Intelligent EMB system health state monitoring method based on integrated sensing and control

Abstract

This invention relates to the field of EMB system detection data processing technology, specifically to a method for monitoring the health status of an intelligent EMB system based on integrated sensing and control. The method includes: collecting detection data generated by the operation of the electromechanical braking system during vehicle braking; the detection data includes motor monitoring data; calculating the single-cycle comprehensive loss of the electromechanical braking system during each braking process based on the motor monitoring data; obtaining the window anomaly index of the corresponding sliding window based on the single-cycle comprehensive loss, and constructing a data feature space in conjunction with the single-cycle comprehensive loss; processing the data feature space using the isolated forest algorithm to obtain isolated factors for each braking process; and classifying the health status of the obtained isolated factors according to a pre-set health status level threshold. This invention can significantly improve the detection sensitivity and robustness for early wear and minor faults.

Description

Technical Field

[0001] This invention relates to the field of EMB system detection data processing technology, specifically to a method for monitoring the health status of an intelligent EMB system based on integrated sensing and control. Background Technology

[0002] The EMB system, or Electromechanical Braking System, uses electric actuators to replace traditional hydraulic pumps and brake fluid to decelerate or stop a vehicle. Because the EMB system eliminates hydraulic components, it reduces maintenance costs, simplifies the system structure, and reduces the overall weight of the vehicle. At the same time, the use of electronic control units enables precise control of the braking force of each tire.

[0003] However, the braking performance of an EMB system relies on friction, which inevitably leads to gradual wear of braking components over time, resulting in a slow decline in braking performance. Therefore, this wear is a gradual, minute change, rather than a sudden failure. Consequently, the changes in its data characteristics are extremely subtle during short-term operation. This poses a challenge to anomaly detection algorithms that rely on finding obvious outliers (such as Isolation Forest), because the wear data does not initially exhibit significant isolation but rather blends into the normal data distribution, making it difficult for anomaly detection algorithms to identify this early, slowly developing anomaly in a timely manner. Summary of the Invention

[0004] To address the technical problem of existing technologies failing to detect early-stage faults in a timely and accurate manner, the present invention aims to provide a health status monitoring method for an intelligent EMB system based on integrated sensing and control. The specific technical solution adopted is as follows:

[0005] A method for monitoring the health status of an intelligent EMB system based on integrated sensing and control, the method comprising:

[0006] Collect detection data generated by the operation of the electromechanical braking system during vehicle braking; the detection data includes motor monitoring data.

[0007] Calculate the single comprehensive loss of the electromechanical braking system during each braking process based on motor monitoring data;

[0008] The window anomaly index of the corresponding sliding window is obtained based on the single comprehensive loss, and the data feature space is constructed by combining the single comprehensive loss.

[0009] The isolated forest algorithm is used to process the data feature space to obtain the isolation factor for each braking.

[0010] The isolated factors are classified according to their health status based on a pre-set health status level threshold.

[0011] Preferably, after collecting the detection data generated by the operation of the electromechanical braking system during vehicle braking, and before calculating the single comprehensive loss of the electromechanical braking system at each braking node based on the motor monitoring data, the method further includes:

[0012] The collected detection data is denoised using a Gaussian filtering algorithm.

[0013] Preferably, the single-cycle comprehensive loss of the electromechanical braking system at each braking node is calculated based on motor monitoring data, including:

[0014] Based on motor monitoring data, the first motor position parameters at the moment of command reversal during motor drive and the second motor position parameters at the moment of clamping force decrease are obtained; the clamping force represents the pressure between the motor-driven brake pads and the vehicle brake disc;

[0015] The motor displacement offset is calculated based on the position parameters of the first motor and the position parameters of the second motor.

[0016] The effective stiffness of the electromechanical braking system during the force loading phase of each braking process was tested.

[0017] Braking wear parameters are obtained by comparing the effective stiffness of the current braking with the effective stiffness of the previous braking in the electromechanical braking system.

[0018] The single comprehensive loss of the electromechanical braking system during each braking process is calculated by combining the motor displacement offset and braking wear parameters.

[0019] Furthermore, the motor displacement offset is calculated based on the position parameters of the first motor and the position parameters of the second motor, including:

[0020] Based on the position parameters of the first motor and the position parameters of the second motor, the absolute value of the motor position difference at the moment of command reversal and the moment of clamping force decrease during the motor driving process is calculated.

