Electronic control method and system for automobile clearance adjuster with torque sensor

Abstract

The application discloses a kind of electric control method and system of automobile gap adjuster with torque sensing, it is related to automobile brake control technical field.The method comprises: establishing the time window benchmark database of torque change rate, generates temperature shrinkage gap abnormality identification;Extract the reference waveform feature vector of brake release torque waveform under different environmental temperature to construct standard fingerprint library, by matching comparison with real-time waveform feature vector, generate non-temperature effect identification;Non-temperature effect identification is determined by triple verification consistency, obtain temperature shrinkage triple confirmation identification, mechanical wear confirmation identification or verification conflict warning identification;If it is verification conflict warning identification, establish adaptive threshold adjustment algorithm and closed-loop control mechanism, generate final gap adjustment control instruction.The application solves the problem of gap adjustment malfunction induced by material thermal shrinkage characteristics in low temperature environment.

Description

Technical Field

[0001] This invention relates to the field of automotive braking control technology, and more specifically, to an electronic control method and system for an automotive clearance adjuster with torque sensing. Background Technology

[0002] Traditional clearance adjustment systems rely on real-time measurements of mechanical clearance values ​​for adjustment. However, in cold environments, especially at night when temperatures drop sharply, metal components such as brake discs experience significant temperature contraction. A key flaw is that the control system only detects an increase in clearance value but fails to distinguish this abnormal increase from genuine wear-induced clearance increases. This leads to erroneous triggering of clearance compensation mechanisms, such as overcompensation. The system misinterprets the temporary clearance increase caused by material contraction as genuine wear, causing the adjustment mechanism to excessively reduce the clearance, resulting in an actual clearance value far below the normal operating range. This brake lock-up caused by system misadjustment under low-temperature conditions can easily lead to vehicle rollover, collisions, and other accidents, especially during the morning rush hour when cold starts and low temperatures combine. Figure 6 The diagram shown is a structural diagram of an existing gap adjuster.

[0003] Most existing automotive braking control technologies avoid misadjustment through mechanical temperature compensation or temperature-based single-variable control, which makes it difficult to avoid misjudgments of wear due to cold contraction, and lacks accurate adjustment schemes for parameters after problems occur.

[0004] Chinese Patent CN114294358B discloses a disc brake and a clearance adjustment mechanism. The clearance adjustment mechanism includes a clamp body and a first housing assembly, which can move relative to the clamp body during braking. The first housing assembly includes a brake pad, a housing, a compensation mechanism, a transmission mechanism, and a reset component. The compensation mechanism includes a pull ring, a ratchet, and a screw. The pull ring rotates and is sleeved on the ratchet. The screw is inserted into the ratchet and threadedly connected to the housing. One end of the screw supports the brake pad. When the ratchet rotates in its rotation direction, it drives the screw to move the brake pad towards the brake disc. When the first housing assembly moves in the braking direction, a first component on the clamp body drives the transmission mechanism to rotate the pull ring in the opposite direction to the ratchet's rotation. The reset component drives the pull ring to rotate in the ratchet's rotation direction. This clearance adjustment mechanism has a simple structure and high reliability. Chinese patent application CN108189819A discloses a vehicle brake-by-wire device with automatic gap adjustment function, including a brake disc with a left friction pad and a right friction pad on each side. During braking, the motor output shaft drives the central gear to rotate, which in turn drives the planetary carrier to rotate, which in turn drives the lead screw to rotate. The lead screw drives the slider to move axially, and the slider pushes the piston to the left through the pawl. This causes the right friction pad to come into contact with the brake disc, while the left friction pad comes into contact with the brake disc, generating braking torque and achieving braking. When the brake is released, the motor reverses, the lead screw and slider return to their initial positions, and the piston and the right friction pad return to their initial positions under the elastic action of the rectangular sealing ring. The friction pads separate from the brake disc, releasing the brake. This invention can achieve vehicle braking and has an automatic gap adjustment function.

[0005] However, the aforementioned existing technologies rely on passive compensation through mechanical structures, failing to perceive the thermal contraction of components. This makes them prone to misinterpreting transient gap increases caused by low-temperature contraction as wear, leading to over-adjustment. Furthermore, the lack of torque change monitoring exacerbates the risk of misadjustment during periods of significant diurnal temperature variation. While electronic control exists, it is based on mechanical motion triggering, lacking a temperature-torque correlation model. Relying on single-position feedback, it cannot identify "false gap increases" at low temperatures. Due to the lack of dynamic calibration, adjustment deviations are prone to occur in cold regions or during temperature fluctuations, posing safety hazards. Additionally, both technologies lack hierarchical storage of historical data and adaptive threshold optimization. With prolonged vehicle use, component aging, or actuator replacement, the system becomes unsuitable for new operating conditions, resulting in decreased adjustment accuracy. Summary of the Invention

[0006] This invention is applicable to all types of vehicles (gasoline, hybrid, and pure electric) with disc or drum brake platforms, and is particularly suitable for urban commuting scenarios in cold winter regions with large day-night temperature differences, as well as operating conditions that easily trigger low-temperature contraction, such as frequent start-stop cycles and restarts after long periods of parking. The system has good compatibility with different brands of brake actuators, different sampling frequencies, and sensor accuracies, and can be deployed as a pre-installed electronic control subsystem in new vehicles or as an aftermarket upgrade for in-service vehicles.

[0007] In low-temperature or rapidly cooling environments, it is difficult to distinguish between the transient gap increase caused by material thermal contraction and the gap increase caused by actual wear, which can easily lead to false triggering and over-adjustment of compensation, and even induce safety risks such as engine lock-up. Therefore, there is a need for an electronic control method and system that can distinguish between temperature effects and mechanical wear online and can adaptively correct thresholds and strategies without affecting braking safety and user experience.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] An electronic control method for automotive clearance adjusters with torque sensing includes:

[0010] A time window benchmark database is established, and the time window benchmark database is tested to obtain temperature contraction candidate identifiers. The temperature contraction candidate identifiers are differentiated to obtain temperature contraction confirmation identifiers. The temperature contraction confirmation identifiers are judged through a multi-level decision tree to obtain temperature contraction gap anomaly identifiers.

[0011] A standard fingerprint database is established, real-time braking release torque waveforms are collected and feature analysis is performed to obtain real-time waveform feature vectors. The real-time waveform feature vectors are combined with the standard fingerprint database to perform similarity calculations to obtain temperature consistency confirmation identifiers and waveform matching success identifiers. Combined with temperature contraction gap anomaly identifiers, non-temperature effect identifiers are obtained.

[0012] Based on non-temperature effect identification, a micro-perturbation excitation signal program is established. Temperature response characteristic parameters are extracted from the micro-perturbation excitation signal program. Material state discrimination is performed on the temperature response characteristic parameters to obtain the material state discrimination results. Triple verification consistency judgment is then performed to obtain the triple verification results.

[0013] Based on the triple verification results, historical data recording units are extracted and abnormal data is identified. The abnormal data is adjusted to obtain the adjusted threshold. Based on the adjusted threshold, the final gap adjustment control command is obtained.

[0014] Furthermore, the method for obtaining the temperature shrinkage confirmation indicator includes:

[0015] The torque change rate feature vector is extracted from the temperature shrinkage candidate identifiers. The feature vector includes the average change rate, the standard deviation of the change rate, and the slope of the change trend.

[0016] The average rate of change, standard deviation of the rate of change, and slope of the trend are calculated using a differential judgment formula to obtain a temperature shrinkage confirmation indicator.

[0017] Furthermore, the method for obtaining the abnormal identification of the temperature contraction gap includes:

[0018] Based on the temperature contraction confirmation flag, the current external ambient temperature and current time are detected through a multi-level decision tree;

[0019] If the current external ambient temperature is lower than the preset low temperature threshold, a low temperature environment identifier is generated; if the current time is during a period of high temperature contraction, a low temperature environment identifier is generated.

[0020] When both the low-temperature environment indicator and the low-temperature environment indicator are generated simultaneously, an abnormal temperature contraction gap indicator is obtained.

[0021] Furthermore, the method for establishing a standard fingerprint database includes:

[0022] The braking release torque waveform is obtained by measuring the braking release torque waveform under different external ambient temperatures. During the torque release phase after each braking operation, the torque sensor continuously samples the waveform to obtain a reference braking release torque waveform.

