Dynamic response method adopting automobile intelligent windshield wiper direct current control

Abstract

The invention provides a dynamic response method adopting automobile intelligent windshield wiper direct current control, which comprises the following steps: generating initial curved surface feature distribution information by analyzing curved surface area friction force and key point resistance distribution, and analyzing a swing angle change rate and contact surface pressure distribution in combination with a real-time position calibration value; obtaining detailed resistance distribution data of each key point; calculating a swing force balance value and a regional resistance compensation value, and determining a forward trajectory correction value and a reverse trajectory adjustment value; a dynamic motion track parameter is generated by combining the speed optimization weight ratio, the track parameter updating rate is updated, and adjusted windshield wiper arm motion control data is obtained; the windshield wiper arm is driven to execute the dynamic track, and if the pressure distribution deviates from a preset threshold value, the swing force balance value is finely adjusted; and analyzing the action of the friction force of the curved surface region on the long-term operation stability, and judging whether the track parameter update rate meets the adaptive requirement or not by combining the region resistance compensation value to obtain a final long-term parameter adjustment range.

Description

[0001] Technical Field This invention relates to the field of information technology, and in particular to a dynamic response method using DC control of intelligent automotive windshield wipers. Background Technology

[0002] As a crucial component for enhancing the safety and comfort of modern vehicles, the intelligent windshield wiper system holds a key position in the research of intelligent driving and vehicle assistance systems. The dynamic response capability of the wiper system directly affects the driver's visibility in adverse weather conditions, especially in complex road conditions. Efficient wiper control technology can significantly reduce the risk of accidents, becoming an indispensable part of ensuring driving safety.

[0003] However, current windshield wiper control methods often have limitations when dealing with complex environments. Many systems struggle to adapt to the different shapes of windshields in various vehicles, and cannot flexibly adjust to dynamic changes during actual use. This limitation can lead to uneven wiping or low efficiency during operation, especially when the glass surface is irregular, traditional control methods cannot guarantee the comprehensiveness and stability of the wiping effect.

[0004] In this field, the core technical challenges lie in accurately controlling the angle changes of the wiper arm and adapting it to the asymmetrical characteristics of the windshield. The wiper arm needs precise angle sensing to determine its position and trajectory during its reciprocating motion. However, because the curvature of the windshield is often asymmetrical, the resistance and contact area experienced by the wiper differ when it swings forward and backward. This difference further makes it difficult for the control system to balance the speed and force of the forward and reverse swings when designing the motion trajectory, resulting in incomplete wiping or excessive wear in some areas. For example, in some vehicles, the wiper may experience a stutter when swinging from left to right due to changes in the glass curvature, while the reverse swing may result in uneven wiping marks due to uneven speed, directly affecting the clarity of the driver's vision.

[0005] Therefore, designing a control strategy that can dynamically adjust the forward and reverse oscillation trajectory of the wipers, taking into account the asymmetrical curved surface characteristics of windshields, has become a key issue in improving the performance of intelligent wiper systems. Solving this problem requires not only attention to the accuracy of angle perception but also continuous adaptation to subtle differences in the glass shape during long-term operation to ensure that the wipers maintain efficient and stable operation under various conditions. Summary of the Invention

[0006] This invention provides a dynamic response method using DC control of automotive intelligent windshield wipers, mainly comprising:

[0007] High-frequency data acquisition is performed on the asymmetric curved surface of the windshield using a sensor array. Real-time analysis of friction and resistance distribution at key points within the curved surface is conducted to generate initial surface feature distribution information, determining the forward oscillation resistance value and the reverse oscillation speed difference on the glass surface. Based on the surface feature distribution information, combined with real-time position calibration values ​​obtained from the wiper arm angle sensing module, the rate of change of oscillation angle and contact surface pressure distribution are analyzed to determine the fluctuation of the forward oscillation resistance value and the reverse oscillation speed difference within the speed deviation range, obtaining detailed resistance distribution data for each key point. For the detailed resistance distribution data of each key point, a preset curved surface region division method is used to divide the glass surface into multiple independent regions. The oscillation force balance value and region resistance compensation value for each region are calculated to determine the forward trajectory correction amount and the reverse trajectory adjustment value. Using the forward trajectory correction amount and the reverse trajectory adjustment value, combined with the speed optimization weight ratio, dynamic motion trajectory parameters for each curved surface region are generated. The trajectory parameter update rate is updated in real time to obtain the adjusted wiper arm motion control data. Based on the adjusted wiper arm motion control data, the wiper arm is driven to execute a dynamic trajectory. The impact of contact surface pressure distribution on wiping uniformity is monitored in real time. If the pressure distribution in a certain area deviates from a preset threshold, the oscillation force balance value is fine-tuned through dynamic feedback adjustment to determine the updated control command. For the updated control command, the operating status of the wiper arm in different curved surface areas is continuously recorded. The effect of friction in the curved surface area on long-term operational stability is analyzed. Combined with the area resistance compensation value, it is determined whether the trajectory parameter update rate meets the adaptability requirements, resulting in the final long-term parameter adjustment range. Using the final long-term parameter adjustment range, the dynamic feedback adjustment value and speed optimization weight ratio are periodically optimized to generate the final motion trajectory control logic adapted to the characteristics of asymmetric curved surfaces, completing the adaptive operation adjustment of the wiper arm on complex curved surfaces.

[0008] Furthermore, the oscillation angle change rate and contact surface pressure distribution are analyzed to determine the fluctuation of the forward oscillation resistance value and the reverse oscillation speed difference within the speed deviation range, obtaining detailed resistance distribution data for each key point. Further, the glass surface is divided into multiple independent regions, and the oscillation force balance value and region resistance compensation value for each region are calculated to determine the forward trajectory correction amount and reverse trajectory adjustment value. Further, dynamic motion trajectory parameters are generated for each curved surface region, and the trajectory parameter update rate is updated in real time to obtain the adjusted wiper arm motion control data. Further, the influence of contact surface pressure distribution on wiping uniformity is monitored in real time. If the pressure distribution in a certain region deviates from a preset threshold, the oscillation force balance value is fine-tuned through dynamic feedback adjustment to determine the updated control command. Further, the effect of surface friction on long-term operational stability is analyzed. Combined with the region resistance compensation value, it is determined whether the trajectory parameter update rate meets the adaptability requirements, obtaining the final long-term parameter adjustment range. Further, the final motion trajectory control logic adapted to the characteristics of asymmetrical curved surfaces is generated to complete the adaptive operation adjustment of the wiper arm on complex curved surfaces.

[0009] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0010] This invention discloses a dynamic response method for DC control of automotive intelligent windshield wipers, aiming to solve the operational problems of uneven wiping and insufficient long-term operational stability caused by uneven distribution of wiper arm movement trajectory and frictional resistance in complex curved surface environments. This invention utilizes high-frequency data acquisition and real-time analysis of curved surface feature distribution, combined with wiper arm angle sensing and dynamic trajectory parameter adjustment, to accurately calculate the resistance compensation value and trajectory correction amount for each region, generating adaptive motion control logic. Simultaneously, this invention ensures uniform pressure distribution on the contact surface through dynamic feedback adjustment and long-term parameter optimization, improving wiping performance and maintaining long-term operational stability. Ultimately, this invention achieves adaptive operation of the wiper arm on asymmetric curved surfaces, significantly improving wiping uniformity and equipment durability, providing an efficient solution for wiper systems in complex curved surface environments. Attached Figure Description

[0011] Figure 1 This is a flowchart of a dynamic response method using DC control of an intelligent automotive windshield wiper according to the present invention. Detailed Implementation

[0012] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0013] like Figure 1 This embodiment of a dynamic response method using DC control of automotive intelligent windshield wipers may specifically include:

[0014] S101 uses a sensor array to collect high-frequency data on the asymmetric curved surface of the windshield, performs real-time analysis on the friction force and key point resistance distribution in the curved surface area, generates initial curved surface feature distribution information, and determines the positive oscillation resistance value and the reverse oscillation speed difference of the glass surface.

[0015] In this embodiment of the application, a corresponding technical process is constructed around the data acquisition and analysis target of the sensor array on the asymmetric curved surface of the windshield, combined with multiple interrelated attributes.

[0016] First, high-frequency data is collected from the asymmetric curved surface of the windshield using a sensor array. Friction values ​​and key points in each region of the surface are continuously monitored to obtain preliminary resistance distribution data. Based on the collected resistance distribution data, this embodiment uses a preset threshold to filter the friction values ​​at key points. If the friction value at a certain key point exceeds the preset threshold, it indicates significant resistance at that point, and is therefore marked as a high-resistance point. The spatial distribution information of high-resistance points is then determined. By processing the spatial distribution information of high-resistance points and combining it with the geometric characteristics of the asymmetric curved surface, the positive resistance value of each region is calculated, thereby obtaining the trend of positive resistance variation on the asymmetric curved surface.

[0017] To address the variation trend of forward resistance on an asymmetric surface, this embodiment further analyzes the fluctuation of reverse velocity. If the fluctuation range of reverse velocity in a certain region exceeds a preset range, that region is recorded as a region of significant velocity difference, and the distribution range of the region of significant velocity difference is obtained. Based on the distribution range of the region of significant velocity difference, and in conjunction with the real-time analysis module, the friction force value and resistance distribution of each region of the surface are dynamically updated to determine whether there are new high resistance points or regions of velocity difference.

