Vehicle wheel side redundancy brake-by-wire system and method based on EMB system

Abstract

The invention provides a vehicle wheel-side redundant brake-by-wire system and method based on an EMB system, and relates to the technical field of vehicle wheel-side redundant brake-by-wire. According to the vehicle wheel-side redundant brake-by-wire system based on the EMB system, main and standby access torque, wheel rotating speed and longitudinal acceleration signals are collected in parallel at different frequencies, and data synchronization is ensured by adopting a timestamp and sliding window alignment technology; secondly, reliability weights of a main path and a standby path are calculated on the basis of Kalman filtering residual hysteresis judgment and in combination with acceleration correlation, rapid and accurate fault recognition is achieved, torque is fused according to the weights, and the required torque of the next period is predicted with the help of a short time sequence model; then estimating a road adhesion coefficient on line, and correcting a dynamic abnormal threshold value and a fusion weight in real time to adapt to different road conditions; and finally, the credibility of a main path is continuously monitored, main and standby switching is completed in a millisecond level through a smooth transition strategy, and continuous and smooth braking output is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle wheel edge redundant brake-by-wire technology, in particular to a vehicle wheel edge redundant brake-by-wire system and method based on an EMB system. BACKGROUND

[0002] As the core development direction of brake-by-wire technology, EMB (electronic mechanical brake system) directly drives brake calipers through motors to achieve precise torque control, and has important application value in new energy vehicles and automatic driving fields. The wheel edge redundancy architecture improves system reliability through main and backup actuators, which is a key design to meet the functional safety requirements of ISO 26262 ASIL-D level. However, in the process of multi-channel collaborative control, three technical problems of real-time anomaly detection of main channel signals, timing synchronization of multi-source heterogeneous data, and smoothness control of redundancy switching have not been effectively solved.

[0003] The prior art usually realizes wheel edge redundancy braking in two ways: one is hardware-level redundancy, that is, a mechanical switch or a static switching valve is used to switch to the backup channel when the main channel fails; the other is software-level redundancy, that is, after the deviation of the main channel feedback signal or the communication anomaly is monitored in real time, the vehicle controller switches the output to the backup channel. However, both of these two ways have limitations: hardware switching often accompanies mechanical hysteresis and impact, resulting in a sudden drop in instantaneous braking force; software switching relies on simple threshold judgment and fixed delay strategy, and cannot balance fault prediction, smooth transition and multi-channel signal fusion, making it difficult to meet the needs of millisecond-level response and high-precision distribution.

[0004] The existing scheme still has shortcomings in brake channel failure detection, main and backup channel data reliability evaluation, and seamless switching smoothness, and cannot guarantee continuous and stable braking output under variable road conditions and high dynamic conditions.

[0005] The above information disclosed in the background section is only used to strengthen the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0006] The purpose of the present application is to provide a vehicle wheel edge redundant brake-by-wire system and method based on an EMB system to solve the problems raised in the background.

[0007] To achieve the above purpose, the present application provides the following technical solutions: A vehicle wheel edge redundant brake-by-wire system based on an EMB system, comprising the following specific steps: A signal acquisition module for asynchronously acquiring and aligning the main and backup channel motor torque signals, wheel speed signals and vehicle body longitudinal acceleration signals of the vehicle to be detected; Anomaly detection module, for predicting the main channel motor torque signal theoretical value based on Kalman filter, calculating the residual value of the theoretical value and its measured value, and determining the anomaly when the residual value exceeds the dynamic anomaly threshold in the sliding window, and calculating the reliability weight of the main and standby channels according to the statistical correlation of the residual value in the same window and the vehicle body longitudinal acceleration signal; Adjustment parameter module, for estimating the adhesion coefficient of the road surface friction by the wheel speed signal and the vehicle body longitudinal acceleration signal using the recursive least squares algorithm, correcting the dynamic anomaly threshold according to the adhesion coefficient, and adjusting the fusion torque distribution coefficient in the redundancy switching module; Fusion calculation module, for weighting and fusing the main and standby channel motor torque signals according to the reliability weight of the main and standby channels to generate a fusion torque, and outputting the demand torque value of the next control period through a short-time sequence prediction model using the data signal in the sliding window as input; Redundancy switching module, for executing the full demand torque value by the main channel when the reliability weight of the main channel is higher than the safety threshold, starting switching when it is lower than the safety threshold, gradually reducing the distribution proportion of the main channel motor torque according to the fusion torque distribution coefficient, synchronously increasing the distribution proportion of the standby channel motor torque, and monitoring the fluctuation intensity of the fusion torque in real time to dynamically adjust the transition window to suppress mutation.

[0008] Further, the signal acquisition module asynchronously receives the following signals through the vehicle CAN / CAN-FD bus: The main channel motor torque signal fed back by the wheel-side main actuator and the standby channel motor torque signal fed back by the wheel-side standby actuator, both with a sampling rate of 500Hz, the wheel speed signal output by the wheel speed sensor with a sampling rate of 100Hz, and the vehicle body longitudinal acceleration signal output by the vehicle body inertia unit with a sampling rate of 1kHz; After the asynchronously collected data is timestamped with the ECU local time, it is stored in a bidirectional circular buffer that can buffer at least W consecutive control period data, and time sequence alignment is performed through a sliding window mechanism: the sliding window contains W consecutive control periods, each control period is , and the window step is one control period long, ensuring that the total window length covers at least the maximum brake response delay of the system ; The time sequence alignment mechanism takes the vehicle body longitudinal acceleration signal as the reference time axis, uses piecewise cubic Hermite interpolation for the wheel speed signal, and uses linear interpolation for the main and standby channel motor torque signals. After interpolation, the timestamp synchronization error of each signal is less than or equal to 5μs. After interpolation, the anomaly detection module reads the aligned signal data in the sliding window.

[0009] Further, the abnormality detection module predicts the theoretical value of the main passage motor torque signal based on the Kalman filter, and calculates the difference between the measured value and the theoretical value as a residual value: wherein i represents the wheel number, k is the index of the control period in the sliding window, and the residual value of each wheel in the abnormality detection module in the kth control period is sequentially indexed in time sequence, and the calculation formula of the residual value is: wherein, is the Kalman filter theoretical value of the main passage motor torque signal of the ith wheel of the vehicle in the kth control period, is the corresponding measured torque value, and the difference between the two is the residual value ; the average residual value in the sliding window is: Only when the average residual value of the main passage of the ith wheel exceeds the dynamic abnormality threshold value in the sliding window, it is determined that the main passage signal of the wheel is abnormal, otherwise the main passage signal is normal.

[0010] Further, according to the residual value and the body longitudinal acceleration signal in the same time window, the logic for calculating the confidence weight of the main passage and the standby passage by statistical correlation analysis is as follows: The calculation formula of the confidence weight of the main passage of the ith wheel in the kth control period is as follows: wherein, is the residual sensitive coefficient in the kth control period, is the Pearson correlation coefficient of the residual sequence and the body longitudinal acceleration sequence in the same window; The calculation formula of the confidence weight of the standby passage of the ith wheel in the kth control period is as follows: The average confidence weight in the sliding window is: , and the same .

[0011] Further, the fusion calculation module calculates the fusion torque of the ith wheel in the kth control period according to the following formula: wherein, and are the measured torque values of the main passage and the standby passage of the ith wheel in the kth control period, respectively. ​​​​​​​​The fused torque sequence of W control cycles in the sliding window, the vehicle longitudinal acceleration signal sequence, and the wheel speed signal sequence are input into the LSTM neural network model to predict and output the required torque value of the next control cycle.

