Vehicle water valve adaptation method and device based on feature self-learning, equipment and medium

Abstract

The application mainly relates to the technical field of automobile engineering. The application discloses a vehicle water valve adaptation method, device, equipment and medium based on feature self-learning, and the method comprises the following steps: in response to a preset trigger condition, a diagnosis excitation signal is sent to a multi-way water valve of a target vehicle, and a response current signal fed back by the multi-way water valve is acquired; physical features are extracted from the response current signal; a target multi-way water valve corresponding to the physical features is screened out from a plurality of multi-way water valves, and the model of the target multi-way water valve is determined; a control parameter set corresponding to the model of the target multi-way water valve is acquired, and the multi-way water valve of the target vehicle is controlled to operate according to the control parameter set. The application can automatically identify the model of a replaced water valve and load adaptation parameters under the condition that there is no hardware identification interface.

Description

Technical Field

[0001] This invention relates to the field of automotive engineering technology, specifically to a method, apparatus, equipment, and medium for adapting vehicle water valves based on feature self-learning. Background Technology

[0002] As a core actuator in the thermal management system of new energy vehicles, the multi-way water valve is responsible for controlling the opening and closing of the coolant flow path and switching between them, enabling coordinated management of multiple thermal circuits such as the battery, motor, and air conditioning. With the trend of integration and intelligence in the development of new energy vehicles, multi-way water valve technology has evolved from simple mechanical valves to electronically controlled valves, from single-function valves to multi-mode switching valves, and from independent components to system integration.

[0003] In existing technologies, some vehicles are designed with hardware identification interfaces for water valves, such as using resistors with different resistance values ​​or coded pins to identify the water valve model. The electronic control unit (ECU) can directly determine the model and call the corresponding parameters by reading this hardware code. However, due to limitations in vehicle wiring space and cost control, many mass-produced models do not have such hardware identification interfaces. When the original model of water valve is discontinued, maintenance personnel can only replace it with a different model. Since the vehicle cannot read the new water valve's identification information through hardware, the ECU still drives the motor according to the parameters of the old model, leading to malfunctions such as opening control deviation, stall protection threshold mismatch, or even valve jamming.

[0004] Another common practice is for maintenance personnel to manually input the water valve model using diagnostic equipment or through host computer software. This method relies on manual operation, is inefficient, and is prone to parameter mismatches due to input errors. Even water valves of the same model from different batches may have individual differences that affect control accuracy. With increased usage time, mechanical wear and insulation aging of the water valves can alter the original control performance, and the existing system lacks adaptive adjustment and fault prediction capabilities. Furthermore, when new water valve models are released, existing vehicles cannot easily obtain support, often requiring hardware replacement or a complete firmware rewrite, resulting in high maintenance costs. Summary of the Invention

[0005] This invention provides a vehicle water valve adaptation method, device, equipment, and medium based on feature self-learning, which can automatically identify the model of the replaced water valve and load adaptation parameters without a hardware identification interface.

[0006] This invention provides a vehicle water valve adaptation method based on feature self-learning, the method comprising: In response to a preset triggering condition, a diagnostic excitation signal is sent to the multi-port water valve of the target vehicle, and the response current signal fed back by the multi-port water valve is acquired. Extract physical features from the response current signal; Select the target multi-way water valve corresponding to the physical characteristics from a plurality of multi-way water valves and determine the model of the target multi-way water valve; Obtain the control parameter set corresponding to the model of the target multi-way water valve, and control the operation of the multi-way water valve of the target vehicle according to the control parameter set.

[0007] Optionally, the step of sending a diagnostic excitation signal to the multi-way water valve of the target vehicle in response to a preset triggering condition, and acquiring the response current signal fed back by the multi-way water valve, includes: Send a drive signal to the multi-way water valve to make the multi-way water valve move; During the movement of the multi-way water valve, the stroke Hall effect data and motor current of the multi-way water valve are collected multiple times. If the motor current is less than the preset stall threshold or the rising edge steepness of the motor current is lower than the preset steepness threshold when the stroke Hall number stops changing, then it is determined that the motor is in a physical limit state and the stroke position is recorded. If the motor current reaches the stall threshold and the rising edge steepness is higher than the steepness threshold when the stroke Hall number stops changing, it is determined to be in an abnormal stuck state. After repairing the abnormal jammed state, the multi-way water valve is controlled to move in the opposite direction to another physical limit, and then the process of repeatedly collecting the stroke Hall number and motor current of the multi-way water valve during its movement is resumed.

[0008] Optionally, the vehicle water valve adaptation method based on feature self-learning further includes: During the operation of the multi-way water valve, the motor current signal is acquired and the ripple characteristic parameters of the motor current signal are extracted; The ripple feature parameters are compared one by one with the ripple feature templates in the preset fault feature library to determine the fault diagnosis result. When a discrete fault diagnosis report indicates a short-circuit fault and the ripple characteristic parameters match the target ripple characteristic template, the diagnosis report result for the short-circuit fault is determined. When a discrete fault diagnosis report indicates a short-circuit fault and the ripple characteristic parameters do not match any of the ripple characteristic templates, it is determined to be a false fault report. When discrete fault diagnosis does not report a fault but the ripple characteristic parameters match the insulation degradation ripple characteristic template, an insulation warning is triggered and maintenance is prompted. When the discrete fault diagnosis does not report a fault but the ripple characteristic parameters match the poor contact ripple characteristic template, it is recorded as an intermittent fault and the fault record data is uploaded to the cloud.

