Vehicle mass estimation method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a vehicle mass estimation method and device, electronic equipment and a storage medium, and relates to the technical field of vehicle parameter estimation. The method comprises the following steps: acquiring original vehicle data required by quality estimation; performing jump processing on the original vehicle data to obtain corrected data; estimating the correction data by using a first estimation algorithm and a second estimation algorithm to obtain a first quality estimation value and a second quality estimation value; and performing adaptive weight fusion based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle. Through jump processing, sudden change and abnormal values in original vehicle data can be effectively corrected, interference on quality evaluation is reduced, robustness under complex working conditions is enhanced, a high-precision and stable quality estimation result is obtained through adaptive weight fusion, and the problem that large errors exist due to the fact that an existing method is insufficient in estimation precision and robustness is solved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle parameter estimation, and in particular to a vehicle mass estimation method, device, electronic device and storage medium. Background Art

[0002] Vehicle mass estimation is a key parameter in vehicle dynamics research, particularly in the areas of intelligent driving and vehicle dynamic control. Real-time vehicle mass estimation directly impacts vehicle performance, control strategies, and safety. For example, accurate vehicle mass information helps control systems such as the Electronic Stability Program (ESP) and Adaptive Cruise Control (ACC) make optimal decisions in complex driving scenarios such as sudden braking, sharp turns, or following a vehicle, thereby reducing the risk of skidding and collisions. Furthermore, for electric or hybrid vehicles, accurate vehicle mass estimation can rationally allocate traction and regenerative braking forces within the energy management system, thereby optimizing energy efficiency and extending the vehicle's range.

[0003] Existing mass estimation methods lack estimation accuracy and robustness when the vehicle undergoes drastic dynamic changes, such as sudden acceleration and braking, resulting in large errors. Summary of the Invention

[0004] The purpose of the embodiments of the present application is to provide a vehicle mass estimation method, device, electronic device and storage medium. Through jump processing, the mutations and outliers in the original vehicle data can be effectively corrected, the interference to the quality assessment can be reduced, and the robustness under complex working conditions can be enhanced. Through adaptive weight fusion, high-precision and stable mass estimation results can be obtained, which solves the problem of insufficient estimation accuracy and robustness of existing methods, resulting in large errors.

[0005] In a first aspect, the present application provides a vehicle mass estimation method, which includes: obtaining original vehicle data required for mass estimation; performing jump processing on the original vehicle data to obtain corrected data; estimating the corrected data using a first estimation algorithm and a second estimation algorithm respectively to obtain a first mass estimation value and a second mass estimation value; and performing adaptive weight fusion based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle.

[0006] In the technical solution of the embodiment of the present application, jump processing is performed on the original vehicle data, which can effectively identify and correct mutations and outliers in the collected original vehicle data, reduce interference with quality assessment, enhance robustness under complex working conditions, and obtain high-precision and stable quality estimation results; the quality estimation values ​​of the two algorithms are adaptively weighted and fused, and the weights of the quality estimation values ​​of the two algorithms are dynamically adjusted, thereby further improving the accuracy and stability of the quality estimation results.

[0007] In some embodiments, the raw vehicle data is subjected to jump processing to obtain corrected data, including: calculating the acceleration change rate based on the raw vehicle data; performing a moving average filter on the raw vehicle data to obtain filtered data; performing an adaptive dynamic threshold calculation based on the raw vehicle data to obtain a dynamic threshold; performing jump detection based on the acceleration change rate, filtered data, and dynamic threshold; if a jump occurs, the raw vehicle data at the previous moment is used as the corrected data; otherwise, the filtered data at the current moment is used as the corrected data. Multiple detections of sudden changes in acceleration and longitudinal driving force are achieved using the acceleration change rate, moving average filtering, and dynamic thresholds, and corrections are made based on the detection results, reducing interference with quality assessment and enhancing robustness under complex operating conditions.

[0008] In some embodiments, the raw vehicle data includes acceleration within a sliding window, and performing adaptive dynamic threshold calculation based on the raw vehicle data to obtain the dynamic threshold includes: calculating the acceleration standard deviation within the sliding window; and obtaining a dynamic acceleration threshold based on the acceleration standard deviation and a preset initial acceleration threshold. The dynamic acceleration threshold is determined based on the acceleration standard deviation and the initial acceleration threshold, and the acceleration threshold can be dynamically adjusted based on data fluctuations.

[0009] In some embodiments, obtaining a dynamic acceleration threshold based on the acceleration standard deviation and a preset initial acceleration threshold includes obtaining a maximum of half the acceleration standard deviation and the initial acceleration threshold as the dynamic acceleration threshold. The maximum of half the acceleration standard deviation and the initial acceleration threshold is used as the dynamic acceleration threshold. When data fluctuates significantly, the dynamic acceleration threshold can be automatically increased to reduce false positives. Conversely, under stable operating conditions, a lower detection sensitivity is maintained.

[0010] In some embodiments, the raw vehicle data includes longitudinal driving force within a sliding window, and performing adaptive dynamic threshold calculation based on the raw vehicle data to obtain the dynamic threshold includes: calculating a driving force standard deviation of the longitudinal driving force within the sliding window; and obtaining a dynamic driving force threshold based on the driving force standard deviation and a preset initial driving force threshold. The dynamic driving force threshold is determined based on the driving force standard deviation and the initial driving force threshold, and the driving force threshold can be dynamically adjusted based on data fluctuations.

[0011] In some embodiments, obtaining a dynamic driving force threshold based on the driving force standard deviation and a preset initial driving force threshold includes: obtaining a maximum value between half the driving force standard deviation and the initial driving force threshold as the dynamic driving force threshold. The maximum value between half the driving force standard deviation and the initial driving force threshold is used as the dynamic driving force threshold. The magnitude of the dynamic driving force threshold is adjusted based on data fluctuations. When data fluctuations are large, the dynamic acceleration threshold is automatically increased to reduce false positives. Conversely, under stable operating conditions, a lower detection sensitivity is maintained.

[0012] In some embodiments, performing jump detection based on the acceleration rate, filtered data, and a dynamic threshold includes: obtaining slope data based on the filtered data and the oldest raw vehicle data within a sliding window; performing jump detection based on the acceleration rate, the difference between the filtered data and the previous raw vehicle data relative to the dynamic threshold, and the slope data. Multiple checks are performed based on the acceleration rate, the change in the difference between the current filtered data and the previous raw vehicle data, and the slope to determine whether a jump has occurred.

