Multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability

Abstract

The present application belongs to the technical field of vehicle engineering, and particularly relates to a multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability. Step 1: a five-axle special vehicle dynamics model is established based on four degrees of freedom of the five-axle special vehicle, including longitudinal, lateral, yaw and roll, and the correctness of the five-axle special vehicle dynamics model is verified; Step 2: a five-axle special vehicle driving risk field model is established based on the relative motion state of surrounding vehicles and the special vehicle and the constant acceleration assumption, and a five-axle special vehicle dynamics constraint model is established based on the vehicle yaw stability and roll safety; Step 3: a MPC trajectory planner is designed with the five-axle special vehicle driving risk field model as a soft constraint and the five-axle special vehicle dynamics constraint model as a hard constraint, and a safe driving trajectory is generated. The method of the present application can generate a vehicle driving trajectory for realizing stable obstacle avoidance, and realize safe and stable high-speed driving of the multi-axle special vehicle.

Description

Technical Field

[0001] This invention belongs to the field of vehicle engineering technology, and specifically relates to a method for planning the safe trajectory of multi-axle special vehicles that takes into account obstacle avoidance stability. Background Technology

[0002] Multi-axle special vehicles are widely used in the transportation sector due to their heavy load-bearing capacity and wide-area mobility. Compared to two-axle light vehicles, multi-axle special vehicles have a higher center of gravity and greater mass, making them more prone to instability and rollover during high-speed maneuvers. Traditionally, the safety and stability control of multi-axle special vehicles relies primarily on driver experience and vehicle motion controllers. By designing a well-designed vehicle motion controller to track the desired vehicle state, yaw stability and anti-rollover control can be achieved to some extent. However, the large inertia of multi-axle special vehicles makes it difficult for their actuator response speed and output range to adapt to complex and changing driving environments. Therefore, relying solely on vehicle motion controllers for stability control often fails to achieve the desired results.

[0003] To achieve stable high-speed driving of multi-axle special vehicles, a safe trajectory planning method considering the vehicle's obstacle avoidance stability during trajectory tracking is designed. When performing trajectory planning, the dynamic performance of the vehicle is taken into account, and a path-velocity co-planning method is adopted to avoid the vehicle controller of the multi-axle special vehicle from exceeding the dynamic constraint boundary during trajectory tracking, which could lead to instability and rollover. Summary of the Invention

[0004] To address the above problems, this invention takes a five-axle special vehicle as the research object, considering the motion state of surrounding vehicles and the vehicle's high-speed maneuverability and obstacle avoidance stability. Based on a dynamic model, it realizes local path-velocity cooperative planning to achieve safe and stable driving of the five-axle special vehicle. First, a dynamic model of the five-axle special vehicle is established considering the four degrees of freedom of the vehicle: longitudinal, lateral, yaw, and roll. The correctness of the model is verified based on the dynamic response of the vehicle model. Second, considering the relative motion state of surrounding vehicles and the special vehicle, a vehicle driving risk field model is established based on the assumption of constant acceleration. At the same time, the dynamic stability domain of the vehicle during obstacle avoidance is considered, and vehicle driving constraints are established based on the phase plane of the center of mass sideslip angle and yaw rate. Third, an MPC trajectory planner is designed, using the risk field model as a soft constraint and vehicle stability as a hard constraint, while also considering trajectory smoothness, to generate feasible driving trajectories. Finally, the correctness of the proposed method is verified through simulation.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A method for planning the safe trajectory of a multi-axle special vehicle that considers obstacle avoidance stability includes:

[0007] Step 1: Establish a dynamic model of a five-axle special vehicle based on its four degrees of freedom (longitudinal, lateral, yaw, and roll), and verify the correctness of the dynamic model.

[0008] Step 2: Based on the five-axle special vehicle dynamics model established in Step 1, and referring to the relative motion states of surrounding vehicles and special vehicles, a driving risk field model for the five-axle special vehicle is established based on the constant acceleration assumption. At the same time, a dynamic constraint model for the five-axle special vehicle is established based on the vehicle's yaw stability and roll safety.

[0009] Step 3: Using the five-axle special vehicle driving risk field model established in Step 2 as soft constraints and the five-axle special vehicle dynamics constraint model as hard constraints, design an MPC trajectory planner to generate a safe driving trajectory.

[0010] Preferably, step 1 includes:

[0011] Step 1.1: Establish a four-degree-of-freedom vehicle dynamics model in the vehicle coordinate system, and establish a five-axle vehicle kinematics model in the geodetic coordinate system;

[0012] Step 1.2: Establish a tire model based on the Dugoff model;

[0013] Step 1.3: Establish a steering model for a five-axle special vehicle;

[0014] Step 1.4: Use Matlab / Simulink software to build a four-degree-of-freedom vehicle model and TruckSim software to build a five-axle vehicle model. By comparing the dynamic response of the two models under the same given input conditions, the accuracy of the four-degree-of-freedom vehicle model is verified.