[0021] Obtain standard backlash quantification parameters for brake pad state transitions within an electromechanical braking system;

[0022] Based on the absolute value of the motor position difference and the standard backlash quantization parameters, the motor displacement offset at the moment of command reversal and the moment of clamping force reduction during each braking period is calculated.

[0023] Furthermore, the effective stiffness of the electromechanical braking system during the force loading phase of each braking process is detected, including:

[0024] Based on motor monitoring data, the clamping force at the beginning and end of the force loading phase, as well as the motor position at the beginning and end of the force loading phase, are obtained.

[0025] The absolute value of the difference between the clamping forces at the beginning and end of the force loading phase, and the motor offset;

[0026] The effective stiffness of the electromechanical braking system during each braking process is obtained by comparing the absolute value of the difference in clamping force with the motor offset.

[0027] Furthermore, after obtaining the clamping forces at the start and end of the force loading phase, and the motor positions at the start and end of the force loading phase, and before calculating the absolute value of the difference between the clamping forces at the start and end of the force loading phase, and the motor offset, the method further includes:

[0028] Based on motor monitoring data, obtain the clamping force detected each time, the temperature-friction relationship coefficient during each braking process, and the actual required friction coefficient;

[0029] The ratio of the actual required friction coefficient to the temperature friction coefficient is used as the compensation coefficient to compensate for each detected clamping force, thus obtaining the compensated clamping force; the clamping force detected each time includes the clamping force at the beginning and end of the force loading stage.

[0030] Furthermore, braking wear parameters are obtained by comparing the current effective stiffness of the electromechanical braking system with the effective stiffness of the previous braking, including:

[0031] Obtain the effective stiffness of the current braking and the effective stiffness of the previous braking in the electromechanical braking system;

[0032] Calculate the difference between the current braking system stiffness and the effective stiffness of the previous braking, and calculate the ratio of the difference to the effective stiffness of the previous braking.

[0033] The braking time is obtained, and the braking wear parameters are calculated by combining the difference with the ratio of the effective stiffness of the previous braking.

[0034] Preferably, the window anomaly index of the corresponding sliding window is obtained based on the single comprehensive loss, and a data feature space is constructed in conjunction with the single comprehensive loss, including:

[0035] Set a sliding window of a certain length, and calculate the cumulative loss within each sliding window based on the sum of the single comprehensive losses;

[0036] Based on the cumulative loss of a certain sliding window and the cumulative loss of the previous sliding window, the window anomaly index is obtained to indicate that a certain sliding window has an abnormal state.

[0037] A data feature space is constructed by combining the window anomaly index of each sliding window with the single comprehensive loss of each braking process.

[0038] Furthermore, the sliding window is stopped a certain number of times to determine its length.

[0039] Preferably, the isolated factors are classified according to a pre-set health status level threshold, including:

[0040] Set a dynamic first health status level threshold and a second health status level threshold, wherein the second health status level threshold is greater than the first health status level threshold.

[0041] If the obtained isolation factor is less than the threshold of the first health status level, then the electromechanical braking system is confirmed to be in a healthy state at this time.

[0042] If the obtained isolation factor is greater than or equal to the first health status threshold and less than the second health status threshold, then the electromechanical braking system is confirmed to be mildly degraded at this time.

[0043] If the isolation factor is greater than or equal to the threshold of the second health status level, then the electromechanical braking system is confirmed to be severely degraded.

[0044] The present invention has the following beneficial effects:

[0045] By analyzing the monitoring data of the intelligent EMB system (i.e., the electromechanical braking system) during vehicle operation, a single comprehensive loss characterizing the progressive wear of the electromechanical braking system during each braking process is constructed based on the motor monitoring data. Furthermore, the window anomaly index of the corresponding sliding window is obtained by using sliding window analysis, which transforms the slow performance drift originally masked by friction nonlinearity into a recognizable pattern in the feature space. Combined with the single comprehensive loss, a data feature space is constructed, thereby significantly improving the sensitivity and robustness of the isolated forest algorithm in detecting early wear and minor faults. Attached Figure Description

[0046] To more clearly illustrate the technical solutions and advantages 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.

