A method and device for controlling a short drop of a frameless window

Abstract

This invention discloses a method and device for controlling the short-close of a frameless car window. The method includes: if the vehicle receives a door opening command, activating the window motor to perform a short-close of the frameless window; inputting real-time window operating status characteristics into a preset stall threshold prediction model to obtain a stall threshold; if the Hall pulse width of the window motor is greater than the stall threshold and the frameless window has not reached the preset short-close position, pausing the window motor to perform the short-close of the frameless window, heating the upper sealing strip of the frameless window, and then restarting the window motor to perform the short-close of the frameless window so that the frameless window reaches the preset short-close position. This method effectively solves the problem of frameless windows failing to close due to freezing in low-temperature environments. By accurately distinguishing between freezing resistance and "actual stall" through an adaptive stall threshold, and combining the operation logic of restarting the short-close after heating and melting the ice, it ensures that the door can be opened normally and improves the success rate of short-close of the frameless window under different operating conditions.

Description

Technical Field

[0001] This invention relates to the field of vehicle control technology, and in particular to a method and device for controlling the short-short descent of frameless vehicle windows. Background Technology

[0002] Currently, with the continuous improvement of automotive intelligence, frameless doors are widely used in mid-to-high-end models due to their aesthetically pleasing and sporty appearance. When opening a frameless door, the window glass needs to lower slightly—a short descent—to detach from the upper weatherstripping, thus preventing interference between the glass and the weatherstripping when the door is opened. The reliability of this short descent directly affects whether the door can be opened normally.

[0003] In existing technologies, window lift control mostly uses Hall effect motors. By comparing the real-time collected Hall pulse width with a fixed stall threshold, it is determined whether the window has stalled. This method can meet basic functional requirements under normal temperature conditions, but in extremely cold winter regions, the window glass is easily frozen to the sealing strip due to ice and snow, which significantly increases the running resistance during short-distance lowering. The fixed stall threshold is prone to misjudgment of stalling in extremely cold weather, causing the door to be unable to open normally.

[0004] Therefore, how to improve the success rate of short-distance descent of frameless windows under different working conditions while ensuring safety has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] To improve the success rate of short-sliding of frameless windows under different working conditions, this invention provides a method and device for controlling short-sliding of frameless windows.

[0006] In a first aspect, embodiments of the present invention provide a method for controlling the short-short descent of a frameless vehicle window, which may include: If the vehicle receives a door opening command, the window motor will be activated to briefly lower the frameless window. Input the real-time window operation status characteristics into the preset congestion threshold prediction model to obtain the congestion threshold; If the Hall pulse width of the window motor is greater than the stall threshold and the frameless window has not reached the preset short-descent position, then the window motor is paused to perform the frameless window short-descent and the upper sealing strip of the frameless window is heated. Then the window motor is restarted to perform the frameless window short-descent so that the frameless window reaches the preset short-descent position.

[0007] In one or more optional embodiments of this application, the following further includes: If the upper sealing strip of the frameless window is heated more than a preset number of times, the window motor will be terminated to allow the frameless window to briefly lower. Send the short-circuit window closure fault to the user's terminal.

[0008] In one or more optional embodiments of this application, heating the upper sealing strip of the frameless window includes: The heating module at the top of the frameless window is activated to heat the upper sealing strip of the window for a preset time; wherein the preset time is negatively correlated with the ambient temperature of the frameless window.

[0009] In one or more optional embodiments of this application, the window operation state characteristics include: The ambient temperature of the frameless window and the voltage, current, and current change rate of the window motor.

[0010] In one or more optional embodiments of this application, the preset stall threshold prediction model is obtained in the following manner: Obtain the historical vehicle window operation status features and corresponding actual stall thresholds to form a historical dataset; Construct an initial random forest regression model; Set the hyperparameters of the initial random forest regression model; The initial random forest regression model is trained based on the historical dataset to obtain the preset blocking threshold prediction model.

[0011] In one or more optional embodiments of this application, the hyperparameters include: Number of decision trees, depth of decision trees, number of split features selected in a single decision tree, and minimum number of leaves.

