Vehicle mass estimation method based on fuzzy logic and weighted least squares

Abstract

The present application belongs to the technical field of vehicle mass online estimation, in particular to a vehicle mass online estimation method based on fuzzy logic and weighted least square method, comprising the following steps: 1) obtaining longitudinal acceleration sensor measurement value signal, longitudinal speed signal, engine output torque signal, gear signal, throttle opening signal, brake pedal signal and steering wheel angle signal from vehicle CAN bus; 2) correcting the measurement value of longitudinal acceleration sensor; 3) calculating equivalent driving force and equivalent acceleration; 4) calculating working condition fitness degree by using fuzzy logic algorithm; 5) estimating vehicle mass by using weighted least square method; and 6) suspending and starting vehicle mass estimation. The structure is reasonable, and the accuracy and working condition adaptability of vehicle mass online estimation can be improved during use.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle mass online estimation, in particular to a vehicle mass online estimation method based on fuzzy logic and weighted least squares. BACKGROUND

[0002] With the further improvement of vehicle intelligence and the increasing demand for active control, accurate vehicle mass information is of great significance and application value for the decision of vehicle dynamics control.

[0003] The existing mass estimation method directly takes the measurement value of the longitudinal acceleration sensor as the longitudinal acceleration of the whole vehicle. However, since the longitudinal acceleration sensor is generally installed on the vehicle body, the pitching motion of the vehicle body will cause an acceleration component of the gravity acceleration in the measurement direction of the sensor, resulting in a deviation between the measurement value of the acceleration sensor and the actual acceleration of the whole vehicle. In addition, the mass estimation method based on longitudinal dynamics is only applicable to simple straight acceleration conditions. However, the actual driving conditions of the vehicle are complex, and the mass estimation error under steering and braking conditions is large. Therefore, it is necessary to eliminate the interference of steering and braking and improve the working condition applicability of the mass estimation method.

[0004] Based on the above problems, a vehicle mass online estimation method based on fuzzy logic and weighted least squares is proposed. SUMMARY

[0005] This section aims to summarize some aspects of the embodiments of the present application and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of the specification to avoid obscuring the purpose of this section, abstract and title. Such simplifications or omissions cannot be used to limit the scope of the present application.

[0006] In view of the problems in the prior art, the present application is proposed.

[0007] Therefore, the purpose of the present application is to provide a vehicle mass online estimation method based on fuzzy logic and weighted least squares, which can improve the accuracy and working condition adaptability of vehicle mass online estimation during use.

[0008] To solve the above technical problems, according to one aspect of the present application, the present application provides the following technical scheme:

[0009] The vehicle mass online estimation method based on fuzzy logic and weighted least squares comprises the following steps:

[0010] Step 1: obtaining the longitudinal acceleration sensor measurement value signal, longitudinal velocity signal, engine output torque signal, gear signal, throttle opening signal, brake pedal signal and steering wheel angle signal from the vehicle CAN bus;

[0011] Step 2: Correcting the measurement of longitudinal acceleration sensor;

[0012] Step 3: Calculating equivalent driving force and equivalent acceleration;

[0013] Step 4: Calculating the working condition fitness degree by using fuzzy logic algorithm;

[0014] Step 5: Estimating the vehicle mass by using weighted least square method;

[0015] Step 6: Suspension and start of vehicle mass estimation.

[0016] As a preferred scheme of the vehicle mass online estimation method based on fuzzy logic and weighted least square method according to the present application, wherein: the step 2 specifically comprises the following steps:

[0017] Correcting the measurement of longitudinal acceleration sensor according to formula (1);

[0018]

[0019] wherein, a xrev is the corrected measurement of longitudinal acceleration sensor, a xsen is the measurement of longitudinal acceleration sensor, v x is the longitudinal speed, is the first order derivative of longitudinal speed, k acc is the acceleration correction coefficient, k gra is the slope correction coefficient, k air is the air resistance correction coefficient.

[0020] As a preferred scheme of the vehicle mass online estimation method based on fuzzy logic and weighted least square method according to the present application, wherein: the step 3 specifically comprises the following steps:

[0021] (1): Defining the equivalent driving force as:

[0022]

[0023] wherein, η t is the transmission system efficiency, M e is the engine output torque, i g is the transmission ratio of gearbox, i0 is the transmission ratio of main reducer, r d is the wheel rolling radius, Θ i is the equivalent rotational mass moment of inertia, C D is the air resistance coefficient, A is the reference area, ρ a is the air density;

[0024] (2): Defining the equivalent acceleration as:

[0025] a res =a xrev +gf R (11)

[0026] Where g is the acceleration due to gravity, f R a is the rolling resistance coefficient. xrev The longitudinal acceleration sensor measurement is the modified value defined by equation (1).

