A method for estimating vehicle state and road surface unevenness

Abstract

The present invention discloses a method for estimating vehicle state and road surface unevenness, comprising the following steps: obtaining the vertical acceleration of the vehicle's center of mass, roll angular velocity, pitch angular velocity, and dynamic deflections of four suspensions in real time through sensors; obtaining the vertical displacement of the vehicle's center of mass, roll angle, and pitch angle based on the vertical acceleration of the center of mass, roll angular velocity, and pitch angular velocity; inputting the above data into the proposed vehicle state and road surface unevenness estimation algorithm to obtain the estimated state variables of the vehicle and the road surface unevenness information of four wheels. The present invention only requires common sensors such as displacement sensors to simultaneously estimate vehicle state variables and road surface unevenness, with low hardware equipment costs; the present invention can be used for state estimation and road surface unevenness estimation of semi-active suspension or active suspension vehicles, and can also be applied to state estimation and road surface unevenness estimation of passive suspension vehicles.

Description

Technical Field

[0001] The present invention relates to the field of vehicle state estimation and road surface unevenness estimation, and particularly to an estimation method for vehicle state and road surface unevenness for an automotive suspension control system. Background Art

[0002] The suspension system of an automobile is located between the vehicle body and the wheels. Its main functions are to carry the vehicle body, transfer the forces and torques acting on the wheels by the road surface to the vehicle body, and to mitigate the continuous impacts caused by road surface unevenness and attenuate the vibrations of the vehicle body. The performance of the suspension system directly determines the ride comfort and handling stability of the automobile.

[0003] With the rapid economic development of our country and the improvement of people's living standards, the comfort and handling stability of automobiles have become the primary considerations when people purchase cars. Semi-active suspensions and active suspensions can adjust the suspension characteristics according to different road conditions, and can effectively balance the comfort and handling stability of automobiles, so they are widely used. However, the accurate estimation of road information is very important for realizing the intelligent control of the suspension under different road surfaces.

[0004] At present, the methods for identifying road surface unevenness mainly include contact measurement, non-contact measurement, and estimation methods based on vehicle responses. Contact measurement is to obtain the road surface elevation using a road surface profile measurement device that keeps in contact with the road surface, which is costly; non-contact measurement is to obtain the road surface elevation through a camera, etc., which is greatly affected by the external environment. Limited by the above factors, these two methods are difficult to be widely applied. The estimation based on vehicle responses, that is, through the relationship between vehicle responses and road surface excitations, inversely deduces the road surface excitations from some vehicle response information that is easy to measure, with low cost and less affected by the environment. At the same time, most suspension control strategies based on modern control theory require the motion state parameters of the vehicle as feedback variables; limited by factors such as high cost, state variables such as the absolute displacement of the tire cannot be directly measured, which limits the application of some suspension control methods. Therefore, it has great engineering significance to study algorithms for estimating road surface unevenness and vehicle state with easily measurable vehicle responses as measurement inputs. Chang Xiaotong et al. provided a "method for identifying road surface unevenness based on Kalman filter theory" in the Chinese invention patent publication CN113353085A, established a half-vehicle model of a passive suspension, designed a road surface unevenness identification algorithm based on Kalman filter theory, and could estimate the road surface unevenness according to the accelerations of the vehicle body and the wheels. However, this solution is based on a simplified vehicle model with a low degree of freedom, ignores the roll motion during vehicle driving, and has a large difference from the motion of an actual vehicle; and it cannot estimate the state variables related to the roll motion of an actual vehicle; at the same time, this solution is based on a vehicle model with a passive suspension and cannot be applied to vehicles equipped with semi-active suspensions or active suspensions. Summary of the Invention

[0005] In order to solve at least one of the problems existing in the prior art, the present invention proposes a method for estimating vehicle states and road unevenness applicable to passive suspension vehicles and semi-active and active suspension control systems. This method uses the vertical acceleration of the center of mass, roll angular velocity, pitch angular velocity, suspension dynamic deflection, and the vertical displacement of the center of mass, roll angle, and pitch angle obtained through data processing as measurement values; then, by inputting the above data into the vehicle state and road unevenness estimation algorithm, the state variables of the vehicle and the road unevenness information of the four wheels can be estimated.