[0012] In one or more optional embodiments of this application, the following further includes: The number of decision trees ranges from 100 to 150. The depth of the decision tree ranges from 5 to 8. The number of split features selected in a single decision tree is 2; The minimum number of leaves is between 5 and 10.

[0013] Secondly, embodiments of the present invention provide a frameless vehicle window short-short-down control device, which may include: The starting module is used to activate the window motor to briefly lower the frameless window if the vehicle receives a door opening operation command. The prediction module is used to input the real-time characteristics of the vehicle window operation status into the preset congestion threshold prediction model to obtain the congestion threshold. The heating module is configured to, if the Hall pulse width of the window motor is greater than the stall threshold and the frameless window has not reached the preset short-descent position, pause the window motor to perform short-descent of the frameless window and heat the upper sealing strip of the frameless window, and then restart the window motor to perform short-descent of the frameless window so that the frameless window reaches the preset short-descent position.

[0014] Thirdly, embodiments of the present invention provide a computer-readable storage medium having a computer program / instruction stored thereon, which, when executed by a processor, implements the frameless window short-descent control method as described above.

[0015] Fourthly, embodiments of the present invention provide a computer program product, including a computer program / instruction, which, when executed by a processor, implements the frameless window short-descent control method as described above.

[0016] Fifthly, embodiments of the present invention provide a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the frameless window short-descent control method as described above.

[0017] The beneficial effects of the above-described technical solutions provided in the embodiments of the present invention include at least the following: This invention provides a frameless window short-descent control method. This method inputs real-time window operating status characteristics into a preset stall threshold prediction model during short-descent of the frameless window. This generates a stall threshold adapted to the current operating conditions. If the real-time Hall pulse width exceeds this stall threshold and the window has not reached the preset short-descent position, the window motor is paused, the upper sealing strip is heated, and the short-descent attempt is repeated until the window successfully descends to the preset position. Compared with existing technologies, this method effectively solves the problem of frameless windows failing to descend due to freezing in low-temperature environments. By using an adaptive stall threshold, it accurately distinguishes between freezing resistance and "actual stall." Combined with the operation logic of heating to melt the ice and then re-descending, it ensures normal door opening and improves the success rate of short-descent under different operating conditions. Simultaneously, this method avoids the defects of fixed stall thresholds causing misjudgment at low temperatures or missed judgment at high temperatures, balancing safety and reliability. Furthermore, through the closed-loop control logic of pausing stall, heating, and re-descending, the short-descent process is automatically completed without user intervention, improving user experience and vehicle intelligence.

[0018] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0019] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0020] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart illustrating the frameless window short-fall control method provided in an embodiment of the present invention. Figure 2 This is an example diagram of a decision tree structure provided in an embodiment of the present invention; Figure 3 A framework diagram of a frameless vehicle window short-descent control method provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the frameless window short-fall control device provided in an embodiment of the present invention. Detailed Implementation

[0021] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0022] The inventors discovered that in existing technologies, window lift control often employs a Hall effect motor solution. This method compares the real-time collected Hall pulse width with a fixed stall threshold to determine if the window is stalled. While this method meets basic functional requirements under normal temperature conditions, in extremely cold winter regions, window glass is prone to freezing onto the sealing strip due to ice and snow, significantly increasing operating resistance during short-distance lowering. The fixed stall threshold is prone to misjudging stalling in extremely cold weather, preventing the door from opening properly. Based on this, the inventors further developed this invention, providing a frameless window short-distance lowering control method and device.

[0023] Example 1 Embodiment 1 of the present invention provides a method for controlling the short-short descent of a frameless vehicle window, referring to... Figure 1 As shown, the method may include the following steps S101-S103: S101: If the vehicle receives a door opening command, the window motor will be activated to briefly lower the frameless window.

[0024] S102: Input the real-time window operation status characteristics into the preset stall threshold prediction model to obtain the stall threshold.

[0025] S103: If the Hall pulse width of the window motor is greater than the stall threshold and the frameless window has not reached the preset short-descent position, then the window motor is paused to perform the frameless window short-descent and the upper sealing strip of the frameless window is heated. Then the window motor is restarted to perform the frameless window short-descent so that the frameless window reaches the preset short-descent position.