[0009] Optionally, the vehicle water valve adaptation method based on feature self-learning further includes: When a system fault is detected, the multi-way water valve is driven to move to a preset initial position; Collect the position feedback signal of the current position of the multi-way water valve and the motor current signal; When the position feedback signal indicates that the initial position has been entered into the tolerance zone and the motor current reaches the target stall current threshold, the multi-way water valve is determined to have successfully returned to its original position.

[0010] Optionally, the vehicle water valve adaptation method based on feature self-learning further includes: Get the current ambient temperature; Based on a preset temperature compensation coefficient table, determine the duty cycle compensation value corresponding to the current ambient temperature; The drive duty cycle in the control parameter set is adjusted using the duty cycle compensation value.

[0011] Optionally, the vehicle water valve adaptation method based on feature self-learning further includes: Obtain the cumulative Hall effect count of the multi-way water valve; Based on the cumulative number of Hall effect sensors and the desired target opening, the angle compensation amount is determined using a preset wear angle correction function. The target Hall number is adjusted using the angle compensation amount, and the multi-way water valve is driven with the adjusted target Hall number to compensate for the angle deviation caused by gear wear.

[0012] Optionally, the vehicle water valve adaptation method based on feature self-learning further includes: Receive percentage opening command; The percentage opening command is converted into the physical angle value corresponding to the model of the multi-way water valve; Based on the mapping relationship between the angle corresponding to the model and the Hall number, the physical angle value is converted into a target Hall number, and the multi-way water valve is driven to reach the target opening degree according to the target Hall number.

[0013] The present invention also provides a vehicle water valve adapter based on feature self-learning, the adapter comprising: The transmission module is used to send a diagnostic excitation signal to the multi-way water valve of the target vehicle in response to a preset trigger condition, and to acquire the response current signal fed back by the multi-way water valve. An extraction module is used to extract physical features from the response current signal; A matching module is used to filter out the target multi-way water valve corresponding to the physical characteristics from a plurality of multi-way water valves and determine the model of the target multi-way water valve; The control module is used to acquire the control parameter set corresponding to the model of the target multi-way water valve, and control the operation of the multi-way water valve of the target vehicle according to the control parameter set.

[0014] The present invention also provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the vehicle water valve adaptation method based on feature self-learning as described in any of the preceding claims.

[0015] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the feature-based self-learning vehicle water valve adaptation method as described in any of the preceding claims.

[0016] The present invention has at least the following beneficial effects: First, the system actively triggers the water valve response using diagnostic excitation signals. When a vehicle's water valve is replaced, the system sends specific instructions to the multi-way water valve via diagnostic equipment, causing it to execute preset actions and acquiring the response current signal from the water valve in real time. This process does not rely on any hardware coding pins or manual input; it obtains information entirely through electrical response. Second, physical features with model-distinguishing characteristics are extracted from the response current signal, such as the number of Hall pulses throughout the stroke, the stall current value, and the shape of the current rise curve. Different water valve models exhibit stable and distinguishable ranges in these features due to differences in mechanical structure and electromagnetic parameters. Third, the extracted physical features are compared and filtered against a multi-model feature library pre-stored locally or in the cloud to automatically determine the model of the currently installed water valve. This matching process requires no manual intervention, avoiding the error risks associated with manual configuration. Finally, the corresponding set of control parameters, including angle range, Hall mapping coefficient, and temperature compensation reference, is automatically loaded based on the identified model, ensuring that the electronic control unit can drive the water valve according to the correct parameters. This technical solution uses electrical excitation and response characteristics as a bridge, replacing hardware identification interfaces and manual configuration methods, and achieving automated, highly reliable model identification and parameter adaptation. Attached Figure Description

[0017] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the technical solutions of the present invention, and do not constitute a limitation on the technical solutions of the present invention.

[0018] Figure 1 This is a flowchart of the steps for a vehicle water valve adaptation method based on feature self-learning. Figure 2 This is a flowchart of the self-learning and type determination steps in a vehicle water valve adaptation method based on feature self-learning. Figure 3 This is a schematic diagram of current ripple fault diagnosis in a vehicle water valve adaptation method based on feature self-learning. Figure 4This is a flowchart of the steps for homing a multi-port water valve in a vehicle water valve adaptation method based on feature self-learning. Figure 5 This is a flowchart illustrating the steps involved in confirming fault location in a vehicle water valve adaptation method based on feature self-learning. Figure 6 This is a curve showing the relationship between temperature and actual duty cycle in a vehicle water valve adaptation method based on feature self-learning. Figure 7 This is a curve showing the relationship between temperature and compensation coefficient in a vehicle water valve adaptation method based on feature self-learning. Figure 8 This is a curve showing the relationship between wear and angle correction in a vehicle water valve adaptation method based on feature self-learning; Figure 9 This is an architecture diagram of a vehicle water valve adaptation system based on feature self-learning. Figure 10 This is a schematic diagram of a vehicle water valve adapter based on feature self-learning. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0020] The researchers of this application discovered several unresolved issues with existing technical solutions. During after-sales replacement, due to the lack of a pre-existing hardware identification interface on the vehicle and significant differences in physical characteristics such as the stroke Hall effect value of different water valve models, the electronic control unit (ECU) still operates according to the original model parameters after replacing the valve with a replacement, easily leading to valve jamming or incomplete execution. The self-learning mechanism lacks robustness; a single scan can easily misjudge the jamming point as the physical endpoint and cannot identify or correct it. Fault diagnosis relies solely on discrete fault codes, lacking in-depth analysis of motor current ripple, making it difficult to detect early hidden dangers such as insulation aging and posing a risk of false alarms. The fail-safe strategy is incomplete, lacking a return-to-position confirmation mechanism, failing to ensure that the valve returns to its safe initial position after a fault. The system does not consider the impact of ambient temperature on drive performance; at low temperatures, increased mechanical resistance may lead to false alarms of jamming faults. Throughout the entire lifecycle, mechanical wear changes gear clearance, causing control accuracy to decrease over time, and existing solutions lack a dynamic correction mechanism. Furthermore, the software architecture has high coupling; upper-level strategies directly call physical angle values, requiring extensive software modifications and re-verification after replacing with different water valve models, resulting in high replacement costs.

