An automatic driving semi-trailer truck transverse trajectory tracking adaptive control method

Abstract

The application discloses a kind of automatic driving semi-trailer truck transverse trajectory tracking adaptive control method, including vehicle parameter adaptive calculation module, vertical load distribution module, tire side deflection stiffness calculation module, weight coefficient adaptive adjustment module, LQR controller solving module and control output constraint module, the method considers that the mass center position change of tractor and trailer caused by trailer load variation, change of moment of inertia, vertical load variation, tire side deflection stiffness variation, the adaptive adjustment of weight coefficient under different vehicle speeds, control output fluctuation caused by positioning jump and signal noise and roll safety, improve the accuracy of semi-trailer vehicle dynamics model;Feedback control output is constrained to be handled, and the oscillation or overshoot caused by positioning jump and signal noise is solved, guarantee the smoothness and stability of vehicle control, provide new ideas for the motion control of automatic driving semi-trailer truck, and it has important value to the design and optimization of automatic driving platform overall architecture.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of automatic driving vehicle motion control, and particularly relates to an automatic driving semi-trailer truck lateral trajectory tracking adaptive control method. BACKGROUND

[0002] The semi-trailer truck is composed of a tractor and a trailer, and has more complex dynamic characteristics than passenger cars. Especially for high-speed heavy-load semi-trailer trucks, when the trailer load changes, the center of mass position and the yaw moment of inertia of the tractor and the trailer will change significantly. At the same time, the tire cornering stiffness of the semi-trailer truck will also change constantly with the change of the vertical load of each axle. In addition, positioning jump and signal noise will seriously affect the smoothness of the control output and the stability of the vehicle, resulting in oscillation or overshoot of the vehicle during the course of lane keeping and lane changing.

[0003] In the field of automatic driving tracking control, the common lateral trajectory tracking control algorithm generally has one or more problems, specifically: (1) the controller designed based on the kinematic model cannot consider the complex dynamic characteristics of the vehicle at high speed; (2) the lateral controller of the semi-trailer truck designed based on the two-degree-of-freedom vehicle dynamics model cannot accurately represent the coupling relationship between the tractor and the trailer; (3) the fixed tire cornering stiffness combined with the vehicle dynamics model cannot accurately reflect the influence of tire dynamics characteristics on the vehicle system; (4) the feedback control output is not constrained, which makes it difficult to ensure the smoothness and stability of the vehicle control. SUMMARY

[0004] In view of the above problems, the application designs an automatic driving semi-trailer truck lateral trajectory tracking adaptive control method to solve the precision problem of the semi-trailer truck and ensure the adaptability and stability of the tracking control.

[0005] In order to achieve the above purpose, the application adopts the following technical scheme:

[0006] An automatic driving semi-trailer truck lateral trajectory tracking adaptive control method, the system of the method comprises a vehicle parameter adaptive calculation module, a vertical load distribution module, a tire cornering stiffness calculation module, a weight coefficient adaptive adjustment module, and an LQR controller solving module.

[0007] The method comprises the following steps:

[0008] Step 1: according to the load of the trailer, the center of mass position and the yaw moment of inertia of the tractor and the trailer are calculated by the vehicle parameter adaptive calculation module;

[0009] Step 2: according to the center of mass position of the tractor and the trailer, the vertical load at each axle is calculated by the vertical load distribution module;

[0010] Step 3: According to the vertical load at each axis and the number of tires, the tire cornering stiffness calculation module is used to update the cornering stiffness of each tire in real time;

[0011] Step 4: According to the vehicle speed, the weight coefficient adaptive adjustment module is used to adjust the weight coefficient of each state variable in real time;

[0012] Step 5: According to the vehicle state variables and the corresponding weight coefficients, the LQR controller solving module is used to calculate the steering wheel angle.

[0013] As a further description of the application, the system of the method further comprises a control output constraint module for constraining and filtering the state error calculation and measurement signals, the control output constraint module comprising feedback state constraints and overall constraints;

[0014] The feedback state constraints are nonlinear adjustment and low-pass filtering;

[0015] The overall constraints are front wheel angle limit constraints calculated based on static roll theory.

[0016] As a further description of the application, in step 1, the vehicle parameter adaptive calculation module inputs the loaded mass of the trailer, and the distance from the overall center of mass of the tractor and trailer to the front and rear axles and the overall yaw moment of inertia of each are calculated according to the moment balance and parallel shift axis theorem.