[0006] To achieve the object of the present invention, a method for estimating vehicle states and road unevenness provided by the present invention is implemented according to the following steps:

[0007] (1) During the driving process of the vehicle, the vertical acceleration of the center of mass, roll angular velocity, pitch angular velocity, and dynamic deflections of the four suspensions of the vehicle are obtained in real time through sensors installed on the vehicle.

[0008] (2) By filtering the vertical acceleration of the center of mass, roll angular velocity, and pitch angular velocity collected by the sensors and performing integral calculations, the vertical displacement of the center of mass, roll angle, and pitch angle of the vehicle can be obtained.

[0009] (3) Input the data directly measured by the above sensors and the data calculated into the proposed vehicle state and road unevenness estimation algorithm to obtain the state variables of the vehicle and the road unevenness information of the four wheels.

[0010] The proposed method for estimating vehicle states and road unevenness is specifically as follows:

[0011] (1) Establish the motion differential equation of the seven-degree-of-freedom model of the whole vehicle. Select the state variables as:

[0012] Select the control variable as the actuation force of the semi-active suspension or active suspension: u = [U fL U fR U rL U rR T , and the disturbance variable as the road excitation of the four wheels: w = [q fL q fR q rL q rR T ; select the measurement variables as the vertical acceleration of the center of mass, roll angular velocity, pitch angular velocity, dynamic deflections of the four suspensions, and the vertical displacement of the center of mass, roll angle, and pitch angle:

[0013]

[0014] Based on the selected state variables, control variables, etc., the state - space equation of the seven - degree - of - freedom vehicle model can be written as follows:

[0015]

[0016] In the above - mentioned state equation, ζ is the process noise, which represents factors such as system uncertainty, such as model error; ν represents the measurement noise generated due to insufficient sensor accuracy. It is assumed that the process noise and the measurement noise are uncorrelated and both satisfy the zero - mean Gaussian distribution, and their variances are S and R respectively.

[0017] (2) Augment the four - wheel road surface excitation into the original state variables, and the augmented state variables are obtained as x a = [x w] T Then, the state equation and the observation equation after augmenting the state variables are as follows:

[0018]

[0019] Discretize the state equation after augmentation to obtain the discretized state - space equation:

[0020]

[0021] In the above formula: A ad = I + A a Δt; B ad = B a Δt; C ad = C a ; D ad = D a . Where Δt is the sampling time of the sensor. ξ k is the process noise of the state equation after augmenting the state variables, and ξ k = [ζ k η k T ; ζ k and η k are respectively the process noise of the state equation before augmentation and the road surface speed excitation. Therefore, the covariance matrix of ξ k can be expressed as Q ak = diag[S k Q k . Where S k is the covariance matrix of the process noise of the state equation before augmentation, that is, S k = E[ζ k ζ k T ; Q k is the covariance matrix of the road surface speed excitation, that is, Q k = E[η​k η k T . Additionally, the measurement noise covariance matrix R k = E[ν k ν k T .

[0022] (3) Based on the Kalman filtering theory, using the above augmented discrete state equation, the prior estimate of the system state at this moment can be obtained according to the posterior state estimate value of the system at the previous sampling moment; and through the observation equation, the system state is corrected by the deviation between the measured value of the sensor and the estimated value of the measured quantity, so as to obtain the posterior state estimate value of the system at this moment. The specific calculation process is as follows:

[0023] a) Perform initialization definitions: Set the initial state of the system x a (0|0), select the initial state error covariance P 0|0 , determine the process noise covariance matrix Q ak = diag[S k Q k and the measurement noise covariance matrix R k = E[ν k ν k T .

[0024] b) Calculate the prior state estimate value at this moment according to the posterior state estimate value at the previous moment:

[0025] x a (k + 1|k)= A ad x a (k|k)+ B ad u(k) (4)

[0026] c) Update the error covariance matrix of the prior state estimate value:

[0027] P k+1|k = A ad P k|k A ad T + Q ak (5)

[0028] d) Calculate the Kalman gain:

[0029] K k = P k+1|k C ad T [C ad P k+1|k C ad T + Rk -1 (6)

[0030] e) Calculate the posterior state estimate at this moment:

[0031] x a (k + 1|k + 1) = x a (k + 1|k) + K k [y k -(C ad x a (k + 1|k) + D ad u(k))] (7)

[0032] f) Update the error covariance matrix of the posterior state estimate:

[0033] P k+1|k+1 = (I - K k C ad )P k+1|k (8)

[0034] By substituting x a (k + 1|k + 1) and P k+1|k+1 into formulas (4) and (5), the above steps can be continuously cycled, thereby estimating the vehicle state and the road surface roughness information of the four wheels at each sampling moment.