[0021] To address the aforementioned technical problems, this technical solution provides a vehicle water valve adaptation method, apparatus, equipment, and medium based on feature self-learning. This method can automatically identify the replaced water valve model and load adaptation parameters even without a hardware identification interface. The following are various embodiments of this technical solution.

[0022] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating the steps of a vehicle water valve adaptation method based on feature self-learning.

[0023] This embodiment provides a vehicle water valve adaptation method based on feature self-learning, including: S101. In response to a preset triggering condition, a diagnostic excitation signal is sent to the multi-port water valve of the target vehicle, and the response current signal fed back by the multi-port water valve is obtained.

[0024] S102. Extract physical features from the response current signal.

[0025] S103. Select the target multi-way water valve with the corresponding physical characteristics from multiple multi-way water valves and determine the model of the target multi-way water valve.

[0026] S104. Obtain the control parameter set corresponding to the model of the target multi-way water valve, and control the operation of the multi-way water valve of the target vehicle according to the control parameter set.

[0027] Understandably, in this embodiment, firstly, the water valve response is actively triggered using a diagnostic excitation signal. When the vehicle's water valve is replaced, the system sends specific instructions to the multi-way water valve through diagnostic equipment, causing it to execute preset actions and collecting the response current signal from the water valve in real time. This process does not rely on any hardware coding pins or manual input; information is obtained entirely through electrical response. Secondly, physical features with model-distinguishing characteristics are extracted from the response current signal, such as the number of Hall pulses throughout the stroke, the stall current value, and the shape of the current rise curve. Different models of water valves exhibit a stable and distinguishable range in these features due to differences in mechanical structure and electromagnetic parameters. Thirdly, the extracted physical features are compared and filtered against a multi-model feature library pre-stored locally or in the cloud to automatically determine the model of the currently installed water valve. This matching process requires no manual intervention, avoiding the error risks associated with manual configuration. Finally, the corresponding set of control parameters, including angle range, Hall mapping coefficient, and temperature compensation reference, is automatically loaded based on the identified model to ensure that the electronic control unit can drive the water valve to operate according to the correct parameters. This technical solution uses electrical excitation and response characteristics as a bridge to replace hardware identification interfaces and manual configuration methods, achieving automated and highly reliable model identification and parameter adaptation.

[0028] In some embodiments, step S101 includes: A drive signal is sent to the multi-way water valve to make it move; during the movement of the multi-way water valve, the stroke Hall value and motor current of the multi-way water valve are collected multiple times; if the motor current is less than the preset stall threshold or the rising edge steepness of the motor current is lower than the preset steepness threshold when the stroke Hall value stops changing, it is determined that it is in a physical limit state and the stroke position is recorded; if the motor current reaches the stall threshold and the rising edge steepness is higher than the steepness threshold when the stroke Hall value stops changing, it is determined that it is in an abnormal jamming state; after repairing the abnormal jamming state, the multi-way water valve is controlled to move in the opposite direction to another physical limit, and the process of collecting the stroke Hall value and motor current of the multi-way water valve multiple times during its movement is returned.

[0029] Understandably, this embodiment avoids misjudging a stuck position as fully engaged by simultaneously monitoring changes in the Hall effect sensor count, motor current value, and current rise steepness. After correction, it automatically reverses to another limit position for relearning. This method improves the reliability and safety of self-learning, ensures the accuracy of water valve stroke range measurement, and provides reliable basic data for subsequent model identification and parameter adaptation.

[0030] In some embodiments, the motor is driven to perform a forward full-stroke scan according to the duty cycle. When the motor current exceeds the stall current threshold at multiple consecutive sampling points, it is determined that the forward physical limit has been reached, and the forward Hall count is recorded. The motor is driven to run in reverse to another physical limit according to the duty cycle, the reverse Hall count is recorded, and the full-stroke Hall count is calculated. Multiple scans are performed to obtain multiple full-stroke Hall counts, and the current rise steepness and current peak value at each stall moment are determined. If the fluctuation of the full-stroke Hall count is lower than the preset fluctuation threshold and the stall current peak value is within the preset range, the average value of all full-stroke Hall counts is taken as the final full-stroke Hall count; otherwise, self-learning is determined to have failed and a mechanical abnormality is reported.

[0031] Please refer to Figure 2 , Figure 2 This is a flowchart of the self-learning and type determination steps in a feature-based vehicle water valve adaptation method.

[0032] like Figure 2 As shown, a short, probing pulse (5% duty cycle, 50ms duration) is applied to the motor to check if it can rotate freely. If the current is abnormal (exceeding 3 times the normal starting current), it is determined to be mechanically stuck, and an error is reported and the process exits.

[0033] The motor is driven from one mechanical limit end to the other with a constant duty cycle. The cumulative number of Hall pulses H(t) and the motor current I(t) are recorded in real time. When the current I(t) exceeds the stall current threshold I_stall_base (the base threshold is calibrated according to room temperature) for three consecutive sampling points, it is determined that the positive physical limit has been reached, the Hall pulse count H_max_raw at this time is recorded, and the drive is stopped immediately.

[0034] Drive the motor in reverse with the same duty cycle to the limit on the other side, record the Hall effect count H_min_raw of the reverse full stroke, and calculate the preliminary Hall effect count H_total_raw = H_max_raw H_min_raw.

