Consistent, the summation range covers all information sources. The basic probability assignments of each information source are fused using the DS synthesis rule to output the confidence level of the stall failure and the fault tracing results. The DS synthesis rule formula is as follows:

[0053] In the formula, Let A represent the basic probability distribution of proposition A after fusion. This represents the basic probability assignment of proposition X by the first information source. This represents the basic probability assignment of proposition Y by the second information source. The summation range covers all proposition combinations that satisfy the intersection condition. The summation range in the denominator covers all proposition combinations whose intersection is an empty set. This is used for normalization to eliminate conflicts.

[0054] 10. Speed ​​and Current MAP Chart Setting Before acquiring operational data, the speed-current mapping is calibrated, and a speed-current MAP with operating condition index is established. During actual operation, the corresponding thresholds are extracted by interpolation based on real-time temperature and voltage to obtain the calibration correction results. It should be noted that the dry-running current and on-load current at the same speed differ under different coolant temperatures and power supply voltages, and an operating condition index MAP needs to be established to improve the accuracy of the judgment.

[0055] A two-dimensional operating condition grid based on temperature and voltage is introduced. The temperature grid nodes are set to -20℃, 0℃, 20℃, 40℃, 60℃ and 80℃, covering the typical operating temperature range of electric vehicles from low temperature start-up to high temperature heat dissipation. The voltage grid nodes are set to 80%, 90%, 100%, 110% and 120% of the rated voltage, covering the voltage output range of the battery pack under SOC changes and load fluctuations.

[0056] Dry-running current and on-load current were tested at each grid node to create a speed-current MAP with operating condition index. In actual operation, the four nearest grid nodes were located based on real-time temperature and voltage, and the threshold current under the current operating condition was calculated using bilinear interpolation, as shown in the following formula: ; In the formula, This represents the interpolated threshold current. This represents the weight of the m-th grid node. This represents the test current at the m-th grid node. Weight The weighting is calculated based on the distance between the real-time temperature and voltage and the grid nodes. The closer the distance, the greater the weight. The formula is as follows: ; In the formula, Indicates real-time temperature. This represents the temperature of the m-th grid node. Indicates real-time voltage. This represents the voltage of the m-th grid node. The distance formula uses Euclidean distance to measure the deviation between the real-time operating condition and the grid node in the temperature-voltage two-dimensional plane. The smaller the distance, the greater the weight, ensuring that the interpolation result is closer to the actual operating condition.

[0057] It should be noted that the aforementioned pre-protection strategy, multi-source signal acquisition, multi-dimensional criterion feature extraction, dynamic weighted fusion model construction, time-series confidence accumulation, dynamic threshold determination and protection interruption, graded protection and recovery strategy, vehicle thermal management linkage, fault diagnosis, and speed-current MAP chart calibration all work together to form a complete anti-dry-run control link from signal acquisition to graded protection. The pre-protection strategy is executed before signal acquisition, blocking the risk of water pump starting without coolant at the source; speed-current MAP chart calibration provides an adaptive judgment benchmark for multi-dimensional criterion feature extraction; multi-source signal acquisition provides operational data for multi-dimensional criterion feature extraction; multi-dimensional criterion feature extraction provides a multi-dimensional criterion feature set for dynamic weighted fusion model construction; dynamic weighted fusion model construction provides dynamic weights for time-series confidence accumulation; time-series confidence accumulation provides fusion confidence for dynamic threshold determination; dynamic threshold determination provides dry-run judgment results for graded protection recovery; and graded protection recovery linkage with vehicle thermal management linkage and fault diagnosis achieves closed-loop control. The data flow between each step is clear, and the logical connection is tight. In addition, the scenario pre-protection strategy and the speed and current MAP chart setting complement each other with the main control link, jointly improving the reliability of the motor and water pump in preventing dry running throughout the entire life cycle.