[0035] Compared with the prior art, the present invention has at least the following advantages:

[0036] 1) The present invention only requires common sensors such as displacement sensors to simultaneously estimate vehicle state variables and road surface roughness, with low hardware equipment costs and less influence from environmental factors;

[0037] 2) The present invention can be used for state estimation and road surface roughness estimation of semi - active suspension or active suspension vehicles, and can also be applied to state estimation and road surface roughness estimation of passive suspension vehicles.

[0038] 3) The proposed vehicle state and road surface roughness estimation algorithm is based on a seven - degree - of - freedom vehicle model. Compared with simplified vehicle models with lower degrees of freedom, this model has higher accuracy and is closer to the actual vehicle situation. Brief Description of the Drawings

[0039] Figure 1 It is a schematic diagram of a seven - degree - of - freedom vehicle model provided by an embodiment of the present invention.

[0040] Figure 2 It is a flowchart of a method for estimating vehicle state and road surface roughness provided by an embodiment of the present invention.

[0041] Figure 3 ​This is the simulation flow chart in the embodiments of the present invention.

[0042] Figure 4 This is a schematic diagram for comparing the estimated value and the true value of the vehicle state in the embodiments of the present invention.

[0043] Figure 5 This is a schematic diagram for comparing the estimated value and the true value of the road surface unevenness in the embodiments of the present invention. Detailed implementation manners

[0044] To make the objectives, technical solutions and advantages of the present invention clearer and more explicit, the following further describes the present invention in detail with reference to the accompanying drawings and by way of examples.

[0045] Please refer to Figure 2 , a method for estimating the vehicle state and road surface unevenness provided by the present invention includes the following steps:

[0046] Step 1: During the driving process of the vehicle, the vertical acceleration of the vehicle's center of mass, roll angular velocity, pitch angular velocity, and four suspension dynamic deflections are obtained in real time through sensors installed on the vehicle.

[0047] Step 2: By filtering the collected vertical acceleration of the center of mass, roll angular velocity, and pitch angular velocity and performing integral calculations, the vertical displacement of the vehicle's center of mass, roll angle, and pitch angle can be obtained.

[0048] Step 3: Input the data directly measured by the above sensors and the calculated data into the proposed vehicle state and road surface unevenness estimation algorithm to obtain the state variables for suspension control of the vehicle and the road surface unevenness information of the four wheels.

[0049] Among them, in some embodiments of the present invention, the proposed algorithm for estimating the vehicle state and road surface unevenness is specifically:

[0050] (1) Establish the motion differential equation of a seven-degree-of-freedom whole vehicle model considering the vertical, roll, and pitch motions of the vehicle body and the vertical motions of the four unsprung masses. The seven-degree-of-freedom whole vehicle model is as Figure 1 shown:

[0051] Vertical motion of the vehicle body:

[0052]

[0053] Roll motion of the vehicle body:

[0054]

[0055] Pitch motion of the vehicle body:

[0056]

[0057] Vertical movement of the left front wheel:

[0058]

[0059] Vertical movement of the right front wheel:

[0060] Vertical movement of the left rear wheel:

[0061]

[0062]

[0063] Vertical movement of the right rear wheel:

[0064]

[0065] Wherein, F sfL 、F sfR 、F srL 、F srR respectively represent the vertical forces of the left front, right front, left rear, and right rear suspension systems:

[0066]

[0067] When the roll angle and pitch angle of the vehicle are small, approximately:

[0068]