[0035] To eliminate jamming interference, three round-trip scans are performed, recording H_total_i for each scan, and analyzing the current rise steepness and peak value at each stall moment. If the fluctuation of the three measurements is less than 2%, and the stall current peak value is stable within I_stall_base ±10%, it is determined to be an effective physical limit, and the average value is taken as the final full-stroke Hall number H_total. If the fluctuation is too large or the current characteristics are abnormal, the error "Self-learning failed - mechanical abnormality" is reported.

[0036] Compare H_total with the pre-stored model feature range (e.g., Type A: 477-530, Type B: 1014-1114, Type C: 1300-1400). If a unique model is matched, the current water valve type is determined.

[0037] If multiple models are matched or no match is found, the feature vectors, including H_total, the stall current I_stall_actual recorded in this learning process, and the relationship between Hall pulse and time, are uploaded to the cloud server. The cloud server performs a similarity match in the water valve feature database and returns the closest model and its complete feature parameters. If no match is found in the cloud, it is treated as an unknown model, and conservative parameters are used to run the program while logging the results.

[0038] Based on the determined water valve type, load the complete control parameter table for that type from local or cloud sources, including: angle-Hall mapping table, basic PWM duty cycle, temperature compensation coefficient, stall current temperature correction coefficient, design life, etc. Simultaneously initialize the cumulative Hall count counter and establish a water valve health record.

[0039] Please refer to Figure 3 , Figure 3 This is a schematic diagram illustrating the current ripple fault diagnosis principle in a vehicle water valve adaptation method based on feature self-learning.

[0040] In some embodiments, based on the principle of current ripple fault diagnosis, a vehicle water valve adaptation method based on feature self-learning further includes the following steps: During the operation of the multi-way water valve, the motor current signal is collected and the ripple characteristic parameters of the motor current signal are extracted. The ripple characteristic parameters are compared one by one with the ripple characteristic templates in the preset fault characteristic library to determine the fault diagnosis result. When the discrete fault diagnosis reports a short circuit fault and the ripple characteristic parameters match the target ripple characteristic template, the short circuit fault diagnosis report result is determined. When the discrete fault diagnosis reports a short circuit fault and the ripple characteristic parameters do not match any ripple characteristic template, it is judged as a false alarm. When the discrete fault diagnosis does not report a fault but the ripple characteristic parameters match the insulation degradation ripple characteristic template, an insulation warning is triggered and maintenance is prompted. When the discrete fault diagnosis does not report a fault but the ripple characteristic parameters match the poor contact ripple characteristic template, it is recorded as an intermittent fault and the fault record data is uploaded to the cloud.

[0041] Understandably, this embodiment achieves accurate fault diagnosis by matching current ripple characteristic analysis with preset fault templates. Specific effects include: distinguishing between true short-circuit faults and false alarms in discrete fault diagnosis; identifying latent faults such as insulation aging and poor contact in advance when traditional diagnoses fail to report faults, triggering insulation warnings or recording intermittent faults and uploading them to the cloud. This method improves the accuracy and foresight of fault diagnosis, achieving an improvement from passive response to proactive early warning, helping to reduce the risk of sudden faults and supporting remote maintenance management.

[0042] Please refer to Figure 4 , Figure 4 This is a flowchart illustrating the steps involved in the repositioning of a multi-port water valve in a vehicle water valve adaptation method based on feature self-learning.

[0043] In some embodiments, a vehicle water valve adaptation method based on feature self-learning further includes the following steps: S201. When a system fault is detected, drive the multi-way water valve to move to the preset initial position.

[0044] S202. Collect the position feedback signal of the current position of the multi-way water valve and the motor current signal.

[0045] S203. When the position feedback signal indicates that the initial position has been entered and the motor current reaches the target stall current threshold, the multi-way water valve is determined to have successfully returned to its original position.

[0046] Understandably, this embodiment uses both position feedback signals and motor current signals to determine whether the homing was successful. Compared to solutions relying solely on position sensors, this method avoids misjudgments of homing caused by sensor drift or position feedback errors, ensuring that the water valve returns to a reliable mechanical zero point before power-off or after a fault. This approach improves the safety and reset reliability of the system after a fault, preventing control deviations or mechanical shocks upon the next power-on due to unclear position status.

[0047] In some embodiments, when the position feedback signal enters the tolerance zone corresponding to the initial position and the motor current does not reach the target stall current threshold within a preset monitoring time, the homing abnormality is determined and the fault is reported.

[0048] Please refer to Figure 5 , Figure 5 This is a flowchart illustrating the steps involved in fault location confirmation for a vehicle water valve adaptation method based on feature self-learning.

[0049] like Figure 5 As shown, when the system determines that the thermal management circuit needs to be disconnected, such as when a serious fault is detected, the vehicle is powered off and in sleep mode, or the user requests to disconnect the air conditioning, a safe return-to-normal procedure is executed: (1) Receive homing command (from fault diagnosis unit, hibernation management or upper layer application).

[0050] (2) The drive motor runs at a preset speed to the initial position (corresponding to 0% opening, generally the fully closed position).

[0051] (3) Monitor Hall position feedback H_feedback and motor current I_motor in real time.

[0052] (4) When H_feedback enters the initial position tolerance band (such as H_target0 ±5 pulse), continue driving and focus on current changes.

[0053] (5) Determine whether I_motor exceeds the corrected stall current threshold I_stall_comp (provided by the dynamic compensation unit) for three consecutive sampling points at this temperature. If it exceeds, it is determined that the mechanical limit has been reached, the return is successful, the drive is stopped and the return completion status is recorded.