[0023] Feature analysis was performed on the reference braking release torque waveform to extract key feature parameters of the torque decay waveform. Key feature parameters include the reference torque return-to-zero time, the number of oscillations of the reference waveform, and the reference decay time constant.

[0024] The reference torque zero-return time, the number of reference waveform oscillations, and the reference decay time constant are combined to obtain the reference waveform feature vector;

[0025] The reference waveform feature vectors obtained under different temperature conditions are combined to obtain a standard fingerprint library.

[0026] Furthermore, the method for obtaining the temperature consistency confirmation identifier includes:

[0027] The reference torque zero-return time, the number of reference waveform oscillations, and the reference decay time constant are calculated to obtain the highest similarity value. When the highest similarity value is greater than the preset similarity threshold, a waveform matching success identifier is generated.

[0028] The similarity between the real-time waveform feature vector and each reference feature vector in the standard fingerprint database is calculated to obtain the waveform matching feature result;

[0029] The temperature deviation is calculated based on the waveform matching characteristics. When the temperature deviation meets the preset temperature consistency conditions, a temperature consistency confirmation mark is obtained.

[0030] Furthermore, the micro-perturbation excitation signal program includes:

[0031] The micro-perturbation excitation signal procedure includes an excitation safety condition check mechanism;

[0032] Incentive safety condition check mechanism, including vehicle stationary status check, parking brake activation status check and system safety mode check;

[0033] The vehicle stationary status check determines whether the vehicle is completely stationary; the parking brake activation status check confirms whether the parking brake system is activated; and the system safety mode check confirms that the vehicle safety system is in normal working condition.

[0034] When the vehicle is completely stationary, the parking brake system is engaged, and the vehicle safety system is functioning normally, a micro-disturbance excitation control signal is sent.

[0035] Furthermore, the method for obtaining the temperature response characteristic parameters includes:

[0036] The response characteristic parameters are obtained by high-frequency sampling of the disturbance excitation control signal using a torque sensor.

[0037] The response characteristic parameters are corrected to obtain the temperature response characteristic parameters.

[0038] Furthermore, the method for obtaining the material state discrimination result includes:

[0039] Obtain the threshold set of response characteristic parameters for low-temperature contraction and the threshold set of response characteristic parameters for mechanical wear.

[0040] The material state discrimination result is obtained by comparing the temperature response characteristic parameters with the threshold set of response characteristic parameters of the low temperature shrinkage state and the threshold set of response characteristic parameters of the mechanical wear state.

[0041] Furthermore, the method for identifying the abnormal data includes:

[0042] Based on the triple verification results, historical data record units are extracted;

[0043] Historical data recording units are stored in layers and time series identification is performed to obtain abnormal patterns.

[0044] An electronic control system for a vehicle clearance adjuster with torque sensing, used to implement the aforementioned electronic control method for a vehicle clearance adjuster with torque sensing, the system comprising:

[0045] Abnormal Temperature Identification Module: This module is used to establish a time window benchmark database, detect the time window benchmark database to obtain temperature contraction candidate identifiers, perform differential judgment on the temperature contraction candidate identifiers to obtain temperature contraction confirmation identifiers, and judge the temperature contraction confirmation identifiers through a multi-level decision tree to obtain temperature contraction gap abnormality identifiers.

[0046] Temperature effect identification module: used to establish a standard fingerprint database, collect real-time braking release torque waveform and perform feature analysis to obtain real-time waveform feature vector, combine the real-time waveform feature vector with the standard fingerprint database to perform similarity calculation, obtain temperature consistency confirmation mark and waveform matching success mark, and combine with temperature contraction gap abnormal mark to obtain non-temperature effect mark;

[0047] Anomaly determination module: Based on non-temperature effect identifiers, it establishes a micro-perturbation excitation signal program, extracts temperature response characteristic parameters from the micro-perturbation excitation signal program, performs material state discrimination on the temperature response characteristic parameters, obtains the material state discrimination result, performs triple verification consistency judgment, and obtains the triple verification result;

[0048] Self-learning module: Based on the triple verification results, it extracts historical data recording units and identifies abnormal data, adjusts the abnormal data to obtain the adjusted threshold, and obtains the final gap adjustment control command based on the adjusted threshold.

[0049] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0050] This invention, based on time-windowed modeling of torque change rate, combines a temperature effect process to generate anomaly markers for temperature contraction gaps, eliminating "pseudo-anomalies" induced by environmental cooling contraction at the source, significantly improving the accuracy and stability of anomaly identification. By constructing a standard fingerprint library of braking release torque waveforms and introducing a temperature consistency verification mechanism, real-time waveforms are evaluated for similarity under reference temperature conditions, enabling online identification of temperature factors and reducing the probability of misjudging low-temperature contraction as mechanical wear. Furthermore, it requires no additional hardware sensors and has good engineering integrability. For non-temperature effect scenarios, a detection strategy combining micro-perturbation excitation satisfying safety constraints and temperature compensation is adopted to extract response features and perform material state discrimination. This more robustly separates the influence of mechanical wear and ambient temperature than a single threshold method, thereby improving diagnostic robustness. Simultaneously, a triple verification consistency judgment and confidence evaluation are introduced, outputting quantifiable decision reliability, providing a clear basis for control strategies, and effectively suppressing the risks of false triggering and over-adjustment. When verification conflicts occur, the system continuously calibrates key thresholds and parameters by relying on hierarchical recording of historical data, adaptive threshold adjustment, and closed-loop optimization control. This adapts to vehicle model differences, seasonal changes, and aging, maintaining stable judgment and control performance over the long term and generating a clear final clearance adjustment control command. In summary, this invention provides a unified, accurate, and continuously learning electronic clearance adjustment control scheme, significantly reducing safety risks such as vehicle lock-up and runaway under low-temperature conditions, improving maintenance efficiency, extending component lifespan, and meeting braking safety requirements under harsh climates and complex operating conditions. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a flowchart illustrating the principle of the electronic control method for an automotive clearance adjuster with torque sensing in this invention.

[0053] Figure 2 This is a flowchart illustrating the judgment process for abnormal temperature contraction gap identification in the electronic control method for an automotive gap adjuster with torque sensing of the present invention.

[0054] Figure 3 This is a flowchart illustrating the determination of non-temperature effect identifiers in the electronic control method for an automotive clearance adjuster with torque sensing according to the present invention.

[0055] Figure 4 This is a flowchart illustrating the determination of the triple verification results in the electronic control method for an automotive clearance adjuster with torque sensing of the present invention.

[0056] Figure 5 This is a functional block diagram of the electronic control system for an automotive clearance adjuster with torque sensing according to the present invention.

[0057] Figure 6 This is a structural diagram of a gap adjuster in the prior art. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Example 1

[0060] Please see Figure 1 As shown, this embodiment provides an electronic control method for an automotive clearance adjuster with torque sensing, including:

[0061] Step S10: Establish a time window benchmark database, detect the time window benchmark database to obtain temperature contraction gap anomaly indicators, perform differential judgment on the temperature contraction gap anomaly indicators to obtain temperature contraction confirmation indicators, and judge the temperature contraction confirmation indicators through a multi-level decision tree to obtain temperature contraction gap anomaly indicators.

[0062] Furthermore, such as Figure 2 As shown, the specific process of step S10 includes:

[0063] Step S11: Establish a time window reference database, detect the time window reference database to determine whether there is a gap abnormal increase signal, when there is a gap abnormal increase signal, perform backtracking analysis on the time window of the gap abnormal increase signal, and obtain temperature shrinkage candidate identifiers after judging the continuity of torque change rate.

[0064] Among them, the time window reference database for torque change rate is obtained by using a preset time reference period. The time window is divided into N time window units. The original torque values ​​within the N time window units are collected and the torque change rate is calculated. The torque change rate obtained from the N time window units is combined with the external ambient temperature and gap status indicators corresponding to each window unit to construct the torque change rate.

[0065] Specifically, the preset time base period The time window is divided into N time window units. Within each time window unit, the raw torque value output by the torque sensor is continuously collected. The time derivative of the raw torque value within each time window unit is calculated to obtain the torque change rate. The time reference period is... This indicates the duration of a complete detection cycle for the brake clearance adjuster; N represents the total number of time window units, and the length of each time window unit is... Satisfying the relation The original torque value represents the instantaneous torque measurement value of the braking system within that time window unit, and the torque change rate represents the magnitude of the change in torque value per unit time.