[0012] Furthermore, the adjustment parameter module uses a recursive least squares algorithm to online identify the road friction adhesion coefficient based on the wheel speed signal and the vehicle longitudinal acceleration signal, and corrects the dynamic abnormality threshold and the fusion torque distribution coefficient according to the following formula: The correction formula of the dynamic abnormality threshold is: ; in, is the basic threshold, is the road surface sensitivity coefficient, is the road friction coefficient estimated based on the recursive least squares method in the kth control cycle, For the Dynamic abnormal threshold of each control cycle; The adjustment of the fusion torque distribution coefficient satisfies: ; in, is the base distribution coefficient, is the attenuation factor, which automatically reduces the proportion of the main channel on low adhesion roads. For the predicted The fusion torque distribution coefficient of the control cycle, For the Road friction adhesion coefficient for a control cycle; In the calculation of the credibility weight, the dynamic adjustment of the residual sensitivity coefficient is: , For the The residual sensitivity coefficient of a control cycle is the smaller the dynamic abnormality threshold, the higher the residual sensitivity.

[0013] Furthermore, the redundant switching module monitors the credibility weight of the primary path and calculates its average credibility weight. When the average credibility weight is lower than a preset safety threshold, the transition phase is entered. Assume the total transition time is , the relative progress of the switching process is defined as Inside, is the time variable, and the exponential smoothing transition function is defined as : ; in, is the smoothing rate coefficient, , The current transition progress. Indicates the start of switching. Indicates that the switch is completed; During the transition period, the required torque value is distributed between the main channel and the backup channel in proportion, where the main channel torque distribution ratio is , the torque distribution ratio of the backup channel is , so that the main passage torque ratio changes from Gradually decays to 0, and the backup path ratio increases from 0 to ; Monitor the fusion torque in real time, calculate the standard deviation of the fusion torque sequence to characterize the fluctuation intensity, and increase or decrease it according to the fluctuation intensity To balance switching speed and smoothness and suppress sudden changes; When the credibility weight of the main path is restored to the safety threshold or above, the redundant switching module uses the opposite exponential smoothing transition function and the same Switch the braking torque distribution back to the main channel to achieve seamless two-way switching.

[0014] The present invention further provides a vehicle wheel-side redundant brake-by-wire method based on an EMB system, the method being used to implement the above-mentioned vehicle wheel-side redundant brake-by-wire system based on an EMB system, comprising: Step 1: Asynchronously collect and align the main and backup channel motor torque signals, wheel speed signals, and body longitudinal acceleration signals of the vehicle to be tested; Step 2: The Kalman filter is used to predict the theoretical value of the main channel motor torque signal. The residual between the theoretical value and the measured value is calculated. If the residual exceeds the dynamic abnormality threshold within the sliding window, an abnormality is determined. The credibility weights of the main and backup channels are calculated based on the statistical correlation between the residual value and the vehicle longitudinal acceleration signal within the same window. Step 3: Using the wheel speed signal and the vehicle longitudinal acceleration signal, a recursive least squares algorithm is used to estimate the road friction adhesion coefficient. Based on this adhesion coefficient, the dynamic abnormality threshold is corrected and the fusion torque distribution coefficient in the redundant switching module is adjusted. Step 4: Perform weighted fusion of the motor torque signals of the active and standby paths according to their credibility weights to generate the fused torque. The data signal within the current sliding window is used as input, and the short-term prediction model is used to output the required torque value for the next control cycle. Step 5: When the credibility of the main channel is higher than the safety threshold, the main channel executes the full required torque value. When it is lower than the safety threshold, the switching is initiated. The distribution ratio of the main channel motor torque is gradually reduced according to the fusion torque distribution coefficient, and the distribution ratio of the backup channel motor torque is simultaneously increased. The fluctuation intensity of the fusion torque is monitored in real time, and the transition time window is dynamically adjusted to suppress sudden changes.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present application effectively eliminates the acquisition error caused by signal timing deviation through high-precision synchronous acquisition and alignment of multi-source signals, and guarantees the accuracy of pedal force signals, main and standby path torque signals, wheel speed and acceleration signals under the same time reference; The double abnormality detection based on Kalman filter residual hysteresis judgment and statistical correlation evaluation not only shortens the fault detection delay, but also avoids misjudgment caused by transient interference; And the residual and mutual information weighted fusion generates a fusion torque, and the LSTM short sequence prediction model is combined to output the demand torque value of the next control period in advance, so that the brake force distribution can accurately track the driver's intention and predict the future dynamic demand, greatly improving the accuracy of torque distribution; The present application uses the road surface friction coefficient identified online by the recursive least square algorithm to dynamically correct the abnormality judgment threshold and the fusion torque distribution coefficient, realizes adaptive control in various road conditions; When the main path fails is detected, the exponential smoothing transition function and the adjustable transition time window make the main and standby path switching complete in a short time, the fusion torque fluctuation amplitude is very small, and the switching impact and jerk are eliminated. The above technical features work together, not only solve the hysteresis and mutation problem of the existing mechanical or software switching, but also significantly improve the safety reliability and passenger comfort of the redundant linear braking system. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The present application is a whole method flowchart; Figure 2 The present application is a residual value-Pearson correlation coefficient-main path reliability weight 3D color mapping surface graph; Figure 3 The present application is a residual value-main path reliability weight vertical line graph; Figure 4 The present application is a residual value-residual sensitivity coefficient fitting curve graph; Figure 5 The present application is a residual value-main path reliability weight emphasis line graph; Figure 6 The present application is a whole system module flowchart. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical scheme and advantages of the present application clearer and more apparent, the present application will be further described in detail below in combination with specific embodiments.

[0018] Unless otherwise defined, technical terms or scientific terms used in the present application shall have the meanings that can be commonly understood by a person with ordinary skill in the art to which this application belongs. The terms "first", "second", and similar terms used in the present application do not denote any order, quantity, or importance, but are used to distinguish different components. The terms "include", "contain", and similar terms mean that the elements or objects before the terms encompass the elements or objects listed after the terms and their equivalents, and do not exclude other elements or objects. The terms "connect" or "connected" or similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The terms "upper", "lower", "left", "right", and the like are used only to indicate relative positional relationships, and when the absolute positions of the described objects change, the relative positional relationships can also change accordingly.