[0026] This invention provides a frameless window short-descent control method. This method inputs real-time window operating status characteristics into a preset stall threshold prediction model during short-descent of the frameless window. This generates a stall threshold adapted to the current operating conditions. If the real-time Hall pulse width exceeds this stall threshold and the window has not reached the preset short-descent position, the window motor is paused, the upper sealing strip is heated, and the short-descent attempt is repeated until the window successfully descends to the preset position. Compared with existing technologies, this method effectively solves the problem of frameless windows failing to descend due to freezing in low-temperature environments. By using an adaptive stall threshold, it accurately distinguishes between freezing resistance and "actual stall." Combined with the operation logic of heating to melt the ice and then re-descending, it ensures normal door opening and improves the success rate of short-descent under different operating conditions. Simultaneously, this method avoids the defects of fixed stall thresholds causing misjudgment at low temperatures or missed judgment at high temperatures, balancing safety and reliability. Furthermore, through the closed-loop control logic of pausing stall, heating, and re-descending, the short-descent process is automatically completed without user intervention, improving user experience and vehicle intelligence.

[0027] In step S101 above, if the vehicle receives a door opening operation command, the window motor is activated to briefly lower the frameless window.

[0028] Specifically, when a user pulls the exterior door handle or triggers the key unlock signal, the body control system detects the door opening command and immediately sends a short-descent trigger signal to the window controller. Upon receiving this signal, the window controller immediately activates the window motor to drive the frameless window downwards from the fully closed position to detach from the weatherstripping on the upper part of the vehicle body. The goal of this short-descent process is to lower the glass to a preset short-descent position, which is typically defined as the minimum descent distance required for the window glass to completely detach from the weatherstripping. This ensures that the window glass does not interfere with the weatherstripping when the door is opened, while minimizing the impact on the sealed environment inside the vehicle.

[0029] In this embodiment, during the frameless window short-short-lowering operation of the window motor, the Hall pulse width of the window motor and the window operating status characteristics are also collected in real time. These window operating status characteristics include the ambient temperature of the frameless window and the voltage, current, and current change rate of the window motor.

[0030] Specifically, the window controller obtains the ambient temperature in real time from the ambient temperature sensor via the vehicle's CAN network or a dedicated sensor interface. The ambient temperature is measured in degrees Celsius (°C) and reflects the freezing risk level of the environment in which the frameless window is currently located.

[0031] Meanwhile, the window controller collects the voltage at the power supply terminal of the window motor in real time through a voltage detection circuit, measured in volts (V). This voltage can characterize the impact of fluctuations in the vehicle's power grid on the output of the window motor and the Hall pulse width. It also collects the current of the window motor in real time through a current sampling resistor or a Hall current sensor, measured in amperes (A). This current directly reflects the real-time load on the window motor during operation. Based on the continuously collected current values, the current change rate per unit time is further calculated, measured in amperes per second (A / s). This current change rate can characterize the dynamic trend of the window's operating resistance, helping to distinguish between sudden stalls (such as impacts with hard objects) and gradual increases in resistance (such as freezing or aging of the rubber strip).

[0032] The aforementioned window operation status characteristic data are continuously collected at a preset sampling frequency to provide real-time input for subsequent stall threshold prediction.

[0033] In addition, the window controller also collects the Hall pulse width of the window motor in real time through a Hall sensor, which is inversely proportional to the rotational speed of the window motor.

[0034] In step S102 above, the real-time window operating status characteristics are input into a preset stall threshold prediction model to obtain the stall threshold. The window operating status characteristics include the ambient temperature of the frameless window and the voltage, current, and current change rate of the window motor.

[0035] Specifically, this can be achieved by combining the real-time collected ambient temperature, voltage, current, and current change rate into a four-dimensional feature vector, which is then input into a pre-trained preset stall threshold prediction model.