[0066] The time window benchmark database contains multiple data recording units. Each data recording unit stores the torque change rate, external ambient temperature, and gap status identifier for the corresponding time window unit. The external ambient temperature represents the average ambient temperature of the external weather within the time window, and the gap status identifier represents the brake gap status identifier at the end of the time window, as shown in Table 1. For example, the brake gap status identifier is 0 for normal, 1 for low temperature contraction, 2 for high temperature expansion, 3 for abnormal increase (warning), 4 for abnormal decrease (jamming), and 5 for irreversible wear.

[0067] Table 1. Criteria for Determining Braking Clearance Status

[0068] serial number Status Name Judgment basis 0 normal 1. The increase in brake clearance is within the preset normal range (0~Amm, where A is the maximum normal increment defined based on brake system design standards, natural wear patterns, and industry safety standards, in mm); 2. The torque change rate has no abrupt change and conforms to the change range under normal operating conditions (±B% / s, where B is the upper limit of normal fluctuations statistically defined by the benchmark database); 3. The ambient temperature is within the normal range (-10℃~40℃, based on the temperature distribution under normal conditions in the benchmark database). 1 Low temperature shrinkage 1. There is an abnormal increase in gap (gap increase > Amm); 2. Backtracking analysis of the abnormal time window shows that the torque change rate has a characteristic abrupt change due to component shrinkage (abrupt change amplitude > C% / s, where C is the low temperature-related abrupt change threshold); 3. The ambient temperature is < -10℃ (below the lower limit of the normal temperature range), and the increase in gap is strongly correlated with low temperature. 2 High temperature expansion 1. The torque change rate exhibits a characteristic abrupt change due to component expansion (abrupt change amplitude < -D% / s, where -D is the high-temperature related negative abrupt change threshold); 2. Ambient temperature > 40℃ (above the upper limit of the normal temperature range); 3. Brake clearance reduction > Emm (E is the upper limit of the normal reduction range defined based on the high-temperature expansion law), which conforms to the characteristics of high-temperature expansion. 3 Abnormal increase (warning) 1. Brake clearance increase > Amm (exceeding the preset normal range); 2. Excluding normal temperature factors such as low temperature contraction (ambient temperature ≥ -10℃); 3. Torque change rate shows a sudden change related to abnormal clearance increase (the change amplitude is > F% / s, where F is the threshold for abnormal increase of non-temperature factors). 4 Abnormal decrease (stagnation) 1. Brake clearance reduction > Gmm (G is the upper limit of the normal reduction range defined based on design standards and temperature expansion law, and G > E); 2. Torque change rate causes abnormal abrupt change due to component jamming (abrupt change amplitude < -H% / s, -H is the negative abrupt change threshold related to jamming, and -H < -D); 3. Clearance reduction exhibits non-temperature-related abnormal shrinkage characteristics. 5 Irreversible wear 1. Brake clearance increase > Imm (I is the maximum allowable wear value based on long-term natural wear patterns, and I > A); 2. The increased clearance is confirmed by testing to be caused by irreversible wear (not due to temperature, jamming, or other reversible factors); 3. The torque change rate shows a continuous abnormal increasing trend (the abrupt change amplitude in more than 3 consecutive time windows is > J% / s, where J is the wear-related continuous abrupt change threshold).

[0069] It is important to note that A, B, C, D, E, F, and J in the table are specific thresholds based on braking system design parameters, benchmark databases, and industry standards, and need to be calibrated according to the specific vehicle model and monitoring system. The abnormal increase in clearance signal refers to whether the increase in clearance exceeds the preset normal range, forming a brake clearance status indicator representing an abnormal increase. For example, the abnormal increase represented by number 3 in Table 1 conforms to the condition of clearance increase exceeding the preset normal range. The increase in clearance refers to the increment in the brake clearance between the brake pads and the brake disc (or brake drum) in the automotive braking system during use. The preset normal range for the increase in clearance is determined based on a comprehensive consideration of braking system design standards, natural wear patterns, temperature effects, benchmark database statistics, and industry safety standards, ensuring safe braking performance and conforming to physical laws.

[0070] Specifically, when the brake clearance monitoring module detects an abnormal increase in clearance, it immediately initiates a time window backtracking analysis program to trace the cause of the abnormal increase in brake clearance. By backtracking and analyzing data such as torque change rate, external ambient temperature, and clearance status indicators in each time window unit, the program clarifies the development process of the abnormal increase, providing a basis for subsequent processing or early warning.

[0071] Time window backtracking analysis refers to starting from the current time window unit and going back M historical time window units, performing a continuous analysis of the torque change rate in these M historical time window units, and calculating the difference in torque change rate between adjacent time window units. Here, M represents the number of historical time window units required for the backtracking analysis.

[0072] A continuity judgment condition for torque change rate is established: when the difference in torque change rate across M consecutive time window units is less than a preset continuity threshold, a temperature contraction candidate identifier is generated for the time window units that meet the condition. When the difference in torque change rate across multiple consecutive time windows is greater than the preset continuity threshold, it indicates a sudden or drastic change, which differs from temperature contraction and is judged as a mechanical wear problem. This is because abnormal clearance caused by mechanical wear manifests as large and rapid fluctuations in torque change rate. The continuity threshold is determined based on the thermal expansion coefficient and mechanical stiffness of the core components of the braking system, combined with bench test data from different temperature ranges and reference to historical fault data of similar systems, ensuring the ability to distinguish between continuous changes caused by temperature and sudden changes caused by mechanical faults. For example, during the winter night from 22:00 to 00:00 the next day, divided into four 30-minute time windows, the ambient temperature drops from 5℃ to -2℃. The torque change rates are -0.5, -0.55, -0.6, and -0.58 Nm / h, respectively. The adjacent differences of 0.05, 0.05, and 0.02 are all less than the continuity threshold of 0.2 Nm / h, which conforms to the continuous change characteristics of temperature contraction. Here, the continuity threshold represents the critical value for judging the continuity of the torque change rate, and the temperature contraction candidate indicator represents the candidate state of gap anomaly suspected to be caused by temperature contraction.

[0073] Step S12: After obtaining the candidate identifier for temperature shrinkage, extract the torque change rate feature vector used for differentiation judgment from the candidate identifier for temperature shrinkage, and obtain the confirmation identifier for temperature shrinkage through the differentiation judgment formula.

[0074] The torque change rate feature vector used for differentiation judgment is based on the obtained temperature contraction candidate identifiers. It extracts features from the torque change rates of the M backtracking time window units upon which the temperature contraction candidate identifiers are generated, producing a torque change rate feature vector. This vector contains three feature components: average change rate, standard deviation of the change rate, and slope of the change rate trend. The average change rate is equal to the arithmetic mean of the torque change rates in the M time window units. The standard deviation of the change rate reflects the dispersion of the torque change rates in the M time window units. The slope of the change rate trend represents the linear trend of the torque change rate over time in the M time window units.

[0075] The differentiation judgment formula is based on the three characteristic components of the torque change rate characteristic vector, and the formula is as follows: The temperature shrinkage confirmation indicator is calculated using a formula. The average rate of change representing the rate of change of torque, i.e. , Let be the rate of change of torque over i time window units; The threshold representing the average rate of change is based on the average distribution of the normal braking torque change rate from a large number of real vehicle tests, the average value of temperature shrinkage from low-temperature environment tests, and the critical value after dynamic correction of the response delay of the brake hydraulic system. The standard deviation of the rate of change, i.e. , The standard deviation threshold is determined through experimental statistics based on the stable characteristics of the torque change rate with low dispersion during temperature contraction, combined with the inherent noise level of the torque sensor. For example, the standard deviation threshold, based on the statistically stable characteristics of torque change during temperature contraction, is 2-3 times higher than the sensor's inherent noise level. K represents the slope of the change rate trend, which is fitted using the least squares method to establish a linear relationship between the torque change rate and time. We obtain , where t is time and b is the intercept. This represents the trend threshold. It compares the gentle linear trend caused by temperature contraction with the steep trend of an abnormal state, and, combined with the theoretical maximum gentle slope calculated based on the time window unit length, takes the median value of the slopes for both types of states. Indicates intersection.