[0019] Embodiments: Please refer to Figures 1-5 The present application provides a technical solution: A vehicle wheel-side redundant brake-by-wire system based on an EMB system, the specific steps comprising: A signal acquisition module for asynchronously acquiring and aligning the main and backup path motor torque signals, wheel speed signals, and vehicle body longitudinal acceleration signals of a vehicle to be detected; The signal acquisition module asynchronously receives the following signals through the vehicle CAN / CAN-FD bus: The main path motor torque signal fed back by the wheel-side main actuator and the backup path motor torque signal fed back by the wheel-side backup actuator, both with a sampling rate of 500Hz, the wheel speed signal output by the wheel speed sensor, with a sampling rate of 100Hz, and the vehicle body longitudinal acceleration signal output by the vehicle body inertia unit, with a sampling rate of 1kHz; The system uses a 500Hz sampling rate to collect main and backup path torque signals in real time, synchronizes with the EMB motor control period, and ensures accurate capture of dynamic responses such as stall and overload; at the same time, it matches the standard frequency of the ABS wheel speed sensor to collect wheel speed signals, optimizes the calculation load while ensuring the accuracy of slip rate detection; and sets up a 20-50Hz high-frequency sampling channel to capture the vehicle body longitudinal acceleration, which is used to identify the braking pitch oscillation characteristics and provide data support for road identification and fault diagnosis; After the asynchronously acquired data is timestamped with the ECU local time, it is stored in a bidirectional circular buffer that can cache at least W consecutive control period data, and time alignment is performed through a sliding window mechanism: the sliding window contains W consecutive control periods, each control period is , the window step is one control period, and the total window length covers at least the maximum brake response delay of the system ; The timing alignment mechanism uses the vehicle longitudinal acceleration signal as the reference time axis, applies piecewise cubic Hermite interpolation to the wheel speed signal, and linear interpolation to the main and backup motor torque signals. After the interpolation is completed, the synchronization error of the signal timestamps is less than or equal to 5μs. After the interpolation is completed, the anomaly detection module is triggered to read the alignment signal data within the sliding window. Linear interpolation requires only one addition and one multiplication (proportional calculation), requiring minimal computation. This is crucial for ECU systems with high real-time requirements, especially when processing active and standby dual-channel signals, saving significant computing resources. Interpolation uses existing discrete sampling point data close in time to the target moment to estimate the target moment, i.e., the signal value at a specific moment on the time axis of the baseline acceleration signal. Based on the relatively gentle change of the torque signal, a straight line is simply used to connect two adjacent original sampling points. Then, according to the position of the target moment on this line, the torque values ​​of the two endpoints are proportionally mixed to obtain the torque estimate at the target moment. Given that torque signals typically change slowly, connecting adjacent sampling points with a straight line within a short sampling interval (relative to the signal change rate) can provide sufficiently accurate intermediate value estimates to meet the requirements of timestamp alignment of less than or equal to 5 μs and subsequent anomaly detection. In areas of slow change, the benefits of high-order interpolation are minimal. The motor torque signal reflects the torque output by the motor. Torque commands are typically issued by a controller, and their changes are often smoother and more linear than the mechanical response of wheel speed. This is especially true within a control cycle. The motor's physical inertia and control response prevent it from experiencing sudden, drastic changes like wheel speed. Torque commands are typically updated at the boundaries of a control cycle, and may remain constant or change slowly according to the command within the control cycle. For rapidly changing and highly nonlinear wheel speed signals, a more complex but more accurate piecewise cubic Hermite interpolation method that can reflect the rate of change is used. For the relatively slowly changing motor torque signal, a simpler and more efficient linear interpolation method is used. The target time point of all interpolation operations is derived from the timestamp of the vehicle longitudinal acceleration signal, ensuring that all signals are ultimately unified to the same time base. The bidirectional circular buffer and sliding window mechanism ensure that the latest, continuous W cycle data is processed and the total length covers the critical delay , so that the aligned data can be effectively used to detect anomalies with delay characteristics.

[0020] An anomaly detection module, which uses a Kalman filter to predict the theoretical value of the main channel motor torque signal and calculate the residual value between the theoretical value and the measured value. An anomaly is determined when the residual value exceeds the dynamic anomaly threshold within a sliding window. The credibility weight of the main and backup channels is calculated based on the statistical correlation between the residual value and the vehicle longitudinal acceleration signal within the same window. The anomaly detection module predicts the theoretical value of the main channel motor torque signal based on the Kalman filter and calculates the difference between the theoretical value and the measured value as the residual value: Assume i represents the wheel number, k is the index of the control cycle in the sliding window, and the indexes are performed in chronological order. The residual value of each wheel in the kth control cycle in the anomaly detection module is The calculation formula is: ; in, is the Kalman filter theoretical value of the main channel motor torque signal of the i-th wheel of the vehicle in the k-th control cycle, is the corresponding measured torque value, and the difference between the two is the residual value ; Average residual value within the sliding window: ; Indicates the theoretical prediction value of the main channel motor torque on the i-th wheel in the k-th control cycle Compared with the actual measured value The deviation between the main path torque sensor reading is directly quantified The torque that should be generated under normal fault-free conditions is predicted based on the Kalman filter. the degree of inconsistency between This indicates a close agreement between the measured values ​​and the model predictions, which is usually a strong indication that the system is working properly, the sensors are accurate, and the model matches the actual operating conditions. A larger value may indicate: a. a fault in the main torque sensor, such as drift, sticking, or excessive noise; b. a discrepancy between the actual torque output of the motor or related drive circuit and the command or model expectations, such as a short circuit, open circuit, or power device failure in the motor winding; c. a deviation in the vehicle model itself, especially in extreme or unmodeled operating conditions; d. improper Kalman filter parameter settings, such as inaccurate process noise or measurement noise covariance. The larger the value, the more torque is actually measured. Significantly smaller than the model prediction , for example, a sensor failure causes a low reading, or the actual output torque of the motor is insufficient, The larger the absolute value of the negative value, the greater the actual measured torque. Significantly greater than the model prediction For example, sensor failure causes the reading to be too high, or the motor unexpectedly outputs too much torque, such as short circuit causes locked torque; in general, The larger, the higher the possibility of the main passage torque signal being abnormal, the greater the deviation of the system operating state from the expected model; In Fixed, The increase will cause Increase, in Fixed, The increase will cause Decrease, The size and sign of the value directly depend on the relative size of the theoretical prediction value and the measured value; Only when the average residual value of the main passage of the ith wheel When the average value exceeds the dynamic abnormal threshold in the sliding window, it is determined that the main passage signal of the wheel is abnormal, otherwise the main passage signal is normal; Such an average value determination avoids false positives caused by individual instantaneous spikes, while ensuring that the detected deviation is persistent and can truly reflect the potential impact on the braking process; This dynamic abnormal threshold is not fixed, but dynamically adjusted, which may change in real time according to factors such as vehicle speed, load, road conditions, estimated signal noise level, etc. The purpose is to adapt the sensitivity of abnormal detection to different working conditions, for example, in low adhesion road or severe dynamic working conditions, the model prediction error, i.e. It may be larger, so raising the threshold can avoid false positives; in stable working conditions, the threshold can be lowered to improve detection sensitivity.