[0018] Specifically, for areas with significant speed differences, real-time friction force and resistance data are extracted for the corresponding regions. These data are then arranged chronologically to construct a time-series dataset, and a sliding window method is used to extract local features. Based on these local features, the real-time friction resistance and its changing trends in different wiping areas of the windshield are identified. According to the resistance change trends, the latest real-time friction force values ​​are input into the resistance distribution model for each region of the asymmetric surface. By validating the real-time dynamic resistance distribution information, the parameters of the resistance distribution model are dynamically adjusted, such as correcting resistance peaks and updating associated boundaries, to optimize the resistance prediction accuracy for each region of the asymmetric surface. Combined with the real-time position calibration values ​​obtained from the wiper arm angle sensing module, the rate of change of the swing angle and the pressure distribution on the contact surface are analyzed. The fluctuations of forward resistance and reverse speed within the speed deviation range are calculated, ultimately generating real-time dynamic resistance distribution data covering the entire surface. The system continuously monitors whether the contact surface pressure distribution deviates from a preset threshold. If a deviation occurs, the swing force balance value in the model is fine-tuned through dynamic feedback adjustment values ​​to ensure that the generated distribution information always closely matches the actual physical state. As an optional embodiment of this application, in order to more accurately identify surface features, the surface features are classified using a support vector machine algorithm by integrating dynamically updated friction force values ​​and resistance distribution data to determine the comprehensive resistance characteristics of each region of the asymmetric surface. Based on the comprehensive resistance characteristics, the difference between forward resistance and reverse velocity is continuously monitored to ultimately obtain matching information between surface resistance and velocity differences.

[0019] In one possible implementation, dynamically updated friction force values ​​and resistance distribution data are used as input data for preprocessing, including noise filtering, feature extraction, standardization, and normalization, to generate a standardized feature dataset. For this standardized feature dataset, a support vector machine (SVM) algorithm is used to classify the resistance characteristics of each region of the asymmetric surface, determining the resistance category for each region, such as high-resistance region, low-resistance region, region with significant speed differences, and comprehensive resistance characteristic category. For each resistance category, resistance feature values ​​are extracted to construct a resistance feature matrix for each region of the asymmetric surface. K-means or hierarchical clustering algorithms are used to perform cluster analysis on the resistance feature matrix to determine the resistance correlation between different regions of the asymmetric surface. Based on the resistance correlation, a comprehensive resistance characteristic distribution map of each region of the asymmetric surface is generated, marking the boundary information of high-resistance and low-resistance regions.

[0020] In one possible implementation, based on the comprehensive drag characteristics, the positive drag value and reverse velocity data of each region of the asymmetric surface are extracted to construct a difference analysis dataset. Statistical analysis of the difference analysis dataset is then performed to determine the distribution pattern of the differences in positive drag value and reverse velocity; for example, under a specific curvature, the proportional relationship that for every 10% increase in drag, the velocity fluctuation increases by 5%.

[0021] Based on the differential distribution pattern, the system continuously updates the resistance and speed matching information for each region of the asymmetric surface. When the sensor detects a drift in resistance characteristics, the system automatically adjusts the expected speed threshold according to existing matching rules. Through dynamic calibration of the matching information, dynamic monitoring results reflecting the current operating status of the wipers are generated. By performing trend analysis on the dynamic monitoring results, the system predicts the direction of change in resistance and speed differences in each region of the asymmetric surface, optimizing subsequent dynamic response strategies such as pre-adjusting trajectory correction amounts to ensure the smoothness of the wipers when crossing different resistance zones. Through this process, the system ensures that on complex asymmetric surfaces, the wiper control commands are always based on the most accurate resistance and speed matching logic.

[0022] The forward oscillation resistance value and the reverse oscillation speed difference are key to determining surface conditions (such as the amount of rainwater or the presence of an oil film) and optimizing the motor control algorithm. Among these, determining the forward oscillation resistance value ( During the forward oscillation process, the motor current I and the speed v fed back by the encoder are sampled in real time by sensors. The torque is then calculated by inversely using the current. After deducting the frictional force of the system's own machinery (no-load resistance), the remaining part is the oscillation resistance value of the glass surface. Determine the difference in reverse oscillation speed ( The speed difference typically refers to the difference between the actual operating speed and the speed commanded by the control, or the asymmetrical difference between forward and reverse speeds at the same coordinate point. Specifically, this is achieved by recording the speed at the same spatial coordinate point during the round trip. instantaneous velocity and Calculate the difference: or If the reverse speed is significantly lower than expected (the speed difference increases), it indicates that the resistance in this area has increased abnormally during the reverse stroke, which usually suggests uneven water film distribution or a shift in the surface clamping force.

[0023] S102, based on the surface feature distribution information and combined with the real-time position calibration value obtained by the wiper arm angle sensing module, analyze the swing angle change rate and contact surface pressure distribution, determine the fluctuation of the forward swing resistance value and the reverse swing speed difference within the speed deviation range, and obtain detailed resistance distribution data for each key point.

[0024] In practical applications, the real-time position calibration data of the wiper arm is obtained through the angle sensing module. Combined with surface feature information, the swing angle is initially recorded to obtain an initial swing angle dataset. For the initial swing angle dataset, the rate of change of the swing angle is calculated, and time series analysis methods, such as the moving average method based on a sliding window, are used to determine the dynamic trend data of the rate of change.

[0025] In one possible implementation, the angle sensing module measures the current physical angle of the wiper arm (e.g., 15°). The system then looks up the corresponding surface features at that angle (e.g., the glass curvature is extremely high at this angle, requiring greater downforce from the wiper to prevent wobbling). The real-time position data is then overlaid with the pre-stored surface features to record the oscillation trajectory under specific surface resistance. This provides an initial baseline for subsequent dynamic compensation (such as adjusting the oscillation speed based on wind resistance or rainfall).

[0026] The surface feature information typically includes: local curvature, i.e. the degree of curvature at different locations on the glass surface; three-dimensional coordinate mapping, i.e., dividing the glass into grid points and recording the precise height (z) and position (x, y) of each location in three-dimensional space; pressure distribution requirements, i.e., the system-preset normal pressure value that varies with position to ensure that the wiper blade can adhere tightly to the glass at different curvature positions; friction coefficient (μ), i.e., the initial friction characteristics based on the glass material and coating (such as water-repellent coating); effective wiping area, i.e., the boundary contour of the glass; and coordinate features of key points such as the edge of the A-pillar, the parking position, or the sensor blind zone.

[0027] Based on the dynamic trend data of the rate of change, combined with the contact surface pressure information collected by the contact pressure sensor, the correspondence between pressure and angle change is analyzed to obtain preliminary mapping data of pressure distribution. From the preliminary mapping data of pressure distribution, relevant information on positive resistance and reverse velocity is extracted, and the difference between the two within the velocity deviation range is calculated to obtain the fluctuation state data of velocity deviation.

[0028] In one possible implementation, the system determines how the pressure distribution dynamically changes with the motion state of the contact surface (characterized by the rate of change of angle) by analyzing the coupling relationship between dynamic trend data represented by the rate of change of angle and contact surface pressure information collected by the contact surface pressure sensor, thereby obtaining mapping data reflecting the relationship between "motion state-pressure distribution".

[0029] In one possible implementation, the system records two types of data—positive resistance and reverse velocity—at the same coordinate point using sensors (such as current sensors, encoders, and six-dimensional force sensors). The relevant information between positive resistance and reverse velocity refers to the mapping of their correspondence within the same spatial location and time window. The positive resistance (… The tangential force required to overcome glass friction, water film resistance, and wind resistance when the windshield wiper swings from its stationary position toward the A-pillar (forward direction) is typically determined by the motor current signal. The reverse velocity () is derived by combining the torque constant. This refers to the actual instantaneous speed of the windshield wiper as it returns (reverses) from the A-pillar to the same location. The correlation between these two values ​​reflects the system's energy conversion efficiency. For example, if the forward resistance in a certain area suddenly increases, it often leads to a decrease in speed during the subsequent reverse travel (i.e., a reduction in reverse speed). This force-speed pairing data is the basis for subsequent diagnosis of the windshield's condition.

[0030] The calculation of the difference between the two within the speed deviation range yields the speed deviation fluctuation data. The specific steps are as follows: First, the system presets a speed curve under an ideal state. (This is typically the speed under moderate rainfall and without an oil film). Speed ​​deviation ( The difference between the actual speed and the ideal speed is denoted as . The speed deviation range is... The deviation is categorized into multiple levels (e.g., 0% to 5% is normal, 5% to 15% is slight fluctuation, and above 15% is significantly abnormal). The second step involves calculating the impact of forward resistance on reverse speed within the selected deviation range using a formula: Where k is a proportionality constant, used to unify the dimensions of force and velocity. and These are samples taken within the same deviation interval. Finally, the calculated... Combined with a timeline, observe its stability across multiple periods. If A large standard deviation indicates high output fluctuation, suggesting localized oil film or dryness on the surface. If... The trend towards stability indicates that although there is resistance, it is evenly distributed.

[0031] In this embodiment, the resistance distribution at each key point is analyzed for the fluctuation data of speed deviation. To improve data accuracy, a support vector machine model is used to classify the resistance distribution and determine the resistance distribution feature values ​​at the key points. Based on the resistance distribution feature values ​​at the key points, detailed resistance distribution data records for each point are generated, thereby obtaining the final comprehensive resistance distribution dataset.

[0032] The support vector machine model is typically trained to identify the following resistance distribution patterns: Category A (Normal Water Film): Uniform resistance distribution, with symmetrical forward and reverse resistance. Category B (Local Dry Friction): A sharp increase in resistance within a certain angle, usually corresponding to stubborn stains or oil films on the glass. Category C (Foreign Object Obstruction): Pulsating resistance within an extremely narrow area (such as fallen leaves or freezing points).

[0033] The key points typically refer to specific locations within the wiper's travel path (such as the 0° starting point, the middle viewing area, and the 90° rotation point). The method for determining their characteristic values ​​is as follows:

[0034] The system uses the angle sensing module to extract a segment of the resistance data stream within a specific angle range (e.g., θ∈[40°,50°]). This range constitutes a key point sampling window. Within this point, the feature value is not a single numerical value, but rather a vector. Among them, static eigenvalues ​​( ): The average frictional resistance at this point. Dynamic characteristic value ( ): The derivative of drag as a function of velocity This reflects the viscosity of the water film at that point. Deviation characteristic value ( ): The difference between the real-time resistance and the pre-stored reference resistance in the initial surface feature distribution information.