[0027] As a preferred embodiment of the online vehicle mass estimation method based on fuzzy logic and weighted least squares described in this invention, the rolling resistance coefficient can be expressed as:

[0028]

[0029] Where f0 is the constant term of the rolling resistance coefficient, f1 is the coefficient of the first term, and f4 is the coefficient of the fourth term. As can be seen from equation (4), the rolling resistance coefficient is determined by the longitudinal velocity.

[0030] As a preferred embodiment of the online vehicle mass estimation method based on fuzzy logic and weighted least squares described in this invention, in step 4, the inputs of the fuzzy logic algorithm include equivalent acceleration, equivalent driving force, vehicle speed, steering wheel angle, rate of change of equivalent acceleration, rate of change of equivalent driving force, and rate of change of steering wheel angle. The input values ​​are converted into fuzzy values ​​through fuzzy membership functions. Based on the corresponding fuzzy rules, the Sugeno model is adopted, and the centroid method is used to defuzzify and obtain the accurate value of the working condition suitability.

[0031] As a preferred embodiment of the online vehicle mass estimation method based on fuzzy logic and weighted least squares described in this invention, step 5 specifically includes the following steps:

[0032] The equivalent driving force and equivalent acceleration obtained in step 3 are weighted using the working condition suitability obtained in step 4, and the output of the least squares method is defined as:

[0033] z = wF res (13)

[0034] Where w is the suitability of the working condition obtained in step 4, and F res The equivalent driving force defined by equation (2);

[0035] The input for least squares is:

[0036]

[0037] Among them, a resThe equivalent acceleration defined by equation (3);

[0038] The parameters to be identified using the least squares method are:

[0039] θ=m (15)

[0040] Where m is the mass of the vehicle to be identified;

[0041] The vehicle mass is estimated using the least squares method (8):

[0042]

[0043] Where λ is the time-varying forgetting factor and P is the recursive covariance matrix.

[0044] As a preferred embodiment of the online vehicle mass estimation method based on fuzzy logic and weighted least squares described in this invention, step 6 includes: determining the size of the recursive covariance matrix P obtained in step 5; if P is less than the first critical value P1, outputting the estimated value obtained in step 5; otherwise, outputting the default initial mass m0; if P is less than the second critical value P2, pausing steps 1-5, and outputting the estimated vehicle mass value at the last moment; if the vehicle parking time exceeds the critical time T0, repeating steps 1-6.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] 1. This invention takes into account the influence of vehicle pitch motion on the measurement value of longitudinal acceleration sensor, and introduces acceleration correction coefficient, slope correction coefficient and air resistance correction coefficient to correct the measurement value of longitudinal acceleration sensor, thereby eliminating the error caused by vehicle pitch motion to mass estimation;

[0047] 2. This invention identifies the driving conditions at each moment by designing reasonable fuzzy logic, obtains the condition suitability, uses the condition suitability as the weight of that moment, the equivalent driving force as the output, and the equivalent acceleration as the input, and substitutes them into the weighted least squares method to eliminate or reduce the weight of conditions with large mass estimation errors, thereby improving the accuracy of vehicle mass estimation and the adaptability of driving conditions. Attached Figure Description

[0048] To more clearly illustrate the technical solutions of the embodiments of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and detailed embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0049] Figure 1This is a schematic diagram of the online vehicle mass estimation method based on fuzzy logic and weighted least squares as described in this invention.

[0050] Figure 2 This is the equivalent acceleration fuzzy membership function of the present invention;

[0051] Figure 3 This is the equivalent driving force fuzzy membership function of the present invention;

[0052] Figure 4 This is the fuzzy membership function for vehicle speed in this invention;

[0053] Figure 5 This is the fuzzy membership function for the steering wheel angle of the present invention;

[0054] Figure 6 This is the fuzzy membership function of the equivalent acceleration rate of change in this invention;

[0055] Figure 7 This is the fuzzy membership function of the equivalent driving force change rate in this invention;

[0056] Figure 8 This is the fuzzy membership function for the rate of change of steering wheel angle in this invention;

[0057] Figure 9 This is a time series diagram of the two vehicle quality identification results of the present invention. Detailed Implementation

[0058] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0059] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0060] This invention provides the following technical solution: an online vehicle mass estimation method based on fuzzy logic and weighted least squares, which can improve the accuracy and adaptability of online vehicle mass estimation during use;

[0061] Example 1

[0062] like Figure 1 As shown, this invention provides an online vehicle mass estimation method based on fuzzy logic and weighted least squares, comprising the following steps:

[0063] Step 1: Obtain the longitudinal acceleration sensor measurement signal, longitudinal speed signal, engine output torque signal, gear signal, throttle opening signal, brake pedal signal, and steering wheel angle signal from the vehicle CAN bus.