[0021] According to the residual value and the body longitudinal acceleration signal in the same time window, the logic for calculating the confidence weight of the main and standby passages is as follows: The confidence weight of the ith wheel main passage in the kth control cycle The calculation formula is as follows: ; Where, is the residual sensitivity coefficient in the kth control cycle, which is used to adjust the influence of the residual value on the confidence weight, is the Pearson correlation coefficient of the residual sequence and the body longitudinal acceleration sequence in the same window; The specific practical significance and reflection of are as follows: it is a value between 0 and 1, indicating the degree of trust of the system in the torque signal of the main passage of the ith wheel in the kth control cycle, and it integrates the size of the current residual and the statistical correlation Two factors are used to dynamically evaluate whether the main channel signal is reliable; Close to 1, it means that the system has high trust in the main channel signal, which happens when the residual is small, that is, Small and the residual is highly correlated with the vehicle body acceleration, i.e. When it is close to 1; Close to 0.5, it means that the system has neutral or medium trust in the main channel signal, which usually occurs when: a. The residual is small but the correlation is also low, that is, When it is close to 0, b. the residual is large but the correlation is high, that is, If the residual is close to 1, it is likely caused by real vehicle dynamics, such as severe acceleration and deceleration, rather than a sensor failure. Close to 0, it means that the system does not trust the main channel signal. This happens when the residual is large. Large and the residual has a low correlation with the vehicle body acceleration, that is, When it is close to 0, this is a strong sign that the sensor or actuator is faulty and the signal is completely unreliable; The larger it is, the more reliable and trustworthy the main path torque signal is considered to be in the kth control cycle, and subsequent systems, such as torque distribution and fault-tolerant control, will be more inclined to use the main path signal; It is a A function that increases and decays rapidly; The bigger, A slight increase will result in Sharp decrease; The bigger, The smaller, The smaller it is, the larger the residual is, and the lower the trust in the main path is. The larger it is, the more sensitive the system is to the residual; The bigger, The bigger, The larger it is, that is, the stronger the correlation between the residual and the vehicle acceleration, the higher the trust in the main channel; In the calculation of the credibility weight, the dynamic adjustment of the residual sensitivity coefficient is: , For the The residual sensitivity coefficient of each control cycle, the smaller the dynamic abnormal threshold, the higher the residual sensitivity; The specific practical significance and reflection of the situation: it is used for the next control cycle , in calculating the main path credibility weight When the absolute value of the residual is adjusted The coefficient of the weight influence intensity, which reflects the system in different detection sensitivities, is given by In this case, the sensitivity of the weight calculation to the residual size needs to be adjusted accordingly to coordinate the weight change with the anomaly judgment logic; high , for good road sections, the threshold is wide, Smaller, in calculation hour, along with Increases and decays slowly, which means that even if the residual Relatively large, but still likely smaller than the relaxed dynamic anomaly threshold , for weight The penalty is relatively light, and the main path weight decreases less dramatically, which matches the relaxed abnormality judgment criteria on good roads; Low , for slippery sections, the threshold is strict, Larger, along with Increases and decays very quickly, which means that even if the residual The absolute value is not large, but it may approach or exceed the tightening dynamic anomaly threshold , which will also affect the weight This results in a significant penalty, causing the main path weight to drop rapidly. This matches the tightened anomaly determination criteria and higher detection sensitivity requirements on the sliding path, ensuring that when the residual approaches or exceeds the low threshold, the weight can quickly reflect the decline in signal credibility. The larger the value, the more the system will calculate the residual in the next cycle. The more sensitive the size change is, the smaller the residual error will increase, which will lead to the main path weight Significantly reduced, The smaller it is, the more the system can control the residual The less sensitive θ is to changes in θ, the larger the residual needs to be to cause a significant decrease in weight.

[0022] The credibility weight of the i-th wheel backup path in the k-th control cycle The calculation formula is: ; It is the complement of the main path credibility weight and is a value between 0 and 1. It represents the system's confidence in the torque signal of the backup path of the i-th wheel in the k-th control cycle. The system's confidence in the main and backup paths is in a trade-off relationship. When the main path credibility is high, the backup path credibility is naturally low. When the main path credibility is low, it means that the backup path may be more reliable, and its credibility is correspondingly improved. Close to 1, it means that the system highly trusts the backup path signal, which directly corresponds to When it is close to 0, the main channel signal is considered extremely unreliable; Close to 0.5, it means that the system has neutral or medium trust in the backup path signal, which corresponds to Close to 0.5; Close to 0, it means that the system does not trust the backup path signal, which corresponds to When it is close to 1, the main channel signal is considered to be very reliable; The larger the value is, the more reliable and trustworthy the backup path torque signal is considered to be in the kth control cycle. When the main path is judged to be abnormal or has a low weight, the system will tend to use the backup path signal for control or diagnosis. The average credibility weight within the sliding window is: , similarly ; A sudden change in the correlation coefficient or residual error in a single cycle may just be a spike caused by transient noise or short pulse interference. Using it to make immediate decisions can easily lead to excessive switching or false alarms. Average weighting can smooth out these short-term spikes and is only sensitive to persistent deviations. In a window of length W, Aggregating information from multiple cycles better represents the overall health of the main channel during that period. A consistently high average indicates a reliable signal most of the time; a declining average can indicate a trending failure risk. Compared to the maximum strategy, the average is less sensitive to single extreme errors; and unlike the minimum strategy, it also doesn't ignore multiple minor anomalies. It can promptly capture anomalies accumulated over multiple cycles while avoiding overreaction to occasional errors. The average weight is directly used for "switching" or "restoring" decisions, which is consistent with the method of inputting each sequence in the sliding window into the LSTM model. Both rely on data statistics in the same period, are more logically unified, and are easier to implement.

[0023] The specific data of some control cycles and main path credibility weights are shown in Table 1.

[0024] Table 1

[0025] As the residual value increases, the credibility weight of the main path shows an exponential decay trend. For example, when the residual value is 0.12, in the first control cycle, the credibility weight of the main path is 0.86, which is highly credible. When the residual value is 1.52, in the 10th control cycle, the credibility weight of the main path drops to 0.12, close to failure. When the residual value is 1.68, in the 14th control cycle, the credibility weight of the main path further drops to 0.09, which is critical failure. This shows the system's sensitive response to abnormal signals: the greater the deviation between the measured torque of the main path and the theoretical prediction value, the lower its credibility evaluation value. This relationship reflects the exponential term in the formula The core role of ——the increase of residual causes the weight to decay exponentially.

[0026] Pearson correlation coefficient The absolute value of Positively correlated with weight: Higher, that is , in the first control cycle, even if the residual value is 0.12, the main channel credibility weight still reaches 0.86. Lower, that is , in the 6th control cycle, even with a residual value of 0.89, the main channel credibility weight dropped to 0.32; this reflects the correlation reward item Compensation mechanism: When the residual is highly correlated with the change in vehicle acceleration, the system determines that the deviation may be due to actual working conditions rather than sensor failure, thereby slowing down weight decay. This design effectively distinguishes between real fault signals and model prediction deviations.

[0027] Residual sensitivity coefficient As a dynamic adjustment parameter, it significantly affects the decay rate: low scene, , in the first control cycle, the decay is gentle, and the residual value needs to reach 1.0 to trigger a significant weight reduction (main channel credibility weight); high scene, , in the 10th control cycle, the attenuation is steep, and when the residual value is 1.52, the main channel credibility weight has dropped to 0.12; Breaking through the critical value of 2.0, at the beginning of the 7th control cycle, the median credibility weight of the main channel dropped from 0.57 to 0.16, and the rate of weight reduction increased by 2.4 times; this change is due to Inversely proportional to the dynamic abnormal threshold: in low adhesion road conditions, Automatically increase to improve detection sensitivity; in high adhesion road conditions, The reduction avoids misjudgment, which reflects the adaptive ability of the system to complex working conditions. Instead of simply relying on a single threshold, the system assesses signal reliability through the dynamic balance of residual error, correlation, and environmental sensitivity, significantly improving the accuracy of fault determination.

[0028] The adjustment parameter module is configured to estimate the adhesion coefficient of the road surface friction by using the recursive least square algorithm based on the wheel speed signal and the vehicle body longitudinal acceleration signal, correct the dynamic abnormal threshold value according to the adhesion coefficient, and adjust the fusion torque distribution coefficient in the redundant switching module. The fusion calculation module calculates the fusion torque of the i-th wheel in the k-th control period According to the following formula: ; Wherein, And are the measured torque values of the i-th wheel main path and the i-th wheel standby path in the k-th control period. The specific actual meaning and reflection of the i-th wheel in the k-th control period: it represents the best estimated torque value of the i-th wheel for vehicle control or subsequent prediction, which is the result of weighted average of the main path measured torque And the standby path measured torque The weight , Directly comes from the real-time reliability of the main and standby paths calculated by the abnormal detection module; When Approaches to 1, Approaches to 0, Approaches to That is, the system highly trusts the main path, and the fusion torque is almost equal to the measured value of the main path, which is the ideal state when the main path is normal; When Approaches to 0, Approaches to 1, Approaches to That is, the system highly trusts the standby path, and the fusion torque is almost equal to the measured value of the standby path, which indicates that the main path may have been determined to be abnormal or extremely unreliable, and the system has been substantially switched to the standby path; When , Both are between 0 and 1, It is the weighted fusion of the torque values of the main and standby paths, which occurs when the system trusts the main path moderately or is uncertain, for example, the residual error is large but the correlation with acceleration is strong Large, or the residual error is small but the correlation is weak Small, the system adopts a soft switching / fusion strategy, smooth transition or comprehensive information of both; The size of the torque output command or estimated value after the wheel is integrated depends on the specific vehicle control requirements, such as acceleration, braking, and maintaining speed. The formula itself does not directly indicate A larger or smaller value indicates the quality of a particular system state, and it represents the torque value selected by the system based on its credibility judgment; The bigger, right The greater the contribution, The bigger, right The greater the contribution; when or It increases itself, if its corresponding weight is not zero, then Tends to increase; fusion torque Dynamically and smoothly follows the torque signal of the path that is currently considered more reliable by the system ( or ), or strike a balance between the two, which isolates the impact of untrustworthy signals and provides more robust input for subsequent control or prediction.