[0035] For example, suppose the system is detecting a critical area in the center of the glass, and the sensor returns a current of 2.5A with a fluctuation frequency of 50Hz for that area. The SVM model calculates that the feature vector at this point falls within the local oil film classification region. Then, the peak resistance at this location is extracted as 1.2 N·m, and the force distribution characteristic value at that location is calculated. (Where w is the weight). Finally, the system obtains a resistance characteristic value of 0.85 (normalized value) for this key point, triggering an instruction to increase water spray or adjust pressure. S103, based on the detailed resistance distribution data of each key point, the glass surface is divided into multiple independent regions using a preset curved surface region division method. The swing force balance value and region resistance compensation value of each region are calculated, and the positive trajectory correction amount and reverse trajectory adjustment value are determined.

[0036] To address the resistance distribution, this embodiment first acquires key point data of the glass surface. Using a pre-defined curved surface region division method, the surface is divided into multiple independent regions, yielding initial division results for each region. Based on these division results, the distribution of the swing force is calculated for each independent region. A pre-defined balance value calculation rule is then used to determine the balance value of the swing force in each region.

[0037] The preset surface region division method can be based on standard division of "A / B / C view area", polar coordinate sector division, feature division based on curvature gradient, or discretized mesh division. For example, the preset surface region division method uses a hybrid division of "angle + curvature". When the wiper arm is at 30°, the system retrieves the current location as "Zone_4 (medium curvature / edge of area A)" through a preset table, and then calls the unique resistance reference value of this area to compare with the real-time feedback of positive resistance, thereby determining whether there is a speed deviation.

[0038] In one possible implementation, based on the division results, the distribution of the oscillation force is calculated for each independent region, i.e., a functional mapping of time, angle, and contact surface pressure. For example, using the motor current feedback I and angle data θ, the instantaneous torque of the wiper in a specific region is calculated. Because the glass is curved, the system needs to decompose the torque into tangential force (the component that overcomes friction) and normal force (the component that keeps the wiper in contact with the glass). For each independent region, the torque values ​​of all sampling points within that region are collected to generate a force histogram for that region. The mean ( This represents the average load in the region. Dispersion ( This reflects the force fluctuations caused by abrupt changes in curvature or surface unevenness (such as oil film) in the region.

[0039] Furthermore, a preset balance value calculation rule is used to determine the balance value of the swaying force in each region. This balance value is a target setpoint pursued by the system within a specific region; it is not fixed but dynamically determined based on the physical characteristics of that region. For example, the initial surface feature distribution information generated in S101 is used as a reference, combined with current environmental data (such as rain sensing frequency and wind resistance component caused by vehicle speed). Environmental corrections are performed. Balance values. It is a point where the brushing efficiency and mechanical wear are optimally balanced. This is typically determined through experimental calibration or simulated annealing algorithms, aiming to find the minimum stable pressure without brush skipping.

[0040] The preset balance numerical calculation rules are typically a set of multi-factor weighted algorithm formulas, the core purpose of which is to compensate for differences and eliminate biases, such as: .in, Curvature compensation factor. The greater the curvature of the region, the higher the equilibrium value is adjusted to compensate for the pressure loss caused by spring deformation. Friction deviation factor. Based on the SVM classification results, if a dry area is identified, the balance value needs to be adjusted by damping to prevent motor overload. Speed ​​adjustment function. It dynamically adjusts based on vehicle speed v. At high speeds, wind pressure compensation is utilized, and the balance value can be appropriately lowered. Weighting coefficients, preset empirical weights (e.g., the highest weight is given to the safe field of vision area A). The specific rule logic is as follows: 1. Consistency rule: If the deviation between the positive and negative resistance of area i exceeds 15%, the balance value needs to be adjusted in the direction of reducing the deviation (i.e., reducing the speed deviation value in the reverse direction). 2. Safety rule: Regardless of the calculation result, the balance value must not be lower than the minimum threshold required to maintain brush clarity. 3. Smoothing rule: Adjacent regions and The fluctuation in the balance value between them must not exceed 10% to ensure the smooth operation of the motor.

[0041] By analyzing the balance values ​​of the swing force, resistance compensation data for each independent region is obtained. If the balance value of a certain region exceeds a preset threshold, it indicates that the resistance in that region is large or the force is uneven. Therefore, resistance compensation is adjusted to obtain the adjusted resistance compensation value. Based on the adjusted resistance compensation value and the direction of motion of the positive trajectory, the correction value of the positive trajectory is calculated. A preset trajectory offset calculation method is used to determine the correction result of the positive trajectory.

[0042] In one possible implementation, the resistance compensation data for each independent region is obtained by analyzing the swing force balance values. This is achieved by calculating the resistance compensation data based on the deviation between the expected value and the feedback value. Specifically, the first step is to calculate the deviation: the system calculates the swing force balance value (target value) for each region. ) and the actual oscillation force (feedback value) fed back by the sensor in real time For comparison: The second step is to convert the compensation amount: the compensation data is usually expressed as a correction value for the motor drive parameters, which includes the following two dimensions:

[0043] Torque compensation: If If the resistance is >0 (too high), the system will calculate the percentage increase in current required. To maintain a constant speed.

[0044] Speed ​​gain: corresponds to the reverse speed difference. If resistance causes speed lag, the compensation data will be converted into the increment of the PWM duty cycle.

[0045] In one possible implementation, when the system detects that the balance value exceeds a preset threshold, the specific execution logic is as follows:

[0046] 1. Graded Response: Level 1 deviation (less than 10%): Only adjust the motor PWM for minor speed compensation. Level 2 deviation (10% to 25%): Determined as abnormal resistance. Invoke resistance compensation data, increase motor current, and may trigger the water spray system (if hardware supports it) to reduce the coefficient of friction. Level 3 deviation (greater than 25%): Determined as an emergency. Reduce oscillation frequency or stop operation to prevent motor overload and burnout.

[0047] 2. Data write-back: After compensation occurs, the system will store the compensation value of the point in a temporary cache as a reference for prediction during the next scan (i.e., pre-compensation).

[0048] The preset threshold setting can be referenced as follows:

[0049] 1. Absolute resistance threshold: This is typically set to 20% to 30% of the motor's rated output torque. If the resistance balance value in a certain area exceeds this limit, it indicates that there may be a physical obstacle (such as snow, ice, or dead branches) at that location, and the system will trigger a protection program instead of blindly increasing the force.

[0050] 2. Rate of change threshold: The volatility between adjacent regions exceeds 15%, or the rate of change of resistance within a single oscillation cycle. This is used to identify the risk of brush skipping. When the value exceeds this threshold, the system determines that the surface characteristics have changed drastically.

[0051] 3. Speed ​​Deviation Threshold: The deviation between the actual speed and the commanded speed reaches 5% to 10%. This is the direct switch that triggers the S102 / S103 compensation mechanism.

[0052] In one possible implementation, the adjusted resistance compensation value is combined with the direction of motion of the positive trajectory to calculate a positive trajectory correction value. A preset trajectory offset calculation method is then used to determine the positive trajectory correction result, where the positive trajectory correction value is... It typically consists of two parts: elastic deformation error caused by resistance and dynamic error caused by velocity hysteresis. Specifically:

[0053] A. Static Deformation Compensation (Based on Resistance Compensation Value): The wiper arm is not a rigid body. When the resistance compensation value increases, the motor output torque increases, which causes the wiper arm to produce a slight twist.

[0054] Calculation formula: .

[0055] in, This is the resistance compensation value determined in the previous step. This is the elastic coefficient of the wiper arm (a preset constant).

[0056] B. Dynamic hysteresis compensation (based on motion direction): In combination with the positive motion direction, the system needs to compensate for the delay error between the sensor collecting the resistance and the motor responding.

[0057] Calculation formula: .

[0058] in, This represents the current angular velocity as the vehicle travels in the forward direction. This refers to the system's control loop response time (e.g., 10ms).

[0059] Final correction value: .

[0060] The preset trajectory offset calculation method typically employs spatial vector superposition or an error transfer function. Its purpose is to transform the calculated values ​​into specific correction results (i.e., the final angular target to be executed by the motor). For example, the target value superposition method, the most direct offset calculation method, works as follows: extract the theoretical target position at the current angle from the surface feature information of S101. Confirm that the current motion is in a positive direction (i.e., the direction of increasing angle). Finally, perform vector synthesis: .

[0061] Note: If the resistance is too great and the wipers "run slow", the correction will "pull" the wipers forward by an offset in the direction of movement, forcing the motor to increase its output to catch up with the trajectory.

[0062] On complex curved surfaces, due to varying resistance in different areas, without this trajectory correction step, the wiper would react slowly in areas of high resistance and overshoot in areas of low resistance. Through a preset trajectory offset calculation method, the system achieves "active predictive control." For example, knowing the resistance is high ahead, it pre-clocks the motor a few degrees to offset the bend in the wiper arm, thus ensuring the wiping trajectory still appears perfectly aligned. As an optional embodiment of this application, after determining the positive trajectory correction result, the reverse trajectory also needs adjustment. Based on the positive trajectory correction result, the direction of movement of the reverse trajectory is analyzed, the adjustment value of the reverse trajectory is calculated, and the adjustment data of the reverse trajectory is obtained through preset trajectory reverse adjustment rules. Specifically, the system compares the resistance distribution characteristic value (from the SVM model) of the region during positive operation, extracts the speed deviation fluctuation state data calculated by S102, and performs rule matching: if the fluctuation is high, it indicates extremely high reverse resistance, and the rule triggers high-power compensation, increasing the reverse trajectory adjustment data; if the fluctuation is low, it indicates a smooth surface, and the rule triggers smooth maintenance, only compensating for mechanical idle. Finally, a set of angle-varying... The sequence, i.e. the adjustment data of the reverse trajectory, is stored in the instruction buffer of the actuator.