[0017] As a further description of the application, in step 2, the vertical load distribution module inputs the distance from the overall center of mass of the tractor and trailer to the front and rear axles, the overall mass, and the longitudinal acceleration, and calculates the vertical load at each axis of the tractor and trailer according to the D'Alembert principle.

[0018] As a further description of the application, in step 3, the tire cornering stiffness calculation module inputs the vertical load at each axis and the number of tires, and outputs the tire cornering stiffness at each axis, which is the sum of the cornering stiffness of each tire.

[0019] As a further description of the application, in step 4, the weight coefficient adaptive adjustment module inputs the vehicle speed, and outputs the weight coefficients of each state variable, including the weight coefficients of lateral position deviation, heading angle deviation, lateral velocity, yaw angular velocity, folding angular velocity, and folding angle.

[0020] As a further description of the application, in step 5, the LQR controller solving module includes a semi-trailer vehicle dynamics model established based on updated vehicle parameters and tire cornering stiffness, and is set to a feedforward + feedback control form.

[0021] As a further description of the application, the established semi-trailer vehicle dynamics model includes three degrees of freedom of lateral velocity, yaw angular velocity, and folding angle of the trailer relative to the tractor.

[0022] Wherein, the feedforward is calculated based on the transfer function of the semi-truck dynamics model, the final value theorem and the trajectory radius.

[0023] As a further description of the present application, in step 5, the steering angle of the steering wheel is obtained from the front wheel angle lookup table of the LQR controller solving module, and the front wheel angle is the sum of the feedforward front wheel angle and the feedback front wheel angle.

[0024] As a further description of the present application, the feedforward front wheel angle is obtained by combining the state equation transfer function of the semi-truck tracking model and the intermediate value theorem, and the feedback front wheel angle is calculated by the feedback gain matrix and the state variable obtained by iterative solving based on the Riccati equation.

[0025] Compared with the prior art, the technical effects of the present application are:

[0026] The present application provides an automatic driving semi-truck lateral trajectory tracking adaptive control method, which improves the accuracy of the semi-truck dynamics model by changing the center of mass position, yaw moment of inertia and tire cornering stiffness, etc. The feedback control output is constrained and processed to solve the oscillation or overshoot caused by the position jump and signal noise, and to ensure the smoothness and stability of the vehicle control. The present application provides a new idea for the motion control of the automatic driving semi-truck, and has important value for the design and optimization of the overall architecture of the automatic driving platform. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 The figure is the overall architecture of the control method of the present application;

[0028] Figure 2 The figure is the structure of the semi-truck in the present application;

[0029] Figure 3 The figure is the dynamics model of the semi-truck in the present application;

[0030] Figure 4 The figure is the trajectory tracking model of the semi-truck in the present application.

[0031] Figure 5 The figure is the curve of the tire cornering stiffness with the vertical load in the present application.

[0032] In the figure, 1. Towing vehicle, 2. Trailer. DETAILED DESCRIPTION

[0033] The present application will be described in detail below with reference to the accompanying drawings:

[0034] An automatic driving semi-truck lateral trajectory tracking adaptive control method, referring to Figures 1-5As shown, the system of the method comprises a vehicle parameter adaptive calculation module, a vertical load distribution module, a tire cornering stiffness calculation module, a weight coefficient adaptive adjustment module, an LQR controller solving module and a control output constraint module.

[0035] The method comprises the following steps:

[0036] Step 1: According to the load of the trailer, the mass center position and the yaw moment of inertia of the tractor and the trailer are calculated by the vehicle parameter adaptive calculation module;

[0037] Step 2: According to the mass center position of the tractor and the trailer, the vertical load at each axle is calculated by the vertical load distribution module;

[0038] Step 3: According to the vertical load at each axle and the number of tires, the cornering stiffness of each tire is updated in real time by the tire cornering stiffness calculation module;

[0039] Step 4: According to the vehicle speed, the weight coefficient of each state variable is adjusted in real time by the weight coefficient adaptive adjustment module;

[0040] Step 5: According to the vehicle state variables and the corresponding weight coefficients, the steering wheel angle is calculated by the LQR controller solving module.