[0076] Step S13: Establish a multi-level decision tree to judge the temperature contraction confirmation flag and obtain the temperature contraction gap abnormality flag.

[0077] The multi-level decision logic tree's judgment logic is as follows: If a low-temperature environment indicator and a time-sensitive indicator are generated simultaneously, based on the generation of a temperature contraction confirmation indicator, an abnormal temperature contraction gap indicator is obtained. A large average rate of change indicates a large overall amplitude of torque change, possibly due to brake pad wear or other mechanical faults. If the standard deviation exceeds a certain threshold, it indicates that the change is unstable within the time window, exhibiting significant fluctuations, which may be a sign of accelerated wear. If the slope of the change trend exceeds a set trend threshold, it indicates a significant upward or downward trend in the change, not a smooth change caused by temperature contraction, but rather by mechanical problems. When the standard deviation exceeds the standard deviation threshold, or the absolute value of the change trend slope exceeds the trend threshold, it indicates that the amplitude and trend of torque change do not conform to the smooth change pattern of temperature contraction, which can be used as an indicator of mechanical wear.

[0078] The low-temperature environment indicator checks whether the current ambient temperature meets the low-temperature requirements. If the current ambient temperature is lower than a preset low-temperature threshold, a low-temperature environment indicator is generated, indicating that the system is currently in a low-temperature environment. The preset low-temperature threshold is determined based on the significant contraction temperature of the core components of the braking system, combined with climate data from the vehicle's primary operating region, and taking into account the impact of brake fluid viscosity changes at low temperatures on torque transmission. The preset low-temperature threshold represents the critical ambient temperature value for initiating temperature contraction monitoring.

[0079] The time-sensitive flag checks whether the current time falls within a period of high temperature contraction. If the current time falls within a preset nighttime or early morning period, a time-sensitive flag is generated, indicating that the current time is within a temperature-sensitive period. The preset nighttime and early morning periods are set based on the diurnal temperature variation patterns in meteorological data. For example, nighttime is from 22:00 to 5:00 the next day (when the ambient temperature continuously drops to its lowest point of the day), and early morning is from 5:00 to 8:00 (when the temperature difference is most drastic). The preset nighttime period represents the time range of low nighttime temperatures, and the preset early morning period represents the time range of drastic early morning temperature variations.

[0080] Step S10 addresses the problem that traditional braking system monitoring methods cannot effectively distinguish between gap abnormalities caused by temperature changes and mechanical wear. By introducing a torque sensor and time window analysis of torque change rate, combined with multi-dimensional information such as ambient temperature and time characteristics, a differentiated judgment mechanism and decision logic tree are constructed. This effectively identifies gap changes caused by temperature contraction, improving the accuracy and reliability of the system's judgment.

[0081] Step S20: Establish a standard fingerprint database, collect real-time braking release torque waveforms and perform feature analysis to obtain real-time waveform feature vectors, combine the real-time waveform feature vectors with the standard fingerprint database to perform similarity calculations, obtain temperature consistency confirmation flags and waveform matching success flags, and combine them with temperature contraction gap anomaly flags to obtain non-temperature effect flags.

[0082] Furthermore, such as Figure 3 As shown, the specific process of step S20 includes:

[0083] Step S21: Establish a standard fingerprint database. When the brake release operation ends, collect the real-time brake release torque waveform and record the real-time external ambient temperature. Perform feature analysis on the real-time brake release torque waveform to obtain the real-time waveform feature vector.

[0084] Specifically, the braking release torque waveform is measured under different external ambient temperatures. During the torque release phase after each braking operation, continuous sampling is performed by a torque sensor to obtain a reference braking release torque waveform. Feature analysis is performed on the reference braking release torque waveform to extract key feature parameters of the torque decay waveform. The key feature parameters include the reference torque return-to-zero time, the number of reference waveform oscillations, and the reference decay time constant. The reference torque return-to-zero time, the number of reference waveform oscillations, and the reference decay time constant are combined to obtain the reference waveform feature vector.

[0085] The standard fingerprint database is obtained by combining the feature vectors of reference waveforms obtained under different temperature conditions. The reference torque zero-return time represents the time required for the torque to decay from its peak value to zero. The number of reference waveform oscillations represents the number of oscillation cycles that occur during the torque decay process. The reference decay time constant represents the exponential decay time constant of the torque decay process. The reference decay time constant is fitted to the torque decay data using the exponential model T(t) = 10・e⁻ᵗ / τ, where T(t) represents the torque value at time t, which is a torque function that changes with time. t represents a specific moment in the torque decay process, and τ represents the decay time constant in seconds (s). For example, when t = 2 seconds, T = 3.68 (approximately 10 / e), and we obtain τ = 2 seconds, meaning the decay time constant is 2 seconds.

[0086] The real-time brake release torque waveform is acquired by the real-time torque waveform acquisition program immediately upon detecting the brake release operation completion signal during actual system operation. The brake operation completion signal indicates the status signal of brake pedal release or brake command cancellation, while the real-time torque waveform data represents the complete torque time series data of the current brake release process.

[0087] The real-time waveform feature vector is obtained by performing feature analysis on real-time torque waveform data, extracting the real-time torque zero-return time, the number of real-time waveform oscillations, and the real-time decay time constant, and combining the real-time torque zero-return time, the number of real-time waveform oscillations, and the real-time decay time constant.

[0088] Step S22: Calculate the similarity between the real-time waveform feature vector and the reference feature vector in the standard fingerprint database to obtain the waveform matching feature result.

[0089] The waveform matching feature result is obtained by calculating the similarity between the real-time waveform feature vector and each reference feature vector in the standard fingerprint database. This yields the similarity of torque zero-return time, oscillation frequency, and decay time constant. A comprehensive similarity evaluation function is constructed to obtain the highest similarity value. The reference ambient temperature corresponding to the highest similarity value is extracted from the fingerprint database. The waveform matching feature result includes the highest similarity value and the corresponding reference ambient temperature in the standard fingerprint database at the time the highest similarity value is obtained. The highest similarity value is determined by comparing the real-time waveform feature vector with the comprehensive similarity scores of all reference feature vectors in the standard fingerprint database; the maximum value of this comparison is the highest similarity value. The similarity evaluation function comprehensively considers the similarity of the three components: torque zero-return time, oscillation frequency, and decay time constant. After normalization, a weighted sum is used to generate the comprehensive similarity score.

[0090] Specifically, the torque zero-return time similarity is based on the relative deviation between the reference torque zero-return time and the real-time torque zero-return time, i.e., according to the formula:

[0091] get.

[0092] The oscillation similarity is based on the degree of difference between the oscillation number of the reference waveform and the oscillation number of the real-time waveform, i.e., according to the formula:

[0093] get.

[0094] The decay time constant similarity is based on the relative deviation between the reference decay time constant and the real-time decay time constant, i.e., according to the formula:

[0095] get.

[0096] Step S23: Temperature effect judgment is performed on the waveform matching feature results to obtain a waveform matching success identifier. The consistency of the reference ambient temperature corresponding to the highest similarity value in the standard fingerprint database with the real-time ambient temperature is verified to obtain a temperature consistency confirmation identifier. Based on the obtained temperature contraction gap anomaly identifier, a non-temperature effect identifier is generated by combining the waveform matching success identifier and the temperature consistency confirmation identifier.

[0097] The waveform matching success flag is generated when the temperature effect judgment condition is met. This condition is determined when the highest similarity value exceeds a preset similarity threshold. The similarity threshold is set by comparing the difference between the standard fingerprint database and real-time data to ensure that temperature effects are not misjudged as mechanical faults. For example, at 25°C, the reference torque zero-return time is 3 seconds, the real-time acquired torque zero-return time is 3.4 seconds, and the similarity threshold is set to ±0.5 seconds. Since the difference of 0.4 seconds is less than the threshold, the system considers the torque zero-return times similar, and the temperature effect is normal.

[0098] Consistency verification involves verifying the consistency between the reference ambient temperature and the real-time ambient temperature in the waveform matching feature results, and calculating the temperature deviation between the two. The aim is to ensure that the accuracy of the waveform matching judgment is not affected and to guarantee the reliability of the temperature effect verification results for the waveform features.