[0015] Preferably, step 1.1 includes:

[0016] Step 1.1.1: Establish a five-axle special vehicle monorail model in the vehicle coordinate system. The four-degree-of-freedom vehicle dynamics equations are expressed as follows:

[0017] Longitudinal motion differential equation:

[0018]

[0019] Lateral motion differential equation:

[0020]

[0021] Differential equation of yaw motion:

[0022]

[0023] Differential equation of roll motion:

[0024]

[0025] Where m is the total vehicle mass; V x V y These represent the longitudinal velocity and lateral velocity of the vehicle along the X-axis and Y-axis in the vehicle coordinate system, respectively. Indicates the yaw rate of the vehicle; I z I is the yaw moment of inertia of the vehicle. x Let F be the vehicle's moment of inertia about its roll center. xi and F yi Let L represent the longitudinal and lateral forces exerted on the vehicle from the ground through the i-th (i = 1, 2, ..., 5) tire in the vehicle coordinate system; i (i = 1, 2, ..., 5) represents the distance from each axle of the vehicle to the center of gravity, θ represents the vehicle roll angle, h is the vertical distance from the vehicle's center of gravity to the ground, and a y Let be the lateral acceleration at the vehicle's center of gravity, expressed as Due to the presence of the suspension, the vehicle body is simultaneously subjected to anti-roll moment, using M R It indicates that its size is K and D represent the stiffness coefficient and damping coefficient of the vehicle body roll, respectively;

[0026] Step 1.1.2: Establish a five-axle vehicle kinematic model in the geodetic coordinate system:

[0027]

[0028] The above formula, The y-axis represents the vehicle's heading angle, and X and Y represent the vehicle's lateral and longitudinal positions in the geodetic coordinate system, respectively.

[0029] Preferably, step 1.2 includes:

[0030] Step 1.2.1: The tire force mapping relationship between the vehicle coordinate system and the tire coordinate system is as follows:

[0031]

[0032] Where, δ i F represents the steering angle of the i-th axis. li F ci This represents the longitudinal and lateral forces acting on the tire along the i-th axis in the tire coordinate system.

[0033] Step 1.2.2: Using the Dugoff tire model, establish the relationship between tire forces and vehicle dynamics:

[0034]

[0035] Among them, Cli C ci Let α represent the longitudinal stiffness and lateral stiffness of the tire on the i-th axis, respectively. i Let S be the tire slip angle of the tire on the i-th axis. i Let F be the tire slip ratio of the i-th axle, μ be the ground adhesion coefficient, and F be the tire slip ratio of the i-th axle. zi Let λ be the longitudinal force of the tire on the i-th axis. i This is the tire force correction factor;

[0036] The tire force calculation based on equation (7) uses the tire slip angle and tire slip ratio as inputs, and the calculation method is as follows:

[0037] Tire slip angle:

[0038]

[0039] Tire slip ratio:

[0040]

[0041] The sign convention is that the first two axes are positive, and the last three axes are negative; ω i The wheel rotation speed, r w v is the tire rolling radius. li The longitudinal velocity of the wheel center is expressed as:

[0042]

[0043] Preferably, step 1.3 specifically includes:

[0044] The steering angle of a five-axle special vehicle is based on the Ackermann steering model. The first, second, fourth, and fifth axles of the five-axle vehicle are the steering axes, and the instantaneous center is located on the extension line of the third axle. For the monorail model, the following relationship is established:

[0045]

[0046] and then:

[0047]

[0048] Preferably, the five-axle special vehicle driving risk field model in step 2 is specifically as follows:

[0049] Step 2.1.1: Based on the bivariate Gaussian distribution function, establish the predicted risk field for five-axle special vehicles:

[0050]

[0051] Among them, X ret Y ret Z represents the relative position between this vehicle and surrounding vehicles; Z is the collision tolerance coefficient, σ x σy As a risk field moderating factor;

[0052] For the i-th vehicle surrounding this vehicle, the relative coordinates are defined as follows, based on the vehicle's relative direction of travel:

[0053]

[0054] Let (x0, y0) represent the heading angle between this vehicle and the i-th surrounding vehicle, and (x0, y0) be the coordinates of this vehicle in the geodetic coordinate system. i y i Let be the position of the i-th vehicle in the geodetic coordinate system;

[0055] Step 2.1.2: Define the coefficient σ as the risk field adjustment factor, and the motion state of the vehicle at a certain moment is... These represent the vehicle's position, speed, acceleration, and heading angle, respectively, with the motion states of surrounding vehicles as follows:

[0056] The formula for calculating the risk field adjustment factor σ is:

[0057]

[0058] In the formula, Here, Δx and Δy are constant coefficients, representing the differences in the X and Y directions between the vehicle and surrounding vehicles, respectively. x0 a y0 V represents the lateral and longitudinal accelerations of the vehicle. x0 V y0 C represents the lateral and longitudinal speeds of the vehicle. x C y Determined based on actual vehicle speed, using the following calculation formula:

[0059]

[0060] In the above formula, Indicates the maximum and minimum speed limits for the road;

[0061] The road boundary is defined as follows:

[0062]

[0063] In the formula, Z is a constant coefficient, Y1 and Y2 are the shortest distances from the vehicle to the two sides of the lane boundary, respectively, and σ y The definition method is the same as that of equation (15);

[0064] Step 2.1.3: To establish a risk field model for the next n time steps, based on the assumption of constant acceleration and the current estimate of the other vehicle's motion state, the motion state of the other vehicle at the next time step is:

[0065]

[0066] Where T is the time interval between two motion moments, X i Y i This represents the lateral and longitudinal positions of the i-th vehicle in the surrounding geodetic coordinate system;

[0067] If the acceleration is a constant value 'a' for a short period of time x a y The motion state predicted at time t for the next n times is as follows:

[0068]

[0069] Preferably, the five-axle special vehicle dynamics constraint model in step 2 includes vehicle stability constraints and vehicle roll constraints;

[0070] The specific vehicle stability constraints are as follows:

[0071] Step 2.2.1: The vehicle dynamics equations are simplified to two degrees of freedom:

[0072]

[0073] In the formula, β is the centroid sideslip angle, and its magnitude is V. y / V x When the vehicle is in a stable state, the rate of change of the sideslip angle is 0, and the maximum adhesion provided by the ground is expressed as: The maximum stable yaw rate is:

[0074]

[0075] μ is the ground adhesion coefficient.

[0076] Considering the center of gravity sideslip angle constraint and the tire force saturation of the non-steering axle, and combining equation (8), we get:

[0077]

[0078] The tire saturation sideslip angle is approximately calculated using the following formula:

[0079]

[0080] Among them, C α For tire lateral stiffness, F z Let μ be the vertical force of the tire and μ be the ground adhesion coefficient. Since the five-axle vehicle body and tire suspension form a statically indeterminate system, the vertical force of the third axle tire is expressed as:

[0081]

[0082] In the formula, m bFor the sprung mass of the vehicle, m w The unsprung mass of the vehicle is defined using the same sign convention as in steps 1, 2, and 3.

[0083] Step 2.2.2: Vehicle roll restraint

[0084] The vehicle roll constraint is based on the offset of the zero-moment point position. If there is a point on the ground that is offset by y relative to the vehicle's center of mass... zm If the sum of the vehicle's weight, inertial force, and the force exerted by the ground on the vehicle body relative to this point produces a zero lateral tilting moment, then the following relationship can be obtained:

[0085]

[0086] Because the body roll angle is small, based on the assumption of a small angle, then:

[0087]

[0088] By limiting the vehicle's center of gravity offset y zm To restrain vehicle roll:

[0089] -y ZM，max ≤y ZM ≤y ZM，max (27)

[0090] Ultimately, the constraints on vehicle driving stability can be expressed as:

[0091]

[0092] Preferably, the MPC trajectory planner design in step 3 is specifically as follows:

[0093] Step 3.1.1: Record the vehicle dynamics model established in Step 1 as follows:

[0094]

[0095] In the formula, ξ represents the state variable of this trajectory planner system, which is... u represents the system control variable, which is...

[0096] Step 3.1.2: The trajectory planner system in Step 3.1.1 is a nonlinear system. We linearize it at the current running time and define the system output as... The trajectory planner system is then represented as:

[0097]

[0098] In the formula, the matrix matrix Let d be a Jacobian matrix. t This is for linearization error;

[0099] Step 3.1.3: Based on the current time t, generate the driving trajectory for the next N time moments. Discretize the above system using the forward Euler method, and then represent the system as:

[0100]

[0101] In the formula, T represents the discrete time, and the linearization error after discretization is expressed as:

[0102]

[0103] Step 3.1.4: Replace the control quantity u(t) with the control increment Δu(t). By directly constraining the control increment, smooth trajectory planning of the vehicle can be achieved. If the control increment Δu(t) = u(t) - u(t-1), then equation (30) is transformed into:

[0104]

[0105] Equation (34) can be expressed as:

[0106]

[0107] If the prediction time domain is H p The control time domain is H c And H p >H c Assuming at the current time t, given the control time domain [t, t+H] c -1] A series of control quantities Δu1, Δu2, ..., Δu M Outside the control time domain [t+H c ,t+H p -1], given a control quantity of 0, the prediction output in the prediction time domain is expressed as:

[0108]

[0109] According to equation (36), the optimal output trajectory is calculated with the current optimized series of input quantities as unknowns.