[0041] The control output constraint module comprises feedback state constraints and overall constraints; the feedback state constraints are obtained by nonlinear adjustment and low-pass filtering to avoid output fluctuations caused by positioning jumps and signal noise; the overall output constraints are overall outputs that limit the front wheel angle by using the front wheel angle limit constraint calculated based on the static roll theory to ensure the roll safety of the vehicle.

[0042] Specifically, the embodiments are analyzed in detail for each module, and the specific implementation content is disclosed as follows:

[0043] 1. Vehicle parameter adaptive calculation module

[0044] The vehicle parameter adaptive calculation module inputs the load mass of the trailer and outputs the distance from the overall mass center of the tractor and the trailer to the front and rear axles and the overall yaw moment of inertia of each, which are calculated based on the moment balance and parallel shift theorem, wherein the fixed parameters are the sprung mass, the unsprung mass, the sprung mass moment of inertia of the tractor and the trailer, and the distance from the center of the sprung mass to each axle.

[0045] Specifically, the sprung mass of the tractor is defined as m s1 , the sprung mass of the trailer is defined as m s2 , the unsprung mass of the front axle of the tractor is defined as m uf , the unsprung mass of the rear axle of the tractor is defined as m ur , and the unsprung mass of the trailer is defined as mut , the sprung mass of the trailer is m t , the yaw moment of inertia of the tractor's sprung mass is I s1 , the yaw moment of inertia of the trailer's sprung mass is I s2 , the distance from the front axle of the tractor to the center of the sprung mass is a s1 , the distance from the rear axle of the tractor to the center of the sprung mass is b s1 , the wheelbase of the tractor is 1, and the distance from the hitch point to the center of the trailer's sprung mass is a s2 , the distance from the rear axle of the trailer to the center of the sprung mass is b s2 , the distance from the center of mass of the tractor to the hitch point is c, the height of the center of mass of the tractor's sprung mass is h1, and the height of the center of mass of the trailer's sprung mass is h2. Define the length of the trailer's sprung mass as 1 t , the width of the trailer's sprung mass is w t , the distance from the front side of the trailer's sprung mass to the hitch point is e.

[0046] Therefore, the total mass of the tractor is m1 = m s1 +m u1 +m u2 , the sum of the unsprung masses of the tractor is m u1 =m uf +m ur , the total sprung mass of the trailer is m ss2 =m s2 +m t , the total mass of the trailer is m2 = m ss2 +m ut .

[0047] (1) Calculate the position of the center of mass of the tractor and the overall yaw moment of inertia

[0048] Determine the distance from the front axle of the tractor to the center of mass of the unsprung mass:

[0049]

[0050] The distance from the center of mass of the tractor's unsprung mass to the center of mass of the sprung mass is:

[0051] l su1 =a u1 -a s1 (2)

[0052] The distance from the front axle of the tractor to the center of mass of the overall mass is:

[0053]

[0054] The distance from the rear axle of the tractor to the center of mass of the overall mass is:

[0055]

[0056] According to the parallel axis theorem, the yaw moment of inertia of the tractor as a whole is:

[0057]

[0058] (2) Calculate the mass center position of the trailer and the yaw moment of inertia of the whole

[0059] The distance from the center of gravity of the trailer load to the hinge point is:

[0060] a ss2 = l t / 2-e (6)

[0061] The distance from the center of gravity of the trailer load to the center of gravity of the spring is:

[0062] l ss2 = a ss2 -a s2 (7)

[0063] According to the new mass center position, the position of the hinge point to the mass center of the trailer is calculated:

[0064]

[0065] According to the new mass center position, the position of the trailer rear axle to the mass center of the trailer is calculated:

[0066]

[0067] The distance between the new mass center of the trailer and the original mass center is:

[0068]

[0069] According to the parallel axis theorem, the moment of inertia of the new spring center of the trailer is:

[0070]

[0071] The distance from the hinge point to the mass center of the trailer as a whole is:

[0072]

[0073] The distance from the non-spring mass of the trailer to the center of gravity of the spring load is:

[0074]

[0075] The distance from the trailer rear axle to the mass center of the whole is:

[0076]

[0077] According to the parallel axis theorem, the yaw moment of inertia of the trailer as a whole is:

[0078]

[0079] 2. Vertical load distribution module

[0080] The vertical load distribution module, input is the distance from the overall center of mass of the tractor and trailer to the front and rear axles, the overall mass and the longitudinal acceleration obtained by the vehicle parameter adaptive calculation module, output is the vertical load at each axle calculated according to the d'Alembert principle, wherein the fixed parameters are the heights of the spring in mass center of the tractor and trailer.