[0099] The temperature consistency confirmation flag is obtained when the temperature deviation meets the preset temperature consistency condition. The temperature consistency condition is based on the degree of influence of the temperature deviation on the characteristic parameters of the torque attenuation waveform. Generally, if the temperature deviation is ≤5°C, the influence is within the acceptable range, and the system continues to make normal judgments.

[0100] Based on the obtained temperature contraction gap anomaly indicator, the waveform matching success indicator and temperature consistency confirmation indicator are used for judgment. When both waveform matching success indicator and temperature consistency confirmation are generated simultaneously, a temperature effect gap anomaly indicator is generated. If no temperature contraction gap anomaly indicator is obtained, or if either the temperature consistency confirmation indicator or the waveform matching success indicator is not met, a non-temperature effect indicator is generated. The non-temperature effect indicator indicates that the current gap anomaly may be caused by actual wear or other mechanical factors, and further verification is required in the subsequent micro-disturbance detection steps.

[0101] Step S20 solves the problem of inaccurately distinguishing between temperature effects and mechanical faults in traditional methods, providing higher diagnostic accuracy and system reliability, thereby improving the safety and maintenance efficiency of the braking system. The system can accurately determine the impact of temperature changes on the braking system and effectively distinguish between gap abnormalities caused by temperature contraction (or expansion) and abnormalities caused by non-temperature effects such as mechanical wear.

[0102] Step S30: Based on the non-temperature effect identifier, establish a micro-perturbation excitation signal program, extract temperature response characteristic parameters from the micro-perturbation excitation signal program, perform material state discrimination on the temperature response characteristic parameters, obtain the material state discrimination result, and perform triple verification consistency judgment to obtain the triple verification result.

[0103] Furthermore, such as Figure 4 As shown, the specific process of step S30 includes:

[0104] Step S31: Based on the non-temperature effect identifier, establish a micro-perturbation excitation signal program, extract temperature response characteristic parameters from the micro-perturbation excitation signal program, perform temperature correlation analysis on the response characteristic parameters, and obtain the temperature response characteristic parameters.

[0105] The micro-disturbance excitation signal program includes an excitation safety condition check mechanism. When the excitation safety condition is met, a micro-disturbance excitation control signal is sent to the brake actuator.

[0106] Specifically, upon obtaining the non-temperature effect indicator, a micro-disturbance excitation signal procedure is initiated to further verify the cause of the gap anomaly. An excitation safety condition check mechanism is established, including vehicle stationary state checks, parking brake activation state checks, and system safety mode checks. Specifically, the vehicle stationary state check confirms the vehicle is completely stationary; the parking brake activation state check confirms the parking brake system is engaged; and the system safety mode check confirms the vehicle safety system is in normal working condition. After all established safety conditions are met, a micro-disturbance excitation control signal is sent to the brake actuator. The micro-disturbance excitation control signal contains the excitation torque amplitude, excitation duration, and excitation waveform type. The excitation torque amplitude represents the magnitude of the small torque excitation applied to the braking system; the excitation duration represents the length of a single micro-disturbance excitation signal; and the excitation waveform type indicates the waveform type of the excitation signal. For example, a sine wave excitation waveform is identified as SW-30Hz, representing a sine wave with a frequency of 30Hz.

[0107] Specifically, before sending the micro-disturbance excitation control signal to the brake actuator, the excitation torque amplitude is set to a tiny percentage value, less than 10% of the normal braking torque, to ensure that the excitation process during its duration does not produce any actual braking effect. The percentage value of the normal braking torque is determined through bench testing to reach the minimum torque threshold for brake pad-disc contact. A value below this minimum torque threshold and less than 10% is selected to ensure no actual braking while maintaining a measurable response. For example, for a vehicle with a normal braking torque of 200-600 Nm, bench testing determines the minimum threshold for brake pad-disc contact to be 30 Nm. The micro-disturbance excitation torque is set to 5% of the normal braking torque, i.e., 10 Nm, which, while having no actual braking effect, can still be detected by the sensor, meeting the testing requirements. The excitation duration is set to a short pulse in the millisecond range to ensure the instantaneousness and safety of the excitation process. The excitation waveform type is set to a waveform and frequency that can distinguish state differences, ensuring no actual braking effect. The excitation waveform type is determined through multiple bench tests, comparing the response sensitivity of waveforms such as sine waves and pulse waves to material states based on the dynamic response characteristics of the braking system.

[0108] During the application of the micro-perturbation excitation control signal and the response observation time after the excitation ends, high-frequency sampling is performed using a torque sensor to obtain response characteristic parameters, including peak response torque, response delay time, and response decay rate. The response observation time represents the complete time window for observing the micro-perturbation excitation response; the peak response torque represents the maximum response amplitude detected by the torque sensor after excitation; the response delay time represents the time delay from the application of the excitation signal to the appearance of the response peak; and the response decay rate represents the rate at which the response signal decays from the peak to a steady state.

[0109] Temperature correlation analysis involves applying temperature compensation to response characteristic parameters based on the current ambient temperature to obtain temperature response characteristic parameters, including peak torque, temperature response delay time, and temperature response decay rate. Specifically, temperature compensation correction is achieved by correlating the response characteristic parameters with the external ambient temperature. By analyzing the impact of different ambient temperatures on these parameters, each characteristic parameter exhibits different trends at different temperatures. The compensation function is derived based on a linear or nonlinear regression model, or by establishing a mathematical model based on the physical characteristics of the braking system. The formula is as follows: .in Represents the temperature response characteristic parameter, The characteristic parameter representing the uncompensated response. The temperature sensitivity coefficient, representing the characteristics of the braking system, is determined by experimentally measuring the degree of change in the braking system's response characteristics (such as peak torque, response delay time, and response fade rate) under different ambient temperatures. For example, if 10 standard braking operations are performed within 5 set temperature ranges, and the temperature increases by 30 degrees Celsius from 5 degrees to 35 degrees Celsius, and the peak torque changes by 6 Nm from 100 Nm to 106 Nm, then the temperature sensitivity coefficient... It is 6 Nm / 30℃, and 0.2 Nm / ℃.

[0110] Represents the current external ambient temperature. Representing the external temperature, temperature response characteristic parameters are generated according to the formula to eliminate the interference of temperature on the response characteristic parameters.

[0111] For example, bench testing revealed that -10℃ causes the peak torque response to be 5 Nm higher than the standard temperature of 25℃. If the peak response measured at -10℃ in real time is 15 Nm, after compensation (15-5=10 Nm), the 10 Nm is compared with the thresholds for low-temperature shrinkage (e.g., threshold ≤ 8 Nm) and mechanical wear (e.g., threshold ≥ 12 Nm) to accurately determine the material state. Material state determination is achieved by comparing the temperature response characteristic parameters with the threshold sets of response characteristic parameters for the low-temperature shrinkage state and the mechanical wear state, respectively, to obtain the low-temperature shrinkage material state identifier and the mechanical wear material state identifier. These are used to distinguish between the low-temperature shrinkage state and the mechanical wear state.

[0112] Step S32: Perform material state discrimination on the temperature response characteristic parameters to obtain the material state discrimination result.

[0113] The material state discrimination involves comparing temperature response characteristic parameters with threshold sets of response characteristic parameters for both the low-temperature shrinkage and mechanical wear states to obtain low-temperature shrinkage and mechanical wear material state identifiers, respectively. The material state discrimination results are used to distinguish between the low-temperature shrinkage and mechanical wear states. The material state discrimination results include low-temperature shrinkage material state identifiers and mechanical wear material state identifiers.

[0114] Specifically, the threshold set of response characteristic parameters for low-temperature contraction includes the low-temperature contraction response peak threshold, the low-temperature contraction response delay threshold, and the low-temperature contraction response decay rate threshold. These thresholds are determined based on the physical property that the material stiffness increases under low-temperature conditions. For example, by applying standardized micro-perturbations at different low temperatures (e.g., -20℃ to 0℃) and collecting data such as response peak torque, response delay time, and response decay rate, it was found that low temperatures reduce the peak value, increase the delay, and decrease the decay rate. Based on these experimental statistical patterns, combined with a safety margin (e.g., a fluctuation of 5% to 10% above or below the response peak threshold), corresponding threshold ranges are set.