[0013] In some embodiments, the filtered data includes filtered acceleration and filtered driving force, the dynamic threshold includes a dynamic acceleration threshold and a dynamic driving force threshold, the slope data includes an acceleration slope and a driving force slope, and the raw vehicle data also includes an accelerator pedal operating state. The jump detection based on the acceleration change rate, the difference between the filtered data and the raw vehicle data at the previous moment relative to the dynamic threshold, and the slope data includes: if the acceleration is greater than a first set value and the accelerator pedal is in an open state, a jump occurs if at least one of the following two conditions is met: Condition 1: The absolute difference between the filtered acceleration and the previous raw acceleration exceeds the dynamic acceleration threshold, the acceleration slope is greater than the second set value, and the acceleration rate of change is greater than the third set value. Condition 2: The absolute difference between the filtered driving force and the previous raw longitudinal driving force exceeds the dynamic driving force threshold, the driving force slope is greater than the fourth set value, and the acceleration rate of change is greater than the third set value. Multiple detection conditions are used to determine the difference between the filtered data and the dynamic threshold, the slope data, and the acceleration rate of change, resulting in more accurate judgment results.

[0014] In some embodiments, the corrected data is estimated using a first estimation algorithm and a second estimation algorithm, respectively, to obtain a first mass estimate and a second mass estimate. This includes: establishing a vehicle dynamics model based on the corrected data; inputting the vehicle dynamics model into an adaptive unscented Kalman filter algorithm to obtain the first mass estimate; and inputting the vehicle dynamics model into a double-forgetting factor recursive least squares algorithm to obtain the second mass estimate. The coordinated use of the two algorithms for mass estimation improves the stability and accuracy of the mass estimation results.

[0015] In some embodiments, inputting the vehicle dynamics model into a dual-forgetting-factor recursive least squares algorithm to obtain a second mass estimate includes adaptively adjusting a forgetting factor in the dual-forgetting-factor recursive least squares algorithm based on an acceleration state variable. Dynamically adjusting the forgetting factor based on acceleration changes can effectively address drastic and complex operating conditions and improve the stability of the estimation results.

[0016] In some embodiments, adaptively adjusting the forgetting factor of the dual-forgetting-factor recursive least squares algorithm based on the acceleration state variable includes: obtaining the acceleration change and maximum acceleration value within a sliding window; and adaptively adjusting the initial forgetting factor value based on the current acceleration magnitude using the acceleration change and maximum acceleration value. The current acceleration magnitude is used to distinguish whether the vehicle is experiencing a period of rapid acceleration change or a period of stable driving, and the forgetting factor is adjusted based on the acceleration change and maximum acceleration value, thereby improving the accuracy and reliability of the estimation results.

[0017] In some embodiments, the adaptive adjustment of the initial value of the forgetting factor based on the current acceleration magnitude using the acceleration change and the maximum acceleration value includes: if the current acceleration is greater than the first acceleration setting value, the forgetting factor is expressed as: ; ; If the current acceleration is less than the second acceleration setting value, the forgetting factor is expressed as: ; ;in, represents the initial value of the forgetting factor, Indicates the set control value. represents the coefficient, represents the acceleration change, Represents the maximum acceleration. During periods of rapid acceleration change, the forgetting factor is reduced to enhance the algorithm's rapid response and improve its adaptability to dynamic conditions. During periods of stable driving, the forgetting factor is increased to enhance the memory of historical data in the estimation results, suppress fluctuations in the estimation results, and improve estimation stability and accuracy.

[0018] In some embodiments, adaptive weighted fusion is performed based on the first mass estimate and the second mass estimate to obtain a mass estimation result for the current vehicle, including: if the adaptive unscented Kalman filter algorithm has reached a convergence state, respectively calculating the confidence levels corresponding to the first mass estimate and the second mass estimate; determining weights corresponding to the first mass estimate and the second mass estimate based on the confidence levels; and fusing the first mass estimate and the second mass estimate based on the weights to obtain a mass estimation result. By effectively combining the adaptive unscented Kalman filter algorithm, which favors global stability, and the double-forgetting factor recursive least squares algorithm, which favors rapid response, the two algorithms ensure both stability and accuracy in the mass estimation result.

[0019] In some embodiments, calculating the confidence levels corresponding to the first quality estimate and the second quality estimate, respectively, includes: calculating a first error covariance and a second error covariance corresponding to the first quality estimate and the second quality estimate, respectively; obtaining the inverse of the first error covariance and the inverse of the second error covariance, and using them as the first confidence level and the second confidence level, respectively. The error covariance reflects the uncertainty of the algorithm regarding the current result; the smaller the error covariance, the greater the confidence level, and the more reliable it is.

[0020] In the second aspect, the present application provides a vehicle mass estimation device, which includes: a data acquisition module for acquiring original vehicle data required for mass estimation; a jump processing module for performing jump processing on the original vehicle data to obtain corrected data; a mass estimation module for estimating the corrected data using a first estimation algorithm and a second estimation algorithm respectively to obtain a first mass estimation value and a second mass estimation value; a fusion module for performing adaptive weight fusion based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle. By performing jump detection on the original vehicle data by the jump processing module, it is possible to effectively identify and correct mutations and abnormal values ​​of the original vehicle data, reduce interference with quality assessment, enhance robustness under complex working conditions, and obtain high-precision and stable mass estimation results; the mass estimation values ​​of the two algorithms are adaptively weighted fused, and the weights of the mass estimation values ​​of the two algorithms are dynamically adjusted, further improving the accuracy and stability of the mass estimation results.

[0021] In a third aspect, the present application provides an electronic device, which includes a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the above-mentioned vehicle communication data encryption method.