[0047] Figure 1 This is a flowchart illustrating a health status monitoring method for an intelligent EMB system based on integrated sensing and control, as provided in one embodiment of the present invention. Detailed Implementation

[0048] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, features, and effects of a health status monitoring method for an intelligent EMB system based on integrated sensing and control proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features or characteristics in one or more embodiments can be combined in any suitable form.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0050] The following description, in conjunction with the accompanying drawings, details a specific scheme for a health status monitoring method for an intelligent EMB system based on integrated sensing and control provided by the present invention.

[0051] This invention provides a method for monitoring the health status of an intelligent EMB system based on integrated sensing and control. By using monitoring sensors installed in the vehicle's monitoring area to detect various data during the operation of the EMB system (i.e., the electromechanical braking system), the health status of the EMB system is monitored. During braking, the EMB system transmits the driver's commands to the braking control components at the vehicle wheels. These components then implement the EMB system commands with precise responses, achieving braking operations and ensuring a stable stop for the vehicle.

[0052] Please see Figure 1 This document illustrates a flowchart of a method for monitoring the health status of an intelligent EMB system based on integrated sensing and control, according to an embodiment of the present invention. The method includes:

[0053] Collect detection data generated by the operation of the electromechanical braking system during vehicle braking; the detection data includes motor monitoring data.

[0054] Calculate the single comprehensive loss of the electromechanical braking system during each braking process based on motor monitoring data;

[0055] The window anomaly index of the corresponding sliding window is obtained based on the single comprehensive loss, and the data feature space is constructed by combining the single comprehensive loss.

[0056] The isolated forest algorithm is used to process the data feature space to obtain the isolation factor for each braking.

[0057] The isolated factors are classified according to their health status based on a pre-set health status level threshold.

[0058] This invention analyzes monitoring data from the intelligent EMB system (i.e., electromechanical braking system) during vehicle braking operations. Based on the motor monitoring data, it constructs a single comprehensive loss characterizing the progressive wear of the electromechanical braking system during each braking process. Furthermore, it utilizes sliding window analysis to obtain the window anomaly index of the corresponding sliding window, transforming the slow performance drift originally masked by frictional nonlinearity into a recognizable pattern in the feature space. Combined with the single comprehensive loss, a data feature space is constructed, thereby significantly improving the sensitivity and robustness of the isolated forest algorithm in detecting early wear and minor faults.

[0059] In one specific embodiment, detection data generated by the operation of the electromechanical braking system during vehicle braking is collected. Specifically, various detection data generated during vehicle braking are collected and recorded based on the sensors within the electromechanical braking system, and the collected detection data is processed in an orderly manner according to the acquisition sequence. The detection data includes motor monitoring data.

[0060] In one specific embodiment, after collecting the detection data generated by the operation of the electromechanical braking system during vehicle braking, and before calculating the single comprehensive loss of the electromechanical braking system at each braking node based on the motor monitoring data, the method further includes:

[0061] The collected detection data is denoised using a Gaussian filtering algorithm.

[0062] This embodiment uses a Gaussian filtering algorithm to reduce noise in the collected detection data, preventing noise from interfering with subsequent data analysis and affecting the accuracy of the detection results. If the collected detection data is free of noise, Gaussian filtering is not necessary, and subsequent data analysis can proceed directly.

[0063] In one specific embodiment, because the wear of the electromechanical braking system is gradual and minute, it manifests in the data as a slow "drift" rather than a clear, sudden "outlier" (i.e., an isolated node). Therefore, directly using the Isolation Forest algorithm to find "outliers" might mistakenly identify this slow drift as part of the normal data cloud, thus failing to trigger an alarm. Therefore, it is necessary to conduct an in-depth physical analysis of the impact mechanism of friction on the dynamic behavior of the EMB system (i.e., the electromechanical braking system) and sensor data during actual operation of the braking device, and based on this, propose targeted signal processing and feature engineering methods to optimize the Isolation Forest algorithm and improve its sensitivity and robustness in detecting early faults.

[0064] In an EMB system, backlash is the mechanical clearance present in the transmission chain (such as gears). When the motor in the EMB system changes from pushing the brake pads to pulling them back—that is, at the instant of transition from braking to release—the motor needs to idle for a short angle or displacement to eliminate the transmission backlash and truly reverse the force. This "idle" displacement is the direct manifestation of backlash. As the EMB system wears down, the gear teeth thin, and the backlash gradually increases. Therefore, by comparing the idle displacement reflected in the monitoring data of the motor at the instant of reversal, quantitative parameters of backlash during brake pad state transitions within the EMB system can be obtained.