[0064] Step 2: To eliminate the influence of vehicle pitch motion on the longitudinal acceleration sensor measurement, considering the inertial force during acceleration, axle load transfer on slopes, and air resistance at high speeds that cause vehicle pitch motion, the longitudinal acceleration sensor measurement is corrected. The corrected longitudinal acceleration sensor measurement is as follows:

[0065]

[0066] Among them, a xre v is the corrected longitudinal acceleration sensor measurement, a xsen The value measured by the longitudinal acceleration sensor, v x For longitudinal velocity, k is the first derivative of the longitudinal velocity. acc To accelerate the correction coefficient, k gra k is the slope correction factor. air This is the air resistance correction factor.

[0067] Step 3: Calculate the equivalent driving force and equivalent acceleration. The equivalent driving force is the sum of all forces independent of the vehicle's mass acting on the vehicle in the longitudinal direction. Specifically, it includes the driving force converted from the engine output torque to the wheel ends, the acceleration resistance converted from the inertial torque of rotating components such as wheels and flywheels to the wheel ends, and the air resistance acting on the entire vehicle. The equivalent driving force is calculated using the following formula:

[0068]

[0069] Where, η t For the efficiency of the transmission system, M e i is the engine output torque. g For the gearbox transmission ratio, i0 is the main reducer transmission ratio, and r is the gear ratio. d Let Θ be the rolling radius of the wheel. i For the equivalent rotating mass moment of inertia, C D Where ρ is the air resistance coefficient, A is the reference area, and ρ is the air resistance coefficient. a This refers to air density.

[0070] Equivalent rotating mass and moment of inertia Θ i The moment of inertia of rotating components in a vehicle's transmission system, such as the engine, flywheel, and wheels, transferred to the wheel ends, can be calculated using the following formula:

[0071]

[0072] Among them, Iw For the moment of inertia of the wheel, I f This refers to the rotational inertia of the engine and flywheel.

[0073] Equivalent acceleration is the acceleration caused by resistances other than the equivalent driving force acting on the vehicle. It includes longitudinal acceleration, the gravitational acceleration component on the slope, and the acceleration caused by rolling resistance, and can be expressed as:

[0074] a res =a x +gsinβ+gf R (20)

[0075] Among them, a x Let g be the longitudinal acceleration of the vehicle, β be the acceleration due to gravity, and f be the longitudinal slope of the road. R This is the rolling resistance coefficient.

[0076] The corrected longitudinal acceleration sensor measurements include vehicle acceleration and road gradient information, i.e.:

[0077] a xrev =a x +g sinβ (21)

[0078] By directly replacing the calculated value with the corrected measured value, the influence of slope estimation error and vehicle longitudinal acceleration calculation error on the mass estimation accuracy can be eliminated, resulting in the final equivalent acceleration.

[0079] a res =a xrev +gf R (twenty two)

[0080] Rolling resistance coefficient f R It can be expressed as the longitudinal velocity v x Related functions:

[0081]

[0082] Where f0 is the constant term of the rolling resistance coefficient, f1 is the coefficient of the first-order term, and f4 is the coefficient of the fourth-order term.

[0083] Step 4: Calculate the suitability of the operating conditions using a fuzzy logic algorithm. The inputs to the fuzzy logic algorithm include equivalent acceleration, equivalent driving force, vehicle speed, steering wheel angle, rate of change of equivalent acceleration, rate of change of equivalent driving force, and rate of change of steering wheel angle.

[0084] First, the input values ​​are converted into fuzzy values ​​using fuzzy membership functions. The fuzzy membership functions for equivalent acceleration axres, equivalent driving force Fres, vehicle speed vx, steering wheel angle sw, equivalent acceleration rate of change daxres, equivalent driving force rate of change dFres, and steering wheel angle rate of change dsw are as follows: Figures 2-8 As shown, S represents small, M represents medium, and L represents large.

[0085] The corresponding fuzzy rules, formulated using expert experience, are shown in the table below.

[0086]

[0087]

[0088] In the table, w represents the suitability of the working conditions, which can be divided into three levels: {S, M, L}, with corresponding fuzzy values ​​of {S:0, M:0.5, L:1}.

[0089] The fuzzy inference system uses the Sugeno model and defuzzifies it using the centroid method to obtain the precise value of the working condition suitability w.

[0090] Step 5: Estimate the quality using weighted least squares. Use the working condition suitability w obtained in Step 4 as a weight for the equivalent driving force F obtained in Step 3. res and equivalent acceleration a xres Weighted calculations are applied, with a higher operating condition suitability (w) indicating greater reliability of the data set, resulting in a higher weighting in the weighted least squares method. Reducing the weight of operating conditions with larger quality estimation errors can improve the accuracy and adaptability of vehicle quality estimation.

[0091] The output of weighted least squares method z = wF res Input quantity The parameter to be identified, θ = m, is used to estimate the vehicle mass m according to the following formula:

[0092]

[0093] Where λ is the time-varying forgetting factor and P is the recursive covariance matrix.