[0115] A set of characteristic parameter thresholds for mechanical wear states was established, including the peak value threshold, delay threshold, and decay rate threshold for the mechanical wear response. These thresholds are determined based on the changes in contact surface and frictional characteristics caused by wear, stemming from the alterations wear brings to the contact surface and frictional characteristics: wear roughens the contact surface and reduces the contact area, leading to an increase in the peak value, a shortening of the delay, and an increase in the decay rate. Different wear levels were simulated on a test bench, and after collecting the response parameters, thresholds conforming to the wear characteristics were set.

[0116] The material state determination criteria are as follows: when the peak torque of the temperature response is less than the peak threshold of the low-temperature shrinkage response, the temperature response delay time is greater than the low-temperature shrinkage response delay threshold, and the temperature response decay rate is less than the low-temperature shrinkage response decay rate threshold, a low-temperature shrinkage material state identifier is generated; when the peak torque of the temperature response is greater than the peak threshold of the mechanical wear response, the temperature response delay time is less than the mechanical wear response delay threshold, and the temperature response decay rate is greater than the mechanical wear response decay rate threshold, a mechanical wear material state identifier is generated. If the determination criteria fail to meet the conditions for generating a low-temperature shrinkage material state identifier or a mechanical wear material state identifier, a verification conflict warning is generated. In other words, the material state determination result includes a low-temperature shrinkage material state identifier, a mechanical wear material state identifier, and a verification conflict warning.

[0117] Step S33: After obtaining the material state discrimination result, perform triple verification consistency judgment to obtain the triple verification result.

[0118] Based on the material condition judgment results, when the triple verification consistency check mechanism is performed, the triple verification results are obtained. The triple verification results include temperature shrinkage triple confirmation mark, mechanical wear confirmation mark and verification conflict warning mark.

[0119] The triple verification consistency determination refers to combining the temperature shrinkage gap anomaly identifier and the non-temperature effect identifier with the material state discrimination result for judgment. When a low-temperature shrinkage material state identifier is generated, the three are combined to generate a temperature shrinkage triple confirmation identifier; when a mechanical wear material state identifier is generated, the three are combined to generate a mechanical wear confirmation identifier; when a verification conflict warning is generated, the three are combined to generate a verification conflict warning identifier. In other words, the triple verification result includes the temperature shrinkage triple confirmation identifier, the mechanical wear confirmation identifier, and the verification conflict warning identifier. When a verification conflict warning is detected as the triple verification result, it is necessary to enter the adaptive learning mode for in-depth analysis. The verification conflict warning identifier indicates that the system has complex gap anomalies.

[0120] Based on the triple verification results, a confidence evaluation mechanism for micro-perturbation detection is established, and a confidence score is calculated. This confidence score reflects the reliability level of the triple verification consistency judgment results. The confidence score is calculated by multiplying the "base score (reflecting the consistency of the three)" and the "step weight (reflecting the historical reliability of each step)". The base score decreases from high to low as the consistency of the three: 0.9 → 0.5 → 0.7 → 0.2. For example, when all three points to the same state, a score of 0.9 is taken; when two points to the same state and the third is irrelevant or neutral, a score of 0.5-0.7 is taken; and when the three are irrelevant, a score of 0.2 is taken. The three refer to the material state discrimination result, the temperature shrinkage gap anomaly indicator, and the temperature effect gap anomaly indicator. The weights are set based on the historical accuracy of each step. For example, the material state discrimination result has a score of 0.4, the temperature shrinkage gap anomaly indicator has a score of 0.3, and the temperature effect gap anomaly indicator has a score of 0.3.

[0121] Step S30 addresses the problem in traditional braking system monitoring methods that cannot accurately distinguish between gap anomalies caused by temperature effects and mechanical wear, providing higher accuracy and reliability. These steps, through micro-perturbation excitation signals and multi-level verification mechanisms, enable the system to accurately differentiate gap anomalies caused by temperature changes (such as low-temperature contraction or high-temperature expansion) and mechanical wear, thereby effectively improving the safety and maintenance efficiency of the braking system.

[0122] Step S40: Based on the triple verification results, extract the historical data recording unit and identify abnormal data, adjust the abnormal data to obtain the adjusted threshold, and obtain the final gap adjustment control command based on the adjusted threshold.

[0123] Furthermore, the specific process of step S40 includes:

[0124] Step S41: Based on the triple verification results, extract historical data record units, implement a hierarchical storage architecture for historical data, and perform time series analysis to identify abnormal data.

[0125] The historical data recording unit combines relevant historical data generated during the triple verification process. Specifically, the historical data in the recording unit is stored hierarchically, analyzed to identify periodic patterns and anomalies, extracts high-confidence features and abnormal data from these patterns, adjusts thresholds for abnormal data, verifies the adjusted results, and corrects unsatisfactory results. Each historical data recording unit includes a timestamp, external ambient temperature, gap anomaly type, and verification result confidence level. The timestamp is determined based on the confidence evaluation mechanism for generating micro-perturbation detection; the external ambient temperature is based on the real-time external ambient temperature during data acquisition; the gap anomaly type is based on the triple verification results, classifying the results; and the verification result confidence level is obtained by classifying the calculated comprehensive confidence level.

[0126] The feature data storage structure is a hierarchical storage architecture for historical data within historical data record units. It includes a daily data layer, a seasonal data layer, and an annual data layer. The daily data layer stores detailed feature data for the most recent 30 days, the seasonal data layer stores statistical feature data for the most recent four seasons, and the annual data layer stores trend feature data for each year.

[0127] Based on a hierarchical storage architecture with a feature-driven data storage structure, time-series identification is performed to identify periodic patterns and anomalies in historical data. Daily temperature changes are extracted from the daily data layer, including 24-hour temperature trends and the temporal distribution characteristics of intermittent adjustment triggers. Seasonal variations are extracted from the seasonal data layer, including temperature contraction patterns across spring, summer, autumn, and winter and corresponding system response characteristics. Annual seasonal variation patterns are extracted from the annual data.

[0128] Analyzing periodic patterns, we extract high-confidence judgment features from successful judgments in historical data to guide subsequent parameter optimization and threshold adjustment. Analyzing abnormal patterns, we extract anomalous data from records with low verification confidence in historical data. Anomalous data includes environmental conditions, time characteristics, and torque characteristics that occurred in the records.

[0129] Step S42: The threshold of abnormal data is adjusted by an adaptive threshold adjustment algorithm. The threshold adjustment is prioritized and adjusted by a priority sorting mechanism to obtain the adjusted threshold.

[0130] The adaptive threshold adjustment algorithm adjusts thresholds in abnormal data. It prioritizes threshold adjustments through a priority ranking mechanism, verifies the effects of adjustments, and progressively corrects thresholds that do not perform well after adjustment.

[0131] Specifically, a priority ranking mechanism for threshold adjustment is established. This mechanism is based on the frequency and impact of low confidence levels in verification results appearing in abnormal data. Thresholds requiring adjustment are determined and prioritized. After obtaining the priority ranking, the thresholds are adjusted to obtain the final adjusted thresholds. The adjustment direction is determined based on the deviation trend between the temperature response characteristic parameters and the thresholds in the abnormal data. The adjustment range is controlled by the deviation rate and safety limits. The frequency is determined by the number of times the threshold has caused verification conflict warnings in historical data. The impact level is based on the degree of harm to braking safety, adjustment accuracy, and system stability caused by the threshold deviation. The adjusted thresholds include the preset continuity threshold for the adjusted time window unit, the similarity threshold for successful waveform fingerprint matching, the set of response characteristic parameter thresholds for the adjusted low-temperature contraction state, and the set of characteristic parameter thresholds for the adjusted mechanical wear state. The deviation rate refers to the degree of difference between the temperature response characteristic parameters and the set thresholds, calculated according to the formula... get.

[0132] The threshold adjustment effectiveness verification mechanism verifies and corrects the adjusted threshold, ensuring that the magnitude of each threshold adjustment is within a safe range and avoiding system instability caused by over-correction. The effectiveness of the adjustment is verified by comparing the system's judgment accuracy before and after the threshold adjustment, and a single adjustment magnitude limit is set to ensure that the single adjustment amount of any threshold does not exceed a preset percentage of its current value.

[0133] Step S43: Establish a closed-loop optimization control mechanism to verify the adjusted threshold and obtain the final gap adjustment control command.