[0054] (6) If H_feedback has reached the initial position but the current has not reached the stall threshold, it indicates that the position sensor may be drifting or the valve may not be in the correct position. In this case, continue driving and increase the current monitoring time. If the stall current is not detected after the maximum protection time (e.g., 2 seconds), the return is abnormal, the fault is reported and safety measures are taken (e.g., disconnect the high-voltage relay).

[0055] Therefore, this embodiment can ensure that the valve reliably returns to its physical initial position, avoiding safety vulnerabilities caused by sensor errors.

[0056] In some embodiments, a vehicle water valve adaptation method based on feature self-learning further includes the following steps: Obtain the current ambient temperature; determine the duty cycle compensation value corresponding to the current ambient temperature based on the preset temperature compensation coefficient table; adjust the drive duty cycle in the control parameter set using the duty cycle compensation value.

[0057] Specifically, the driving PWM duty cycle is corrected in real time according to the ambient temperature T, so as to ensure that the motor has consistent driving torque at different temperatures. The compensation formula is: Duty_comp = Duty_base × K_temp(T) Wherein, Duty_base is the basic duty cycle (calibrated value at normal temperature), and K_temp(T) is the temperature compensation coefficient, which is obtained through laboratory calibration. A typical curve is shown in Figure 6 , 7 , and segmented linear interpolation is adopted: T ≤ -40℃: K_temp = 1.8; -40℃<T ≤ -20℃: K_temp decreases linearly from 1.8 to 1.3; -20℃<T ≤ 20℃: K_temp decreases linearly from 1.3 to 1.0; 20℃<T ≤ 55℃: K_temp is kept at 1.0; T>55℃: K_temp decreases linearly from 1.0 to 0.9; In addition, the stall current threshold also needs to be corrected with temperature: the lubricating grease is viscous at low temperature, and the stall current will increase, therefore I_stall_comp = I_stall_base × K_I_temp(T), wherein K_I_temp(T) is calibrated according to experiments.

[0058] It can be understood that, in this embodiment, the corresponding duty cycle compensation value is obtained by querying a preset temperature compensation coefficient table, and the original control parameters are corrected, so as to offset the influence of factors such as changes in motor winding resistance and changes in lubricating oil viscosity on the valve movement speed and到位 accuracy at different temperatures. This measure improves the control consistency and response stability of the water valve over the full temperature range, avoids problems such as slow action at low temperature or overshoot at high temperature, and enhances the adaptability of the system to harsh environmental working conditions.

[0059] In some embodiments, a vehicle water valve adaptation method based on feature self-learning further comprises the following steps: acquiring the cumulative running Hall count of a multi-way water valve; determining an angle compensation amount through a preset wear angle correction function according to the cumulative running Hall count and an expected target opening; adjusting the target Hall count by using the angle compensation amount, and driving the multi-way water valve with the adjusted target Hall count, so as to compensate the angle deviation caused by gear wear.

[0060] Specifically, as the cumulative running Hall count N increases, the wear of the transmission mechanism leads to an increase in clearance, and the actual angle corresponding to the same percentage opening will be smaller. For this reason, a wear compensation function f_wear(N) is established to correct the target angle: θ_target_comp = θ_target × (1 + f_wear(N)) A typical curve for f_wear(N) is as follows: Figure 8 As shown, an exponentially decaying saturation form is adopted: f_wear(N) = a × (1 e^(-N / N0)) Where 'a' is the maximum wear compensation coefficient (e.g., 0.05, i.e., maximum compensation of 5%), and 'N0' is the wear time constant (e.g., 500,000 Hall effect time). The compensated angle θ_target_comp is then converted into the actual driving position through angle-Hall effect mapping.

[0061] Each time the system learns, it can recalibrate the wear coefficient: if the measured full-stroke Hall number H_total drifts from the initial value (e.g., wear causes the magnetic ring to loosen), the wear model parameters are updated.

[0062] Therefore, temperature compensation ensures reliable start-up and operation of the valve across the entire temperature range of -40℃ to 85℃; wear compensation maintains the valve's opening accuracy within ±2% throughout its entire lifespan, significantly improving the user experience.

[0063] Understandably, this embodiment can assess the wear degree of the water valve gear based on the cumulative Hall effect count and dynamically calculate the angle compensation amount using a preset wear angle correction function. During the target opening drive process, the target Hall effect count after adjusting this compensation amount is used for drive, thereby offsetting the deviation between the actual angle and the commanded angle caused by gear wear. This method achieves the maintenance of opening control accuracy throughout the entire life cycle of the water valve, avoids the expansion of valve positioning error caused by the accumulation of mechanical wear, extends the effective service life of the water valve, and maintains the control quality of the air conditioning and thermal management system.

[0064] In some embodiments, a vehicle water valve adaptation method based on feature self-learning further includes the following steps: Receive percentage opening command; convert the percentage opening command into the physical angle value corresponding to the model of the multi-way water valve; based on the mapping relationship between the angle corresponding to the model and the Hall number, convert the physical angle value into the target Hall number, and drive the multi-way water valve to achieve the target opening according to the target Hall number.

[0065] Specifically, all thermal management function modules (such as air conditioning controller and battery manager) only issue percentage opening instructions P_target (0%~100%), such as "Battery cooling mode requires water valve opening of 50%".

[0066] Based on the mechanical stroke angle range corresponding to the current water valve type, for example, type A: 0°~130°, type B: 0°~150°, type C: 0°~200°, convert P_target to the target physical angle θ_target: θ_target = (P_target / 100) × θ_max_type Where θ_max_type is the maximum mechanical angle of this type of water valve.