[0063] Wherein, the reverse trajectory adjustment value It is based on the benchmark after positive correction, and then performs a secondary correction incorporating the hysteresis effect. Specifically, the system determines this through an angle sensor. (Assuming positive is positive and negative is negative). At this point, the system identifies the rubber strip's flip point. Because the rubber strip generates a momentary resistance peak at the instant of flipping, and the direction of movement is opposite to the positive direction, the previous positive trajectory correction becomes invalid at this moment. The formula for calculating the reverse trajectory adjustment value is: .in, (Forward Mapping Reference): Retrieves the correction amount during the forward motion of this region as a preliminary reference. (Speed ​​Deviation Compensation): In conjunction with the reverse speed difference mentioned in S102, if the actual reverse speed is slower than the commanded speed, this value increases. (Damping Correction): Fine-tuning is performed to address the "viscous resistance" generated by the residual water film on the scraping path during reverse motion.

[0064] The preset trajectory reverse adjustment rules typically include the following core principles:

[0065] Rule 1: Symmetry Deviation Criterion: Due to the idle stroke of mechanical connecting parts and the overturning of rubber strips, the resistance of the reverse stroke is usually not equal to that of the forward stroke. A proportional coefficient is preset when calculating adjustment data. (Typically between 0.85 and 1.1). .

[0066] Rule 2: Boundary Buffer Rule: When the windshield wiper moves from the A-pillar (forward end point) to the stop position (reverse start point), the initial acceleration is relatively large. Within the initial 0-15° range of the reverse stroke, a gradual adjustment is used. The purpose is to prevent excessive adjustment from causing a jarring sensation or abnormal noise from the wiper during the reversal.

[0067] Rule 3: Dynamic Zeroing Rule: Regardless of any corrections made during the process, when the angle sensing module indicates the wiper is nearing the stop position, the adjusted data must be forcibly converged to zero. The purpose is to ensure that each cycle returns to the physical origin, preventing accumulated errors from causing the wiper wiping range to shift.

[0068] To further optimize the trajectory adjustment scheme, the adjustment data from the reverse trajectory, combined with the resistance distribution and key point information of each independent region, is used to optimize the trajectory adjustment values, thus determining the final trajectory adjustment scheme. With the final trajectory adjustment scheme obtained, trajectory execution parameters for each independent region are generated based on the curved surface characteristics of the glass surface, yielding a complete trajectory optimization result.

[0069] In one possible implementation, the core of optimizing the trajectory adjustment using the Support Vector Machine (SVM) algorithm is to find an optimal function f such that: the optimized adjustment = initial adjustment + SVM prediction residual. The goal is to minimize the mean square error between the actual wiper trajectory and the theoretical CAD trajectory under any curvature and resistance. Specifically, the SVM algorithm optimizes the dynamic gain of the compensation coefficient. Through extensive experimental data, SVM learns the complex manifold between "resistance-angle-deformation," and the system no longer rigidly executes K... Instead of using a formula like F, the SVM outputs a trajectory offset weight that most closely approximates the actual physical situation in real time based on the current sensor characteristics. This optimization method effectively solves the trajectory drift problem caused by uneven water film on the glass surface and wear of the wiper blades.

[0070] In one possible implementation, obtaining the final trajectory adjustment scheme involves generating trajectory execution parameters for each independent region based on the characteristics of the curved surface area of ​​the glass surface, thus obtaining a complete trajectory optimization result. Specifically, the system extracts the basic trajectory template for each region from a pre-stored database according to the curved surface area division method. For example, the template corresponding to Zone 5 (high curvature edge) typically has a slower angular velocity and a higher current limit. Combining the trajectory adjustment value optimized by SVM previously, the nominal path is superimposed with the correction offset. Based on the resistance compensation data, the required compensation torque is calculated. Finally, the above calculation results are converted into data messages readable by the controller, generating a list of trajectory points containing four-dimensional information {Pos, Vel, Force, Time}. Dynamic update: If the SVM detects a sudden change in surface friction (such as entering a dry friction zone), the system will modify the "acceleration slope" parameter for that second in real time to reduce the impact.

[0071] The curved surface characteristics of the glass typically include geometric curvature features, dynamic friction features, and kinematic constraints. The trajectory execution parameters are direct input vectors to the motor controller. They determine whether the wiper blades provide a firm or soft feel, and whether the speed is fast or slow, when passing through this area. Common parameter sets include: target position sequence (…). ): The optimal wiper angle per millisecond after optimization and correction. Dynamic current limit ( / Torque): The maximum torque allowed for the motor to output in this region. PWM frequency and duty cycle: Electrical signal parameters that directly control the motor power. Acceleration / deceleration slope (S-Curve): Controls the smoothness of the wiper entering and exiting this region. S104 generates dynamic motion trajectory parameters for each curved surface region by combining the positive trajectory correction amount and the reverse trajectory adjustment value with the speed optimization weight ratio, and updates the trajectory parameter update rate in real time to obtain the adjusted wiper arm motion control data.

[0072] In this embodiment, by collecting initial data of the forward trajectory and feedback data of the reverse trajectory, and combining this with the geometric characteristics of the curved surface region, preliminary correction values ​​and adjustment data are calculated to obtain the basic trajectory parameters for each region. Based on the preliminary basic trajectory parameters, combined with the speed optimization target and weight ratio allocation rules, dynamic motion control commands are generated to determine the motion path adjustment scheme for each curved surface region. Through the dynamic motion control commands, the changing trend of trajectory parameters is analyzed in real time to obtain the update requirements of the parameter refresh frequency and determine the frequency value adapted to the current motion state.

[0073] In one possible implementation, based on preliminary trajectory parameters, combined with the speed optimization target and weight allocation rules, dynamic motion control commands are generated to determine the motion path adjustment scheme for each curved area. Specifically, the previous trajectory correction results and real-time feedback are used as inputs, and the processing intensity of the current fluctuation is determined according to the weight allocation rules (e.g., if currently in area A, accuracy is prioritized). A PWM command is output to drive the motor to run smoothly according to the S-curve set by the "optimization target".

[0074] Speed ​​optimization is not just about increasing frequency, but also about maintaining stability under complex drag conditions. It primarily includes three core objectives: 1. Instantaneous stability: Ensuring minimal angular velocity fluctuations when the wipers traverse areas of different curvatures or points of sudden drag changes, preventing any visible jerking. 2. Minimizing reversing impact: Reducing mechanical collision energy and noise at the endpoints of forward / reverse switching (such as the A-pillar) by smoothing the deceleration curve. 3. Maximizing efficiency: Maintaining efficient wiping in critical areas of visibility (A-zone), while appropriately reducing speed in non-critical edge areas based on drag conditions to balance power consumption.

[0075] Since the above objectives are sometimes contradictory (e.g., pursuing efficiency may increase impact), the system needs to find a balance point through weight allocation rules. This is typically expressed using a weighted scoring function J: The specific allocation rules are as follows:

[0076] Dimensions: (Tracking accuracy); Weight name: Position weight; Allocation logic: When driving in area A or at high speed, this weight is set to the highest (approximately 0.6) to prioritize ensuring that the trajectory does not deviate.

[0077] Dimensions: (Smoothness); Weight name: Dynamic weight; Allocation logic: This weight is increased in high curvature edge areas or during reverse switching, allowing a little speed to be sacrificed for smoother operation.

[0078] Dimensions: (Energy consumption / lifespan); Weight name: Economic weight; Allocation logic: When the motor temperature is detected to be too high or in a dry scraping state, this weight is increased and the output intensity of the motor is reduced.

[0079] The system will dynamically adjust based on the SVM classification results. For example, if it is identified as "heavy rainfall", the efficiency weight will be greatly increased; if it is identified as "oil film / dryness", the smoothness weight will be increased to prevent skipping.

[0080] In one possible implementation, the system uses dynamic motion control commands to analyze the changing trends of trajectory parameters in real time, obtains the update requirements for parameter refresh frequency, and determines the frequency value suitable for the current motion state. The judgment logic is typically based on the error change rate and the environmental entropy value. The system selects the specific frequency value (e.g., switching between 100Hz, 500Hz, and 1kHz) through a multi-level triggering mechanism. Specifically, the frequency value can be determined using the following methods:

[0081] 1. State change rate analysis method: The system calculates key trajectory parameters (such as angle deviation). The first and second derivatives of the rate of change of current (dI / dt) are calculated using the following formulas: If the resistance changes drastically over time (large second derivative), the system determines that it is in an unstable state and automatically increases the refresh rate to the highest level (e.g., refreshes once every 1ms). If the change approaches zero, the frequency is reduced.

[0082] 2. Key Point Prediction Method: Combining the region division results of S103, when the angle sensing module indicates that the wiper is about to enter a high curvature zone or commutation buffer zone, the system triggers the frequency refresh requirement in advance. Rule setting: Before entering the edge zone by 10°, the sampling and correction frequency is forcibly increased by 2 to 3 times.

[0083] 3. Error Convergence Status Judgment: If the residual between the actual trajectory and the optimized trajectory remains low and stable, the frequency value remains at the baseline level. Once the residual exceeds the preset fluctuation threshold (e.g., the deviation fluctuation rate is greater than 5%), the frequency is immediately determined to be insufficient, and upward adaptation is initiated.

[0084] Through this dynamic refresh mechanism, the system achieves on-demand allocation of computing resources, providing millisecond-level ultra-fast response at critical moments (when resistance is high and the trajectory is prone to deviation), and maintaining low-energy and stable operation during normal times (smooth regions). Considering unforeseen situations that may occur during actual movement, if the trend of trajectory parameter changes exceeds a preset threshold range, the execution logic of motion control is adjusted according to the update requirements of the refresh frequency, resulting in optimized control data. Using this optimized control data, the actual movement state of the wiper arm is matched in real time to obtain adjusted execution commands and determine the final motion trajectory scheme.