[0134] The closed-loop optimization control mechanism integrates and verifies the adjusted threshold obtained from the adaptive threshold adjustment algorithm, forming a complete adaptive learning feedback loop. Specifically, the closed-loop optimization control mechanism periodically compares the system's judgment accuracy before and after threshold adjustment with the output results of the adaptive threshold adjustment algorithm. This verifies the effectiveness of the adjustment and generates a performance evaluation report. Automatic adjustments are made when the performance evaluation results decline or the learning effect is poor. The performance evaluation report includes accuracy, response time, and system stability metrics. The accuracy metric compares the system's judgment results with the actual situation using the formula: The accuracy rate is calculated to ensure the judgment is correct. The response time metric evaluates the time delay from receiving an abnormal signal to generating a gap adjustment command. The system stability metric evaluates the system's stability over long-term operation, ensuring that fluctuations in judgment accuracy do not exceed a predetermined range. System stability is assessed by calculating the standard deviation of the confidence level over a period of time using the adaptive threshold adjustment algorithm.

[0135] Specifically, the system performance evaluation index system involves periodically conducting a comprehensive assessment of the adaptive learning effect of the adaptive threshold adjustment algorithm, for example, once a month. This is achieved through evaluations of accuracy, response time, and system stability. The accuracy index is based on the statistical ratio of the output results from the adaptive threshold adjustment algorithm to the actual fault detection results, measured by the proportion of correct judgments to the total number of judgments. For example, the accuracy rate for mechanical wear judgment is 76%, and the accuracy rate for low-temperature shrinkage judgment is 90%. The response time index is based on the time taken from the detection of an abnormal signal to the output of the adjustment command; for example, 2.5 seconds meets the preset threshold of ≤3 seconds. The system stability index is based on the degree of result fluctuation; for example, the output results of the adaptive threshold adjustment algorithm are periodically tested for 30 days, and the calculated confidence standard deviation for 30 consecutive days is 0.08 (≤0.1). After threshold adjustment, the accuracy fluctuation is 2% (≤5%). A performance evaluation report is then generated, which includes the accuracy, response time, and system stability indices.

[0136] An adaptive adjustment mechanism is established to automatically adjust learning strategy parameters when performance evaluation results show a decline in system performance or poor learning effectiveness. For example, during a regular monthly testing period, if the accuracy falls below a preset threshold (e.g., below 85%), the response time exceeds the specified range (e.g., exceeding 3 seconds), system stability decreases (e.g., standard deviation exceeds 0.1), the number of threshold adjustments is too high (e.g., more than 10 times in a month), the frequency of validation conflict warnings increases (e.g., more than 3 times in a month), or the deviation between prediction and actual results increases (e.g., error exceeds 10%), the system may need to enter adaptive learning mode for optimization if any of these indicators exceed the preset thresholds.

[0137] By combining the historical data recording unit, adaptive threshold adjustment algorithm, and effect verification mechanism into a complete process, and integrating all the analysis results in steps S10-S43, a final gap adjustment control command is generated. This command clearly indicates whether to execute the gap adjustment operation and the specific adjustment parameters. For example, if the external environment is 4℃, the gap anomaly type is measured as low-temperature shrinkage, the material state judgment result is a low-temperature shrinkage material state identifier, and the verification result confidence level is 0.92 (high confidence); the system state is normal; the current gap state is abnormally increased (warning), and the final gap adjustment control command is executed. The result indicates that the operation is executed; the adjustment type is determined to be temperature compensation correction; the adjustment parameters include an adjustment range of +0.15mm, a target gap value of 1.25mm (the baseline gap value is 1.1mm), and a temperature compensation coefficient of 0.2Nm / °C (based on the current ambient temperature and historical data). The verification results show successful temperature consistency verification, successful waveform matching verification, and a triple verification result of a triple confirmation of temperature shrinkage. In the execution conditions, the safety status check indicates the vehicle is stationary, the parking brake is enabled, and the system safety mode is normal.

[0138] Step S40 addresses the technical challenge of enabling the system to learn and continuously optimize the adjustment strategy under different environmental conditions. Through hierarchical storage of historical data, adaptive threshold adjustment, and closed-loop optimization control, the system can cope with complex scenarios, achieving dynamic optimization of the threshold and adaptability of the control strategy. This enhances the system's robustness under complex operating conditions and ensures the accuracy and reliability of the final gap adjustment command.

[0139] Example 2

[0140] This embodiment, based on embodiment 1, provides an electronic control system for an automotive clearance adjuster with torque sensing, as shown in Figure 5, including:

[0141] Abnormal Temperature Identification Module: This module is used to establish a time window benchmark database, detect the time window benchmark database to obtain temperature contraction candidate identifiers, perform differential judgment on the temperature contraction candidate identifiers to obtain temperature contraction confirmation identifiers, and judge the temperature contraction confirmation identifiers through a multi-level decision tree to obtain temperature contraction gap abnormality identifiers.

[0142] Temperature effect identification module: used to establish a standard fingerprint database, collect real-time braking release torque waveform and perform feature analysis to obtain real-time waveform feature vector, combine the real-time waveform feature vector with the standard fingerprint database to perform similarity calculation, obtain temperature consistency confirmation mark and waveform matching success mark, and combine with temperature contraction gap abnormal mark to obtain non-temperature effect mark;

[0143] Anomaly determination module: Based on non-temperature effect identifiers, it establishes a micro-perturbation excitation signal program, extracts temperature response characteristic parameters from the micro-perturbation excitation signal program, performs material state discrimination on the temperature response characteristic parameters, obtains the material state discrimination result, performs triple verification consistency judgment, and obtains the triple verification result;

[0144] Self-learning module: Based on the triple verification results, it extracts historical data recording units and identifies abnormal data, adjusts the abnormal data to obtain the adjusted threshold, and obtains the final gap adjustment control command based on the adjusted threshold.

[0145] In the abnormal temperature identification module, the process of establishing a time window benchmark database, detecting the time window benchmark database to obtain temperature contraction candidate identifiers, performing differential judgment on the temperature contraction candidate identifiers to obtain temperature contraction confirmation identifiers, and judging the temperature contraction confirmation identifiers through a multi-level decision tree to obtain temperature contraction gap abnormality identifiers, including:

[0146] Step S11: Establish a time window reference database, detect the time window reference database, determine whether there is a gap abnormal increase signal, when there is a gap abnormal increase signal, perform backtracking analysis on the time window of the gap abnormal increase signal, and obtain temperature shrinkage candidate identifiers after judging the continuity of torque change rate.

[0147] Step S12: After obtaining the candidate identifier for temperature shrinkage, extract the torque change rate feature vector used for differentiation judgment from the candidate identifier for temperature shrinkage, and obtain the confirmation identifier for temperature shrinkage through the differentiation judgment formula.

[0148] Step S13: Establish a multi-level decision tree to judge the temperature contraction confirmation flag and obtain the temperature contraction gap abnormality flag.

[0149] In the temperature effect identification module, the process involves establishing a standard fingerprint database, collecting real-time braking release torque waveforms and performing feature analysis to obtain real-time waveform feature vectors, combining these feature vectors with the standard fingerprint database for similarity calculation to obtain temperature consistency confirmation and waveform matching success identifiers, and combining these with temperature contraction gap anomaly identifiers to obtain non-temperature effect identifiers, including:

[0150] Step S21: Establish a standard fingerprint database. When the brake release operation ends signal is detected, collect the real-time brake release torque waveform and record the real-time external ambient temperature. Perform feature analysis on the real-time brake release torque waveform to obtain the real-time waveform feature vector.

[0151] Step S22: Calculate the similarity between the real-time waveform feature vector and the reference feature vector in the standard fingerprint database to obtain the waveform matching feature result;

[0152] Step S23: Perform temperature effect judgment on the waveform matching feature results to obtain a waveform matching success identifier. Verify the consistency between the reference ambient temperature corresponding to the highest similarity value in the standard fingerprint database and the real-time ambient temperature to obtain a temperature consistency confirmation identifier. Based on the temperature contraction gap abnormality identifier, combine the waveform matching success identifier and the temperature consistency confirmation identifier to generate a non-temperature effect identifier.