[0067] Then, based on the angle-Hall mapping relationship of this type of water valve, θ_target is converted into the target Hall pulse number H_target.

[0068] This demonstrates that regardless of the type of water valve installed at the vehicle's bottom layer, the upper-layer application software requires no modification, greatly reducing software maintenance costs and after-sales replacement workload.

[0069] Understandably, this embodiment can convert the percentage opening command issued by the upper-level controller into the physical angle value corresponding to the current water valve model, and then calculate the target Hall number based on the inherent angle-Hall number mapping relationship of that model to drive the water valve. This method achieves automatic conversion from general percentage commands to specific model physical parameters, shielding the differences in stroke range and mapping characteristics between different water valve models. This allows upper-level units such as air conditioning controllers to obtain accurate opening control simply by issuing a unified percentage opening request without needing to know the specific water valve model, thus reducing the coupling and adaptation complexity of the upper-level system.

[0070] Please refer to Figure 9 , Figure 9 This is an architecture diagram of a vehicle water valve adaptation system based on feature self-learning.

[0071] The present invention provides a multi-way water valve full life cycle adaptive control system comprising: Cloud server 100: Deployed in a remote data center, it contains a water valve feature database 101 and an OTA upgrade management module 102. The water valve feature database 101 stores feature parameters of multi-port water valves of different brands and models, including at least: full-stroke Hall effect number range, rated operating voltage, normal operating current range, stall current threshold, mechanical angle and Hall effect pulse correspondence, temperature-duty cycle compensation reference curve, and cumulative Hall effect number over design life. The OTA upgrade management module 102 is responsible for secure communication with the vehicle terminal, providing incremental updates and downloads of the feature database.

[0072] Vehicle terminal 200: integrated into the vehicle controller (VCU) or a separate thermal management domain controller, including a communication module 201, a local storage module 202, a control core module 203, a drive output module 204, and a signal acquisition module 205.

[0073] Communication module 201: Supports communication with other controllers in the vehicle via CAN / LIN / Ethernet, and data interaction with cloud server 100 via 4G / 5G or Wi-Fi. Supports local diagnostic communication of diagnostic instruments via OBD interface.

[0074] Local storage module 202: Stores a subset of water valve features that may be compatible with this vehicle model downloaded from the cloud, as well as the vehicle's water valve health profile (including current water valve type, cumulative Hall effect count, most recent self-learning results, wear coefficient, etc.) generated during the self-learning process.

[0075] Control core module 203: adopts a high-performance microcontroller (MCU) to run the core control algorithm of this invention, including a self-learning and type determination unit 2031, a fault diagnosis and prediction unit 2032, a homing control unit 2033, a dynamic compensation unit 2034, and a normalization interface unit 2035.

[0076] Drive output module 204: Includes H-bridge drive circuit or pre-drive chip, drives multi-way water valve motor according to PWM duty cycle signal output by control core module 203, and has current sampling amplification circuit to provide real-time feedback of motor current.

[0077] Signal acquisition module 205: Acquires pulse signals from the Hall sensor built into the multi-way water valve and converts them into position information; at the same time, it acquires the temperature value from the ambient temperature sensor (located near the valve body or provided by the vehicle network).

[0078] Multi-way water valve actuator 300: includes DC motor 301, reduction gear set 302, valve core assembly 303, and Hall position sensor 304. Hall sensor 304 outputs pulse signal to signal acquisition module 205 to determine the absolute position and relative movement of valve core.

[0079] Diagnostic instrument 400: During after-sales maintenance, it connects to the vehicle terminal 200 via the OBD interface, can send diagnostic commands (such as 31 service routine control) to trigger the self-learning process, and read the water valve health report.

[0080] In this embodiment, the self-learning and type determination unit 2031 is responsible for extracting the physical characteristics of the new water valve after it is installed on the vehicle or in the absence of a hardware ID, by driving the water valve to move throughout its entire stroke, and comparing it with the cloud feature library to determine the water valve model.

[0081] When the on-board terminal 200 detects that the multi-way water valve actuator 300 is powered on for the first time, or determines that a new water valve has been replaced by resetting the cumulative operating Hall effect count to zero, it automatically triggers self-learning. After the water valve is replaced by after-sales personnel, the service personnel send the 31 service (routine control) command through the diagnostic tool 400 to force entry into self-learning mode. When the cumulative operating Hall effect count reaches a preset threshold (e.g., every 100,000 Hall effect points), a short-stroke calibration is automatically triggered to update the wear coefficient.

[0082] In this embodiment, the fault diagnosis and prediction unit 2032 integrates traditional discrete fault codes and current ripple analysis to achieve multi-dimensional redundant diagnosis. The diagnostic function integrated within the drive output module 204 detects faults such as SCB (short circuit to ground), SCG (short circuit to power supply), and OL (open circuit) in real time, and reports them to the control core module 203 via SPI or digital signals.

[0083] During motor operation, the signal acquisition module 205 continuously acquires the motor current I(t) at a sampling rate of 10kHz. The control core module 203 performs FFT transformation or wavelet analysis on the current signal to extract the energy and waveform morphology of characteristic frequency bands. Various fault modes are pre-calibrated in the laboratory to establish a "fault-ripple feature library," for example: Normal operation: The current ripple exhibits fundamental frequency components related to Hall commutation, with few harmonics.

[0084] Short circuit to ground: The current spikes instantaneously, and the ripple exhibits extremely low-frequency, large-amplitude impacts.

[0085] Open circuit: Current drops to zero, ripple disappears.

[0086] Insulation degradation / leakage: 50Hz / 100Hz power frequency interference (if the vehicle power supply has ripple) or high frequency spikes occur.