[0085] The preset threshold range is typically divided into a stability threshold and a response threshold, set based on motion control theory (such as the deviation limit of PID):

[0086] Trajectory deviation threshold: If the difference between the actual position and the commanded position exceeds this range, it indicates that the current drag compensation is no longer sufficient to maintain trajectory accuracy.

[0087] Force gradient threshold: An excessively rapid increase in resistance indicates that a hard obstacle may have been encountered or that an extreme dry friction zone has been entered.

[0088] Velocity volatility threshold: An excessively large coefficient of variation in instantaneous speed indicates that the windshield wipers are experiencing physical fluctuations.

[0089] The method for adjusting the execution logic of the motion control can be: 1. Switching the control gain: dynamically increasing the PID controller's... (proportionality factor) and (Differential coefficients). To address trends exceeding the threshold, increase the actuator's rigidity for a more rapid response, forcefully suppressing trajectory deviation. 2. Introduce feedforward prediction and frequency alignment: Combined with refresh frequency update requirements, synchronously raise the control frequency to the highest level (e.g., 1kHz). Under high-frequency sampling, the execution logic switches from hysteresis feedback to feedforward compensation based on SVM prediction. The system anticipates the resistance in the next millisecond and pre-releases current. 3. Trajectory envelope limitation: Temporarily abandon the maximum speed target and switch to smoothing-first logic. Limit the maximum derivative of acceleration to smooth abrupt step signals. Even with high resistance, ensure the motor passes at a constant speed.

[0090] The motion control execution logic is adjusted according to the refresh rate update requirements to obtain optimized control data. For example, the SVM optimization amount, refresh rate increment, and corrected PID gain are fused together. Then, the instruction is rewritten: the original instruction maintains a speed of 60° / s. The optimized instruction reduces the speed to 45° / s, but increases the current limit by 30% and forces real-time angle calibration with a 1ms period. Finally, a highly interference-resistant instruction stream is output to drive the motor to forcefully overcome resistance zones or smoothly commutate.

[0091] As an optional embodiment of this application, to further improve control accuracy, after obtaining the final motion trajectory scheme, the trajectory parameters are fine-tuned using a pre-established decision tree model, combined with real-time feedback data from the curved surface region, to obtain a control output adapted to the current environment. Through the above control output, the operating status of the wiper arm is continuously monitored, and deviation data during dynamic motion is corrected in real time to determine the optimal motion control result.

[0092] The specific construction process of the pre-established decision tree model is as follows: In a laboratory or real-vehicle environment, different glass surface states (such as localized oil film, dry glass, and high wind resistance) are artificially set. Under these states, experienced engineers record the most suitable fine-tuning instructions. These historical records constitute the training set. Then, the ID3 or C4.5 algorithm (based on information entropy), or the CART algorithm (based on Gini impurity), is used to automatically calculate which feature can most effectively separate normal and abnormal states. For example, the model might find that region identifiers are the first priority split point because the strategies for edge regions are often completely different from those for the center region. The algorithm continues to split downwards until each leaf node can provide a clear fine-tuning coefficient. To prevent the model from becoming too sensitive (overfitting) during real-vehicle operation, some overly fragmented decision branches need to be removed, retaining only the robust core logic.

[0093] In one possible implementation, after obtaining the final motion trajectory scheme, the trajectory parameters are fine-tuned using a pre-established decision tree model, combined with real-time feedback data from the curved surface region. The output of the decision tree is a correction vector, which is fused with the original control commands generated by S104 to obtain a control output adapted to the current environment. .in, These are the fine-tuning coefficients selected by the decision tree based on the current environment (resistance, curvature, rainfall).

[0094] S105, based on the adjusted wiper arm motion control data, drives the wiper arm to execute a dynamic trajectory, monitors the influence of the contact surface pressure distribution on the wiping uniformity in real time, and if the pressure distribution in a certain area is detected to deviate from the preset threshold, the swing force balance value is finely adjusted through dynamic feedback adjustment value to determine the updated control command.

[0095] The pressure data of the wiper arm's contact surface during dynamic trajectory execution is acquired through a sensor array. The pressure value of each area is recorded in real time to obtain a preliminary pressure distribution map. Based on the preliminary pressure distribution map, a pre-established pressure distribution analysis model is used to determine whether the pressure value of each area exceeds a preset threshold range. If the pressure value of a certain area exceeds the preset threshold, that area is marked as a deviation area, and the range of areas that need adjustment is determined.

[0096] Specifically, once the pre-established pressure distribution analysis model runs and acquires real-time pressure data, the following judgment logic will be executed: Obtain the current area Real-time pressure value Then a threshold comparison is performed: when If insufficient pressure is detected, the system will trigger the resistance compensation logic in S104, attempting to increase the fit stability by adjusting trajectory parameters (such as deceleration). If a pressure overload is detected, the system will reduce the motor output torque and may fine-tune the trajectory through a decision tree to avoid high-pressure points. Finally, the map is updated, and the points exceeding the limit are recorded in the pressure anomaly matrix for pre-adjustment in the next sweep. The pre-established pressure distribution analysis model is typically a hybrid model based on physical modeling and real-time data mapping, constructed as follows:

[0097] A. Static Benchmark Modeling (Finite Element Analysis): During the development phase, finite element analysis was performed on the wiper arm's spring mechanism and rubber strip using Computer-Aided Engineering (CAE) software. The theoretical pressure distribution of the wiper arm under different glass curvatures (from flat to extremely curved) was simulated. This distribution defines the standard pressure value that each region should have under ideal conditions. .

[0098] B. Dynamic Feature Extraction (Real-time Mapping): Motor current (for estimated torque), wiper arm angle, and surface features determined by S101 are collected as input variables. A transfer function is established to incorporate tangential resistance. Transformed into normal pressure : .in, The current surface friction coefficient is identified by SVM.

[0099] C. Data Fusion and Model Training: The model is constructed using neural networks or multinomial regression algorithms to align real-time acquired current and position signals with static baseline maps. The model can automatically compensate for wind pressure caused by vehicle speed and elasticity attenuation caused by rubber strip aging.

[0100] The preset threshold setting typically consists of two parts: an absolute pressure threshold and a regional consistency threshold. The absolute pressure threshold (ensuring basic functionality) includes a lower limit threshold (…). The linear pressure is typically set at 15-20 g / cm. Below this value, the wipers will become unstable at high speeds due to wind force, resulting in incomplete wiping. Upper limit threshold ( The pressure is typically set at 45-60 g / cm (linear pressure). Exceeding this value will cause a sharp increase in friction, which may lead to motor overheating, abnormal noise from the rubber strip flipping, or scratches on the glass surface. The deviation range for the zone consistency threshold (pressure distribution gradient) is ±15%~20%. Judgment logic: Compare the pressure difference between two adjacent independent zones (such as Zone 1 and Zone 2). If the pressure change exceeds 20%, it indicates that the wiper is at risk of being suspended or bumped when passing through the curvature switching point.

[0101] For the marked deviation areas, corresponding wiper arm motion parameters such as instantaneous angular displacement, angular velocity, angular acceleration, torque / current feedback, and end linear velocity are extracted from real-time monitoring data. The correlation between the deviation areas and dynamic trajectory execution is analyzed to obtain the adjustment direction of the motion parameters corresponding to the deviation areas. Based on the adjustment direction of the motion parameters corresponding to the deviation areas, the adjustment value for oscillation force balance is calculated. A dynamic feedback adjustment mechanism is used to fine-tune the driving force of the wiper arm, and the adjusted motion control parameters are determined.

[0102] In one possible implementation, the corresponding wiper arm motion parameters are extracted from real-time monitoring data, and the correlation between the deviation area and the dynamic trajectory execution is analyzed to obtain the adjustment direction of the motion parameters corresponding to the deviation area. Specifically, firstly, the deviation area in the pressure distribution map (e.g., Zone 3 with low pressure) is aligned with the motion parameters extracted at that moment on the time axis. Then, the system consults the historical feature database to determine the type of relationship between the two: 1. Positive correlation: excessive speed → lift generation → pressure reduction (common in high-speed driving). 2. Negative correlation: excessive acceleration → inertial vibration → instantaneous pressure surge (common at reversing points). 3. Nonlinear correlation: angle at the edge of high curvature → mechanical structure interference → pressure deviation. Finally, the sensitivity factor (S) is calculated. By calculating the sensitivity of pressure changes to changes in a certain motion parameter, it is possible to determine which parameter is causing the deviation.

[0103] Furthermore, based on the sensitivity factor obtained from the correlation analysis, the system determines the adjustment direction according to the principles of compensating for bias and regressing to the median. Specifically, when the correlation analysis result indicates speed sensitivity, an increase in speed leads to lower pressure than expected. When the correlation analysis result is a torque bottleneck type, where excessive resistance causes speed lag and pressure surge, increase the driving torque and decrease the acceleration to steadily push through the high resistance zone; when the correlation analysis result is a position interference type, where the pressure continuously deviates in a specific angular area, correct the trajectory offset by finely adjusting the angle to avoid curvature abrupt change points and allow the rubber strip to realign; when the correlation analysis result is an inertial jitter type, where the pressure at the reversing point oscillates violently, optimize the deceleration slope, decelerate earlier, and smoothly enter the reversing zone.

[0104] In one possible implementation, the direction of adjustment is determined by the motion parameters corresponding to the deviation region, the adjustment value for balancing the swing force is calculated, and a dynamic feedback adjustment mechanism is used to fine-tune the driving force of the wiper arm. The core formula for the adjustment value in the determined motion control parameters is: .in, (Region Gain): A coefficient determined based on the characteristics of the region divided by S103. The gain is usually set larger for high curvature edge regions to achieve fast feedback. (Pressure Deviation Item): The difference between the target pressure value and the actual pressure currently analyzed. (Trend Compensation Coefficient): Adjusts direction based on motion parameters. If the direction adjustment indicates that deceleration is required, this coefficient will generate a damping torque.