[0153] In the anomaly determination module, based on non-temperature effect identifiers, a micro-perturbation excitation signal program is established. Temperature response characteristic parameters are extracted from the micro-perturbation excitation signal program, and material state discrimination is performed on the temperature response characteristic parameters to obtain the material state discrimination result. A triple verification consistency determination is then performed to obtain the triple verification result, including:

[0154] Step S31: Based on the non-temperature effect identifier, establish a micro-perturbation excitation signal program, extract temperature response characteristic parameters from the micro-perturbation excitation signal program, perform temperature correlation analysis on the response characteristic parameters, and obtain the temperature response characteristic parameters.

[0155] Step S32: Perform material state discrimination on the temperature response characteristic parameters to obtain the material state discrimination result;

[0156] Step S33: After obtaining the material state discrimination result, perform triple verification consistency judgment to obtain the triple verification result.

[0157] In the self-learning module: based on the triple verification results, historical data recording units are extracted and abnormal data is identified. The abnormal data is adjusted to obtain an adjusted threshold. Based on the adjusted threshold, the final gap adjustment control command is obtained, including:

[0158] Step S41: Based on the triple verification results, extract historical data record units, implement a hierarchical storage architecture for historical data, and perform time series analysis to identify abnormal data.

[0159] Step S42: The threshold of abnormal data is adjusted by an adaptive threshold adjustment algorithm, and the threshold adjustment is prioritized and adjusted by a priority sorting mechanism to obtain the adjusted threshold.

[0160] Step S43: Establish a closed-loop optimization control mechanism to verify the adjusted threshold and obtain the final gap adjustment control command.

Claims

1. An electronic control method for an automotive clearance adjuster with torque sensing, characterized in that, The method includes: A time window benchmark database is established, and the time window benchmark database is tested to obtain temperature contraction candidate identifiers. The temperature contraction candidate identifiers are differentiated to obtain temperature contraction confirmation identifiers. The temperature contraction confirmation identifiers are judged through a multi-level decision tree to obtain temperature contraction gap anomaly identifiers. A standard fingerprint database is established, real-time braking release torque waveforms are collected and feature analysis is performed to obtain real-time waveform feature vectors. The real-time waveform feature vectors are combined with the standard fingerprint database to perform similarity calculations to obtain temperature consistency confirmation identifiers and waveform matching success identifiers. Combined with temperature contraction gap anomaly identifiers, non-temperature effect identifiers are obtained. Based on non-temperature effect identification, a micro-perturbation excitation signal program is established. Temperature response characteristic parameters are extracted from the micro-perturbation excitation signal program. Material state discrimination is performed on the temperature response characteristic parameters to obtain the material state discrimination results. Triple verification consistency judgment is then performed to obtain the triple verification results. Based on the triple verification results, historical data recording units are extracted and abnormal data is identified. The abnormal data is adjusted to obtain the adjusted threshold. Based on the adjusted threshold, the final gap adjustment control command is obtained.

2. The electronic control method for an automotive clearance adjuster with torque sensing according to claim 1, characterized in that, The method for obtaining the temperature shrinkage confirmation indicator includes: The torque change rate feature vector is extracted from the temperature shrinkage candidate identifiers. The feature vector includes the average change rate, the standard deviation of the change rate, and the slope of the change trend. The average rate of change, standard deviation of the rate of change, and slope of the trend are calculated using a differential judgment formula to obtain a temperature shrinkage confirmation indicator.

3. The electronic control method for an automotive clearance adjuster with torque sensing according to claim 2, characterized in that, The method for obtaining the abnormal identification of temperature contraction gap includes: Based on the temperature contraction confirmation flag, the current external ambient temperature and current time are detected through a multi-level decision tree; If the current external ambient temperature is lower than the preset low temperature threshold, a low temperature environment identifier is generated; if the current time is during a period of high temperature contraction, a low temperature environment identifier is generated. When both the low-temperature environment indicator and the low-temperature environment indicator are generated simultaneously, an abnormal temperature contraction gap indicator is obtained.

4. The electronic control method for an automotive clearance adjuster with torque sensing according to claim 3, characterized in that, The method for establishing a standard fingerprint database includes: The braking release torque waveform is obtained by measuring the braking release torque waveform under different external ambient temperatures. During the torque release phase after each braking operation, the torque sensor continuously samples the waveform to obtain a reference braking release torque waveform. Feature analysis was performed on the reference braking release torque waveform to extract key feature parameters of the torque decay waveform. Key feature parameters include the reference torque return-to-zero time, the number of oscillations of the reference waveform, and the reference decay time constant. The reference torque zero-return time, the number of reference waveform oscillations, and the reference decay time constant are combined to obtain the reference waveform feature vector; The reference waveform feature vectors obtained under different temperature conditions are combined to obtain a standard fingerprint library.

5. The electronic control method for an automotive clearance adjuster with torque sensing according to claim 4, characterized in that, The method for obtaining the temperature consistency confirmation identifier includes: The reference torque zero-return time, the number of reference waveform oscillations, and the reference decay time constant are calculated to obtain the highest similarity value. When the highest similarity value is greater than the preset similarity threshold, a waveform matching success identifier is generated. The similarity between the real-time waveform feature vector and each reference feature vector in the standard fingerprint database is calculated to obtain the waveform matching feature result; The temperature deviation is calculated based on the waveform matching characteristics. When the temperature deviation meets the preset temperature consistency conditions, a temperature consistency confirmation mark is obtained.

6. The electronic control method for an automotive clearance adjuster with torque sensing according to claim 5, characterized in that, The micro-perturbation excitation signal program includes: The micro-perturbation excitation signal procedure includes an excitation safety condition check mechanism; Incentive safety condition check mechanism, including vehicle stationary status check, parking brake activation status check and system safety mode check; The vehicle stationary status check determines whether the vehicle is completely stationary; the parking brake activation status check confirms whether the parking brake system is activated; and the system safety mode check confirms that the vehicle safety system is in normal working condition. When the vehicle is completely stationary, the parking brake system is engaged, and the vehicle safety system is functioning normally, a micro-disturbance excitation control signal is sent.

7. The electronic control method for an automotive clearance adjuster with torque sensing according to claim 6, characterized in that, The method for obtaining the temperature response characteristic parameters includes: The response characteristic parameters are obtained by high-frequency sampling of the disturbance excitation control signal using a torque sensor. The response characteristic parameters are corrected to obtain the temperature response characteristic parameters.

8. The electronic control method for an automotive clearance adjuster with torque sensing according to claim 7, characterized in that, The method for obtaining the material state discrimination result includes: Obtain the threshold set of response characteristic parameters for low-temperature contraction and the threshold set of response characteristic parameters for mechanical wear. The material state discrimination result is obtained by comparing the temperature response characteristic parameters with the threshold set of response characteristic parameters of the low temperature shrinkage state and the threshold set of response characteristic parameters of the mechanical wear state.

9. The electronic control method for an automotive clearance adjuster with torque sensing according to claim 8, characterized in that, The method for identifying abnormal data includes: Based on the triple verification results, historical data record units are extracted; Historical data recording units are stored in layers and time series identification is performed to obtain abnormal patterns.

10. An electronic control system for an automotive clearance adjuster with torque sensing, used to implement the electronic control method for an automotive clearance adjuster with torque sensing as described in any one of claims 1-9, characterized in that, The system includes: Abnormal Temperature Identification Module: This module is used to establish a time window benchmark database, detect the time window benchmark database to obtain temperature contraction candidate identifiers, perform differential judgment on the temperature contraction candidate identifiers to obtain temperature contraction confirmation identifiers, and judge the temperature contraction confirmation identifiers through a multi-level decision tree to obtain temperature contraction gap abnormality identifiers. Temperature effect identification module: used to establish a standard fingerprint database, collect real-time braking release torque waveform and perform feature analysis to obtain real-time waveform feature vector, combine the real-time waveform feature vector with the standard fingerprint database to perform similarity calculation, obtain temperature consistency confirmation mark and waveform matching success mark, and combine with temperature contraction gap abnormal mark to obtain non-temperature effect mark; Anomaly determination module: Based on non-temperature effect identifiers, it establishes a micro-perturbation excitation signal program, extracts temperature response characteristic parameters from the micro-perturbation excitation signal program, performs material state discrimination on the temperature response characteristic parameters, obtains the material state discrimination result, performs triple verification consistency judgment, and obtains the triple verification result; Self-learning module: Based on the triple verification results, it extracts historical data recording units and identifies abnormal data, adjusts the abnormal data to obtain the adjusted threshold, and obtains the final gap adjustment control command based on the adjusted threshold.

Images