[0058] Reference Figure 2 Based on the same inventive concept, this application also provides a method and system for preventing dry running of a motor and water pump, comprising: The signal acquisition module is used to acquire the current, speed, temperature, voltage and duty cycle signals of the motor and water pump to obtain operating data; The feature extraction module is used to extract current harmonic features, speed-voltage-duty cycle features, temperature rise rate coupling features, and torque power features from the operating data to obtain a multi-dimensional criterion feature set; The weight configuration module is used to establish a dynamic weighted fusion model. It configures the weights of each feature in the multi-dimensional criterion feature set according to the vehicle's operating conditions to obtain dynamic weights. The confidence accumulation module is used to introduce a time-series confidence accumulation mechanism. Based on dynamic weights, it performs weighted integral accumulation on the confidence of each feature output in the multidimensional criterion feature set within a preset time window to obtain the fused confidence. The threshold determination module is used to set a dynamic threshold based on the vehicle's operating conditions, compare the fusion confidence level with the dynamic threshold, and trigger a protection interruption when the fusion confidence level exceeds the dynamic threshold to obtain the dry running determination result. The graded protection module is used to execute graded protection and recovery strategies based on the dry running determination results, linking the vehicle's thermal management and diagnosis to obtain the motor and water pump anti-dry running control results.

[0059] Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A dry-running prevention control method for a motor water pump, characterized by, Includes the following steps: The current, speed, temperature, voltage, and duty cycle signals of the motor and water pump are acquired to obtain operating data; Extract current harmonic features, speed-voltage-duty cycle features, temperature rise rate coupling features, and torque power features from the operating data to obtain a multi-dimensional criterion feature set; A dynamic weighted fusion model is established, and the weights of each feature in the multi-dimensional criterion feature set are assigned according to the vehicle operating conditions to obtain dynamic weights. A time-series confidence accumulation mechanism is introduced, which calculates and accumulates the weighted integral of the confidence scores of each feature output in the multidimensional criterion feature set within a preset time window based on dynamic weights, and obtains the fusion confidence score. A dynamic threshold is set based on the vehicle's operating conditions. The fusion confidence level is compared with the dynamic threshold. When the fusion confidence level exceeds the dynamic threshold, a protection interruption is triggered to obtain the dry running judgment result. Based on the dry running determination result, a graded protection and recovery strategy is implemented, which is linked to the vehicle's thermal management and diagnosis to obtain the motor and water pump anti-dry running control result.

2. The method for preventing dry running of a motor-driven water pump according to claim 1, characterized in that, By extracting current harmonic characteristics, speed-voltage-duty cycle characteristics, temperature rise rate coupling characteristics, and torque-power characteristics from the operating data, a multi-dimensional criterion feature set is obtained, including: Frequency domain decomposition is performed on the current signal in the operating data to extract high-frequency harmonic components and calculate the total harmonic distortion rate to obtain the current harmonic characteristics. The speed, voltage, and duty cycle signals in the operating data are mapped in three dimensions to obtain the speed-voltage-duty cycle characteristics. The temperature and current signals in the operating data are coupled and analyzed. A thermal inertia compensation coefficient is introduced to correct the measured temperature rise rate to an equivalent steady-state temperature rise rate, thus obtaining the temperature rise rate coupling characteristics. Power analysis is performed on the current and speed signals in the operating data. Threshold current and threshold power are updated in real time by identifying motor parameters online to obtain torque power characteristics.

3. The method for preventing dry running of a motor-pump according to claim 1, characterized in that, A dynamic weighted fusion model is established, and weights are assigned to each feature in the multi-dimensional criterion feature set based on the vehicle's operating conditions to obtain dynamic weights, including: Calculate the feature reliability index based on the historical misclassification rate and sensitivity of each feature; The weights of each feature are updated in real time based on the feature reliability index to obtain dynamic weights.

4. The method for preventing dry running of a motor-driven water pump according to claim 1, characterized in that, A time-series confidence accumulation mechanism is introduced. Based on dynamic weights, the confidence scores of each feature output in the multidimensional criterion feature set are weighted and integrated within a preset time window to obtain the fused confidence score, including: Within a preset time window, the confidence scores of each feature output in the multidimensional criterion feature set are weighted and summed according to dynamic weights to obtain the instantaneous fusion confidence score. The instantaneous fusion confidence is obtained by integrating and accumulating the results within a time window.