[0065] In this embodiment, the calculation of the single comprehensive loss of the electromechanical braking system at each braking node based on motor monitoring data includes:

[0066] Based on motor monitoring data, obtain the command reversal during motor drive ( )time First motor position parameters and decrease in clamping force ( )time Second motor position parameters The clamping force refers to the pressure between the motor-driven brake pads and the vehicle brake disc.

[0067] The motor displacement offset is calculated based on the position parameters of the first motor and the position parameters of the second motor. ;

[0068] The effective stiffness of the electromechanical braking system during the force loading phase of each braking process was tested. ;

[0069] Braking wear parameters are obtained by comparing the current effective stiffness of the electromechanical braking system with the effective stiffness of the previous braking. ;

[0070] The comprehensive loss of the electromechanical braking system during each braking process is calculated by combining the motor displacement offset and braking wear parameters. .

[0071] In this embodiment, the motor displacement offset is calculated based on the first motor position parameters and the second motor position parameters, including:

[0072] According to the position parameters of the first motor Second motor position parameters The absolute value of the motor position difference at the moment of command reversal and the moment of clamping force reduction during the motor drive process is calculated.

[0073] Obtaining standard backlash quantization parameters for brake pad state transitions within an electromechanical braking system ;

[0074] Based on the absolute value of the motor position difference and the standard backlash quantization parameters, the motor displacement offset at the command reversal moment and the clamping force reduction moment during each braking period is calculated. .

[0075] In this embodiment of the invention, Backlash refers to the gap between two or more gears in the meshing state. It can be specifically expressed as 20 arc minutes. The higher the equipment precision requirement, the smaller the arc minute requirement.

[0076] It should be noted that, for ease of calculation, all indicator data involved in the calculation in this embodiment of the invention have undergone data preprocessing to eliminate the influence of dimensions. The specific methods for eliminating the influence of dimensions are well known to those skilled in the art and are not limited here.

[0077] Specifically, this embodiment provides the following formula for calculating the motor position difference coefficient:

[0078]

[0079] In the formula, This represents the motor position difference coefficient, which is a dimensionless data point. This is the absolute value of the motor position difference during the state transition when the motor reverse command is in effect. The motor position difference value represents the motor's offset angle. The larger the absolute value, the greater the motor displacement offset. This is represented as a preset standard backlash quantization parameter. It should be understood that this standard backlash quantization parameter is a positive number. This represents the ratio of the current backlash quantization parameter to the standard backlash quantization parameter. By scaling the backlash quantization parameter for each braking operation, the segmentation of feature data is improved.

[0080] In this embodiment, during vehicle braking, each manipulated mechanical component operates according to its respective function, increasing the pressure between the brake pads and the vehicle's brake disc, thereby increasing the friction between the brake pads and the brake disc, thus achieving the braking process. However, during vehicle braking, the wear of the brake pads will become increasingly severe with the increase of usage time. But because the wear of the brake pads is gradual and minor, it will not have a significant impact on the vehicle's braking effect in the short term, and the detection system will mistakenly identify it as normal data. However, brake wear will cause the motor to require a longer motor stroke to achieve the same effect when the brake pads contact the brake disc, and will also cause changes in the transmission stiffness within the electromechanical braking system.

[0081] The effective stiffness of the electromechanical braking system, i.e., the ratio of the change in clamping force to the change in motor displacement, can serve as a proxy indicator of wear. As the brake pads wear, their contact surface changes, and the stiffness of the entire "motor-drive chain-brake pad-brake disc" system will change accordingly, typically showing a decrease. Therefore, the effective stiffness K of the electromechanical braking system can be detected based on the force loading stage of each braking operation.

[0082] In this embodiment, detecting the effective stiffness of the electromechanical braking system during the force loading phase of each braking process includes:

[0083] Based on motor monitoring data, the clamping force at the beginning and end of the force loading phase, as well as the motor position at the beginning and end of the force loading phase, are obtained.

[0084] The absolute value of the difference between the clamping forces at the beginning and end of the force loading phase, and the motor offset;

[0085] The effective stiffness of the electromechanical braking system during each braking process is obtained by comparing the absolute value of the difference in clamping force with the motor offset.