[0069] In the above formula, M cb is the sprung mass, M wf 、M wr are the unsprung masses; J x 、J y are the roll and pitch moments of inertia of the vehicle body; Z cb 、φ、 are the displacement of the center of mass, roll angle, and pitch angle respectively, are the second derivatives of Z cb 、φ、 respectively; Z fL 、Z fR 、Z rL 、Z rR are the displacements of the unsprung masses of the left front, right front, left rear, and right rear suspensions respectively, are the second derivatives of Z fL 、Z fR 、Z rL 、Z rR respectively; q fL 、q fR 、q rL 、q rR are the road excitations of the four wheels; L f 、L ris the distance from the centroid to the front axle and the rear axle; d is the track width; K sf and K sr are the suspension spring stiffnesses respectively; C sf and C sr are the basic damping coefficients of the shock absorbers respectively; K tf and K tr are the stiffnesses of the front and rear wheels respectively; U fL and U fR and U rL and U rR are the actuating forces of the left front, right front, left rear, and right rear suspensions in the semi-active suspension. Z1, Z2, Z3, and Z4 are the absolute displacements at the connections between the left front, right front, left rear, and right rear suspensions and the vehicle body respectively.

[0070] Convert the seven-degree-of-freedom vehicle model's motion differential equation in formulas (9a)-(11) into the form of a state-space equation. The state variables are selected as: Select the control variable: u = [U fL U fR U rL U rR T , and the disturbance variable w = [q fL q fR q rL q rR T ; Select the measurement variables as: x, u, w, and y represent the state variable, control variable, disturbance variable, and measurement variable respectively. The superscript T represents the transpose, are the first derivatives of Z fL , Z fR , Z rL , Z rR , Z cb respectively;

[0071] According to the selected state variables, control variables, disturbance variables, and measurement variables, the state-space equation of the seven-degree-of-freedom vehicle model can be written

[0072]

[0073] In the formula: is the first derivative of the state variable x; A is a 14×14-dimensional system state coefficient matrix; B is a 14×4-dimensional system control coefficient matrix; G is a 14×4-dimensional system disturbance coefficient matrix; C is a 10×14-dimensional output state coefficient matrix; D is a 10×4-dimensional output control coefficient matrix.

[0074] ​​Among them: ζ in the above state equation is the process noise, representing factors such as system uncertainty, such as model error; ν represents the measurement noise generated due to insufficient sensor accuracy. It is assumed that the process noise and the measurement noise are uncorrelated and both satisfy a zero-mean Gaussian distribution, with variances S and R respectively. The matrices in the state equation are specifically as follows:

[0075] (2) Augment the four-wheel road surface excitation into the original state variables, thereby obtaining the augmented state variable x a =[x w] T , then the state equation and the observation equation after augmenting the state variables are as follows:

[0076]

[0077] In the above formula: is the first derivative of the state variable x a ; A a is an 18×18-dimensional system state coefficient matrix; B a is an 18×4-dimensional system control coefficient matrix; C a is a 10×18-dimensional output state coefficient matrix; D a is a 10×4-dimensional output control coefficient matrix.

[0078] Among them, B a =[B O 4×4 T , C a =[C O 10×4 , D a =D. ξ = [ζ η] T , ζ is the process noise of the state equation before augmenting the state variables; η is the four-wheel road surface speed excitation, that is

[0079] Discretize the state equation after augmentation to obtain the discretized state space equation:

[0080]

[0081] x a (k + 1) represents the state variable at the (k + 1)-th sampling moment; x a (k) represents the state variable at the k-th sampling moment; u(k) represents the control variable at the k-th sampling moment; y(k) represents the measurement variable at the k-th sampling moment.

[0082] The matrices in the above formula: A ad =I + A a Δt; B ad =B a Δt; C​ad = C a ; D ad = D a . Among them, Δt is the sampling time of the sensor. ξ k is the process noise of the state equation after augmenting the state variables, and ξ k = [ζ k η k T ; ζ k and η k are respectively the process noise of the state equation before augmentation and the road surface speed excitation. Therefore, the covariance matrix of ξ k can be expressed as Q ak = diag[S k Q k . Among them, S k is the process noise covariance matrix of the state equation before augmentation, that is, S k = E[ζ k ζ k T , and the meaning of E is to find the mean value; Q k is the covariance matrix of the road surface speed excitation, that is, Q k = E[η k η k T . In addition, the measurement noise covariance matrix R k = E[ν k ν k T .