[0022] In a fourth aspect, the present application provides a readable storage medium, in which computer program instructions are stored. When the computer program instructions are read and executed by a processor, the above-mentioned vehicle communication data encryption method is executed. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 A flow chart of a vehicle mass estimation method provided in an embodiment of the present application; Figure 2 A flowchart of the jump processing provided in an embodiment of the present application; Figure 3 A flow chart for obtaining a dynamic acceleration threshold value provided in an embodiment of the present application; Figure 4 A flowchart for obtaining a dynamic driving force threshold value provided in an embodiment of the present application; Figure 5 A specific flow chart of jump detection provided in an embodiment of the present application; Figure 6 A flowchart for obtaining the first quality estimation value and the second quality estimation value provided in an embodiment of the present application; Figure 7 A flowchart of adaptive adjustment of the forgetting factor provided in an embodiment of the present application; Figure 8 A flowchart for obtaining quality estimation results provided in an embodiment of the present application; Figure 9 A confidence calculation flow chart provided in an embodiment of the present application; Figure 10 A flowchart of a specific implementation of a vehicle mass estimation method based on jump value preprocessing and adaptive fusion provided in an embodiment of the present application; Figure 11 A structural block diagram of a vehicle mass estimation device provided in an embodiment of the present application; Figure 12 A schematic diagram of the jump processing module principle provided in an embodiment of the present application.

[0025] icon: 110 - data acquisition module; 120 - jump processing module; 130 - quality estimation module; 140 - fusion module. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0027] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0028] Under complex vehicle operating conditions such as start-stop, sudden acceleration, and sudden braking, the collected acceleration and driving force signals are prone to sudden changes and anomalies. Existing mass estimation methods lack the ability to dynamically adjust to vehicle state changes. During operation, they are unable to make real-time adjustments based on the vehicle's driving state, especially under these conditions. This leads to significant errors in specific operating conditions (such as sudden acceleration and braking). In particular, when the vehicle undergoes drastic dynamic changes, the estimation accuracy and robustness are insufficient, resulting in significant errors in the estimation results.

[0029] To address the above technical issues, an embodiment of the present application provides a vehicle mass estimation method that detects and corrects jumps in raw vehicle data. This method then uses the corrected data as input into two algorithms to obtain two mass estimates. These two mass estimates are then adaptively weighted and fused to obtain a mass estimation result for the current vehicle. This method effectively corrects for mutations and outliers in the raw vehicle data, effectively reducing the impact of mutations in the raw vehicle data on the estimation results, reducing interference with quality assessment, and enhancing robustness under complex operating conditions. Furthermore, adaptive weighted fusion is employed to obtain highly accurate and stable mass estimation results.

[0030] Please see Figure 1 , Figure 1 A flow chart of a vehicle mass estimation method provided in an embodiment of the present application, the method comprising the following steps: S110: Acquire original vehicle data required for mass estimation; S120: Perform jump processing on the original vehicle data to obtain corrected data; S130: Estimating the corrected data using a first estimation algorithm and a second estimation algorithm, respectively, to obtain a first quality estimation value and a second quality estimation value; S140: Perform adaptive weight fusion based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle.

[0031] For example, raw vehicle data, i.e., data required for mass estimation, includes but is not limited to acceleration, speed, longitudinal driving force, and accelerator pedal status (whether it is on or not, True / False). Raw vehicle data can be real-time vehicle data acquired via an actual vehicle bus (such as CAN) or simulated data acquired via the Carsim simulation platform. Corrected data, i.e., data used to input the algorithm to obtain the first and second mass estimates at the current moment, refers to the data after the jump if no jump occurs. If a jump occurs, the corrected data at the current moment refers to the raw vehicle data at the previous moment.

[0032] The original vehicle data is processed for jumps, effectively identifying and suppressing sudden changes and outliers from sensors or signal acquisition, reducing interference with the estimation algorithm. This allows the algorithm to maintain stable output of high-precision estimation results under conditions such as start-stop, rapid acceleration, and braking. Adaptively weighting the estimation results of the two algorithms leverages their respective strengths, improving the accuracy and stability of mass estimation.

[0033] Please see Figure 2 , Figure 2 This is a flow chart of jump processing. In some embodiments, jump processing is performed on original vehicle data to obtain corrected data, including: S150: Calculating acceleration change rate based on original vehicle data; S160: Perform moving average filtering on the original vehicle data to obtain filtered data; S170: Calculating an adaptive dynamic threshold based on the original vehicle data to obtain a dynamic threshold; S180: Perform jump detection based on acceleration change rate, filtered data, and dynamic threshold; S190: If a jump occurs, the original vehicle data at the previous moment is used as the correction data; otherwise, the filtered data at the current moment is used as the correction data.

[0034] Set the time interval between each two received data to t, the acceleration change rate can be expressed as: ratio_a=(a2-a1) / t; Among them, a2 represents the acceleration at the current moment, and a1 represents the acceleration at the previous moment.

[0035] The raw vehicle data is smoothed using a moving average filter to obtain the filtered acceleration filtered_a_x and filtered driving force filtered_F_X at the current moment, effectively reducing the impact of sensor noise.

[0036] Corrected data refers to the original vehicle data at the previous moment if a jump occurs, and the filtered data at the current moment otherwise.

[0037] Jump detection is performed based on acceleration change rate, filtered data and dynamic threshold. The robustness under complex working conditions is enhanced through jump processing.

[0038] Please see Figure 3 , Figure 3 The flowchart for obtaining a dynamic acceleration threshold value is as follows. In some embodiments, adaptive dynamic threshold value calculation is performed based on raw vehicle data to obtain the dynamic threshold value, including: S171: Calculate the acceleration standard deviation within the sliding window; S172: Obtain a dynamic acceleration threshold based on the acceleration standard deviation and a preset initial acceleration threshold.

[0039] The dynamic threshold includes a dynamic acceleration threshold and a dynamic driving force threshold. The original vehicle data includes the acceleration within the sliding window and the longitudinal driving force within the sliding window. Since the original vehicle data in this application is acquired in real time, a sliding window can be used to store the latest historical data, and the original vehicle data within the sliding window can be jump processed, such as the length of the sliding window N=10. In addition, a queue or array can be used to store the original vehicle data, and the original vehicle data in the queue or array is updated in real time, without any limitation here. Calculate the standard deviation of the original vehicle data (acceleration) within the sliding window. The standard deviation reflects the fluctuation of the data in the sliding window. The dynamic acceleration threshold can be obtained based on the fluctuation of the data.

[0040] In some embodiments, obtaining the dynamic acceleration threshold based on the acceleration standard deviation and a preset initial acceleration threshold includes: obtaining a maximum value of half the acceleration standard deviation and the initial acceleration threshold as the dynamic acceleration threshold.

[0041] For example, the initial acceleration threshold initial_threshold_a is set to 0.05, and the maximum value between half of the acceleration standard deviation std_a_x of the acceleration data recent_a_x in the sliding window and the initial acceleration threshold is taken as the dynamic acceleration threshold. The specific code is as follows: std_a_x=std(recent_a_x); threshold_a = max(initial_threshold_a,0.5 std_a_x); %Dynamic acceleration threshold.