[0081] Specifically, the axle load is calculated based on the d'Alembert principle, wherein the vertical load at the articulation point is:

[0082]

[0083] The spring in mass load at the front axle of the tractor is:

[0084]

[0085] The spring in mass load at the rear axle of the tractor is:

[0086]

[0087] The spring in mass load at the rear axle of the trailer is:

[0088]

[0089] Therefore, the vertical load at the front axle of the tractor is:

[0090] F zf = F zfs + m uf g (20)

[0091] The vertical load at the rear axle of the tractor is:

[0092] F zr = F zrs + m ur g (21)

[0093] The vertical load at the rear axle of the trailer is:

[0094] F zt = F zts + m ut g (22)

[0095] Wherein, in the formula, a x is the longitudinal acceleration, and g is the acceleration of gravity.

[0096] 3. Tire cornering stiffness calculation module

[0097] The tire cornering stiffness calculation module has an input of vertical load and tire number at each axle and an output of tire cornering stiffness at each axle, which is the sum of tire cornering stiffness, assuming that each tire cornering stiffness is a cubic polynomial function of vertical load.

[0098] When the tire model is determined, the tire cornering stiffness is mainly affected by the vertical load and the road adhesion coefficient. For a semi-trailer vehicle, the vertical load of the wheel will be significantly different due to the change of the trailer load mass. Therefore, the present application assumes that the tire cornering stiffness is a cubic polynomial function of the vertical load, i.e.:

[0099]

[0100] where k represents the tire cornering stiffness, F z represents the vertical load of the corresponding wheel, and p0, p1, p2, and p3 are fitting parameters of the cubic polynomial.

[0101] The calibration values and fitting values under different vertical loads are shown in Table 1. Figure 5 Therefore, the cubic polynomial of the present application can accurately reflect the mapping relationship between the tire cornering stiffness and the vertical load. In addition, the idea of the tire cornering stiffness calculation module of the present application can also be realized based on two-dimensional table interpolation.

[0102] 4. Weight coefficient adaptive adjustment module

[0103] The weight coefficient adaptive adjustment module has an input of vehicle speed and an output of weight coefficients of state variables such as lateral position deviation, heading angle deviation, lateral velocity, yaw rate, folding angle velocity, and folding angle, with a mapping relationship of two-dimensional table or nonlinear function, which is obtained by interpolation or fitting of calibration parameters at different speeds.

[0104] Since the dynamics of the semi-trailer vehicle is different at different speeds, the present embodiment determines the weight coefficients of state variables such as vehicle position deviation, heading angle deviation, lateral velocity, yaw rate, folding angle velocity, and folding angle based on fitting functions or two-dimensional tables; for example, when the control performance under the condition of 5km / h to 80km / h is concerned, after determining the control variable weight coefficient R, the optimal state variable weight coefficients Q1 to Q6 at 5km / h, 10km / h, 20km / h, 30km / h, 40km / h, 50km / h, 60km / h, 70km / h, and 80km / h can be preferably calibrated, and then the coefficients of the fitting function are obtained offline, and further online interpolation calculation or offline table interpolation is used.

[0105] 5. LQR controller solving module

[0106] The LQR controller solving module includes a semi-trailer vehicle dynamics model established based on the updated vehicle parameters and tire cornering stiffness, and is set as a feedforward+feedback control form; the established semi-trailer vehicle dynamics model includes three degrees of freedom of lateral velocity, yaw angular velocity and folding angle of the trailer relative to the tractor; wherein the feedforward is obtained based on the transfer function of the semi-trailer vehicle dynamics model, the final value theorem and the trajectory radius calculation.