[0029] The fused torque sequence of W control cycles within the sliding window, along with the vehicle longitudinal acceleration signal sequence and wheel speed signal sequence, is input into the LSTM neural network model to predict and output the required torque value for the next control cycle. The LSTM neural network is a model that is good at learning long-term dependencies in time series data. In this invention, it uses the historical data sequence of W consecutive control cycles in the sliding window, inputs the fusion torque sequence of W control cycles in the sliding window, the vehicle longitudinal acceleration signal sequence, and the wheel speed signal sequence, and its purpose is to predict the next control cycle. The required torque value may be a driver request, a higher-level controller target, or the torque required to meet a certain performance indicator; This predictive model integrates historical control inputs (fused torque), vehicle dynamic responses (acceleration), and actuator / system states (wheel speed) to infer future demands. It captures the vehicle's dynamic response to these inputs. The predicted demand torque value serves as feedforward or expected input for subsequent control modules, such as torque distribution and motor control, helping to improve system response speed and accuracy. Using fused torque as input ensures that the torque data fed into the prediction model has undergone credibility assessment and possible fault-tolerance processing, eliminating the impact of unreliable sensor readings and improving the accuracy and robustness of the prediction model. Vehicle torque demand and historical states, especially acceleration response and wheel speed changes, have complex time correlations and lag effects, such as inertia. Using LSTM can effectively model this time series dynamic characteristic, resulting in more accurate predictions than models that only use current data.

[0030] The parameter adjustment module uses a recursive least squares algorithm to online identify the road friction adhesion coefficient based on the wheel speed signal and the vehicle body longitudinal acceleration signal. , and correct the dynamic abnormality threshold and fusion torque distribution coefficient according to the following formula: The specific practical significance and reflection of is: It is a value between 0 (such as ice surface) and 1 (such as dry asphalt), which represents the maximum available friction coefficient between the tire and the road surface estimated by the system in the kth control cycle. The ultimate performance of the vehicle's longitudinal dynamics (acceleration, braking) is fundamentally limited by the road adhesion conditions. It is a real-time online estimation of this key physical environment parameter.

[0031] The wheel speed signal reflects the wheel speed and slip rate, and the body longitudinal acceleration signal reflects the actual acceleration and deceleration of the vehicle. The recursive least squares algorithm iteratively updates by minimizing the model prediction error. The core idea is to use the real-time relationship between wheel speed and acceleration to continuously adjust the tire force-slip relationship according to a model that describes the relationship between tire force and slip rate, such as the simplified form of the magic formula. The value of makes the vehicle acceleration predicted by the model consistent with the actual measured vehicle acceleration as much as possible; When the actual measured acceleration is When the predicted acceleration deviates greatly from the wheel speed calculation, the recursive least squares algorithm will adjust To reduce this bias, the algorithm has the ability to "forget" old data (through the forgetting factor) and pay more attention to recent dynamics; It is a key environmental parameter inferred by the system in real time based on wheel behavior (wheel speed) and vehicle overall response (acceleration), and is the cornerstone of all subsequent parameter adjustments; The correction formula of the dynamic abnormality threshold is: ; in, is the basic threshold, is the road surface sensitivity coefficient, is the road friction coefficient estimated based on the recursive least squares method in the kth control cycle, For the Dynamic abnormal threshold of each control cycle; The specific practical significance and reflection of the situation: it is for the next control cycle Set to determine the main path residual The threshold value of whether the standard is exceeded is determined by the current estimated road adhesion coefficient. Dynamic adjustment, which reflects the system's tolerance for torque signal anomalies on different road surfaces; high , Large, such as dry and good road sections, the system allows the main channel to measure the torque A larger deviation from the model-predicted torque may trigger an abnormality determination. This is because on good roads, the model is generally more accurate and the vehicle dynamics are more stable. Smaller deviations are more likely to indicate problems with the sensor or actuator itself. At the same time, a larger threshold can avoid false alarms during normal operating fluctuations, such as heavy acceleration / braking. Low , Small, such as icy and slippery roads, the system has stricter requirements on the consistency of the main channel signal. Even a small deviation may trigger a concern or abnormal judgment. This is because on low-adhesion roads: a. Vehicle dynamics are complex and severe (prone to slip and lock), and the model prediction itself may be more inaccurate, resulting in the absolute value of the residual. The baseline becomes larger. If a fixed threshold for high-adhesion roads is still used, a large number of "false normals" will be generated, and real faults will be missed. b. Any torque sensor failure or abnormal actuator output, even if the absolute deviation is small, can easily cause vehicle loss of control on low-adhesion roads. A lower threshold is needed to improve detection sensitivity and detect potential problems as early as possible. The larger the value is, the higher the system's tolerance for the residual error of the main channel torque signal in the next cycle, and the higher the threshold for determining abnormality (the more difficult it is to trigger abnormality determination). The smaller the value, the lower the system's tolerance for residuals, the lower the threshold for determining anomalies (the easier it is to trigger an anomaly determination), and the higher the detection sensitivity. It is a benchmark threshold set under typical or calibrated conditions, such as medium adhesion dry road. is a positive coefficient, controlling The intensity of the threshold adjustment, The larger the value, the greater the impact of road adhesion changes on the threshold. On good roads, the threshold is relaxed to prevent false alarms, and on slippery roads, the threshold is tightened to improve sensitivity and ensure safety. regulating the aggressiveness of this adaptive strategy; The adjustment of the fusion torque distribution coefficient satisfies: ; in, is the base distribution coefficient, is the attenuation factor, which automatically reduces the proportion of the main channel on low adhesion roads. For the predicted The fusion torque distribution coefficient of the control cycle, For the predicted Road friction adhesion coefficient for a control cycle; The specific practical meaning and reflection of the situation: it is a value between 0 and The value between is used for the next predicted control cycle , dynamically adjust the degree to which the fusion strategy tends to use the main channel torque, which reflects the system's reduced trust in the main channel under low-adhesion road conditions; It is obtained through online identification using the recursive least squares algorithm. The algorithm has built-in mechanisms such as historical data weight decay or Kalman filter, and has already done some smoothing on noise and mutation. Therefore, it is directly used to identify the latest Updating the threshold and allocation coefficient can achieve both robustness and real-time performance without the need for outer window averaging; and are control parameters used for adaptive adjustment of the system. They are not directly used for instantaneous switch judgment, but as the target threshold or allocation ratio of the next cycle. They should reflect the latest working conditions to maintain accurate matching of the system error level or road surface conditions without excessive lag. Smoothing within the identification algorithm, and If window averaging is performed again, a double layer of filtering will be formed, causing the parameters to respond too slowly to changes in road conditions, losing the ability to adapt to sudden changes in road adhesion in real time, and reducing safety. and Relying on the latest online estimation results to ensure that the control parameters are highly consistent with the current road conditions, the functional positioning and timeliness requirements of the two are completely different. Therefore, in this invention, only the residuals and weights are averaged, and the latest values ​​of the dynamic thresholds and allocation coefficients are directly used; high , such as dry and good road sections, near ,The system tends to maintain or allow a higher basic weight of the main route. On a good road, the main route is usually more reliable and should be used first; Low , like a slippery road with ice and snow, Significantly smaller than The system proactively reduces the baseline weight or upper limit of the main path. This is because: a. Vehicle stability on low-adhesion roads is sensitive and has low fault tolerance; b. The main path is the main path, and once a fault occurs, the damage is greater; c. Even if the main path does not report a fault, the performance of its sensors or actuators under extreme working conditions may be more susceptible to impact; d. By reducing ,The system pre-biases the backup path at the architectural level, or limits the maximum weight of the main path in the fusion, which is equivalent to making a plan for the failure of the main path in advance on the sliding path, thereby improving the redundancy safety level; The larger it is, the higher the weight or proportion of the main path torque set or allowed by the system in the fusion strategy of the next cycle is. The smaller it is, the lower the weight or proportion of the main path set or allowed by the system foundation is, and the more likely it is to rely on the backup path; On high adhesion road, When it is close to 1, the ideal benchmark weight or proportion of the main channel is is a positive coefficient, which controls when the road adhesion decreases. The rate of decay, The larger it is, the faster it decays and the more aggressive it reacts to low-adhesion roads. Indicates the degree of loss of road adhesion. The lower, The bigger, the increases and decreases rapidly; On slippery roads, the system strategically reduces its default reliance on the main path, prepares for possible failures, and improves the overall robustness of the system under adverse conditions. Control the intensity of this strategy.