[0042] In this way, when the data fluctuates greatly, the dynamic acceleration threshold can be automatically increased to reduce misjudgment; conversely, a lower detection sensitivity is maintained under stable working conditions.

[0043] Please see Figure 4 , Figure 4 This is a flow chart for obtaining a dynamic driving force threshold. In some embodiments, the original vehicle data includes the longitudinal driving force within a sliding window. An adaptive dynamic threshold calculation is performed based on the original vehicle data to obtain the dynamic threshold, including: S173: Calculating the driving force standard deviation of the longitudinal driving force within the sliding window; S174: Obtaining a dynamic driving force threshold based on the driving force standard deviation and a preset initial driving force threshold.

[0044] The standard deviation of the data (longitudinal driving force) within the sliding window is calculated. This calculation is conventional and will not be detailed here. The standard deviation measures the fluctuation of the longitudinal driving force within the sliding window. This fluctuation is used to determine the dynamic driving force threshold, improving the accuracy of the jump detection results.

[0045] In some embodiments, obtaining the dynamic driving force threshold based on the driving force standard deviation and a preset initial driving force threshold includes: obtaining a maximum value between half of the driving force standard deviation and the initial driving force threshold as the dynamic driving force threshold.

[0046] Here, the initial driving force threshold initial_threshold_F can be set to 300 for example. The maximum value between half of the standard deviation std_F_X of the longitudinal driving force recent_F_X in the sliding window and the initial driving force threshold is taken as the dynamic driving force threshold. The specific code is as follows: std_F_X=std(recent_F_X); threshold_F = max(initial_threshold_F, 0.5 std_F_X); % Dynamic driving force threshold.

[0047] The larger the driving force standard deviation, that is, the greater the fluctuation, the dynamic driving force threshold will also increase accordingly to reduce misjudgment; conversely, under stable working conditions, a lower detection sensitivity will be maintained.

[0048] Please see Figure 5 , Figure 5 This is a specific flow chart of jump detection. In some embodiments, jump detection is performed based on acceleration change rate, filtered data, and a dynamic threshold, including: S181: Obtain slope data based on the filtered data and the oldest original vehicle data in the sliding window; S182: Perform jump detection based on the acceleration change rate, the difference between the filtered data and the original vehicle data at the previous moment relative to the dynamic threshold, and the slope data.

[0049] The filtered data includes filtered acceleration and filtered driving force. The slope data includes acceleration slope slope_a_x and driving force slope slope_F_X. The slope data reflects the changing trend of the data within the sliding window and can be obtained based on the filtered data and the oldest original vehicle data within the sliding window. The specific formula is as follows: slope_a_x=(filtered_a_x-recent_a_x) / N; slope_F_X =(filtered_F_X-recent_F_X) / N; Where N is the length of the sliding window, recent_a_x is the oldest acceleration in the sliding window, filtered_a_x is the filtered acceleration at the current moment, recent_F_X is the oldest longitudinal driving force in the sliding window, and filtered_F_X is the filtered driving force at the current moment.

[0050] The accuracy of the jump detection results is ensured through multiple tests of the acceleration change rate, the difference between the filtered data and the original vehicle data at the previous moment relative to the dynamic threshold, and the slope data.

[0051] In some embodiments, jump detection is performed based on the acceleration change rate, the difference between the filtered data and the original vehicle data at the previous moment relative to the dynamic threshold, and the slope data, including: If the acceleration is greater than the first set value and the accelerator pedal is in the open state, a jump occurs if at least one of the following two conditions is met: Condition 1: The absolute difference between the filtered acceleration and the original acceleration at the previous moment exceeds the dynamic acceleration threshold, the acceleration slope is greater than the second set value, and the acceleration change rate is greater than the third set value; Condition 2: The absolute difference between the filtered driving force and the original longitudinal driving force at the previous moment exceeds the dynamic driving force threshold, the driving force slope is greater than the fourth set value, and the acceleration change rate is greater than the third set value.

[0052] For example, when the acceleration is greater than 0.2 and the accelerator pedal is in the open state, A: the absolute difference between the filtered acceleration and the raw acceleration at the previous moment exceeds the dynamic acceleration threshold, the acceleration slope is greater than 0.01, and the acceleration change rate ratio_a>0.02; B: The absolute difference between the filtered driving force and the raw longitudinal driving force at the previous moment exceeds the dynamic driving force threshold, the driving force slope is greater than 10, and the acceleration change rate ratio_a is greater than 0.02; If at least one condition is met, it is considered that a jump has occurred, the current data point will be skipped, and the original vehicle data of the previous moment will continue to be used; otherwise, the current filtering result will be accepted.

[0053] The specific code is as follows: if a_x>0.2&&ratio_a>0.02&&gas# The accelerator pedal is on, the acceleration is greater than 0.2, and the acceleration change rate is greater than 0.02; % Detect the jump of acceleration and driving force if (abs(filtered_a_x - prev_a_x)>threshold_a&&abs(slope_a_x)>0.01) || (abs(filtered_F_X-prev_F_X)>threshold_F&&abs(slope_F_X)>10) % If a jump occurs, skip the current data point and use the previous value; a_x_prev = prev_a_x; F_X_prev = prev_F_X; else % No jump occurs, update processing to the current input value; a_x_prev = filtered_a_x; F_X_prev = filtered_F_X.

[0054] By setting multiple detection conditions such as moving average filtering, dynamic threshold, slope detection, etc., it is possible to accurately detect sudden changes and outliers in the original vehicle data, correct the original vehicle data, reduce interference with quality assessment, enhance robustness under complex working conditions, and provide more stable and robust basic data for online estimation of vehicle quality.

[0055] Please see Figure 6 , Figure 6 This is a flowchart for obtaining the first quality estimate and the second quality estimate. In some embodiments, the correction data is estimated using the first estimation algorithm and the second estimation algorithm respectively to obtain the first quality estimate and the second quality estimate, including: S131: Establishing a vehicle dynamics model based on the correction data; S132: Inputting the vehicle dynamics model into an adaptive unscented Kalman filter algorithm to obtain a first mass estimate; S133: Input the vehicle dynamics model into a double-forgetting factor recursive least squares algorithm to obtain a second mass estimation value.