[0036] The preset congestion threshold prediction model is a random forest regression model that integrates multiple decision trees. When the preset congestion threshold prediction model receives a four-dimensional feature vector, each decision tree within the model independently makes a judgment and prediction, and each outputs a predicted value based on the splitting rule of that decision tree. The preset congestion threshold prediction model arithmetically averages the predicted values ​​output by all decision trees, and uses the average value as the final congestion threshold output.

[0037] The stall threshold output is in milliseconds (ms), representing the Hall pulse width threshold that the window controller should use under the current ambient temperature, voltage, current, and rate of change of current. The entire model prediction process is completed within milliseconds to meet the real-time requirements of the window short-fall control.

[0038] In this embodiment of the application, the above-mentioned preset stall threshold prediction model can be obtained through the following steps S1021-S1024: S1021: Obtain the operating status features of multiple historical vehicle windows and the corresponding actual stall thresholds to form a historical dataset.

[0039] Specifically, before training the model, a large amount of historical data needs to be collected as training samples. Each training sample contains a set of input features and corresponding output labels. The input features are the core parameters affecting the stall threshold, namely the historical window operating status characteristics, including ambient temperature, voltage, current, and current change rate. The output labels are the actual measured stall thresholds under the corresponding operating conditions of the historical window operating status characteristics.

[0040] Data acquisition can be achieved through both bench calibration experiments and real-vehicle road tests. For bench calibration experiments, the window regulator mechanism is placed in an environmental test chamber with different temperature gradients (e.g., -30℃ to 25℃) and different voltages (e.g., 9V to 16V). Under different operating conditions, the window operation is controlled to artificially induce stalling, and the Hall pulse width value at the moment of stalling is recorded as the true stall threshold under that condition. For real-vehicle road tests, data acquisition equipment is installed on the test vehicle to record natural stalling events occurring during actual use (e.g., winter freezing, rubber strip aging, etc.). Simultaneously, the ambient temperature, voltage, current, current change rate, and the true Hall pulse width value at the moment of stalling are collected.

[0041] Using the methods described above, a large amount of data covering different operating conditions such as different temperatures and voltages is collected to form the original historical dataset.

[0042] In addition, to ensure data quality, the original historical dataset needs to be preprocessed. The preprocessing steps may include at least outlier removal and feature standardization.

[0043] Outlier removal employs the 3σ principle, deleting outlier data (such as instantaneous voltage spikes, abnormal current jumps, calibration errors, etc.) that deviate more than three standard deviations from the mean to avoid interfering with model training. Feature standardization normalizes the four input features—ambient temperature, voltage, current, and rate of change of current—to eliminate the influence of different units and numerical ranges. The standardization formula is shown below:

[0044] In the formula, These are the standardized eigenvalues. These are the original eigenvalues. The mean of the corresponding feature. is the standard deviation of the corresponding feature.

[0045] After preprocessing, the historical dataset is obtained.

[0046] S1022: Construct the initial random forest regression model.

[0047] Specifically, after preparing the historical dataset, an initial random forest regression model needs to be constructed as the basic architecture for the pre-set blocking threshold prediction model. The core of model construction is determining the model's basic framework and algorithmic logic, including: selecting random forest as the ensemble learning algorithm type, setting decision trees as the base learners, defining the model's input layer to accept four features (ambient temperature, voltage, current, and rate of change of current), defining the model's output layer to output a scalar value (the blocking threshold), and configuring the model's internal ensemble strategy to take the arithmetic mean of all decision tree predictions. Once the above framework is determined, an initial random forest regression model, which has not yet been trained and whose internal parameters are in their initial state, is obtained, awaiting subsequent steps to input training data for parameter learning.

[0048] Random forest is an ensemble learning algorithm whose core idea is to train multiple decision trees simultaneously, allowing each decision tree to make predictions independently, and finally taking the average of the predictions from all decision trees as the final output.

[0049] The randomness of random forests manifests in two aspects: first, sample randomness, meaning that the training data for each decision tree is randomly selected from the training set through bootstrap sampling, resulting in slightly different training samples for each tree; second, feature randomness, meaning that when splitting a node, each tree does not use all input features, but randomly selects a subset from all input features to determine the split. These two random mechanisms cause some differences in the prediction results of each decision tree. The ensemble random forest regression model effectively offsets the prediction bias of individual decision trees, ensuring prediction accuracy while minimizing the risk of overfitting.