[0032] The redundant switching module is used to execute the full required torque value by the main channel when the credibility weight of the main channel is higher than the safety threshold. When it is lower than the safety threshold, the switching is initiated. The distribution ratio of the main channel motor torque is gradually reduced according to the fusion torque distribution coefficient, and the distribution ratio of the backup channel motor torque is simultaneously increased. The fluctuation intensity of the fusion torque is monitored in real time, and the transition time window is dynamically adjusted to suppress sudden changes. The redundant switching module monitors the credibility weight of the main path and calculates its average credibility weight. When the average credibility weight is lower than the preset safety threshold, the system enters the transition phase. This means that the system determines that the reliability of the main path torque signal of the i-th wheel has become dangerously low and is not sufficient to safely assume the full torque control task, reflecting that the main path may have failed (based on the continuous exceeding of the standard judgment of the abnormality detection module) or its credibility weight has exceeded the threshold. The sharp drop due to large residuals and low correlation indicates that its signal quality is extremely poor; This is a hard signal, indicating that the redundant switching mechanism must be started to begin the transfer of control responsibility from the primary path to the backup path, the preset safety threshold is a set threshold value, for example, 0.3 or 0.4, below which means that it is too risky to continue to rely on the primary path; the lower the preset safety threshold is set, the higher the tolerance of the system to the failure of the primary path (the switching is less sensitive), and the higher the preset safety threshold is set, the easier it is to trigger the switching; Let the total transition time be , the relative progress of the switching process is defined as , where is the time variable, and the exponential smoothing transition function is defined as : ; where is the smoothing rate coefficient, , is the current transition progress, indicates the start of switching, indicates the completion of switching; represents the continuous time elapsed since the start of the current switching action, when the switching is triggered, the switching start point, when , it indicates that the planned total switching time is exhausted, and the switching end point; The specific actual meaning and reflection of are as follows: it is a value between 0 and asymptotically approaching 1, within the fixed total transition time , it defines the completion degree or intensity of the switching process over time, which reflects that the system adopts a nonlinear, slow-to-fast transition strategy, at the beginning of the switching , it is small and changes slowly, and at the end , it approaches 1 and changes rapidly; is close to 0, i.e. , when it is close to 0, the switching has just started and the change to the system is very small, gradually increases, the switching process continues, and the influence on the system gradually deepens, when , at the end of the set transition period , the switching has completed most of it and is close to the completion state; , the greater the current switching progress, the higher the intensity of the current switching progress, i.e. the system state is closer to the state of completely switching to the backup path, which directly controls the rate of change of the torque distribution ratio; is the smoothing rate coefficient, which controls the shape and rate of change of the transition curve, the greater it is, the more gently the curve rises initially and the more steeply it rises later, is a calibrated value, ensuring that the state of completion is reached within ; represents the proportion of the current time to the total planned switching time ; increases, decreases, increases, with its growth being non-linear, small (at the beginning of switching), very slow growth (small slope), large (at the later stage of switching), relatively fast growth (slope becomes larger); slow decrease of the main path torque at the beginning of switching to leave response time for the actuator and vehicle dynamics, avoiding initial mutation; speed up at the later stage to complete the switching as soon as possible to reduce the risk window period; During the transition period, the demand torque value is proportionally distributed between the main path and the standby path, with the main path torque distribution proportion being , and the standby path torque distribution proportion being , so that the main path torque proportion gradually decays from to 0, and the standby path proportion synchronously rises from 0 to ; The specific practical significance and reflection of : it represents the proportion of the demand torque of the i-th wheel that is allocated to the main path execution during the switching transition period, which reflects that the responsibility of the main path is being gradually unloaded; At the beginning of switching , , the main path assumes its baseline distribution proportion of torque from the adjustment parameter module, which is consistent with the fusion strategy basis before switching triggering or in the normal state; during switching , , as increases, linearly decreases, and the proportion assumed by the main path gradually decreases; at the end of switching , , the main path completely exits torque execution and no longer assumes any torque output task; The larger , the greater the proportion of torque that the main path still needs to assume during the transition period, decreases to 0 indicating that the main path is completely unloaded; The specific practical significance and reflection of : it represents the proportion of the demand torque of the i-th wheel that is allocated to the standby path execution during the switching transition period, which reflects that the responsibility of the standby path is being gradually loaded; Switch start , , the main path assumes its reference allocation ratio of torque, which is consistent with the fusion strategy basis before the switch trigger or in the normal state; in the middle of the switch , , as increases, linearly, the proportion of the standby path gradually increases; at the end of the switch , , the standby path assumes full torque execution task; is the reference weight of the main path at normal or switch start, the larger the value, the higher the proportion of the main path at normal, and the lower the starting proportion of the standby path ; the larger the value, the greater the proportion of torque assumed by the standby path during the transition period, increasing to 1 means that the standby path takes over completely; Although and are directly controlled by linearly, itself is a linear time schedule , while is an exponential function, which means that the rate of change of the torque allocation ratio / is linear and uniform in time , here mainly used to define the concept of transition intensity or possibly used in adaptive adjustment logic, but not directly involved in torque allocation calculation, the torque allocation ratio is a linear function of time schedule ; The core logic is to linearly reduce the torque execution ratio of the main path from to 0 at a constant rate within a fixed planned transition time , while increasing the execution ratio of the standby path from to 1, which ensures the smoothness of the transfer of torque command between paths; Real-time monitoring of fusion torque, calculating the standard deviation of fusion torque sequence to represent the fluctuation intensity, and increasing or decreasing to balance the switching speed and smoothness, and to suppress mutations; The specific practical significance and reflection of the standard deviation of the fusion torque sequence: it is calculated in a small time window, such as several control periods, within the transition period of the switch, which reflects the size of the torque command fluctuation or oscillation caused by the current switching process, and it quantifies the disturbance intensity of the switching action on the system output; The smaller the standard deviation of the fusion torque sequence is, the smoother the switching process is, and the smaller the output fluctuation of the fusion torque is, and the smoother the vehicle dynamic experience is; The larger the standard deviation of the fusion torque sequence is, the significant torque fluctuation or oscillation caused by the switching process is, which may be due to: a. the large difference in characteristics of the master and standby channel torque sensors / actuators, b. the switching speed is too fast, which exceeds the response capability of the actuator or vehicle dynamics, c. external disturbances, such as road bumps, exacerbate the switching impact; The larger the standard deviation of the fusion torque sequence is, the more intense the disturbance caused by the current switching process to the system is, the worse the smoothness is, and the higher the potential risk is; The specific actual meaning and reflection of the adjusted total transition time : it is the new planned total transition time for subsequent switching or the remaining part of the current switching, which is dynamically calculated by the system according to the currently monitored standard deviation of the fusion torque sequence, and it reflects that the system adaptively balances the switching speed and smoothness; The smaller the standard deviation of the fusion torque sequence is, may be reduced or maintained, indicating that the current switching speed can be acceptable to be slower, and the system attempts to speed up the switching speed to complete the switching faster to reduce the time window of the unreliable state of the master channel; The larger the standard deviation of the fusion torque sequence is, increases, indicating that the current switching speed is too fast, which exceeds the smooth response capability of the system, and the system actively prolongs the switching time to slow down the torque transfer rate to suppress fluctuations and improve smoothness; The larger the adjusted total transition time is, the longer time the system plans to use to complete the subsequent switching process or the remaining part of the current switching, and the slower the switching speed is, in order to obtain better smoothness, The smaller the adjusted total transition time is, the faster the switching speed is; When the credibility weight of the master channel recovers to the safety threshold and above, the redundant switching module returns the braking torque distribution to the master channel with an opposite exponential smoothing transition function and the same two-way seamless switching is realized; The specific actual meaning and reflection of the master channel credibility weight recovery, i.e., the credibility weight of the master channel is greater than or equal to the preset safety threshold: this indicates that the master channel torque signal of the i-th wheel has recovered to a safe level at the current time, which reflects that the master channel may recover from failure, such as transient disturbance disappearing, failure being eliminated, or its signal quality significantly improving, residual error becoming small, and correlation improving; it indicates that this is a signal indicating that the control right can be safely switched back to the master channel; The return process is to use the same base or current adjusted As the total length of the plan, it ensures that the return process and the cut-out process are symmetrical in time scale; The most reasonable and safe implementation of the opposite exponential smoothing transition function is: a. Define a new return progress , whose value is from 0 to 1, b. Define the return transition function , in order to achieve the effect of "opposite", the more common engineering practice is: A. Load the main path: the torque distribution ratio of the main path Gradually increases from 0 to , which can be defined as , so that the load is slow at the beginning of the return , and fast later , which conforms to the general engineering principle of slow first and fast later; B. Unload the standby path: the torque distribution ratio of the standby path Gradually decreases from 1 to , which can be defined as , so that the unloading is slow at the beginning of the return , and fast later ; No matter scheme A or B, the form of the transition function Itself does not change, only the object or starting point of its application changes, that is, the main load and the standby unloading, so "opposite" is more likely to refer to the reversal of the operation object, that is, from the "unload main load standby" of the cut-out to the "load main unload standby" of the return, rather than the mathematical function form negation, the function form remains unchanged, which ensures the similarity of the switching dynamics; The meaning of two-way seamless switching: refers to the mechanism of triggering based on credibility weight, exponential smoothing transition, torque ratio linear transfer, and time length adaptive adjustment, which ensures that the switching in both directions is smooth, impact-free and controlled, like "seamless" connection.