[0056] The vehicle dynamics model is established based on the corrected longitudinal driving force and acceleration: ; in, m represents the total mass of the vehicle, represents acceleration, represents the longitudinal driving force, represents the air density, represents the drag coefficient, A represents the frontal area, represents the longitudinal velocity, f represents the rolling resistance coefficient, g represents the acceleration due to gravity, Indicates the slope (this application is for quality assessment on a flat road, so here is 0).

[0057] For nonlinear state estimation based on Kalman filtering, for example, the Adaptive Unscented Kalman Filter (AUKF) algorithm can be used. This algorithm can perform nonlinear estimation of vehicle states and can be used to handle nonlinear dynamics during vehicle operation, especially in complex driving conditions such as cornering and braking, enabling accurate tracking of mass estimation. The AUKF algorithm is directly used here, primarily leveraging its powerful nonlinear modeling capabilities to compensate for the shortcomings of the improved RLS-MFF algorithm in terms of dynamic applicability.

[0058] For weighted estimation based on recursive least squares, for example, a double forgetting factor recursive least squares algorithm (RLS-MFF) can be used. The algorithm is based on the recursive least squares (RLS) method and uses a double forgetting factor ( 、 ) dynamically adjusts the weight of historical data to achieve real-time weighted estimation of quality. This algorithm can effectively cope with complex changes in different driving environments and improve estimation stability. In the conventional RLS-MFF algorithm, the forgetting factor is generally used as a fixed parameter, for example, set to =0.99 and =0.9, in order to better adapt to the changes in different working conditions, here it is improved to an adaptive parameter adjustment strategy based on acceleration state variables, which can dynamically adjust the parameters according to the acceleration state variables. 、 The dual-algorithm collaborative mechanism of the AUKF algorithm and the improved RLS-MFF algorithm not only takes into account the stability of the global estimation, but also has fast convergence and online adaptive learning capabilities, greatly improving the accuracy and stability of quality estimation.

[0059] In some embodiments, inputting the vehicle dynamics model into a double forgetting factor recursive least squares algorithm to obtain a second mass estimate includes adaptively adjusting a forgetting factor of the double forgetting factor recursive least squares algorithm based on an acceleration state variable.

[0060] Specifically, the size of the forgetting factor is dynamically adjusted according to the real-time size of the acceleration. Compared with the fixed forgetting factor in the traditional RLS-MFF algorithm, adaptive adjustment can provide more suitable parameters for the algorithm under different working conditions, enabling the algorithm to better track the dynamic characteristics of the vehicle, and then more accurately estimate the vehicle mass in different states, allowing the algorithm to adapt to different working conditions and keep the output results stable and reliable.

[0061] Please see Figure 7 , Figure 7 FIG. 4 is a flowchart of adaptive adjustment of the forgetting factor. In some embodiments, adaptive adjustment of the forgetting factor of the double forgetting factor recursive least squares algorithm based on the acceleration state variable includes: S134: Obtain the acceleration change and the maximum acceleration value within the sliding window; S135: Based on the current acceleration magnitude, the initial value of the forgetting factor is adaptively adjusted using the acceleration change and the maximum acceleration value.

[0062] The data in the sliding window is constantly changing, so the acceleration change can be calculated based on the acceleration of two consecutive moments (the current moment and the previous moment) ( ), and the maximum acceleration at the current moment; the current acceleration magnitude represents the state of the vehicle, such as strong acceleration or smooth driving. The initial value of the forgetting factor can be adaptively adjusted according to the current state of the vehicle to change the size of the forgetting factor, thereby improving the accuracy and reliability of the second mass estimate.

[0063] In some embodiments, based on the current acceleration magnitude, the initial value of the forgetting factor is adaptively adjusted using the acceleration change and the maximum acceleration value, including: If the current acceleration is greater than the first acceleration setting value, the forgetting factor is expressed as: ; ; If the current acceleration is less than the second acceleration setting value, the forgetting factor is expressed as: ; ; in, represents the initial value of the forgetting factor, Indicates the set control value. represents the coefficient, represents the change in acceleration, Indicates the maximum acceleration.

[0064] If the current acceleration is greater than the first acceleration setting value, that is, the acceleration is more obvious, indicating that it is in a fierce driving or strong acceleration stage; if the current acceleration is less than the second acceleration setting value, that is, a slight speed change, it means that it is in a stable driving or slow acceleration stage. For example, the coefficient k Set to 0.5, (lambda_1 in the following code), The initial value of (lambda_2 in the following code) Set to 0.99 and 0.9 respectively, 0.005, the maximum acceleration Corresponding to a_max in the following code, the specific code is as follows: % Dynamically adjust the forgetting factor according to the acceleration a_x, coefficient k (k in the following code) is set to 0.5 if a_x>1# When the current acceleration a_x is greater than 1 m / s² (more obvious acceleration), it indicates that the vehicle is in a fierce driving or strong acceleration phase; lambda_1 = 0.995- k (|Δa| / a_max); lambda_2 = 0.905- k (|Δa| / a_max); else if a_x<0.5# When the current acceleration a_x is less than 0.5 m / s² (slight speed change), it indicates that the vehicle is in a stable driving or slow acceleration phase; lambda_1 = 0.985 + k (|Δa| / a_max); lambda_2 = 0.895 + k (|Δa| / a_max).

[0065] During the intense change phase (high acceleration), reduce 、 The value of increases the weight of the new data on the estimation results, so that the algorithm can quickly respond to changes in vehicle status and improve its adaptability to dynamic conditions. In the stable driving stage (low acceleration), increase 、 The value of can enhance the memory of historical data on the estimation results, suppress the fluctuation of the estimation results, and improve the stability and accuracy of the estimation.

[0066] Please see Figure 8 , Figure 8 FIG. 1 is a flow chart for obtaining a mass estimation result. In some embodiments, adaptive weight fusion is performed based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle, including: S141: If the adaptive unscented Kalman filter algorithm is in a converged state, respectively calculating the confidences corresponding to the first quality estimate and the second quality estimate; S142: Determine a weight based on the confidence level; S143: The first quality estimation value and the second quality estimation value are fused based on the weights to obtain a quality estimation result.

[0067] If the adaptive unscented Kalman filter (AUCF) algorithm has converged, the quality estimates of the two algorithms are weighted and fused. If the AUCF algorithm has not converged, only the estimate from the double-forgetting factor recursive least squares algorithm is used as the final quality estimate. Because different algorithms have different system modeling and error handling capabilities, the RLS-MFF algorithm tends to be fast response, while the AUKF algorithm tends to be globally stable. Therefore, a heterogeneous fusion mechanism based on error covariance perception is adopted to effectively combine the advantages of both algorithms.