[0105] The dynamic feedback adjustment mechanism is a real-time closed-loop control logic, the core of which lies in the infinite loop of "real-time sampling - deviation calculation - command correction - resampling". This mechanism consists of the following three parts:

[0106] 1. Closed-loop circuit

[0107] Feedback sources: motor current sensor, encoder position, pressure analysis model output.

[0108] Execution frequency: Use the frequency value determined in S104 (such as 500Hz or 1kHz), that is, make adjustments 500 to 1000 times per second.

[0109] 2. Adaptive adjustment logic

[0110] The mechanism considers not only the current deviation but also its trend: Proportional adjustment (P): The larger the deviation, the greater the adjustment. Derivative adjustment (D): Predicts the deviation trend. If the pressure is decreasing rapidly, the mechanism intervenes early, providing an adjustment value before reaching the threshold. Integral adjustment (I): Eliminates persistent small deviations (such as constant deviations caused by minor wear of the rubber strip).

[0111] 3. Dynamic switching strategy

[0112] This mechanism dynamically switches and adjusts weights based on the current motion state: During the stable period, a "low gain, high smoothness" strategy is adopted to save energy and ensure quiet operation. During the fluctuating period (entering the deviation region), it automatically switches to a "high response, feedforward compensation" strategy, using the results of decision tree or SVM optimization to quickly suppress resistance fluctuations.

[0113] For example, suppose the wiper is running at the edge of the A-pillar (Zone 6), and the system detects that the pressure is lower than the warning value (signs of drifting): 1. Direction Extraction: Analysis shows that the adjustment direction is to increase the normal pressure and reduce the angular velocity. 2. Adjustment Value Calculation: The formula shows that a 0.5 N·m increase in drive torque compensation is needed, and the PWM duty cycle should be reduced by 5% to slow down the speed. 3. Feedback Adjustment: The dynamic mechanism adds this adjustment value to the current motor command. 4. Convergence Monitoring: The pressure is read again in the next millisecond. If the pressure rises back to within the threshold, the adjustment value is gradually reduced until the system returns to equilibrium. In this embodiment, to ensure the uniformity of the wiping effect, a new control command is generated based on the adjusted motion control parameters to drive the wiper arm to execute the updated dynamic trajectory, and the changes in the contact surface pressure data are recorded in real time to obtain an updated pressure distribution map. For the updated pressure distribution map, by comparing it with the preset threshold range, it is determined whether the wiping uniformity of each area meets the expectations. If there are still areas exceeding the preset threshold, the above adjustment process is repeated until the pressure distribution meets the preset standard, and the final control command is determined. After obtaining the final control command, it is transmitted to the wiper arm drive module to complete the execution of dynamic trajectory and optimization of pressure distribution, and record the data log of the entire adjustment process for subsequent analysis.

[0114] S106, in response to the updated control commands, continuously records the operating status of the wiper arm in different curved areas, analyzes the effect of friction in the curved area on long-term operational stability, and, in conjunction with the area resistance compensation value, determines whether the trajectory parameter update rate meets the adaptability requirements, thus obtaining the final long-term parameter adjustment range.

[0115] The system acquires operational status data of the wiper arm across multiple curved surfaces, including position and velocity sequences. To accurately assess operational smoothness, a sliding window method is used to process the operational data, calculating the standard deviation of the velocity sequence within each window to obtain a smoothness index. Simultaneously, based on the positional information of the curved surfaces, the system retrieves the corresponding resistance values ​​from a pre-defined resistance mapping table.

[0116] The preset resistance mapping table typically exists in the form of a two-dimensional or three-dimensional matrix, and its core dimensions include:

[0117] Index Key: *Location Information: Typically, the horizontal axis is the swing angle θ of the wiper arm, and the vertical axis is the discrete sampling point d (from root to tip) on the wiper blade. Zone ID: Corresponds to the independent zone number (Zone ID) divided in S103.

[0118] Numerical Items: Nominal Resistance Value: The standard frictional torque of the glass under clean and wet conditions. Resistance Threshold Range: The upper and lower limits of the allowable resistance fluctuation at this location. Curvature Correction Coefficient: A preset mechanical compensation weight for the geometry of this point (such as large edge curvature).

[0119] A preset resistance mapping table is typically compiled from data from the following three sources:

[0120] 1. Offline factory calibration: In a laboratory environment, a high-precision sensor is used to perform a full-surface scan of a standard glass, recording the resistance data under undisturbed conditions.

[0121] 2. Physical Model Simulation: The contact pressure between the curved glass surface and the wiper blade at different angles is calculated using a CAD model and theoretical formulas. (in Simulation data is generated for the empirical friction coefficient.

[0122] 3. Self-learning initialization (S101 stage): The initial feature distribution generated in real time by the system through "scanning once" during the first run or self-test cycle will also be written back and updated to this table.

[0123] The preset resistance mapping table eliminates resistance differences caused by the shape of the glass itself. For example, the resistance at the edge of the glass is naturally greater than at the center due to its curvature. Without this table, the system might misinterpret the high resistance at the edge as a foreign object; with this table, the system knows that the resistance is due to the normal shape, thus maintaining smooth operation.

[0124] The stability index and resistance value are input into a linear regression model, which outputs the coefficient of friction's effect on stability. If the coefficient is greater than a preset threshold, the regional resistance compensation value is deemed insufficient, triggering a trajectory parameter update. Based on the difference between the coefficient and the current compensation value, a proportional-integral controller is used to calculate the trajectory parameter update rate. It is then determined whether the update rate is within a preset adaptability range. If it is, the adjustment domain is determined based on the update rate and long-term constraints.

[0125] The preset threshold is determined through the following methods: 1. Motor stall test: Determine the maximum resistance torque that the motor can overcome under different PWM duty cycles to determine the efficiency inflection point. 2. Robustness calibration: Apply media of different viscosities (such as machine oil or detergent) to glass and test the system at what action coefficient will "brush skipping" or "stalling" occur, thereby setting a safety threshold. 3. Dynamic response simulation: Simulate high-speed wind resistance conditions, calculate the compensation lag angle, and set the critical point that causes the trajectory deviation to exceed 2° as the threshold. Common numerical ranges are as follows: 1. Nominal ratio threshold, usually set between 1.2 and 1.5. If the action coefficient (e.g., the electric distance of the added force) is greater than 1.5, it indicates that although the system has attempted compensation, the resistance is still increasing, or the speed deviation has not decreased, therefore it is judged as "insufficient compensation value". 2. Percentage deviation threshold, set at 15%-20%. When the deviation rate between the actual compensation effect and the expected target value exceeds 20%, the judgment of insufficient compensation is triggered. 3. Physical limit thresholds, such as torque margin thresholds, indicate that if the currently compensated torque accounts for more than 85% of the motor's maximum allowable torque and the action coefficient is still increasing, then the compensation potential is exhausted and the compensation value is absolutely insufficient.

[0126] In one possible implementation, the step of determining whether the update rate is within a preset adaptability range is performed. If it is, an adjustment domain is determined based on the update rate and long-term constraints. Specifically, the displacement of the parameter from the mapping table baseline value is calculated based on the current update rate. .Will Evaluate within the constraint model; formula example: .in This is a smoothing factor (calculated based on long-term constraints). If long-term operating data shows that the rubber strip is aging, The system will automatically shrink and narrow the adjustment domain. The calculation results are mapped to a multi-dimensional adjustment domain, for example: torque adjustment domain: [0.8, 1.2] N·m; speed adjustment domain: [40°, 65°] / s; time adjustment domain (frequency): [200Hz, 800Hz]. The adjustment domain command is output, and the final determined adjustment domain is sent to the subsequent execution module of S106. This means that in the next scraping cycle, the system will quickly find the optimal solution within this set range based on real-time resistance, without deviating from this safe range.

[0127] The adaptive range refers to the range within which the system considers the current parameter update frequency to be within a controllable and healthy range. It distinguishes between normal environmental changes and sudden failures / extreme interference. It is typically defined by the number of parameter updates or the magnitude of change per unit time. Typical ranges are as follows: (Numerical example:) For example, [5Hz, 50Hz].

[0128] Below the range (less than 5Hz): The update rate is too low, and the system considers the current environment to be extremely stable, so there is no need for frequent adjustments.

[0129] Within the specified range (5Hz to 50Hz): The system determines that it is currently in a dynamic adaptive state, such as entering a light rain zone or a zone of stable speed change. At this time, the system will consider this adjustment to be effective and sustainable.

[0130] Above the range (greater than 50Hz): The update rate is too high, which is judged as system oscillation or sampling noise. It may be due to sensor failure or encountering a very large foreign object. The system will stop updating and trigger protection.

[0131] The long-term constraints are protective criteria set by the system to prevent over-adjustment, ensuring that the system's evolution meets the lifespan and stability requirements of the physical hardware. They mainly include the following three dimensions: 1. Mechanical Lifespan Constraint: Frequent switching of torque or speed by the motor can lead to heat and wear. The constraint stipulates that the root mean square change of the control command must not exceed 20% of the nominal value within 10 minutes. 2. Convergence Constraint: This is an algorithm-level requirement. The direction of long-term adjustments must be to converge the error, not diverge it. If the error increases for five consecutive cycles, the long-term constraint will forcibly lock the parameters. 3. Consistency Constraint (Historical Memory): The current adjustment cannot conflict with the original surface features stored in S101. For example, if the glass is convex, the long-term constraint does not allow adjusting a local area to a concave resistance model.

[0132] S107, through the final long-term parameter adjustment range, periodically optimizes the dynamic feedback adjustment value and speed optimization weight ratio, generates the final motion trajectory control logic that adapts to the characteristics of asymmetric curved surfaces, and completes the adaptive operation adjustment of the wiper arm on complex curved surfaces.