[0033] Please refer to Figure 6 , the application further provides a vehicle wheel edge redundant line control brake method based on an EMB system, which is used to execute the vehicle wheel edge redundant line control brake system based on the EMB system. Step 1: asynchronously collect and align the main and standby path motor torque signals, wheel speed signals and vehicle body longitudinal acceleration signals of the vehicle to be detected; Step 2: predict the theoretical value of the main path motor torque signal based on the Kalman filter, calculate the residual value of the theoretical value and the measured value, and determine the abnormality when the residual value exceeds the dynamic abnormal threshold in the sliding window, and calculate the credibility weight of the main and standby paths according to the statistical correlation of the residual value and the vehicle body longitudinal acceleration signal in the same window; Step 3: Estimate the adhesion coefficient of the road surface friction by using the recursive least square algorithm through the wheel speed signal and the body longitudinal acceleration signal, correct the dynamic anomaly threshold value according to the adhesion coefficient, and adjust the fusion torque distribution coefficient in the redundant switching module; Step 4: Weighted fusion of the main and standby channel motor torque signals according to the reliability weight of the main and standby channels to generate the fusion torque, and output the required torque value of the next control cycle through the short-time sequence prediction model with the data signal in the current sliding window as the input; Step 5: When the main channel reliability is higher than the safety threshold, the main channel executes the full required torque value, and when it is lower than the safety threshold, the switching is started, the distribution proportion of the main channel motor torque is gradually reduced according to the fusion torque distribution coefficient, the distribution proportion of the standby channel motor torque is simultaneously increased, and the fluctuation intensity of the fusion torque is monitored in real time to dynamically adjust the transition window to suppress mutation.

[0034] The above formulas are all dimensionless numerical calculations, the formulas are obtained by software simulation of a large amount of data to obtain a formula of the latest real situation, and the preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0035] The above embodiments can be realized wholly or partially by software, hardware, firmware or any combination thereof. When realized by software, the above embodiments can be realized in the form of a computer program product wholly or partially. Those skilled in the art can realize that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software methods depends on the specific application and design constraints of the technical solutions.

[0036] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, which can be located in one place or distributed on multiple network units. Part or all of the units can be selected to achieve the purpose of the embodiments according to actual needs.

[0037] The above is merely a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. A vehicle wheel-side redundant brake-by-wire system based on an EMB system, characterized in that: Specifically include: A signal acquisition module is used to asynchronously acquire and align the main and backup channel motor torque signals, wheel speed signals, and body longitudinal acceleration signals of the vehicle to be tested; An anomaly detection module, which uses a Kalman filter to predict the theoretical value of the main channel motor torque signal and calculate the residual value between the theoretical value and the measured value. An anomaly is determined when the residual value exceeds the dynamic anomaly threshold within a sliding window. The credibility weight of the main and backup channels is calculated based on the statistical correlation between the residual value and the vehicle longitudinal acceleration signal within the same window. a parameter adjustment module for estimating the adhesion coefficient of road friction using a recursive least squares algorithm based on the wheel speed signal and the vehicle longitudinal acceleration signal, correcting the dynamic abnormality threshold based on the adhesion coefficient, and adjusting the fusion torque distribution coefficient in the redundant switching module; The fusion calculation module is used to perform weighted fusion of the motor torque signals of the main and backup paths according to their credibility weights to generate fused torque. The module uses the data signal in the sliding window as input and outputs the required torque value for the next control cycle through the short-time series prediction model. The redundant switching module is used to execute the full required torque value on the main channel when the credibility weight of the main channel is higher than the safety threshold, and initiate switching when it is lower than the safety threshold. The distribution ratio of the main channel motor torque is gradually reduced according to the fusion torque distribution coefficient, and the distribution ratio of the backup channel motor torque is simultaneously increased. The fluctuation intensity of the fusion torque is monitored in real time, and the transition time window is dynamically adjusted to suppress sudden changes.