[0087] Poor contact: The current drops intermittently, and the ripple envelope shows random pulses.

[0088] In this embodiment, the water valve feature database 101 of the cloud server 100 not only stores static feature parameters, but also continuously collects self-learning data and fault data reported by vehicles. Through big data analysis, the feature database is continuously optimized (such as updating the travel range and adding new models). The OTA upgrade management module 102 can periodically or when there is a need for adaptation to new vehicle models, push incremental feature database packages to the vehicle terminal 200, so that vehicles in use can also identify the latest water valve models.

[0089] In addition, this technical solution also provides a complete after-sales replacement scenario as an example to describe the workflow of the above-mentioned multi-way water valve full life cycle adaptive control system: (1) The maintenance personnel replaced the original Type A water valve (which is no longer in production) with a Type B water valve. The vehicle does not have a reserved hardware ID identification interface, so the model number cannot be read directly.

[0090] (2) The maintenance personnel send the 31 service self-learning command through the diagnostic instrument 400, and the vehicle terminal 200 enters the self-learning mode.

[0091] (3) Self-learning and type determination unit 2031 execution Figure 2 The process was followed, and the Hall effect count for the entire stroke was measured to be 1050, the stall current was 1.2A (at room temperature), and the current rise curve was recorded.

[0092] (4) The local storage module 202 has three types of data: A, B and C. 1050 falls within the range of type B (1014-1114), and the stall current is consistent with the nominal value of type B. Therefore, it is determined to be a type B water valve.

[0093] (5) Load the complete parameter table of the Type B water valve from local storage, including the angle range of 0-150°, the angle-Hall mapping linear coefficient, the temperature compensation reference, etc.

[0094] (6) Initialize the cumulative Hall counter and record information such as the replacement date and model.

[0095] Normal operation: ① The air conditioner controller requests 50% opening.

[0096] ② The normalized interface unit 2035 calculates the target angle θ_target = 50% × 150° = 75°.

[0097] ③ The wear compensation submodule calculates f_wear=0 and θ_target_comp=75° based on the current cumulative Hall effect count N (initially 0).

[0098] ④ The target Hall pulse count is obtained by angle-Hall mapping, and the drive output module 204 drives the motor to the target position with the duty cycle after current temperature compensation.

[0099] ⑤ During operation, the fault diagnosis unit 2032 continuously monitors the current ripple, and everything is normal.

[0100] Several months later, the cumulative Hall effect count reached 800,000, and the wear compensation coefficient gradually increased. During a certain operation, current ripple analysis revealed high-frequency spikes, which were determined to be a trend of insulation aging, triggering an "insulation warning" to remind the user to check during the next maintenance.

[0101] A year later, the D-type water valve was launched. With cloud-based push notifications of feature library updates, vehicles can download these updates via OTA (Over-The-Air) updates, enabling self-learning and adaptation for the D-type water valve.

[0102] Before each vehicle power-off, the return control unit 2033 executes the return procedure to ensure the valve returns to zero. On one occasion, due to sensor drift, the Hall feedback showed that the valve was in the zero position but the current did not reach the stall threshold. The system determined that the return was abnormal, recorded the fault, and maintained drive until it was truly in the correct position to prevent incorrect positioning on the next power-on.

[0103] Understandably, the combination of cloud database, local self-learning, and normalized interface allows for automatic identification of water valve models without hardware IDs, eliminating the need for upper-level software modifications. This significantly improves after-sales compatibility and reduces replacement costs by over 90%. The self-learning mechanism, employing a dual closed-loop system combining stroke and current verification, effectively eliminates misjudgments caused by stalling, increasing the self-learning success rate to 99.5%. The integration of current ripple analysis and discrete fault codes drastically reduces false alarm rates and enables predictive maintenance. A dual confirmation mechanism using backflip current and Hall effect position ensures reliable repositioning after a fault, enhancing functional safety. Dynamic compensation for temperature and duty cycle enables reliable valve startup across the entire temperature range, eliminating low-temperature startup failures. Dynamic correction of wear and angle keeps the opening error within ±2% throughout the entire lifespan, ensuring a consistent user experience. Furthermore, the normalized interface decouples the application layer from the hardware, significantly reducing software maintenance costs and shortening the development cycle.

[0104] Please refer to Figure 10 , Figure 10 This is a schematic diagram of a vehicle water valve adapter based on feature self-learning.

[0105] This embodiment also provides a vehicle water valve adapter based on feature self-learning, including: The transmission module 501 is used to send a diagnostic excitation signal to the multi-way water valve of the target vehicle in response to a preset trigger condition, and to acquire the response current signal fed back by the multi-way water valve.

[0106] Extraction module 502 is used to extract physical features from the response current signal.

[0107] The matching module 503 is used to select the target multi-way water valve with corresponding physical characteristics from multiple multi-way water valves and determine the model of the target multi-way water valve.

[0108] The control module 504 is used to obtain the control parameter set corresponding to the model of the target multi-way water valve, and control the operation of the multi-way water valve of the target vehicle according to the control parameter set.

[0109] It will be understood by those skilled in the art that all or some of the steps and apparatuses in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. As is known to those skilled in the art, communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0110] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0111] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement any of the above-mentioned vehicle water valve adaptation methods based on feature self-learning.

[0112] It is understood that the content of the above method embodiments is applicable to the embodiments of this electronic device. The specific functions implemented by the embodiments of this electronic device are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0113] This application also provides a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to implement the feature-based self-learning vehicle water valve adaptation method as described in any of the above specific embodiments.