[0094] Step 6: Pausing and Starting Vehicle Mass Estimation. First, determine if the covariance matrix P in the weighted least squares method is less than the first critical value P1. If P < P1, the estimated vehicle mass is relatively accurate, and this estimated value is output; otherwise, the default initial mass m0 is output. Next, determine if the covariance matrix P is less than the second critical value P2. If P < P2, the estimated vehicle mass has converged. Since the vehicle mass does not change significantly during driving, steps 1-5 can be paused, and the final estimated vehicle mass is output. After the vehicle stops, passengers may get on or off, or cargo may be loaded or unloaded, potentially changing the mass. Therefore, if the vehicle's stopping time exceeds the critical time T0, steps 1-6 are repeated.

[0095] Figure 9 The figure shows the vehicle mass estimation results of a GAC ​​SUV on a circular track. As can be seen from the figure, after the vehicle starts, the estimated vehicle mass value quickly approaches the true value of 2387 kg, converges to 2373 kg, and the mass estimation pauses with an error of 14 kg. When the vehicle stops for longer than the threshold, the mass estimation restarts, and the second vehicle mass estimation result converges to 2394 kg with an error of only 7 kg. This shows that the vehicle mass estimation method of the present invention has the advantages of high accuracy, fast convergence, and strong applicability to various working conditions.

Claims

1. A method for online vehicle mass estimation based on fuzzy logic and weighted least squares, characterized in that: Includes the following steps: Step 1: Obtain the longitudinal acceleration sensor measurement value signal, longitudinal speed signal, engine output torque signal, gear signal, throttle opening signal, brake pedal signal, and steering wheel angle signal from the vehicle CAN bus; Step 2: Correct the measured value of the longitudinal acceleration sensor according to equation (1); （1）； in, The corrected longitudinal acceleration sensor measurement. The value is measured by the longitudinal acceleration sensor. For longitudinal velocity, The first derivative of the longitudinal velocity, To accelerate the correction factor, This is the slope correction factor. This is the air resistance correction factor; Step 3: Calculate the equivalent driving force and equivalent acceleration; Step 4: Calculate the suitability of the working condition using a fuzzy logic algorithm. The inputs to the fuzzy logic algorithm include equivalent acceleration, equivalent driving force, vehicle speed, steering wheel angle, rate of change of equivalent acceleration, rate of change of equivalent driving force, and rate of change of steering wheel angle. The input values ​​are converted into fuzzy values ​​using fuzzy membership functions. Based on the corresponding fuzzy rules, the Sugeno model is used, and the centroid method is used for defuzzification to obtain the accurate value of the suitability of the working condition. Step 5: Estimate the vehicle mass using the weighted least squares method; Step 6: Pausing and starting vehicle mass estimation: The equivalent driving force and equivalent acceleration obtained in step 3 are weighted using the working condition suitability obtained in step 4, and the output of the least squares method is defined as: （5）； in, The suitability of the working conditions obtained in step 4, The equivalent driving force defined by equation (2); The input for least squares is: （6）； in, The equivalent acceleration defined by equation (3); The parameters to be identified using the least squares method are: （7）； in, The quality of the vehicle to be identified; The vehicle mass is estimated using the least squares method (8): （8）； in, A time-varying forgetting factor, Let be the recursive covariance matrix.

2. The online vehicle mass estimation method based on fuzzy logic and weighted least squares as described in claim 1, characterized in that: Step 3 specifically includes the following steps: (1): Define the equivalent driving force as: （2）； in, For the efficiency of the transmission system, For engine output torque, For the gearbox ratio, Main reducer transmission ratio, The radius of the wheel's rolling motion. The moment of inertia is the equivalent rotating mass. The air drag coefficient, For reference area, air density; (2): Define the equivalent acceleration as: （3）； in, It is the acceleration due to gravity. The rolling resistance coefficient, The longitudinal acceleration sensor measurement value is defined by equation (1).

3. The online vehicle mass estimation method based on fuzzy logic and weighted least squares as described in claim 2, characterized in that: The rolling resistance coefficient can be expressed as: （4）； in, This is the constant term for the rolling resistance coefficient. The coefficient of the first-order term, As the coefficient of the fourth term, it can be seen from equation (4) that the rolling resistance coefficient is determined by the longitudinal velocity.

4. The online vehicle mass estimation method based on fuzzy logic and weighted least squares as described in claim 1, characterized in that: Step 6 includes: determining the covariance matrix recursively obtained in step 5. The size, if Less than the first critical value Output the estimated value obtained in step 5; otherwise, output the default initial mass. ;like Less than the second critical value Pause steps 1-5 and output the estimated vehicle mass at the last moment; if the vehicle's stopping time exceeds the critical time... Repeat steps 1-6.

Images