[0050] To facilitate understanding by those skilled in the art, the decision tree structure in the random forest regression model is briefly introduced here: (Refer to...) Figure 2 As shown, each decision tree consists of three types of nodes: root node, internal nodes, and leaf nodes. The root node is the entry point for all samples, where all samples first converge. Then, based on the splitting features and thresholds of the root node, binary classification begins. Internal nodes are the decision nodes, where each internal node performs binary classification on the samples again based on the selected splitting features and thresholds to divert the samples to different child nodes. Leaf nodes are the final destinations, where samples no longer split after reaching a leaf node. The average of the true blocking thresholds of all training samples that previously fell into that leaf node is taken as the predicted blocking threshold for new samples falling into that leaf node.

[0051] S1023: Set the hyperparameters of the initial random forest regression model. These hyperparameters should include at least the number of decision trees, the depth of the decision trees, the number of features selected per split in a single decision tree, and the minimum number of leaves.

[0052] Specifically, before model training, the hyperparameters of the initial random forest regression model need to be set to ensure that the trained model predicting the pre-set stall threshold can adapt to the embedded hardware platform of the window controller while maintaining prediction accuracy, meeting the millisecond-level real-time control requirements of the window descent process. Therefore, the setting of each hyperparameter needs to comprehensively consider prediction accuracy, model complexity, and the computational resource limitations of the vehicle controller.

[0053] The hyperparameters should include at least the number of decision trees, the depth of the decision trees, the number of split features selected per decision tree, and the minimum number of leaves. The specific settings are as follows: The number of decision trees should range from 100 to 150. Too few decision trees will lead to underfitting of the model, insufficient prediction accuracy, and an inability to accurately reflect the changing patterns of stall thresholds under different operating conditions. Too many decision trees will increase computational complexity, leading to increased inference time and failing to meet the millisecond-level real-time requirements of the window descent process. Experiments have verified that a range of 100 to 150 decision trees achieves a good balance between prediction accuracy and computational efficiency.

[0054] The depth of a decision tree ranges from 5 to 8 layers. The depth is defined as the maximum path length from the root node to the bottommost leaf node. A shallow depth (less than 5 layers) leads to insufficient learning ability, underfitting, and low prediction accuracy. An excessively deep depth (more than 8 layers) increases complexity and computation time, and also increases the risk of overfitting, reducing the model's ability to generalize to unknown scenarios. Experiments have shown that limiting the decision tree depth to 5 to 8 layers ensures both predictive ability and controlled inference time.

[0055] The number of features selected for splitting a single decision tree is set to 2. According to the rule, when splitting a node in each decision tree, the number of randomly selected features is the square root of the total number of features. The preset stall threshold prediction model has four input features (ambient temperature, voltage, current, and current change rate), and the square root is determined to be 2. Therefore, the number of features selected for splitting a single decision tree is set to 2. This setting effectively reduces the overfitting risk of a single decision tree and improves the overall robustness and generalization ability of the model.

[0056] The minimum number of leaves should range from 5 to 10. The minimum number of leaves refers to the minimum number of training samples contained in each leaf node. Setting this value too small will cause the decision tree to overfit the detailed noise in the training set, reducing generalization ability; setting it too large will result in an overly coarse decision tree with underfitting. Validation on a debug set shows that controlling the minimum number of leaves within 5 to 10 achieves a balance between fitting accuracy and generalization ability, while reducing unnecessary computation and adapting to the embedded platform requirements of the window controller.

[0057] S1024: Train the initial random forest regression model based on the historical dataset to obtain a prediction model for the preset blocking threshold.

[0058] Specifically, after completing the construction and hyperparameter setting of the initial random forest regression model, the initial random forest regression model needs to be trained using historical datasets so that it learns the mapping relationship between the four input features and the blocking threshold.