[0107] The LQR controller solving module inputs real-time vehicle position deviation, heading angle deviation, lateral velocity, yaw angular velocity, folding angle velocity and folding angle and the corresponding weight coefficients, and outputs a steering wheel angle. Wherein the steering wheel angle is obtained by looking up the table of the front wheel angle calculated by the control solution, the front wheel angle is the sum of the feedforward front wheel angle and the feedback front wheel angle, and the feedback front wheel angle is calculated by the feedback gain matrix and the state variables based on the iterative solution of the Riccati equation, and the specific calculation is disclosed as follows:

[0108] In consideration of the lateral motion, yaw motion and folding motion of the semi-trailer vehicle, the following 3-degree-of-freedom dynamics model is established:

[0109] m1a y1 =F yf +F yr -F hy (24)

[0110]

[0111] m2a y2 =F yt +F hy (26)

[0112]

[0113] In the formula, m1 is the mass of the tractor, m2 is the mass of the trailer, I z1 is the yaw moment of inertia of the tractor, I z2 is the yaw moment of inertia of the trailer, a1 is the distance from the front axle of the tractor to its center of mass, b1 is the distance from the rear axle of the tractor to its center of mass, c is the distance from the hinge point to the center of mass of the tractor, a2 is the distance from the hinge point to the center of mass of the trailer, b2 is the distance from the rear axle of the trailer to its center of mass, a y1 is the lateral acceleration of the tractor, a y2 is the lateral acceleration of the trailer, γ1 is the yaw angular velocity of the tractor, γ2 is the yaw angular velocity of the trailer, F yf is the lateral force of the front axle of the tractor, F yr is the lateral force of the rear axle of the tractor, F yt is the lateral force of the rear axle of the trailer, F hy is the interaction force at the hinge point.

[0114] When the folding angle is small, it can be considered that:

[0115] v x1 = v x2 = v x (28)

[0116] In the formula, v x1 is the longitudinal speed of the towing vehicle, v x2 is the longitudinal speed of the trailer, and v x is the longitudinal speed.

[0117] According to the geometric and motion relationship (the folding angle of the present embodiment is the angle between the trailer and the towing vehicle), the relationship between the lateral angular velocity of the trailer and the lateral angular velocity of the towing vehicle can be expressed as:

[0118]

[0119] In the formula, θ is the relative rotation angle between the towing vehicle and the trailer, that is, the folding angle.

[0120] At the hinge point, there are

[0121]

[0122] The lateral acceleration of the towing vehicle and the trailer is respectively:

[0123]

[0124]

[0125] According to the acceleration formula, the lateral velocity differential of the trailer can be expressed as:

[0126]

[0127] Simultaneous integration on the left and right can be obtained:

[0128]

[0129] In the formula, v y1 is the lateral speed of the towing vehicle, and v y2 is the lateral speed of the trailer.

[0130] The motion characteristics of the towing vehicle are similar to those of a two-axle vehicle, which are considered as the front axle as the steering axle and the rear axle as the non-steering axle. According to the coordinate system, the side slip angle of the front and rear tires of the towing vehicle is:

[0131]

[0132]

[0133] In the formula, δf for the front wheels of the tractor.

[0134] The expression for the trailer rear tire side slip angle is:

[0135]

[0136] According to equation (24), we have:

[0137] F hy = F yf + F yr - m1a y1 (38)

[0138] Substituting the above equation into equations (25) to (27), we have:

[0139]

[0140] m2a y2 = F yt + F yf + F yr - m1a y1 (40)

[0141]

[0142] After rearranging, we have:

[0143]

[0144] m1a y1 + m2a y2 = F yf + F yr + F yt (43)

[0145]

[0146] where,

[0147]

[0148] In the equation, k f , k r and k t are the side slip stiffness of each tire, respectively.

[0149] Rearranging equations (35) to (45), we have the state equation form of the dynamics model of the semi-trailer truck:

[0150]

[0151] where,

[0152]

[0153]

[0154] After sorting, we get:

[0155]

[0156] Where, A-M -1 G, B = M -1 F.

[0157] In order to express the relationship between the semi-trailer vehicle and the reference trajectory, the trajectory tracking model is established as shown in Figure 4 . Where, e y is the lateral distance from the mass center of the tractor to the preview point d p on the desired trajectory; e yo is the distance from the mass center of the tractor to the nearest point d of the desired trajectory; L is the preview distance; ψ is the actual heading angle of the tractor; ψ d is the desired heading angle.

[0158] According to the relative position relationship between the vehicle and the reference trajectory, the single-point preview-based trajectory tracking model is:

[0159]

[0160] In the formula, e y is the lateral position deviation; e ψ is the heading angle deviation; κ d is the road curvature of the desired trajectory.