[0083] (3) Based on the Kalman filter theory, using the above augmented discrete state equation, the prior estimate of the system state at this moment can be obtained according to the posterior state estimate value of the system at the previous sampling moment; and through the observation equation, the estimated values of the measured quantities (centroid acceleration, roll angular velocity, pitch angular velocity, suspension dynamic deflection, centroid vertical displacement, roll angle, pitch angle) can be calculated. Then, the system state is corrected through the deviation between the measured value of the sensor and the estimated value of the measured quantity, so as to obtain the posterior state estimate value of the system at this moment.

[0084] In some embodiments of the present invention, the specific calculation process is as follows:

[0085] a) Perform initialization definition: Set the initial state of the system x a (0|0), select the initial state error covariance P 0|0 , determine the process noise covariance matrix Q ak = diag[S k Q k and the measurement noise covariance matrix R k ​= E[ν k ν k T .

[0086] b) Calculate the prior state estimate at this time based on the posterior state estimate at the previous time:

[0087] x a (k + 1|k) = A ad x a (k|k) + B ad u(k) (4)

[0088] c) Update the error covariance matrix of the prior state estimate:

[0089] P k+1|k = A ad P k|k A ad T + Q ak (5)

[0090] d) Calculate the Kalman gain:

[0091] K k = P k+1|k C ad T [C ad P k+1|k C ad T + R k -1 (6)

[0092] e) Calculate the posterior state estimate at this time:

[0093] x a (k + 1|k + 1) = x a (k + 1|k) + K k [y k -(C ad x a (k + 1|k) + D ad u(k))] (7)

[0094] f) Update the error covariance matrix of the posterior state estimate:

[0095] P k+1|k+1 =(I - K k C ad )P k+1|k (8)

[0096] By comparing x a (k + 1|k + 1) with P k+1|k+1 ​Substituting into Formula (4) and Formula (5) can achieve continuous cycling of the above steps, thereby estimating the vehicle state at each sampling moment and the road surface unevenness information of the four wheels.

[0097] x a (k|k) represents the posterior state estimate value at the k-th sampling moment, and P k|k represents the error covariance matrix of the posterior state estimate value at the k-th sampling moment, and R k represents the measurement noise covariance matrix at the k-th sampling moment, and K k represents the Kalman gain at the k-th sampling moment, and I represents the identity matrix.

[0098] In some embodiments of the present invention, it further includes Step 4: Through the above steps, the estimated value of the vehicle state and the four-wheel road surface unevenness information can be obtained. However, the road surface unevenness information obtained at this time is the relationship between the road surface unevenness and time; the vehicle travel distance can be calculated through the vehicle speed, so that the relationship between the road surface unevenness and the longitudinal distance can be obtained.

[0099] In some embodiments of the present invention, a specific embodiment given by using the above method:

[0100] Please refer to Figure 3 , and verify the effectiveness of this method through the co-simulation of CarSim and Simulink. The parameters of the simulation vehicle are shown in Table 1.

[0101] For convenience, this embodiment uses a passive suspension vehicle model to verify the effectiveness of this method, that is Figure 1 the suspension in fL U fR U fL U fL are all 0. If this method is to be used for a semi-active or active suspension vehicle, set U fL U fR U fL U fL to the corresponding actuator force values. The specific process of the simulation is as shown in Figure 3 . Establish a Class C road surface in CarSim as the input of the CarSim vehicle model. Set the vehicle speed to 20 m / s. Directly export the vehicle centroid acceleration, roll angular velocity, pitch angular velocity, suspension dynamic deflection, and centroid vertical displacement, roll angle, and pitch angle from CarSim as measurement values.

[0102] Table 1 Simulation vehicle model parameters

[0103]

[0104] Among them, the sampling time of the simulation is set to 5 ms. In addition, the process noise covariance matrix in the simulation is Q ak = diag[S k Q k , where S k is mainly the variance of the error between the seven-degree-of-freedom vehicle model and the CarSim vehicle model; since the error between the two models is small, a small value can be taken: S k = 10 -6 ; Q k represents the variance of the road surface speed excitation, Q k = 0.1. The diagonal elements of the measurement noise covariance matrix are set to 10 -4 .

[0105] Import the above-mentioned measured quantities into the vehicle state and road surface unevenness estimation algorithm built in Simulink, and the vehicle state estimation value and the road surface unevenness estimation value can be obtained. Finally, the estimated value is compared with the true value, and the comparison results are as shown in Figure 4 and Figure 5 . It can be seen that the estimated value obtained by the method of the present invention coincides highly with the true value, effectively demonstrating the effectiveness of the method of the present invention.