[0059] Before the training process begins, the historical dataset obtained in step S1021 is first divided into three parts in a 7:2:1 ratio: 70% is used as the training set for model parameter learning; 20% is used as the debugging set for model hyperparameter optimization and bias correction; and 10% is used as the test set for final performance verification. Random sampling is used during the division to ensure that the three parts of the data have a consistent distribution and can cover different combinations of operating conditions such as temperature, voltage, and current.

[0060] During the model training phase, historical vehicle window operation status features from the training set are used as input, and the corresponding actual blocking thresholds are used as supervision labels. These are then fed into the initial random forest regression model for training. The core objective of training is to enable the model to learn the mapping relationship between input features and output thresholds, minimizing the deviation between the model's predictions and the actual values ​​for the training set samples. During training, each decision tree randomly selects samples from the training set based on Bootstrap sampling, and randomly selects a subset of features at each node split. The trees grow progressively according to the hyperparameters set in step S1023 above, until all decision trees are constructed.

[0061] During training, after each round or certain training phase, the current model is validated on a test set, and the deviation between the predicted stall threshold and the actual stall threshold in the test set is calculated. The test set is not used for direct learning of model parameters; it is only used to evaluate the model's performance on unseen data. The purpose is to prevent the model from overfitting the training set and to improve its predictive ability for unknown operating conditions.

[0062] After training, the historical window operation status features from the test set are input into the preset congestion threshold prediction model to obtain the predicted congestion threshold. This predicted congestion threshold is then compared with the actual congestion threshold in the test set, and the root mean square error (RMSE) is calculated as the evaluation metric for the final performance of the preset congestion threshold prediction model. The RMSE of the preset congestion threshold prediction model on the test set is required to be no more than 5 milliseconds to ensure that the model's prediction accuracy meets the actual requirements of window short-descent control. The test set was not used throughout the training process, thus objectively reflecting the model's final generalization ability.

[0063] After completing the above training, validation, and testing steps, the model is solidified into a set of deployable algorithm parameters and structure files. This can then be used as a preset congestion threshold prediction model and deployed to the embedded software of the vehicle window controller for subsequent real-time congestion threshold prediction. Since the random forest model only needs to execute predetermined decision paths and read node stored values ​​after training, without requiring complex calculations, it is suitable for deployment and application on automotive embedded platforms with limited computing power.

[0064] In step S103 above, if the Hall pulse width of the window motor is greater than the stall threshold and the frameless window has not reached the preset short-descent position, the window motor is paused to perform the frameless window short-descent and the upper sealing strip of the frameless window is heated. Then, the window motor is restarted to perform the frameless window short-descent so that the frameless window reaches the preset short-descent position.

[0065] Specifically, during the short-term lowering of the window, the window controller continuously compares the real-time collected Hall pulse width with the stall threshold obtained in step S102. When the real-time Hall pulse width is detected to exceed the stall threshold at the corresponding moment, it indicates that the window's operating resistance has exceeded the normal range allowed under the current operating conditions, suggesting a possible stall due to freezing.

[0066] At this point, the window controller first determines whether the frameless window has descended to the preset short-descent position. If it has, the short-descent of the frameless window is completed normally. If not, the anti-freeze protection logic is activated, and the window controller immediately sends a pause command to the window motor to stop its operation, preventing overheating or damage to mechanical components caused by forced operation under excessive resistance. Simultaneously, the window controller sends a start command to the heating module located near the upper sealing strip of the frameless window to heat the upper sealing strip and melt any existing ice.

[0067] After the heating continues for a preset time, the window controller restarts the window motor to perform a short-term descent of the frameless window and executes step S102 to obtain a new stall threshold. The real-time Hall pulse width is compared with the new stall threshold. If the Hall pulse width drops below the new stall threshold, it indicates that the ice has melted and the resistance has returned to normal, and the window continues to descend until it reaches the preset short-term descent position. If the Hall pulse width is still greater than the new stall threshold, the steps of pausing the window motor, heating the sealing strip, and restarting the frameless window short-term descent are repeated until the frameless window reaches the preset short-term descent position or reaches the preset maximum number of heating cycles.

[0068] In this embodiment of the application, it is further included that if the number of times the upper sealing strip of the frameless window is heated is greater than a preset number, the window motor is terminated from short-term lowering of the frameless window, and the short-term lowering fault of the window is sent to the user's operating terminal.