[0086] This embodiment provides the following formula for calculating the effective stiffness of an electromechanical braking system:

[0087]

[0088] In the formula, , These represent the start of the force loading stage ( ), and end ( The clamping force at that time; This indicates the offset of the motor position at the beginning and end of the force loading phase; , These represent the absolute value of the difference in clamping force during the beginning and end of the force loading phase, and the motor offset, respectively. This indicates normalization processing. Adding 0.1 to the denominator is a safety value set to prevent the denominator from being 0. The specific setting of this value depends on the actual scenario, and the resulting calculation error is within the allowable range of the scenario.

[0089] This embodiment obtains the effective stiffness during the current braking process by comparing the change in clamping force with the motor offset, and uses the norm function for normalization to avoid the influence of dimensions.

[0090] In this embodiment, due to the friction pair formed by the brake pads tightening and the brake disc, the friction temperature generated during the friction process gradually increases. As the temperature rises, the hardness of the protrusions on the metal contact surface between the brake pads and the brake disc (which are generally metal products) decreases due to the thermal softening effect. When the two surfaces slide relative to each other, the softened protrusions are more easily sheared, resulting in a decrease in the coefficient of friction.

[0091] Therefore, to ensure braking effectiveness, the EMB system needs to compensate for the clamping force at the current temperature; specifically, this is achieved by comparing the friction coefficient at the current temperature. The coefficient of friction required in practice The compensation coefficient is obtained, and the clamping force after compensation is analyzed in conjunction with the actual clamping force detected. .

[0092] Therefore, in this embodiment, after obtaining the clamping force at the beginning and end of the force loading phase, and the motor position at the beginning and end of the force loading phase, and before calculating the absolute value of the difference between the clamping forces at the beginning and end of the force loading phase, and the motor offset, the method further includes:

[0093] Based on motor monitoring data, obtain the clamping force detected each time and the temperature-friction coefficient during each braking process. and the actual required coefficient of friction ;

[0094] The ratio of the actual required friction coefficient to the temperature-friction coefficient is used as a compensation coefficient to compensate for the clamping force detected each time, thus obtaining the compensated clamping force. The clamping force detected each time includes the clamping force at the beginning and end of the force loading phase.

[0095] In this embodiment of the invention, a mathematical model of friction coefficient versus temperature is constructed by obtaining the friction coefficient of the braking device at different temperatures during operation in the laboratory. Then, the existing temperature is substituted into the model to obtain the temperature-friction relationship coefficient at the current temperature. The actual required friction coefficient is derived by using a dynamic model to obtain the traction force required under the current operating conditions. These are all existing technologies, and will not be further limited or elaborated upon. It should be noted that in real-world scenarios, there is no situation where the coefficient of friction is 0; therefore, this invention assumes that the coefficient of friction is greater than 0.

[0096] This embodiment specifically provides the calculated and supplemented clamping force. The calculation formula is as follows:

[0097]

[0098] In the formula, This indicates the clamping force being measured at the current moment. Indicates the actual required coefficient of friction Coefficient of friction with temperature The ratio of .

[0099] This embodiment introduces the friction coefficient into existing detection data to obtain the clamping force required during the current braking process. Based on the obtained compensated clamping force... This is then corrected by substituting it into the formula for calculating effective stiffness, thereby eliminating the effective stiffness after the temperature effect is achieved. This allows for more precise monitoring of the health status of the electromechanical braking system.

[0100] In this embodiment, braking wear parameters are obtained by comparing the effective stiffness of the current braking with the effective stiffness of the previous braking, including:

[0101] Obtain the effective stiffness of the current braking and the effective stiffness of the previous braking in the electromechanical braking system;

[0102] Calculate the difference between the current braking system stiffness and the effective stiffness of the previous braking, and calculate the ratio of the difference to the effective stiffness of the previous braking.

[0103] The braking time is obtained, and the braking wear parameters are calculated by combining the difference with the ratio of the effective stiffness of the previous braking. .