[0068] In some embodiments, before the steps of respectively calculating the error covariance and confidence corresponding to the first quality estimation value and the second quality estimation value, the method further includes: obtaining a quality estimation residual based on the first quality estimation value at the current moment and the first quality estimation value at the previous moment: if the quality estimation residual is less than a set threshold, the adaptive unscented Kalman filtering algorithm is in a convergence state.

[0069] The first mass estimation value mass_aukf obtained by the AUKF algorithm is used to obtain the mass estimation residual Δmass_aukf. The mass estimation residual is used to determine whether the AUKF algorithm has entered the convergence state. If the mass estimation residual is less than the set threshold, it means that it has converged. The first mass estimation value is in a relatively stable state and can be used as a quality estimation, ensuring the stability and accuracy of the estimation result.

[0070] Please see Figure 9 , Figure 9This is a confidence calculation flow chart. In some embodiments, respectively calculating the confidences corresponding to the first quality estimate and the second quality estimate includes: S144: Calculate a first error covariance and a second error covariance corresponding to the first quality estimation value and the second quality estimation value respectively; S145: Obtain the inverse of the first error covariance and the inverse of the second error covariance, and use them as the first confidence level and the second confidence level, respectively.

[0071] The error covariance (first error covariance var_aukf, second error covariance var_rls) reflects the algorithm's "uncertainty" of the current quality estimate. Code example: # Calculate the covariance of each estimation error var_rls = calculate_mass_var_rls(); var_aukf = calculate_mass_ var_aukf(); # Assign weights based on confidence weight_rls = 1 / var_rls; weight_aukf = 1 / var_aukf.

[0072] The inverse of the error covariance is used to represent the confidence level (the first estimate corresponds to the confidence level weight_aukf, and the second estimate corresponds to the confidence level weight_rls). The smaller the error covariance, the greater the confidence level, the more credible it is, the greater the weight should be, and the more accurate the fusion result will be.

[0073] In some embodiments, determining the weight based on the confidence level includes: normalizing the first confidence level to obtain a first weight; and normalizing the second confidence level to obtain a second weight. For example: For example, the Softmax function can be used to normalize the above confidence. The code example is: normalized_weights =softmax([weight_rls, weight_aukf]).

[0074] To avoid fusion instability caused by large differences in the absolute values ​​of the weights, normalization is performed to prevent a single estimate from completely dominating the fusion process. The error covariance between the first and second estimates at the current moment is used to adjust the weights of each algorithm in real time, improving the algorithm's dynamic responsiveness.

[0075] In some embodiments, fusing the first quality estimation value and the second quality estimation value based on the weight to obtain a quality estimation result includes: The quality estimation result is expressed as: ; in, represents the first quality estimate, represents the first weight corresponding to the first quality estimate, represents the second quality estimate, Indicates a second weight corresponding to the second quality estimation value.

[0076] Perform weighted fusion processing on mass_rls and mass_aukf obtained by calculating the improved RLS-MFF and AUKF algorithms to obtain the final quality estimation result. Sample code: # Fusion estimation results fused_mass=normalized_weights[0] mass_rls+normalized_weights[1] mass_aukf.

[0077] Since the original vehicle data is acquired in real time, the jump detection and quality estimation process adopts a recursive calculation structure, which does not require the full amount of historical data and is suitable for the real-time processing requirements of the vehicle control system. This method has good real-time performance and online adaptability.

[0078] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be described clearly and completely below. Figure 10 , Figure 10 The flowchart of the specific implementation of the vehicle mass estimation method based on jump value preprocessing and adaptive fusion includes the following steps: S201: Data acquisition and input: Acquire key data such as acceleration a, longitudinal driving force Tq, velocity v, etc. as input sources for mass estimation; S202: Jump value preprocessing: After data is received, jump value preprocessing is performed immediately to identify and correct abnormal jumps in the acceleration and driving force signals, filter out noise interference, and improve data quality. If no jump occurs, the memory variables are updated. Specifically, persistent variables (prev_a_x, prev_F_X) are used to store the acceleration and longitudinal driving force data at the previous moment. S203: Mass estimation algorithm fusion processing: The pre-processed data is used to construct a vehicle dynamics model, and the improved RLS-MFF algorithm is combined with the AUKF algorithm to achieve high-precision and real-time estimation of vehicle mass.

[0079] This method is also highly scalable, allowing for the integration of additional information sources (such as model-based estimates and inertial navigation system outputs) to build a more comprehensive multi-source perception system. Accurate and reliable vehicle mass estimation provides key parameters for control systems (such as ESP and ACC) and energy management (particularly in electric and hybrid vehicles). It can also assist in identifying abnormal vehicle loads, monitoring vehicle health, and providing safety warnings, ultimately enhancing vehicle intelligence and operational safety.

[0080] Please see Figure 11 , Figure 11 This is a structural block diagram of a vehicle mass estimation device provided in this application. It should be understood that the device is Figure 1 The method embodiment executed in the embodiment corresponds to the method, and can execute the steps involved in the aforementioned method. The specific functions of the device can be found in the description above. To avoid repetition, detailed description is appropriately omitted here. The device includes but is not limited to: A data acquisition module 110 is used to acquire original vehicle data required for mass estimation; The jump processing module 120 is used to perform jump processing on the original vehicle data to obtain corrected data; A quality estimation module 130 is configured to estimate the correction data using a first estimation algorithm and a second estimation algorithm to obtain a first quality estimation value and a second quality estimation value; The fusion module 140 is configured to perform adaptive weight fusion based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle.

[0081] In the technical solution of the embodiment of the present application, jump detection is used to accurately identify and correct sudden changes in acceleration and driving force signals, providing more stable and robust basic data for subsequent online estimation of vehicle quality; adaptive weight fusion is used to fuse the results of the two algorithms, making the results more stable and accurate.

[0082] According to some embodiments of the present application, the data acquisition module 110 is specifically used to obtain data required for mass estimation, such as acceleration, speed, longitudinal driving force, and accelerator pedal working status (whether it is on).