5. The method for preventing dry running of a motor-driven water pump according to claim 1, characterized in that, A dynamic threshold is set based on the vehicle's operating conditions. The fusion confidence level is compared with the dynamic threshold. When the fusion confidence level exceeds the dynamic threshold, a protection interruption is triggered, resulting in a dry-run determination result, including: Dynamic thresholds are set based on the overall vehicle operating conditions; When the fusion confidence exceeds the dynamic threshold and the number of consecutive abnormal frames reaches a preset value, a protection interrupt is triggered and the abnormal counter is incremented to obtain the dry-run judgment result.

6. The method for preventing dry running of a motor-driven water pump according to claim 1, characterized in that, Based on the dry-turn determination result, a graded protection and recovery strategy is implemented, including: The severity assessment value is calculated based on the degree to which the fusion confidence exceeds the dynamic threshold and the proportion of the current deviating from the expected current range. The target safe speed is determined based on the severity assessment value, and the speed is reduced to the target safe speed to obtain the warning current limiting result. If the abnormality persists after the warning and current limiting, the machine is restarted in a stepped speed increase manner after shutdown. The machine stops at each speed step and detects the current-speed coupling characteristics. If the current is normal and the speed is stable, it proceeds to the next step until the test speed is reached, and the shutdown-restart cycle result is obtained. After the shutdown and restart cycle failed, the water pump was locked and shut down, resulting in a fault lockout.

7. The method for preventing dry running of a motor-driven water pump according to claim 6, characterized in that, Based on the fault location results, the vehicle's thermal management is linked, including: The thermal hazard level is calculated based on the degree of approximation between the real-time temperature and the extreme temperature of each heat source in the vehicle. The coolant distribution ratio and radiator fan speed distribution are dynamically adjusted according to the thermal hazard level to obtain the overall vehicle thermal management linkage result.

8. The method for preventing dry running of a motor-driven water pump according to claim 6, characterized in that, Based on the fault location results, perform fault diagnosis, including: The duty cycle signal in the running data is detected to obtain the PWM time characteristics; Collect LIN cable fault codes; The PWM time characteristics, LIN line fault codes, and current and temperature signals in the operating data are weighted and fused to output the dry run fault confidence and fault source tracing results, thus obtaining the fault diagnosis results.

9. The method for preventing dry running of a motor-driven water pump according to claim 1, characterized in that, Before acquiring the current, speed, temperature, voltage, and duty cycle signals of the motor and water pump, a scenario pre-protection strategy is executed, including: The domain controller sends a factory mode configuration word to the motor and water pump controller. The motor and water pump controller replies with an acknowledgment frame. The domain controller sends an enable lock command. The motor and water pump controller cuts off the power supply path to the motor drive circuit, thus obtaining the factory mode lock result. When the domain controller detects that the engine or motor is idling and the coolant temperature is below the cold engine threshold, it limits the maximum speed of the motor and water pump to the safe exhaust speed, thus obtaining the pre-protection result for the idling condition. After the vehicle is powered off and replenished with fluid, it is powered on again. The motor and water pump enter a low-speed exhaust mode. The current-speed coupling characteristics during the exhaust process are collected. Once the exhaust is completed, the pre-protection state is released, and the exhaust self-test results for the maintenance scenario are obtained.

10. A control system for preventing dry running of a motor-driven water pump, characterized in that, include: The signal acquisition module is used to acquire the current, speed, temperature, voltage and duty cycle signals of the motor and water pump to obtain operating data; The feature extraction module is used to extract current harmonic features, speed-voltage-duty cycle features, temperature rise rate coupling features, and torque power features from the operating data to obtain a multi-dimensional criterion feature set; The weight configuration module is used to establish a dynamic weighted fusion model. It configures the weights of each feature in the multi-dimensional criterion feature set according to the vehicle's operating conditions to obtain dynamic weights. The confidence accumulation module is used to introduce a time-series confidence accumulation mechanism. Based on dynamic weights, it performs weighted integral accumulation on the confidence of each feature output in the multidimensional criterion feature set within a preset time window to obtain the fused confidence. The threshold determination module is used to set a dynamic threshold based on the vehicle's operating conditions, compare the fusion confidence level with the dynamic threshold, and trigger a protection interruption when the fusion confidence level exceeds the dynamic threshold to obtain the dry running determination result. The graded protection module is used to execute graded protection and recovery strategies based on the dry running determination results, linking the vehicle's thermal management and diagnosis to obtain the motor and water pump anti-dry running control results.

Images