[0104] This embodiment provides a method for calculating brake wear parameters. The calculation formula is as follows:

[0105]

[0106] In the formula, For the first i Corrected effective stiffness during the second braking phase Indicates the first i The corrected effective stiffness during the second braking is the same as that during the third braking. i The difference between the corrected effective stiffness at the -1st braking event and the first braking event. i The ratio of the corrected effective stiffness at the -1st braking event is used to reflect the wear quantification parameter of the electromechanical braking system during use, and is also the first... i The degree of deviation of the corrected effective stiffness from the "healthy" state during the second braking; A positive value indicates a decrease in stiffness, which may mean that the brake pads are wearing out more severely; This indicates the duration of the current braking operation. This is used to analyze the degree of friction loss during the current braking process. The larger the size, the greater the degree of wear and tear; This represents the activation function, that is, using The activation function is normalized. Adding 0.1 to the denominator prevents calculation problems caused by a denominator of 0. The specific value is set according to the actual scenario, and the resulting calculation error is within the acceptable range of the scenario. It should be noted that the brake wear parameter in this embodiment is a dimensionless data point, used only to represent the brake wear condition based on its numerical value. In the actual calculation, all parameters are also dimensionless data, analyzed only based on their numerical values. This represents a normalization function, which can be specifically, for example, a maximum / minimum value normalization function. Indicates the first Brake wear parameters during the braking process.

[0107] In this embodiment, since backlash is an inherent anomaly of the electromechanical braking system, as wear increases, when the motor switches states in response to the braking process issued by the EMB system, the gap between the brake pads and the brake disc increases due to wear, thereby exacerbating the backlash condition within the system.

[0108] Therefore, in this embodiment, the single comprehensive loss of the electromechanical braking system during each braking process is calculated by combining the motor displacement offset and braking wear parameters. .

[0109] The specific formula for calculating the comprehensive loss per instance is as follows:

[0110]

[0111] In the formula, Indicates the first The overall loss during a single braking process; , These represent the weighting coefficients for backlash and wear, respectively. PCA was used on historical data. and The two features are linearly combined, with the coefficients of the first principal component used as weights. Specifically, historical back gap variations are first collected and standardized. and wear change The data is then processed by calculating the covariance matrix of the standardized data and performing eigenvalue decomposition. Next, the eigenvector corresponding to the largest eigenvalue is selected as the first principal component. Finally, the absolute value of the coefficients of the eigenvector is normalized and used as the weights. Indicates the first Motor displacement offset during the second braking process; Indicates the first Braking wear parameters during a single braking process; overall loss per braking cycle This means that by weighting and integrating the backlash parameter and wear parameter under different weights, the characteristic differences of different parameters in constructing the feature function can be obtained; The larger the value, the more severe the wear and tear on the braking components within the EMB system.

[0112] Based on the above steps, the single-cycle comprehensive loss during a single braking process compared to the previous braking process is obtained. However, since the wear caused by brake pad friction is small and gradual, it is difficult to effectively identify the influence of the EMB system on the vehicle braking process by only observing the difference in the comprehensive loss of a single braking process in adjacent braking processes.

[0113] Furthermore, the detection signals of the EMB system, especially the current and position signals, are sampled at a high frequency (e.g., 1kHz) to ensure the precise control requirements of the braking system. During a complete vehicle braking process, from pressing the pedal to the pedal being fully released after the vehicle stops, the time interval will have a high data value when sampled at a high frequency. Using the average data is difficult to effectively reflect the detailed changes during the vehicle braking process.

[0114] Therefore, a sliding window is needed to divide the data into windows during the braking process, enabling the identification of characteristic stages and moments such as the force loading stage and motor braking and steering during braking. Furthermore, a fixed length is set... A sliding window (e.g., 50 braking operations) can effectively identify abnormal states during vehicle braking by detecting the cumulative loss and wear trend of the friction pair in adjacent braking processes within the window.

[0115] In this embodiment, the window anomaly index of the corresponding sliding window is obtained based on the single comprehensive loss, and a data feature space is constructed by combining the single comprehensive loss, including:

[0116] A sliding window of a certain length is set, and the cumulative loss within each sliding window is calculated based on the sum of the single comprehensive losses; the sliding window is defined by a certain number of braking cycles, such as 50 braking cycles as one sliding window.

[0117] Based on the cumulative loss of a certain sliding window and the cumulative loss of the previous sliding window, the window anomaly index is obtained to indicate that a certain sliding window has an abnormal state.

[0118] A data feature space is constructed by combining the window anomaly index of each sliding window with the single comprehensive loss of each braking process.

[0119] In this embodiment, the cumulative loss within each sliding window is calculated based on the sum of the single comprehensive losses. ,in .