[0069] Specifically, during the frameless window's short-descent process, the window controller has a built-in counter to record the number of retries to heat the upper sealing strip. Each time the window motor is paused and the heating module is started, the counter automatically increments. After one heating cycle, when the window motor is restarted for the frameless window's short-descent, if the real-time Hall pulse width is still greater than the corresponding stall threshold and the window has not yet reached the preset short-descent position, and the counter has counted three times, it indicates that after multiple heating attempts, the resistance cannot be overcome, and the frameless window cannot descend normally to the preset short-descent position. At this point, the window controller determines it as an unrecoverable short-descent fault, immediately terminates the window motor operation, and stops attempting to descend further.

[0070] Simultaneously, the window controller generates a short-short-down window fault message and sends it to the user's operating system via the vehicle bus, including the instrument panel, central control display, and the user's mobile client. The message displays a text or icon indicating that the frameless window short-short-down function is malfunctioning and suggests checking the door seals for severe freezing or mechanical jamming. Upon receiving the fault message, the user can promptly take measures such as manual de-icing or contacting repair services to avoid overheating of the motor or damage to mechanical components due to repeated attempts.

[0071] In step S103 above, the upper sealing strip of the frameless window can be heated in the following way: Activating the heating module at the top of the frameless window heats the upper sealing strip of the window for a preset time. This preset time is negatively correlated with the ambient temperature of the frameless window.

[0072] Specifically, the heating module can be a heating wire embedded in the upper frame of the frameless window. This heating wire is arranged along the upper edge of the window and is in close contact with the upper sealing strip. When the window controller issues a heating command, it supplies power to the heating wire. After the heating wire is energized, it generates heat, which locally heats the upper sealing strip in contact with it through heat conduction, causing the temperature of the sealing strip to rise and the surface ice layer to melt.

[0073] Meanwhile, the preset heating time is not fixed but dynamically adjusted according to the current ambient temperature: the lower the ambient temperature, the thicker or more firmly the ice layer, the longer the heating time required; when the ambient temperature is higher, the ice layer is thinner or only slightly frozen, and the required heating time is correspondingly shorter. For example, a temperature-time mapping table can be preset in the window controller. When the ambient temperature is below -20℃, the preset heating time is 30 seconds; when the ambient temperature is between -10℃ and -20℃, the preset heating time is 20 seconds; when the ambient temperature is between 0℃ and -10℃, the preset heating time is 10 seconds; and when the ambient temperature is above 0℃, a shorter heating time can be set.

[0074] By using this heating preset time control method that is negatively correlated with ambient temperature, it ensures that there is enough time to melt the ice layer under extreme low temperature conditions, while avoiding excessive heating time in light freezing conditions that may affect user experience or cause unnecessary energy consumption.

[0075] To facilitate understanding by those skilled in the art, a clearer and more complete explanation of the frameless window short-sliding control method is provided here in conjunction with a framework diagram: (Refer to...) Figure 3 As shown, the window controller continuously monitors whether it receives a door opening command. If no command is received, it continues to wait; if a command is received, it activates the window motor to perform a short-term descent of the frameless window.

[0076] During the short-descent process, the Hall pulse width is monitored in real time to see if it exceeds the stall threshold. If it remains below the threshold, the frameless window short-descent ends after reaching the preset short-descent position. If the Hall pulse width is less than or equal to the stall threshold at any moment, and the window has not yet reached the preset short-descent position, the window motor is paused, and the heating module is activated to heat the upper sealing strip. After heating is complete, the window motor is restarted to continue the short-descent process.

[0077] Before each heating cycle, it checks whether the number of heating cycles exceeds the preset number (three times). If it does not exceed the preset number, it continues to pause the window motor operation, heat, and restart the frameless window short-lowering cycle. If it exceeds the preset number, it terminates the frameless window short-lowering and sends the window fault information to the user's terminal.