[0161] In order to ensure the adaptability of the preview control, the higher the vehicle speed, the farther the road information needs to be concerned. Therefore, the adaptive preview distance is defined as:

[0162] L = v x Δt, L min ≤ L ≤ L max (49)

[0163] In the formula, Δt is the preview time; L min and L max are the lower limit and upper limit of the preview distance, respectively.

[0164] Combined with formula (47) and formula (48), the state variables of the system are defined as The control variable is u = δ f , the disturbance variable is w = -κ d v x , and the output variable is y = [e y , e ψ ] TAt this time, the state equation of the semi-trailer vehicle tracking model can be obtained as follows:

[0165]

[0166] wherein,

[0167]

[0168]

[0169] wherein, a ij = A(ij), b i = B(i);

[0170] In the present application, the control variable in formula (50) is the feedback front wheel steering angle, and the feedback gain matrix K and the state variable x are obtained by iterative solution of the Riccati equation, i.e.:

[0171] δ f = -Kx (51)

[0172] The front wheel steering angle δ ff is obtained by combining the transfer function form of formula (50) and the mean value theorem, i.e. on the basis of the optimal control amount δ f , a feedforward link is added, and the steady-state value of the lateral deviation amount when the system tends to be stable is 0. Therefore, the control amount of the actual front wheel steering angle is:

[0173] δ = δ f + δ ff (52)

[0174] According to the actual front wheel steering angle, the steering wheel steering angle can be obtained by table lookup, and the lateral trajectory tracking control of the semi-trailer truck is realized.

[0175] 6. Control output constraint module

[0176] Based on the present application, considering the control output fluctuation caused by positioning jump and signal noise, the present application performs constraint and filtering processing on the state error calculation and the measurement signal. Preferably, in order to avoid frequent fluctuation of the vehicle near the trajectory line during the trajectory tracking process, the present application performs nonlinear constraint on the lateral position deviation term, i.e. the first term in the feedback front wheel steering angle:

[0177]

[0178] wherein, δ f1K1 is the first item of the feedback gain matrix, x1 is the first item of the state matrix, tanh is a hyperbolic tangent function, and λ1 is a transition coefficient for reducing the fluctuation of the lateral position deviation. It should be noted that other technical means that can realize the idea of formula (53) in the present application also belong to the protection scope of the present application.

[0179] In order to reduce the influence degree of the yaw rate signal noise, the present application preferably uses a first-order low-pass filter for constraint processing, that is:

[0180]

[0181] In the formula, δ f4 K4 is the fourth item of the feedback gain matrix, x4 is the fourth item of the state matrix, k is a gain coefficient, τ is a time coefficient, and s is a Laplace operator.

[0182] In order to ensure the roll safety of the semi-trailer truck, the actual front wheel steering angle is constrained in the present application. According to the static roll theory, the lateral load transfer rate is defined as:

[0183]

[0184] In the formula, LTR is the lateral load transfer rate, h1 is the height of the center of mass of the tractor, and B1 is the wheelbase of the tractor.

[0185] Since the influence of the vehicle suspension system, road slope and the estimation accuracy of the center of mass height is ignored, the rollover warning LTR value can be selected as 0.8, and the lateral acceleration threshold value at this time is:

[0186]

[0187] The steady-state lateral acceleration can be approximately expressed as a y1 = v x γ1, and the steady-state yaw rate γ can be expressed as:

[0188] γ1 = G γ δ (57)

[0189] In the formula, G γ is the yaw rate steady-state gain obtained according to formula (46).

[0190] Therefore, the actual front wheel steering angle constraint considering the roll safety is:

[0191]

[0192] In the formula, sign is a sign function.

[0193] It should be noted that if the mass center height of the trailer can be accurately estimated, the lateral load transfer rate of the towing vehicle and the lateral load transfer rate of the trailer should be jointly constrained, and the minimum value of the two is selected as the output constraint of the whole. In addition, if other terms except the lateral position deviation term and the yaw angular velocity term are subject to similar disturbances, the logic of the present application can be applied to constraint processing.

[0194] The above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application. Other modifications or equivalent replacements to the technical solutions of the present application made by those skilled in the art should be covered in the scope of claims of the present application as long as they do not depart from the spirit and scope of the present application.