[0083] According to some embodiments of the present application, the jump processing module 120 is specifically used to: calculate the acceleration change rate based on the original vehicle data; perform moving average filtering on the original vehicle data to obtain filtered data; perform adaptive dynamic threshold calculation based on the original vehicle data to obtain a dynamic threshold; perform jump detection based on the acceleration change rate, filtered data and dynamic threshold: if a jump occurs, use the original vehicle data at the previous moment, otherwise use the filtered data at the current moment, such as Figure 12As shown, it is a schematic diagram of the jump processing module 120, which performs jump detection based on the acceleration change rate ratio_a, acceleration a, longitudinal driving force Tq, and accelerator pedal working state Gas / gas as input parameters, thereby determining whether the acceleration a and the longitudinal driving force Tq have jumped, and obtaining corrected data of the acceleration a and the longitudinal driving force Tq.

[0084] According to some embodiments of the present application, the specific dynamic threshold acquisition includes a dynamic acceleration threshold and a dynamic driving force threshold, wherein the specific acquisition process of the dynamic acceleration threshold is: The acceleration standard deviation within the sliding window is calculated; and a dynamic acceleration threshold is obtained based on the acceleration standard deviation and a preset initial acceleration threshold.

[0085] According to some embodiments of the present application, the dynamic acceleration threshold may be specifically determined by obtaining a maximum value between half of the acceleration standard deviation and the initial acceleration threshold and using the maximum value as the dynamic acceleration threshold.

[0086] According to some embodiments of the present application, the specific process of obtaining the dynamic driving force threshold is: calculating the driving force standard deviation of the longitudinal driving force within the sliding window; and obtaining the dynamic driving force threshold based on the driving force standard deviation and a preset initial driving force threshold.

[0087] According to some embodiments of the present application, the dynamic driving force threshold may be specifically determined by obtaining a maximum value between half of the driving force standard deviation and the initial driving force threshold as the dynamic driving force threshold.

[0088] According to some embodiments of the present application, the jump processing module 120 is specifically used to: obtain slope data based on the filtered data and the oldest original vehicle data in the sliding window; perform jump detection based on the acceleration change rate, the difference between the filtered data and the original vehicle data at the previous moment relative to the dynamic threshold, and the slope data.

[0089] According to some embodiments of the present application, the specific detection condition is: if the acceleration is greater than a first set value and the accelerator pedal is in the open state, a jump occurs if at least one of the following two conditions is met: Condition 1: The absolute difference between the filtered acceleration and the previous raw acceleration exceeds the dynamic acceleration threshold, the acceleration slope is greater than the second set value, and the acceleration change rate is greater than the third set value; Condition 2: The absolute difference between the filtered driving force and the original longitudinal driving force at the previous moment exceeds the dynamic driving force threshold, the driving force slope is greater than the fourth set value, and the acceleration change rate is greater than the third set value.

[0090] According to some embodiments of the present application, the mass estimation module 130 is specifically used to: establish a vehicle dynamics model based on the corrected data; input the vehicle dynamics model into an adaptive unscented Kalman filter algorithm to obtain a first mass estimation value; input the vehicle dynamics model into a double forgetting factor recursive least squares algorithm to obtain a second mass estimation value.

[0091] According to some embodiments of the present application, the specific adjustment method of the double forgetting factor is: The acceleration change and the maximum acceleration value within the sliding window are obtained; based on the current acceleration magnitude, the initial value of the forgetting factor is adaptively adjusted using the acceleration change and the maximum acceleration value.

[0092] According to some embodiments of the present application, the specific calculation formula of the forgetting factor is: If the current acceleration is greater than the first acceleration setting value, the forgetting factor is expressed as: ; ; If the current acceleration is less than the second acceleration setting value, the forgetting factor is expressed as: ; ; in, represents the initial value of the forgetting factor, Indicates the set control value. represents the coefficient, represents the change in acceleration, Indicates the maximum acceleration.

[0093] According to some embodiments of the present application, the fusion module 140 is specifically used to: if the adaptive unscented Kalman filter algorithm is in a convergence state, then respectively calculate the confidence corresponding to the first quality estimation value and the second quality estimation value; determine the weight based on the confidence; and fuse the first quality estimation value and the second quality estimation value based on the weight to obtain a quality estimation result.

[0094] According to some embodiments of the present application, the convergence judgment condition is: obtaining a quality estimation residual based on the first quality estimation value at the current moment and the first quality estimation value at the previous moment: if the quality estimation residual is less than a set threshold, the adaptive unscented Kalman filter algorithm is in a convergence state.

[0095] According to some embodiments of the present application, the specific calculation method of the confidence is: respectively calculate the first error covariance and the second error covariance corresponding to the first quality estimate value and the second quality estimate value; obtain the inverse of the first error covariance and the inverse of the second error covariance, and use them as the first confidence level and the second confidence level, respectively.

[0096] According to some embodiments of the present application, the specific calculation method of the weight is: normalizing the first confidence to obtain the first weight; normalizing the second confidence to obtain the second weight.

[0097] According to some embodiments of the present application, the calculation formula for the quality estimation result is: ;in, represents the first quality estimate, represents the first weight corresponding to the first quality estimate, represents the second quality estimate, Indicates a second weight corresponding to the second quality estimation value.

[0098] The present application provides an electronic device, which includes a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the method in any of the aforementioned optional implementations.

[0099] The present application provides a readable storage medium, which stores computer program instructions. When the computer program instructions are read and executed by a processor, the method in any of the aforementioned optional implementations is executed.

[0100] The storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0101] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions and operations of the devices, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or can be implemented using a combination of dedicated hardware and computer instructions.

[0102] In addition, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0103] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard drives, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0104] The foregoing is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included within the scope of protection of the present application. It should be noted that similar reference numerals and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined or explained in subsequent figures.

[0105] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

[0106] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

Claims

1. A vehicle mass estimation method, characterized in that: The method comprises: Obtaining raw vehicle data required for mass estimation; Performing jump processing on the original vehicle data to obtain corrected data; Estimating the correction data using a first estimation algorithm and a second estimation algorithm respectively to obtain a first quality estimation value and a second quality estimation value; Adaptive weight fusion is performed based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle.

2. The vehicle mass estimation method according to claim 1, characterized in that: The step of performing jump processing on the original vehicle data to obtain corrected data includes: calculating an acceleration rate of change based on the raw vehicle data; Performing moving average filtering on the original vehicle data to obtain filtered data; Performing adaptive dynamic threshold calculation based on the original vehicle data to obtain a dynamic threshold; Performing jump detection based on the acceleration change rate, filtered data, and a dynamic threshold; If a jump occurs, the original vehicle data at the previous moment is used as the revised data; otherwise, the filtered data at the current moment is used as the revised data.