[0133] By collecting operational data of the wiper arm on a complex curved surface, initial values ​​of long-term parameters and parameter ranges are obtained to determine the basic adjustment benchmark under the operating environment. Based on the initial values ​​and operating environment data, the dynamic feedback and the changing trend of the feedback values ​​are analyzed to obtain deviation information under real-time operating conditions. For the deviation information, the ratio between speed weight and motion trajectory is adjusted. If the deviation exceeds a preset threshold, the speed weight is dynamically corrected to obtain the corrected trajectory parameters.

[0134] In one possible implementation, the system acquires initial values ​​for long-term parameters and parameter ranges by collecting operational data of the wiper arm on a complex curved surface, thus determining the basic adjustment benchmark under the operating environment. Specifically, the system first establishes a multi-dimensional coordinate system using the operational data collected by S107. The data includes consistency data on the wiper's current consumption and oscillation speed on a standard curved surface under different ambient temperatures and vehicle speeds (wind resistance). Basic adjustment benchmark. It is typically defined as the optimal combination of control parameters under moderate rainfall, low speed, and standard temperature conditions. The system does not determine the baseline based on a single sweep, but rather through weighted calculations using accumulated long-term parameters, such as moving averages or Kalman filters, to eliminate incidental noise such as foreign objects and localized dryness encountered during a single sweep. This determined baseline represents the optimal performance state of the current mechanical hardware (e.g., motor efficiency, spring tension). The baseline is primarily determined based on the intersection of the following three core indicators: 1. Power plateau point: By analyzing current data, the optimal power output fluctuation rate ( ) is found when overcoming the normal pressure of the curved surface. 1. The smallest set of PWM command sequences; this point is defined as the force adjustment reference. 2. Trajectory overlap: Compare the actual running trajectory with the ideal surface CAD trajectory of S101. Criterion: Within the long-term parameter range, find a value that minimizes the trajectory residual. The minimum parameter offset, which serves as the position adjustment reference. 3. Environmental compensation coefficient: The system calculates the slope of the impact of different environmental factors on operation, for example: The baseline adjustment reference includes these basic zero-point compensations to ensure that the control algorithm starts working from a pre-compensated reference point when the environment changes.

[0135] Establishing a baseline adjustment point is essentially a form of system self-calibration. Without this step, the system will experience control drift due to the slow aging of the rubber strips or drastic changes in ambient temperature. By establishing a baseline, the system ensures that its core algorithm always starts and operates in a calibrated, safe, and efficient state, regardless of the external environment.

[0136] In one possible implementation, the ratio of speed weight to trajectory is adjusted based on deviation information. If the deviation exceeds a preset threshold, the speed weight is dynamically corrected to obtain the corrected trajectory parameters. Essentially, the ratio of speed weight to trajectory is the weight allocation of the objective function in the control algorithm. Since the system cannot simultaneously achieve perfect constant speed and zero trajectory deviation, a priority setting is required: speed weight (…). The trajectory weight determines the system's adherence to "reaching the target angle on time." A high weight means that even with significant resistance, the current must be increased to maintain the speed. The weighting determines the system's adherence to "fitting the curved glass surface / path accuracy." A high weight means that if a trajectory deviation is detected, the system will prioritize position calibration even at a reduced speed. The formulaic expression of the weighting relationship is as follows: .in, It's a speed error. It is a deviation in trajectory position.

[0137] The process of adjusting the ratio between velocity weight and motion trajectory is based on the importance of the region and the real-time resistance status: under normal conditions, and The system is in a balanced state (e.g., 50:50). When entering complex curved surfaces / high-resistance areas, the system will automatically reduce... ,improve This allows for a slight slowdown (becoming "smoother") in exchange for preventing the wiper arms from bouncing or hovering over sharply undulating surfaces. The system accelerates when entering the core of the field of vision (Area A). This ensures rapid rainwater removal and maintains visual clarity, while increasing tolerance for minor trajectory fluctuations.

[0138] The preset threshold typically refers to the maximum tolerable position deviation. Once this value is exceeded, the system determines that simple fine-tuning is ineffective and dynamic correction is necessary. For example, the absolute position deviation threshold is usually set to ±1.5° to ±2.5°. Dynamic hysteresis threshold: the actual swing angle lags behind the theoretical model by more than 15ms-20ms. Sudden resistance change threshold: the motor feedback torque suddenly exceeds 30% of the base reference (determined in S107).

[0139] Furthermore, if the deviation exceeds a preset threshold, the system will initiate a weight reset logic to dynamically correct the speed weight. Specifically, if the excessive deviation is due to excessive resistance causing the speed pursuit failure, the system will forcibly reduce the speed weight. Reducing the speed setpoint by 10% to 20% lightens the motor load, allowing the mechanical structure time to re-adhere to the glass under resistance, thus reducing trajectory deviation. The system transfers the power allocation originally allocated to the speed loop to the position loop by calling the adjustment domain in S106 to complete gain reconstruction. For example, instantaneously increasing the proportional gain of the position loop (…). As a result, the wipers become "harder," pressing harder against the glass surface to correct the trajectory. Combined with the refresh rate requirement mentioned in S104, the frequency and weights are refreshed in tandem. For example, at the moment of weight correction, the parameter refresh rate is increased to the highest level (e.g., 1kHz). Once the deviation falls below the threshold, the original speed weight is slowly restored.

[0140] By using the corrected trajectory parameters, control logic is generated for asymmetric and complex surfaces, determining the wiper arm's operating strategy in different surface regions. Based on the operating strategy, real-time adjustments are made to adaptive operation and operational regulation to obtain the actual operating path of the wiper arm on the complex surface, resulting in path-optimized control commands. These optimized control commands drive the wiper arm to perform adaptive operation. If a deviation exists between the operating path and the preset trajectory, a dynamic feedback mechanism is used for fine-tuning to achieve final operational stability.

[0141] In one possible implementation, control logic is generated for asymmetric and complex surfaces using the modified trajectory parameters. For example, the left and right / front and back resistance vectors of asymmetric surfaces are inconsistent. The allotropic compensation logic sets different PID parameter sets (asymmetric gain) for the forward and reverse strokes to counteract the unidirectional effects of gravity and wind resistance. The pressure distribution of complex surfaces (large curvature) varies greatly. The pressure closed-loop logic directly locks the motor current feedback to the pressure balance adjustment value, and the logic becomes constant pressure priority, with speed as the controlled variable. The refresh rate requirement is high in the variable region (transition zone). The frequency conversion sampling logic automatically interrupts the controller to instantly switch the sampling frequency from 100Hz to 1kHz.

[0142] In one possible implementation, the system performs real-time adjustments to adaptive operation and operation regulation based on the operating strategy, obtains the actual operating path of the wiper arm on the complex curved surface, and obtains the path-optimized control command. Specifically, at the instant the wiper starts or enters a new area, the system weights the target speed. Position weight Basic torque The frequency setting value is loaded into the controller's register. Then, the system starts driving the motor, collecting current, angle, and encoder pulses at a preset frequency (e.g., 1kHz), and comparing the actual current / speed with the expected values ​​of the strategy in real time. If a deviation occurs, the PWM duty cycle is immediately adjusted according to the dynamically corrected logic based on the speed weight, achieving adaptive oscillation. During the process of the wiper sweeping across complex curved surfaces, the system records the trajectory coordinates at each moment using high-precision sensors. ) and the corresponding resistance value The system specifically records the differences in the wiper's behavior at the same physical position during forward and reverse sweeping. This is the actual operating path, which contains all the information about the real physical environment (wind resistance, uneven glass, rubber strip deformation). Mathematical analysis of the collected actual path identifies abrupt jumps that cause vibration or uneven pressure. It calculates the speed curve at which the motor consumes the least power to overcome the path resistance. The actual path is fitted using Bézier curves or least squares methods to remove glitches caused by sensor noise or transient interference. Finally, the smoothed path is reconstructed into a three-dimensional command flow of {position, speed, torque}. The resistance abrupt changes detected in this run are converted into feedforward compensation for the next run. For example, if unexpected high resistance is encountered at 30°, the optimized command will increase the current in advance when reaching 29.5° next time. Finally, the controller's lookup table is updated, enabling the wiper to enter a near-muscle memory-like high-efficiency operating state.

[0143] As an optional embodiment of this application, to further enhance the system's intelligence level, a support vector machine algorithm is used to model the correlation between long-term parameters and dynamic feedback, analyze the operational adjustment requirements on complex curved surfaces, and determine the optimal control logic for the wiper arm under different environments. In this way, the system can automatically adjust its control strategy according to environmental changes, achieving efficient and stable operation of the wiper arm on complex curved surfaces.

[0144] In one possible implementation, the goal of using the Support Vector Machine (SVM) algorithm to model the correlation between long-term parameters and dynamic feedback is to find an optimal hyperplane that accurately classifies different operational adjustment requirements or predicts the optimal control gain. In the modeling process, a Gaussian radial basis function is typically chosen, followed by the establishment of a structural risk function. The model seeks a balance by minimizing the following objective function: .in, This determines the stability of the control logic. As a penalty factor, if If the value is too large, the model will become oversensitive (causing the wipers to vibrate frequently); if it is too small, the model will be sluggish.

[0145] After modeling, SVM determines the optimal control logic in the following ways: First, feature space partitioning: SVM divides the operating environment into multiple demand quadrants. Quadrant A (high speed + high curvature): identified as high-pressure fit requirement; Quadrant B (dry friction + low speed): identified as anti-vibration requirement. Second, parameter prediction: When real-time data is input into the model, SVM uses the established hyperplane to calculate in real-time which optimal control mode the current state is closest to. For example, Output 1: determines the adjustment direction (e.g., increasing normal force compensation by 5%). Output 2: sets the parameter boundaries (i.e., the adjustment domain in S106).

[0146] It should be noted that the above examples are merely some specific embodiments of the present invention. Obviously, the present invention is not limited to the above embodiments and many variations are possible. All variations that can be directly derived or conceived by those skilled in the art from the content disclosed in this invention should be considered within the scope of protection of this invention.