Claims

1. A method for adaptive control of lateral trajectory tracking of an autonomous semi-trailer truck, characterized in that: The system of the method comprises a vehicle parameter adaptive calculation module, a vertical load distribution module, a tire cornering stiffness calculation module, a weight coefficient adaptive adjustment module, and an LQR controller solving module. The method comprises the following steps: Step 1: according to the load of the trailer, the mass center positions and the yaw moment of inertia of the tractor and the trailer are calculated by the vehicle parameter adaptive calculation module; Step 2: according to the mass center positions of the tractor and the trailer, the vertical loads at each axle are calculated by the vertical load distribution module; Step 3: according to the vertical loads at each axle and the number of tires, the cornering stiffness of each tire is updated in real time by the tire cornering stiffness calculation module; Step 4: according to the vehicle speed, the weight coefficients of each state variable are adjusted in real time by the weight coefficient adaptive adjustment module; Step 5: according to the vehicle state variables and the corresponding weight coefficients, the steering wheel angle is calculated by the LQR controller solving module.

2. The automatic driving semi-trailer truck lateral trajectory tracking adaptive control method according to claim 1, characterized in that: The system of the method further comprises a control output constraint module for constraining and filtering the state error calculation and the measurement signal, wherein the control output constraint module comprises a feedback state constraint and an overall constraint; The feedback state constraint is a nonlinear adjustment and a low-pass filter, and the overall constraint is a front wheel angle limit constraint calculated based on the static roll theory.

3. The automatic driving semi-trailer truck lateral trajectory tracking adaptive control method according to claim 1, characterized in that: In step 1, the load mass of the trailer is input into the vehicle parameter adaptive calculation module, and the distance from the mass center of the tractor and the trailer to the front and rear axles and the yaw moment of inertia of each whole body are calculated according to the moment balance and the parallel shift theorem.

4. The automatic driving semi-trailer truck lateral trajectory tracking adaptive control method according to claim 1, characterized in that: In step 2, the distance from the mass center of the tractor and the trailer to the front and rear axles, the overall mass and the longitudinal acceleration are input into the vertical load distribution module, and the vertical loads at each axle of the tractor and the trailer are calculated according to the d'Alembert principle.

5. The adaptive control method for lateral trajectory tracking of an autonomous semi-trailer truck according to claim 1, wherein: In step 3, the vertical loads at each axle and the number of tires are input into the tire cornering stiffness calculation module, and the tire cornering stiffness at each axle, i.e., the sum of the cornering stiffness of each tire, is output.

6. The automatic driving semi-trailer truck lateral trajectory tracking adaptive control method according to claim 1, characterized in that: In step 4, the vehicle speed is input into the weight coefficient adaptive adjustment module, and the weight coefficients of each state variable, including the lateral position deviation, the heading angle deviation, the lateral velocity, the yaw angular velocity, the folding angular velocity and the folding angle, are output.

7. The automatic driving semi-trailer truck lateral trajectory tracking adaptive control method according to claim 1, characterized in that: In step 5, the LQR controller solving module comprises a semi-trailer vehicle dynamics model established based on the updated vehicle parameters and the tire cornering stiffness, and is set to a feedforward+feedback control form.

8. The automatic driving semi-trailer truck lateral trajectory tracking adaptive control method according to claim 7, characterized in that: The established semi-trailer vehicle dynamics model comprises three degrees of freedom, i.e., the lateral velocity, the yaw angular velocity and the folding angle of the trailer relative to the tractor. The feedforward is calculated based on the transfer function of the semi-trailer vehicle dynamics model, the final value theorem and the trajectory radius.

9. The automatic driving semi-trailer truck lateral trajectory tracking adaptive control method according to claim 8, characterized in that: In step 5, the steering wheel angle is obtained by looking up the front wheel angle table of the LQR controller solving module, and the front wheel angle is the sum of the feedforward front wheel angle and the feedback front wheel angle.

10. The automatic driving semi-trailer truck lateral trajectory tracking adaptive control method according to claim 9, characterized in that: The feedforward front wheel angle is obtained by combining the state equation transfer function of the semi-trailer tracking model and the intermediate value theorem, and the feedback front wheel angle is calculated by using the feedback gain matrix and the state variables obtained by iterative solving based on the Riccati equation.

Images