[0114] This application also discloses a computer program product, including a computer program or computer instructions, which are stored in a computer-readable storage medium. The processor of the computer device reads the computer program or computer instructions from the computer-readable storage medium and executes the computer program or computer instructions, causing the computer device to perform the feature-based self-learning vehicle water valve adaptation method as described in any of the preceding embodiments.

[0115] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0116] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices. It should be understood that in this application, “at least one” means one or more, and “more than one” means two or more.

[0117] In the several embodiments provided in this application, it should be understood that the disclosed apparatus, devices, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0118] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0119] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0120] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0121] Although the description of this application has been quite detailed and particularly focused on several of the described embodiments, it is not intended to limit itself to any of these details or embodiments or any particular embodiment. Rather, it should be considered as effectively covering the intended scope of this application by referring to the appended claims and taking into account the prior art, which provides for a broad possible interpretation of these claims. Furthermore, the foregoing description of this application with respect to embodiments foreseeable by the inventors is intended to provide a useful description, and non-substantial modifications to this application that have not yet been foreseen may still represent equivalent modifications.

Claims

1. A vehicle water valve adaptation method based on feature self-learning, characterized in that, The method includes: In response to a preset triggering condition, a diagnostic excitation signal is sent to the multi-port water valve of the target vehicle, and the response current signal fed back by the multi-port water valve is acquired. Extract physical features from the response current signal; Select the target multi-way water valve corresponding to the physical characteristics from a plurality of multi-way water valves and determine the model of the target multi-way water valve; Obtain the control parameter set corresponding to the model of the target multi-way water valve, and control the operation of the multi-way water valve of the target vehicle according to the control parameter set.

2. The method according to claim 1, characterized in that, The process of responding to a preset trigger condition by sending a diagnostic excitation signal to the multi-way water valve of the target vehicle and acquiring the response current signal fed back by the multi-way water valve includes: Send a drive signal to the multi-way water valve to make the multi-way water valve move; During the movement of the multi-way water valve, the stroke Hall effect data and motor current of the multi-way water valve are collected multiple times. If the motor current is less than the preset stall threshold or the rising edge steepness of the motor current is lower than the preset steepness threshold when the stroke Hall number stops changing, then it is determined that the motor is in a physical limit state and the stroke position is recorded. If the motor current reaches the stall threshold and the rising edge steepness is higher than the steepness threshold when the stroke Hall number stops changing, it is determined to be in an abnormal stuck state. After repairing the abnormal jammed state, the multi-way water valve is controlled to move in the opposite direction to another physical limit, and then the process of repeatedly collecting the stroke Hall number and motor current of the multi-way water valve during its movement is resumed.

3. The method according to claim 1, characterized in that, The method further includes: During the operation of the multi-way water valve, the motor current signal is acquired and the ripple characteristic parameters of the motor current signal are extracted; The ripple feature parameters are compared one by one with the ripple feature templates in the preset fault feature library to determine the fault diagnosis result. When a discrete fault diagnosis report indicates a short-circuit fault and the ripple characteristic parameters match the target ripple characteristic template, the diagnosis report result for the short-circuit fault is determined. When a discrete fault diagnosis report indicates a short-circuit fault and the ripple characteristic parameters do not match any of the ripple characteristic templates, it is determined to be a false fault report. When discrete fault diagnosis does not report a fault but the ripple characteristic parameters match the insulation degradation ripple characteristic template, an insulation warning is triggered and maintenance is prompted. When the discrete fault diagnosis does not report a fault but the ripple characteristic parameters match the poor contact ripple characteristic template, it is recorded as an intermittent fault and the fault record data is uploaded to the cloud.

4. The method according to claim 1, characterized in that, The method further includes: When a system fault is detected, the multi-way water valve is driven to move to a preset initial position; Collect the position feedback signal of the current position of the multi-way water valve and the motor current signal; When the position feedback signal indicates that the initial position has been entered into the tolerance zone and the motor current reaches the target stall current threshold, the multi-way water valve is determined to have successfully returned to its original position.

5. The method according to claim 1, characterized in that, The method further includes: Get the current ambient temperature; Based on a preset temperature compensation coefficient table, determine the duty cycle compensation value corresponding to the current ambient temperature; The drive duty cycle in the control parameter set is adjusted using the duty cycle compensation value.

6. The method according to claim 1, characterized in that, The method further includes: Obtain the cumulative Hall effect count of the multi-way water valve; Based on the cumulative number of Hall effect sensors and the desired target opening, the angle compensation amount is determined using a preset wear angle correction function. The target Hall number is adjusted using the angle compensation amount, and the multi-way water valve is driven with the adjusted target Hall number to compensate for the angle deviation caused by gear wear.

7. The method according to claim 1, characterized in that, The method further includes: Receive percentage opening command; The percentage opening command is converted into the physical angle value corresponding to the model of the multi-way water valve; Based on the mapping relationship between the angle corresponding to the model and the Hall number, the physical angle value is converted into a target Hall number, and the multi-way water valve is driven to reach the target opening degree according to the target Hall number.

8. A vehicle water valve adapter based on feature self-learning, characterized in that, The device includes: The transmission module is used to send a diagnostic excitation signal to the multi-way water valve of the target vehicle in response to a preset trigger condition, and to acquire the response current signal fed back by the multi-way water valve. An extraction module is used to extract physical features from the response current signal; A matching module is used to filter out the target multi-way water valve corresponding to the physical characteristics from a plurality of multi-way water valves and determine the model of the target multi-way water valve; The control module is used to acquire the control parameter set corresponding to the model of the target multi-way water valve, and control the operation of the multi-way water valve of the target vehicle according to the control parameter set.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the vehicle water valve adaptation method based on feature self-learning as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the vehicle water valve adaptation method based on feature self-learning as described in any one of claims 1 to 7.

Images