[0106] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for estimating vehicle state and road surface unevenness, characterized in that, It includes the following steps: (1) During the driving process of the vehicle, the vertical acceleration of the vehicle's center of mass, roll angular velocity, pitch angular velocity, and the dynamic deflections of the four suspensions are obtained in real time through sensors installed on the vehicle. (2) Based on the vertical acceleration of the center of mass, roll angular velocity, and pitch angular velocity, the vertical displacement of the vehicle's center of mass, roll angle, and pitch angle are calculated. (3) The data directly measured by the above sensors and the calculated data are input into the vehicle state and road surface unevenness estimation algorithm to obtain the state variables of the vehicle and the road surface unevenness information of the four wheels. Among them, the vehicle state and road surface unevenness estimation algorithm is specifically as follows: (3.1) Establish the motion differential equation of a seven-degree-of-freedom vehicle model, and select the state variables as: , , , , are respectively , , , the first-order derivatives of , , , are respectively the unsprung mass displacements of the left front, right front, left rear, and right rear suspensions, is the first-order derivative of is the center-of-mass displacement, , are respectively the roll angle and pitch angle, , are respectively , the first-order derivatives of, with the superscript T being the device; select the control variable , , , , are the actuating forces of the left front, right front, left rear, and right rear suspensions in the semi-active suspension, and the disturbance variable is the four-wheel road surface excitation ; select the measured variables as the center-of-mass vertical acceleration, roll angular velocity, pitch angular velocity, the dynamic deflections of the four suspensions, and the center-of-mass vertical displacement, roll angle, and pitch angle: , , , , are the absolute displacements at the connections between the left front, right front, left rear, and right rear suspensions and the vehicle body, respectively; According to the selected state variables, control variables, disturbance variables, and measurement variables, write the state space equation of the seven-degree-of-freedom vehicle model: In the formula, is the state variable 's first derivative, is the state coefficient matrix, is the process noise; is the control coefficient matrix; is the perturbation coefficient matrix, is the output state coefficient matrix, is the output control coefficient matrix, represents the measurement noise generated due to the insufficient accuracy of the sensor; (3.2) Augment the four-wheel road surface excitation into the original state variables, and the augmented state variables are obtained as , then the state equation and the observation equation after augmenting the state variables are as follows: Wherein: is the state variable is the first derivative of; is the system state coefficient matrix; is the system control coefficient matrix; is the output state coefficient matrix; is the output control coefficient matrix; Discretize the augmented state equation to obtain the discretized state space equation: where: represents the state variable at the k +(1)th sampling instant; represents the state variable at the k th sampling instant; represents the control variable at the k th sampling instant; represents the measurement variable at the k th sampling instant; ; ; ; , is the sampling time of the sensor, is the process noise of the state equation after augmenting the state variables, and ; and are respectively the process noise of the state equation before augmentation and the road surface speed excitation; therefore, 's covariance matrix is expressed as , where is the process noise covariance matrix of the state equation before augmentation, that is ; is the covariance matrix of the road surface speed excitation, that is , E is to find the mean value. In addition, the measurement noise covariance matrix ; (3.3) Based on the Kalman filtering theory, use the augmented discrete state equation, and according to the a priori state estimate value of the system at the previous sampling moment, perform a priori estimation on the system state at this moment; and through the observation equation, correct the system state through the deviation between the actual measured value of the sensor and the estimated value of the measured quantity, so as to obtain the a posteriori state estimate value of the system at this moment.

2. The estimation method of vehicle state and road surface unevenness according to claim 1, characterized in that In step (1), the method for obtaining the measurement parameters is to measure the dynamic deflection of the vehicle suspension through a displacement sensor, and measure the vertical acceleration of the vehicle's center of mass, roll angular velocity, and pitch angular velocity through an inertial measurement unit.

3. The estimation method of vehicle state and road surface unevenness according to claim 1, characterized in that: In the said step (2), the method for obtaining the three parameters of the vertical displacement of the center of mass, roll angle, and pitch angle is: through filtering processing and integral operation on the signals of the vertical acceleration of the center of mass, roll angular velocity, and pitch angular velocity collected by the sensor.