[0078] Example 2 Based on the same inventive concept, embodiments of the present invention also provide a frameless vehicle window short-lowering control device, referring to... Figure 4 As shown, the device includes: The starting module 101 is used to start the window motor to briefly lower the frameless window if the vehicle receives a door opening operation command. The prediction module 102 is used to input the real-time window operation status characteristics into the preset congestion threshold prediction model to obtain the congestion threshold. The heating module 103 is used to pause the window motor from performing a short-distance descent of the frameless window and heat the upper sealing strip of the frameless window if the Hall pulse width of the window motor is greater than the stall threshold and the frameless window has not reached the preset short-distance descent position, and then restart the window motor to perform a short-distance descent of the frameless window so that the frameless window reaches the preset short-distance descent position.

[0079] Example 3 Based on the same inventive concept, embodiments of the present invention also provide a computer-readable storage medium storing a computer program / instruction thereon, which, when executed by a processor, implements the frameless window short-descent control method as described in Embodiment 1 above.

[0080] Example 4 Based on the same inventive concept, embodiments of the present invention also provide a computer program product, including a computer program / instruction, which, when executed by a processor, implements the frameless window short-descent control method as described in Embodiment 1 above.

[0081] Example 5 Based on the same inventive concept, embodiments of the present invention also provide a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the frameless window short-descent control method as described in Embodiment 1 above.

[0082] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0083] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0084] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0085] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0086] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for controlling the short-short descent of a frameless vehicle window, characterized in that, The method includes: If the vehicle receives a door opening command, the window motor will be activated to briefly lower the frameless window. Input the real-time window operation status characteristics into the preset congestion threshold prediction model to obtain the congestion threshold; If the Hall pulse width of the window motor is greater than the stall threshold and the frameless window has not reached the preset short-descent position, then the window motor is paused to perform the frameless window short-descent and the upper sealing strip of the frameless window is heated. Then the window motor is restarted to perform the frameless window short-descent so that the frameless window reaches the preset short-descent position.

2. The method according to claim 1, characterized in that, Also includes: If the upper sealing strip of the frameless window is heated more than a preset number of times, the window motor will be terminated to allow the frameless window to briefly lower. Send the short-circuit window closure fault to the user's terminal.

3. The method according to claim 1, characterized in that, The heating of the upper sealing strip of the frameless window includes: The heating module at the top of the frameless window is activated to heat the upper sealing strip of the window for a preset time; wherein the preset time is negatively correlated with the ambient temperature of the frameless window.

4. The method according to claim 1, characterized in that, The characteristics of the vehicle window's operating status include: The ambient temperature of the frameless window and the voltage, current, and current change rate of the window motor.

5. The method according to claim 1, characterized in that, The preset stall threshold prediction model is obtained in the following manner: Obtain the historical vehicle window operation status features and corresponding actual stall thresholds to form a historical dataset; Construct an initial random forest regression model; Set the hyperparameters of the initial random forest regression model; The initial random forest regression model is trained based on the historical dataset to obtain the preset blocking threshold prediction model.

6. The method according to claim 5, characterized in that, The hyperparameters include: Number of decision trees, depth of decision trees, number of split features selected in a single decision tree, and minimum number of leaves.

7. The method according to claim 6, characterized in that, Also includes: The number of decision trees ranges from 100 to 150. The depth of the decision tree ranges from 5 to 8. The number of split features selected in a single decision tree is 2; The minimum number of leaves is between 5 and 10.

8. A frameless vehicle window short-short-down control device, characterized in that, include: The starting module is used to activate the window motor to briefly lower the frameless window if the vehicle receives a door opening operation command. The prediction module is used to input the real-time characteristics of the vehicle window operation status into the preset congestion threshold prediction model to obtain the congestion threshold. The heating module is configured to, if the Hall pulse width of the window motor is greater than the stall threshold and the frameless window has not reached the preset short-descent position, pause the window motor to perform short-descent of the frameless window and heat the upper sealing strip of the frameless window, and then restart the window motor to perform short-descent of the frameless window so that the frameless window reaches the preset short-descent position.

9. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instruction is executed by the processor, it implements the frameless window short-descent control method according to any one of claims 1-7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the frameless window short-descent control method according to any one of claims 1-7.

Images