2. The EMB-based vehicle wheel-side redundant brake-by-wire system according to claim 1, characterized in that: The signal acquisition module asynchronously receives the following signals via the vehicle CAN / CAN-FD bus: The sampling rate of the main motor torque signal fed back by the wheel-side main actuator and the backup motor torque signal fed back by the wheel-side backup actuator are both 500Hz. The wheel speed signal output by the wheel speed sensor has a sampling rate of 100Hz, and the vehicle longitudinal acceleration signal output by the vehicle inertia unit has a sampling rate of 1kHz. After the asynchronously collected data is timestamped by the ECU local time, it is stored in a bidirectional circular buffer that can cache at least W consecutive control cycle data. Timing alignment is performed through a sliding window mechanism: the sliding window contains W consecutive control cycles, and each control cycle is , the window step length is the length of a control cycle, ensuring the total window length At least cover the system's maximum brake response delay ; The timing alignment mechanism uses the vehicle body longitudinal acceleration signal as the reference time axis, adopts piecewise cubic Hermite interpolation for the wheel speed signal, and adopts linear interpolation for the main and backup channel motor torque signals. After the interpolation is completed, the synchronization error of the timestamps of each signal is less than or equal to 5μs. After the interpolation is completed, the anomaly detection module is triggered to read the alignment signal data in the sliding window.

3. The EMB-based vehicle wheel-side redundant brake-by-wire system according to claim 2, characterized in that: The anomaly detection module predicts the theoretical value of the main channel motor torque signal based on the Kalman filter and calculates the difference between the theoretical value and the measured value as the residual value: Assume i represents the wheel number, k is the index of the control cycle in the sliding window, and the indexes are performed in chronological order. The residual value of each wheel in the kth control cycle in the anomaly detection module is The calculation formula is: ； in, is the Kalman filter theoretical value of the main channel motor torque signal of the i-th wheel of the vehicle in the k-th control cycle, is the corresponding measured torque value, and the difference between the two is the residual value ; Average residual value within the sliding window: ; Only when the average residual value of the main channel of the i-th wheel When the dynamic abnormality threshold is exceeded within the sliding window, the main path signal of the wheel is determined to be abnormal; otherwise, the main path signal is maintained normal.

4. The EMB-based vehicle wheel-side redundant brake-by-wire system according to claim 3, characterized in that: Based on the residual value and the vehicle longitudinal acceleration signal in the same time window, the credibility weights of the primary and backup paths are calculated using statistical correlation analysis. The logic is as follows: The credibility weight of the i-th wheel main channel in the k-th control cycle The calculation formula is as follows: ； in, is the residual sensitivity coefficient in the kth control cycle, is the Pearson correlation coefficient between the residual sequence and the vehicle longitudinal acceleration sequence in the same window; The credibility weight of the i-th wheel backup path in the k-th control cycle The calculation formula is: ; Average credibility weight within the sliding window: , similarly .

5. The EMB-based vehicle wheel-side redundant brake-by-wire system according to claim 4, characterized in that: The fusion calculation module calculates the fusion torque of the i-th wheel in the k-th control cycle Calculate as follows: ； in, and are the measured torque values ​​of the main channel and backup channel of the i-th wheel in the k-th control cycle respectively; The fused torque sequence of W control cycles in the sliding window, the vehicle longitudinal acceleration signal sequence, and the wheel speed signal sequence are input into the LSTM neural network model to predict and output the required torque value of the next control cycle.

6. The EMB-based vehicle wheel-side redundant brake-by-wire system according to claim 5, characterized in that: The parameter adjustment module uses a recursive least squares algorithm to online identify the road friction coefficient based on the wheel speed signal and the vehicle longitudinal acceleration signal, and corrects the dynamic abnormality threshold and the fusion torque distribution coefficient according to the following formula: The correction formula of the dynamic abnormality threshold is: ； in, is the basic threshold, is the road surface sensitivity coefficient, is the road friction coefficient estimated based on the recursive least squares method in the kth control cycle, For the Dynamic abnormal threshold of each control cycle; The adjustment of the fusion torque distribution coefficient satisfies: ； in, is the base distribution coefficient, is the attenuation factor, which automatically reduces the proportion of the main channel on low adhesion roads. For the predicted The fusion torque distribution coefficient of the control cycle, For the Road friction adhesion coefficient for a control cycle; In the calculation of the credibility weight, the dynamic adjustment of the residual sensitivity coefficient is: , For the The residual sensitivity coefficient of a control cycle is the smaller the dynamic abnormality threshold, the higher the residual sensitivity.

7. The EMB-based vehicle wheel-side redundant brake-by-wire system according to claim 6, characterized in that: The redundant switching module monitors the credibility weight of the primary path and calculates its average credibility weight. When the average credibility weight is lower than a preset safety threshold, the module enters the transition phase. Assume the total transition time is , the relative progress of the switching process is defined as Inside, is the time variable, and the exponential smoothing transition function is defined as : ； in, is the smoothing rate coefficient, , The current transition progress. Indicates the start of switching. Indicates that the switch is completed; During the transition period, the required torque value is distributed between the main channel and the backup channel in proportion, where the main channel torque distribution ratio is , the torque distribution ratio of the backup channel is , so that the main passage torque ratio changes from Gradually decays to 0, and the backup path ratio increases from 0 to ; Monitor the fusion torque in real time, calculate the standard deviation of the fusion torque sequence to characterize the fluctuation intensity, and increase or decrease it according to the fluctuation intensity To balance switching speed and smoothness and suppress sudden changes; When the credibility weight of the main path is restored to the safety threshold or above, the redundant switching module uses the opposite exponential smoothing transition function and the same , switching the braking torque distribution back to the main channel to achieve seamless two-way switching.

8. A vehicle wheel-side redundant brake-by-wire method based on an EMB system, characterized by: The method is used to implement the vehicle wheel-side redundant brake-by-wire system based on the EMB system according to any one of claims 1 to 7, and the specific steps include: Step 1: Asynchronously collect and align the main and backup channel motor torque signals, wheel speed signals, and body longitudinal acceleration signals of the vehicle to be tested; Step 2: The Kalman filter is used to predict the theoretical value of the primary motor torque signal. The residual between the theoretical value and the measured value is calculated. If the residual exceeds the dynamic abnormality threshold within the sliding window, an abnormality is determined. The credibility weight of the primary and backup paths is calculated based on the statistical correlation between the residual value and the vehicle longitudinal acceleration signal within the same window. Step 3: Using the wheel speed signal and the vehicle longitudinal acceleration signal, a recursive least squares algorithm is used to estimate the road friction adhesion coefficient. Based on this adhesion coefficient, the dynamic abnormality threshold is corrected and the fusion torque distribution coefficient in the redundant switching module is adjusted. Step 4: Perform weighted fusion of the motor torque signals of the active and standby paths according to their credibility weights to generate the fused torque. The data signal within the current sliding window is used as input, and the short-term prediction model is used to output the required torque value for the next control cycle. Step 5: When the credibility of the main path is higher than the safety threshold, the main path executes the full required torque value. When it is lower than the safety threshold, the switching is initiated. The distribution ratio of the main path motor torque is gradually reduced according to the fusion torque distribution coefficient, and the distribution ratio of the backup path motor torque is simultaneously increased. The fluctuation intensity of the fusion torque is monitored in real time, and the transition time window is dynamically adjusted to suppress sudden changes.

Images