4. The estimation method of vehicle state and road surface unevenness according to claim 1, characterized in that The vehicle state and road surface unevenness estimation algorithm is based on a seven-degree-of-freedom vehicle model, considering the vertical, roll, and pitch motions of the vehicle body and the vertical motions of the four unsprung masses.

5. The estimation method of vehicle state and road surface unevenness according to claim 1, characterized in that, This method can not only be used for semi-active suspension or active suspension vehicles, but also be used for the estimation of the state of passive suspension vehicles and road surface unevenness by setting the control variable to 0.

6. The estimation method of vehicle state and road surface unevenness according to claim 1, wherein The a priori estimation state equation and observation equation of the vehicle state and road surface unevenness estimation algorithm are obtained by augmenting the disturbance variable, that is, the four-wheel road surface excitation, in the state equation of the seven-degree-of-freedom vehicle model to the state variables and then discretizing.

7. A method for estimating a vehicle state and road surface unevenness according to claim 1, characterized in that The vehicle state and road surface unevenness estimation algorithm is a cyclic process: based on the a posteriori state estimate value of the system at the previous sampling moment, a priori estimation on the system state at this moment can be performed based on the state equation; and the system state can be corrected through the deviation between the actual measured value of the sensor and the estimated value of the measured quantity, so as to obtain the a posteriori state estimate value of the system at this moment.

8. The estimation method of vehicle state and road surface unevenness according to claim 1, characterized in that In step (3.1), the motion differential equation of the seven-degree-of-freedom vehicle model is Vertical motion of the vehicle body: Roll motion of the vehicle body: Pitch motion of the vehicle body: Vertical motion of the left front wheel: Vertical motion of the right front wheel: Vertical motion of the left rear wheel: Vertical motion of the right rear wheel: Among them, , , , respectively represent the vertical forces of the left front, right front, left rear, and right rear suspension systems: When the roll angle and pitch angle of the vehicle are relatively small, approximately: In the above formula, is the sprung mass, , are the unsprung masses; , are the roll and pitch moments of inertia of the vehicle body; , , are the displacement of the center of mass, roll angle, and pitch angle respectively, , , are respectively , , the second derivatives of; , , , are the unsprung mass displacements of the left front, right front, left rear, and right rear suspensions respectively, , , , are respectively , , , the second derivatives of; , , , are the road excitations of the four wheels respectively; , are the distances from the center of mass to the front and rear axles; is the track width; , are the suspension spring stiffnesses respectively; , are the basic damping coefficients of the shock absorbers respectively; , are the stiffnesses of the front and rear wheels respectively; , , , are the actuating forces of the left front, right front, left rear, and right rear suspensions in the semi-active suspension; , , , are the absolute displacements at the connections between the left front, right front, left rear, and right rear suspensions and the vehicle body respectively.

9. The estimation method of vehicle state and road surface unevenness according to claim 1, characterized in that The road surface unevenness information obtained in step (3) is the relationship between road surface unevenness and time. Further, the driving distance of the vehicle can be calculated based on the vehicle speed, so that the relationship between road surface unevenness and longitudinal distance can be obtained.

10. A method for estimating a vehicle state and road surface unevenness according to any one of claims 1-9, characterized in that, In step (3.3), it specifically includes the following sub-steps: 3.3.1) Perform initialization definitions: Set the initial state of the system , select the initial state error covariance , determine the process noise covariance matrix and the measurement noise covariance matrix ; 3.3.2) Calculate the prior state estimate value at this moment according to the posterior state estimate value at the previous moment: 3.3.3) Update the error covariance matrix of the prior state estimate value: 3.3.4) Calculate the Kalman gain: 3.3.5) Calculate the posterior state estimate value at this moment: 3.3.6) Update the error covariance matrix of the posterior state estimate value: represents the posterior state estimate at the k th sampling time, represents the error covariance matrix of the posterior state estimate at the k th sampling time, represents the measurement noise covariance matrix at the k th sampling time, represents the Kalman gain at the k th sampling time, represents the identity matrix; By substituting and into Formula (4) and Formula (5), the continuous cycle of the above steps can be realized, so as to estimate the vehicle state at each sampling moment and the road surface roughness information of the four wheels.

Images