Claims

1. A dynamic response method using DC control of automotive intelligent windshield wipers, characterized in that, include: By acquiring high-frequency data from the asymmetric curved surface of the windshield, analyzing the frictional force and key point resistance distribution in the curved surface region, generating initial curved surface feature distribution information, and determining the positive oscillation resistance value and the reverse oscillation speed difference of the glass surface; Based on the surface feature distribution information and combined with the real-time position calibration value obtained by the angle sensing module, the swing angle change rate and contact surface pressure distribution are analyzed to determine the fluctuation of the forward swing resistance value and the reverse swing speed difference within the speed deviation range, and detailed resistance distribution data of each key point are obtained. Based on the resistance distribution data, the glass surface is divided into multiple independent regions, and the swing force balance value and regional resistance compensation value of each region are calculated to determine the positive trajectory correction amount and the reverse trajectory adjustment value. Based on the positive trajectory correction and reverse trajectory adjustment values, combined with the speed optimization weight ratio, dynamic motion trajectory parameters for each curved area are generated, and the trajectory parameter update rate is updated in real time to obtain the adjusted wiper arm motion control data. Based on the adjusted wiper arm motion control data, the wiper arm is driven to execute a dynamic trajectory. If the pressure distribution in a certain area is detected to deviate from the preset threshold, the swing force balance value is finely adjusted to determine the updated control command. For the updated control commands, the operating status of the wiper arm in different curved areas is recorded, the effect of friction in the curved areas on long-term operational stability is analyzed, and combined with the area resistance compensation value, it is determined whether the trajectory parameter update rate meets the adaptability requirements, and the final long-term parameter adjustment range is obtained. By adjusting the long-term parameters, the dynamic feedback adjustment value and speed optimization weight ratio are periodically optimized to generate the final motion trajectory control logic that adapts to the characteristics of asymmetric curved surfaces, thus completing the adaptive operation adjustment of the wiper arm on complex curved surfaces.

2. The dynamic response method using DC control of automotive intelligent windshield wipers according to claim 1, characterized in that, The analysis of the frictional force and key point resistance distribution in the curved surface region generates initial curved surface feature distribution information, determines the positive oscillation resistance value and the reverse oscillation velocity difference of the glass surface, including: By monitoring the frictional force values ​​and key points in various regions of the asymmetric curved surface, preliminary resistance distribution data are obtained; Based on the resistance distribution data, high resistance points whose friction values ​​exceed a preset threshold are selected, and the spatial distribution information of high resistance points is determined. By processing the spatial distribution information of high resistance points and combining it with the geometric characteristics of asymmetric surfaces, the positive resistance value of each region is calculated, and the variation trend of positive resistance on asymmetric surfaces is obtained. Based on the aforementioned trend, the fluctuation of the reverse velocity is analyzed, and the areas where the fluctuation amplitude exceeds the preset range are recorded as areas of significant velocity difference, thus determining the distribution range of areas of significant velocity difference. Based on the distribution range of areas with significant speed differences, the friction force and resistance distribution data of each region of the asymmetric surface are dynamically updated to determine whether there are new high resistance points or areas with speed differences. By integrating the friction force values ​​and resistance distribution data, and using a classification algorithm to classify the surface features, the comprehensive resistance characteristics of each region of the asymmetric surface are determined. Based on the comprehensive resistance characteristics, the difference between the positive resistance value and the reverse speed is continuously monitored to generate the final matching information between the surface resistance and the speed difference.

3. The dynamic response method using DC control of automotive intelligent windshield wipers according to claim 1, characterized in that, Based on the surface feature distribution information and combined with the real-time position calibration value obtained by the angle sensing module, the swing angle change rate and contact surface pressure distribution are analyzed to determine the fluctuation of the forward swing resistance value and the reverse swing speed difference within the speed deviation range, thus obtaining detailed resistance distribution data for each key point, including: The real-time position calibration data of the wiper arm is obtained by the angle sensing module. Combined with the surface feature information, the swing angle is recorded to obtain the initial swing angle dataset. For the aforementioned swing angle dataset, the rate of change of the swing angle is calculated, and time series analysis is used to determine the dynamic trend data of the rate of change. Based on the dynamic trend data and the contact surface pressure information, the correspondence between pressure and angle changes is analyzed to obtain preliminary mapping data of pressure distribution. Based on the preliminary mapping data, the difference between the forward resistance and the reverse speed within the speed deviation range is calculated to obtain the fluctuation state data of the speed deviation. For the fluctuating state data of velocity deviation, a support vector machine model is used to classify the resistance distribution and determine the resistance distribution characteristic values ​​at key points. Based on the resistance distribution characteristic values ​​of key points, detailed resistance distribution data records for each point are generated, resulting in the final comprehensive resistance distribution dataset.

4. The dynamic response method using DC control of automotive intelligent windshield wipers according to claim 1, characterized in that, The calculation of the swing force balance value and regional resistance compensation value for each region, and the determination of the positive trajectory correction amount and reverse trajectory adjustment value, includes: For each independent region, a preset balance value calculation rule is used to determine the balance value of the swing force in each region; By analyzing the balance value of the swing force, the resistance compensation data of each independent region is obtained. If the balance value of a certain region exceeds the preset threshold, the resistance compensation is adjusted to obtain the adjusted resistance compensation value. Based on the adjusted resistance compensation value and the direction of motion of the positive trajectory, the correction value of the positive trajectory is calculated, and the correction result of the positive trajectory is determined. Based on the correction results of the forward trajectory, the direction of motion of the reverse trajectory is analyzed, the adjustment value of the reverse trajectory is calculated, and the adjustment data of the reverse trajectory is obtained.

5. The dynamic response method using DC control of automotive intelligent windshield wipers according to claim 1, characterized in that, The process involves generating dynamic motion trajectory parameters for each curved surface region based on the forward trajectory correction and reverse trajectory adjustment values, combined with a speed optimization weight ratio. The trajectory parameter update rate is updated in real time to obtain the adjusted wiper arm motion control data, including: By collecting initial data of the forward trajectory and feedback data of the reverse trajectory, and combining the geometric characteristics of the curved surface region, preliminary correction values ​​and adjustment data are calculated to obtain the basic trajectory parameters of each region. Based on the trajectory parameters, combined with the speed optimization target and weight ratio allocation rules, dynamic motion control commands are generated to determine the motion path adjustment scheme for each surface region. By analyzing the changing trends of trajectory parameters through dynamic motion control commands, the update requirements of parameter refresh frequency are obtained, and the frequency value adapted to the current motion state is determined. If the trend of trajectory parameters exceeds the preset threshold range, the execution logic of motion control is adjusted according to the update requirements of the refresh frequency to obtain optimized control data.

6. The dynamic response method using DC control of automotive intelligent windshield wipers according to claim 1, characterized in that, The process involves driving the wiper arm to execute a dynamic trajectory based on the adjusted wiper arm motion control data. If the pressure distribution in a certain area deviates from a preset threshold, the swing force balance value is fine-tuned to determine the updated control command, including: The sensor array acquires the contact surface pressure data of the wiper arm during the dynamic trajectory execution process, records the pressure value of each area in real time, and obtains a preliminary pressure distribution map. Based on the preliminary pressure distribution map, a pre-established pressure distribution analysis model is used to determine if the pressure value in a certain area exceeds a preset threshold, and then that area is marked as a deviation area. For the marked deviation areas, the corresponding wiper arm motion parameters are extracted, the correlation between the deviation areas and the dynamic trajectory execution is analyzed, and the adjustment direction of the motion parameters corresponding to the deviation areas is obtained. By adjusting the direction of motion parameters corresponding to the deviation area, calculating the adjustment value for oscillation force balance, and using a dynamic feedback adjustment mechanism to fine-tune the driving force of the wiper arm, the adjusted motion control parameters are determined.

7. The dynamic response method using DC control of automotive intelligent windshield wipers according to claim 1, characterized in that, The updated control commands are used to continuously record the operating status of the wiper arm in different curved areas, analyze the effect of friction in the curved areas on long-term operational stability, and, in conjunction with the area resistance compensation value, determine whether the trajectory parameter update rate meets the adaptability requirements, thus obtaining the final long-term parameter adjustment range, including: The operating status data of the wiper arm in multiple curved areas are obtained. The standard deviation of the velocity sequence in each window is calculated using the sliding window method to obtain the stability index. From the preset resistance mapping table, obtain the corresponding resistance value based on the location information of the curved area; Input the stability index and resistance value into the linear regression model, and output the coefficient of friction on stability. If the coefficient is greater than the preset threshold, the trajectory parameters will be updated. The update rate of trajectory parameters is calculated using a proportional-integral controller based on the difference between the action coefficient and the current compensation value. Determine whether the update rate is within the preset adaptive range. If it is, determine the adjustment domain based on the update rate and long-term constraints.

8. The dynamic response method using DC control of automotive intelligent windshield wipers according to claim 1, characterized in that, The process involves periodically optimizing the dynamic feedback adjustment value and speed optimization weight ratio within the final long-term parameter adjustment range to generate the final motion trajectory control logic adapted to the characteristics of asymmetric curved surfaces. This completes the adaptive operation adjustment of the wiper arm on complex curved surfaces, including: By collecting the operating data of the wiper arm on a complex curved surface, the initial values ​​of long-term parameters and parameter ranges are obtained, and the basic adjustment benchmark under the operating environment is determined. Based on the initial values ​​and operating environment data, analyze the dynamic feedback and the changing trend of the feedback values ​​to obtain deviation information under real-time operating conditions; Based on the deviation information, the ratio between the velocity weight and the motion trajectory is adjusted. If the deviation exceeds the preset threshold, the velocity weight is dynamically corrected to obtain the corrected trajectory parameters. By using the corrected trajectory parameters, control logic is generated for asymmetric and complex surfaces, and the operating strategy of the wiper arm in different surface regions is determined.

Images