3. The vehicle mass estimation method according to claim 2, characterized in that: The original vehicle data includes acceleration within a sliding window, and the adaptive dynamic threshold calculation based on the original vehicle data to obtain the dynamic threshold includes: Calculating the acceleration standard deviation within the sliding window; A dynamic acceleration threshold is obtained based on the acceleration standard deviation and a preset initial acceleration threshold.

4. The vehicle mass estimation method according to claim 3, characterized in that: The obtaining of a dynamic acceleration threshold based on the acceleration standard deviation and a preset initial acceleration threshold comprises: A maximum value between half of the acceleration standard deviation and the initial acceleration threshold is obtained and used as the dynamic acceleration threshold.

5. The vehicle mass estimation method according to claim 2, characterized in that: The original vehicle data includes a longitudinal driving force within a sliding window, and the adaptive dynamic threshold calculation based on the original vehicle data to obtain the dynamic threshold includes: calculating a driving force standard deviation of the longitudinal driving force within the sliding window; A dynamic driving force threshold is obtained based on the driving force standard deviation and a preset initial driving force threshold.

6. The vehicle mass estimation method according to claim 5, characterized in that: The obtaining of a dynamic driving force threshold based on the driving force standard deviation and a preset initial driving force threshold includes: A maximum value between half of the driving force standard deviation and the initial driving force threshold is obtained and used as the dynamic driving force threshold.

7. The vehicle mass estimation method according to claim 2, characterized in that: The performing jump detection based on the acceleration change rate, the filtered data and the dynamic threshold comprises: Obtaining slope data based on the filtered data and the oldest original vehicle data within the sliding window; Jump detection is performed based on the acceleration change rate, the difference between the filtered data and the original vehicle data at the previous moment relative to the dynamic threshold, and the slope data.

8. The vehicle mass estimation method according to claim 7, characterized in that: The filtered data includes filtered acceleration and filtered driving force, the dynamic threshold includes a dynamic acceleration threshold and a dynamic driving force threshold, the slope data includes an acceleration slope and a driving force slope, and the original vehicle data also includes an accelerator pedal operating state. The jump detection based on the acceleration change rate, the difference between the filtered data and the original vehicle data at the previous moment relative to the dynamic threshold, and the slope data includes: If the acceleration is greater than the first set value and the accelerator pedal is in the open state, a jump occurs if at least one of the following two conditions is met: Condition 1: The absolute difference between the filtered acceleration and the original acceleration at the previous moment exceeds the dynamic acceleration threshold, the acceleration slope is greater than the second set value, and the acceleration change rate is greater than the third set value; Condition 2: The absolute difference between the filtered driving force and the original longitudinal driving force at the previous moment exceeds the dynamic driving force threshold, the driving force slope is greater than the fourth set value, and the acceleration change rate is greater than the third set value.

9. The vehicle mass estimation method according to claim 1, characterized in that: The estimating the correction data using the first estimation algorithm and the second estimation algorithm respectively to obtain the first quality estimation value and the second quality estimation value includes: establishing a vehicle dynamics model based on the correction data; Inputting the vehicle dynamics model into an adaptive unscented Kalman filter algorithm to obtain a first mass estimate; The vehicle dynamics model is input into a double-forgetting factor recursive least squares algorithm to obtain a second mass estimation value.

10. The vehicle mass estimation method according to claim 9, characterized in that: Inputting the vehicle dynamics model into a double-forgetting factor recursive least squares algorithm to obtain a second mass estimate includes: The forgetting factor of the double forgetting factor recursive least squares algorithm is adaptively adjusted based on the acceleration state variable.

11. The vehicle mass estimation method according to claim 10, characterized in that: The adaptively adjusting the forgetting factor of the double-forgetting-factor recursive least squares algorithm based on the acceleration state variable includes: Get the acceleration change and maximum acceleration within the sliding window; Based on the current acceleration magnitude, the initial value of the forgetting factor is adaptively adjusted using the acceleration variation and the maximum acceleration value.

12. The vehicle mass estimation method according to claim 11, characterized in that: The adaptive adjustment of the initial value of the forgetting factor based on the current acceleration magnitude and utilizing the acceleration change and the maximum acceleration value includes: If the current acceleration is greater than the first acceleration setting value, the forgetting factor is expressed as: ； ； If the current acceleration is less than the second acceleration setting value, the forgetting factor is expressed as: ； ； in, represents the initial value of the forgetting factor, Indicates the set control value. represents the coefficient, represents the acceleration change, Indicates the maximum value of the acceleration.

13. The vehicle mass estimation method according to claim 1 or 9, characterized in that: The performing adaptive weight fusion based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle includes: If the adaptive unscented Kalman filter algorithm is in a converged state, respectively calculating confidences corresponding to the first quality estimate and the second quality estimate; Determining weights corresponding to the first quality estimation value and the second quality estimation value respectively based on the confidence level; The first quality estimation value and the second quality estimation value are fused based on the weight to obtain a quality estimation result.

14. The vehicle mass estimation method according to claim 13, characterized in that: The respectively calculating the confidences corresponding to the first quality estimation value and the second quality estimation value includes: Calculating a first error covariance and a second error covariance corresponding to the first quality estimation value and the second quality estimation value respectively; The inverse of the first error covariance and the inverse of the second error covariance are obtained and used as the first confidence level and the second confidence level, respectively.

15. A vehicle mass estimation device, characterized in that: The device comprises: A data acquisition module, used to obtain raw vehicle data required for mass estimation; a jump processing module, configured to perform jump processing on the original vehicle data to obtain corrected data; A quality estimation module, configured to estimate the correction data using a first estimation algorithm and a second estimation algorithm, respectively, to obtain a first quality estimation value and a second quality estimation value; A fusion module is used to perform adaptive weight fusion based on the first mass estimation value and the second mass estimation value to obtain a mass estimation result of the current vehicle.

16. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to perform the vehicle mass estimation method according to any one of claims 1 to 14.

17. A readable storage medium, characterized in that The readable storage medium stores computer program instructions, and when the computer program instructions are read and executed by a processor, the vehicle mass estimation method according to any one of claims 1 to 14 is executed.

Images