[0120] The window anomaly index provided in this embodiment The calculation formula is as follows:

[0121]

[0122] In the formula, Indicates the window anomaly index; Indicates the first The cumulative loss within each sliding window; Indicates the first The cumulative loss within each sliding window; This represents the ratio of the cumulative loss difference between two adjacent sliding windows, and the coefficient of variation of the cumulative loss within the sliding window. A larger coefficient indicates a greater difference in the cumulative loss within the adjacent sliding windows, suggesting that the cumulative loss is more significant in the first sliding window. The possibility of abnormal conditions is higher during the second braking operation. Adding 0.1 to the denominator is to prevent calculation problems caused by a denominator of 0. The specific setting of its value is based on the actual scenario, and the resulting calculation error is within the allowable range of the scenario.

[0123] This embodiment focuses on obtaining the window anomaly index reflected by each sliding window in the above steps. ( (This refers to the numbering of the sliding window). Simultaneously, it combines the changing relationships of characteristic data segments within each sliding window with the single comprehensive loss of a single braking process. Constructing a data feature space ,in Indicates the first The window anomaly index of a sliding window. Indicates the first The window anomaly index of each sliding window, and Less than , This is used for outlier analysis.

[0124] In one specific embodiment, the isolated forest algorithm is used to process the data feature space to obtain the isolation factor for each braking.

[0125] Specifically, the detection data from each braking process under the operating state of the EMB system are processed using the above method to obtain the data feature space for each braking process. Then, the isolation factor of each sample point (i.e., the data feature space) is calculated using the Isolation Forest algorithm. , The larger the value, the more likely it is to be an abnormal state.

[0126] Among them, the provided isolation factor The formula expression is as follows:

[0127]

[0128] in, This indicates that the isolated forest algorithm will be applied. Indicates the first i The isolated factor after the second braking. It should be noted that... The specific value range is the normalized range, and the outlier scores with values ​​of [0,1] in the Isolation Forest algorithm can be directly used as the isolation factor.

[0129] In this embodiment, the obtained isolated factors are classified according to their health status based on a pre-set health status level threshold, including:

[0130] Set a dynamic first health status threshold Second health status level threshold And the second health status level threshold Greater than the first health status level threshold ;

[0131] If the obtained isolation factor is less than the threshold of the first health status level (i.e. If the result is positive, then the electromechanical braking system is confirmed to be in a healthy state.

[0132] If the obtained isolation factor is greater than or equal to the threshold of the first health status level And less than the threshold for the second health status level. (i.e. If the result is positive, then the electromechanical braking system is confirmed to be slightly degraded at this time.

[0133] If the isolation factor is greater than or equal to the threshold of the second health status level (i.e. If the result is positive, then the electromechanical braking system is confirmed to be severely degraded.

[0134] Additionally, it should be noted that the threshold for the first health status level... Second health status level threshold It can adaptively update based on feature distribution, thereby dynamically setting the threshold for the first health status level. Second health status level threshold Optionally, the first health status level threshold. Specifically, it could be, for example, 0.5; the second health status level threshold. For example, it can be 0.8.

[0135] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0136] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for monitoring the health status of an intelligent EMB system based on integrated sensing and control, characterized in that: The method includes: Collect detection data generated by the operation of the electromechanical braking system during vehicle braking; the detection data includes motor monitoring data. Calculate the single comprehensive loss of the electromechanical braking system during each braking process based on motor monitoring data; The window anomaly index of the corresponding sliding window is obtained based on the single comprehensive loss, and the data feature space is constructed by combining the single comprehensive loss. The isolated forest algorithm is used to process the data feature space to obtain the isolation factor for each braking. The isolated factors are classified according to their health status based on a pre-set health status level threshold. Methods for calculating single-instance comprehensive loss include: Based on motor monitoring data, the first motor position parameters at the moment of command reversal during motor drive and the second motor position parameters at the moment of clamping force decrease are obtained; the clamping force represents the pressure between the motor-driven brake pads and the vehicle brake disc; The motor displacement offset is calculated based on the position parameters of the first motor and the position parameters of the second motor. The effective stiffness of the electromechanical braking system during the force loading phase of each braking process was tested. Braking wear parameters are obtained by comparing the effective stiffness of the current braking with the effective stiffness of the previous braking in the electromechanical braking system. The single comprehensive loss of the electromechanical braking system during each braking process is calculated by combining the motor displacement offset and braking wear parameters. Methods for constructing data feature spaces include: Set a sliding window of a certain length, and calculate the cumulative loss within each sliding window based on the sum of the single comprehensive losses; Based on the cumulative loss of a certain sliding window and the cumulative loss of the previous sliding window, the window anomaly index is obtained to indicate that a certain sliding window has an abnormal state. A data feature space is constructed by combining the window anomaly index of each sliding window with the single comprehensive loss of each braking process.

2. The health status monitoring method for an intelligent EMB system based on integrated sensing and control as described in claim 1, characterized in that: After collecting the detection data generated by the operation of the electromechanical braking system during vehicle braking, and before calculating the single comprehensive loss of the electromechanical braking system at each braking node based on the motor monitoring data, the method further includes: The collected detection data is denoised using a Gaussian filtering algorithm.

3. The health status monitoring method for an intelligent EMB system based on integrated sensing and control as described in claim 1, characterized in that: The motor displacement offset is calculated based on the position parameters of the first motor and the position parameters of the second motor, including: Based on the position parameters of the first motor and the position parameters of the second motor, the absolute value of the motor position difference at the moment of command reversal and the moment of clamping force decrease during the motor driving process is calculated. Obtain standard backlash quantification parameters for brake pad state transitions within an electromechanical braking system; Based on the absolute value of the motor position difference and the standard backlash quantization parameters, the motor displacement offset at the moment of command reversal and the moment of clamping force reduction during each braking period is calculated.

4. The health status monitoring method for an intelligent EMB system based on integrated sensing and control as described in claim 1, characterized in that: The effective stiffness of the electromechanical braking system during the force loading phase of each braking process is detected, including: Based on motor monitoring data, the clamping force at the beginning and end of the force loading phase, as well as the motor position at the beginning and end of the force loading phase, are obtained. The absolute value of the difference between the clamping forces at the beginning and end of the force loading phase, and the motor offset; The effective stiffness of the electromechanical braking system during each braking process is obtained by comparing the absolute value of the difference in clamping force with the motor offset.

5. The health status monitoring method for an intelligent EMB system based on integrated sensing and control as described in claim 4, characterized in that: After acquiring the clamping forces at the start and end of the force loading phase, and the motor positions at the start and end of the force loading phase, and before calculating the absolute value of the difference between the clamping forces at the start and end of the force loading phase, and the motor offset, the method further includes: Based on motor monitoring data, obtain the clamping force detected each time, the temperature-friction relationship coefficient during each braking process, and the actual required friction coefficient; The ratio of the actual required friction coefficient to the temperature friction coefficient is used as the compensation coefficient to compensate for each detected clamping force, thus obtaining the compensated clamping force; the clamping force detected each time includes the clamping force at the beginning and end of the force loading stage.

6. The health status monitoring method for an intelligent EMB system based on integrated sensing and control as described in claim 1, characterized in that: Braking wear parameters are obtained by comparing the current effective stiffness of the electromechanical braking system with the effective stiffness of the previous braking event, including: Obtain the effective stiffness of the current braking and the effective stiffness of the previous braking in the electromechanical braking system; Calculate the difference between the current braking system stiffness and the effective stiffness of the previous braking, and calculate the ratio of the difference to the effective stiffness of the previous braking. The braking time is obtained, and the braking wear parameters are calculated by combining the difference with the ratio of the effective stiffness of the previous braking.

7. The health status monitoring method for an intelligent EMB system based on integrated sensing and control as described in claim 1, characterized in that: The sliding window is stopped a certain number of times to determine its length.

8. The health status monitoring method for an intelligent EMB system based on integrated sensing and control as described in claim 1, characterized in that: The isolated factors are classified according to their health status based on a pre-set health status level threshold, including: Set a dynamic first health status level threshold and a second health status level threshold, wherein the second health status level threshold is greater than the first health status level threshold. If the obtained isolation factor is less than the threshold of the first health status level, then the electromechanical braking system is confirmed to be in a healthy state at this time. If the obtained isolation factor is greater than or equal to the first health status threshold and less than the second health status threshold, then the electromechanical braking system is confirmed to be mildly degraded at this time. If the isolation factor is greater than or equal to the threshold of the second health status level, then the electromechanical braking system is confirmed to be severely degraded.

Images