[0110] Preferably, the optimal driving trajectory generation in step 3 specifically involves:

[0111] Step 3.2.1: Using the driving risk field model and dynamic constraint model from Step 2 as constraints, the planner iteratively solves the optimal problem J in the prediction time domain within a solution time T. The optimization problem J is defined as:

[0112]

[0113] In the formula, Tra represents the driving trajectory. ref For reference driving trajectory input, V x V is the longitudinal velocity of the vehicle. x ref For reference vehicle speed, P is the motion risk field generated in the prediction time domain based on the observation results of the motion state of other vehicles and the road boundary, calculated by equations (13)-(19);

[0114] Step 3.2.2: κ is the curvature of the vehicle's trajectory at a certain moment in the predicted time domain, calculated as follows:

[0115]

[0116] In the formula, X and Y are cubic spline fitting curves of the trajectory points output by the planner. Their boundary conditions are given by constraining the yaw angle state of the vehicle at the starting and ending points. Then, at time t, the output position of the planner is:

[0117]

[0118] The parameters in the formula are the fitting coefficients of the cubic spline curve, which have no physical meaning.

[0119] Step 3.2.3: If the maximum permissible lateral acceleration of the vehicle is |a lat | max Then, the vehicle speed is constrained as follows:

[0120]

[0121] In the formula, v sign The speed limit for the lane involves vehicle ride comfort and passenger comfort, and combined with the safety constraints of equation (27), the optimization problem is expressed as:

[0122]

[0123] In the prediction time domain H p Inside, the planner generates a continuous driving trajectory of the vehicle by solving the optimization problem (42), and the lower controller realizes the automatic driving of the vehicle on the structured road by tracking the driving trajectory in real time.

[0124] Compared with the prior art, the beneficial effects of the present invention are:

[0125] (1) The traditional path-velocity decoupled trajectory planning method cannot guarantee the obstacle avoidance stability of the vehicle when cruising at high speed. The present invention adopts the path-velocity collaborative planning method, which takes into account the dynamic constraints of the vehicle itself and the driving state of the surrounding vehicles, and generates a vehicle driving trajectory that achieves stable obstacle avoidance, so as to realize the safe and stable high-speed driving of multi-axle special vehicles.

[0126] (2) Traditional trajectory planning methods do not consider the dynamic characteristics of surrounding vehicles when establishing the risk field, which cannot guarantee the safety of vehicle driving. This invention is based on the binary Gaussian distribution, considers the kinematic characteristics of surrounding vehicles when driving, establishes a risk field model, generates the risk field at N future times, and uses this as a constraint to generate a safe vehicle driving trajectory. Attached Figure Description

[0127] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0128] In the attached diagram:

[0129] Figure 1 This is a five-axle vehicle monorail model;

[0130] Figure 2 shows a comparison of vehicle models: a. Steering wheel angle input; b. Model center of gravity sideslip angle response under different excitations; c. Yaw rate response under different excitations;

[0131] Figure 3 shows the obstacle risk field: a. obstacle risk field; b. road boundary risk field; c. vehicle motion risk field;

[0132] Figure 4 This is a schematic diagram of stability constraints;

[0133] Figure 5 For trajectory planning, a timeline diagram is needed;

[0134] Figure 6 To plan and verify the process;

[0135] Figure 7 The distribution of vehicle locations around the straight-ahead road;

[0136] Figure 8 The vehicle trajectory at a speed of 18 m / s;

[0137] Figure 9 The vehicle trajectory at a speed of 25 m / s;

[0138] Figure 10 shows the spatiotemporal diagrams of trajectory planning at different vehicle speeds: a. Spatiotemporal diagram at a vehicle speed of 18 m / s; b. Spatiotemporal diagram at a vehicle speed of 25 m / s;

[0139] Figure 11 shows the vehicle phase trajectory and roll angle at different vehicle speeds: a. Vehicle phase trajectory at 18 m / s, b. Vehicle roll angle at 18 m / s; c. Vehicle phase trajectory at 25 m / s, d. Vehicle roll angle at 25 m / s.

[0140] Figure 12 The vehicle's trajectory under curved road conditions;

[0141] Figure 13 A spatiotemporal diagram of a vehicle operating under curve conditions;

[0142] Figure 14(a) shows the phase trajectory of the vehicle when it is driving on a curve, and Figure 14(b) shows the roll angle of the vehicle when it is driving on a curve.

[0143] Figure 15 Comparison of vehicle trajectories under different constraints;

[0144] Figure 16(a) shows the phase trajectory of the vehicle when it is traveling without constraints, and Figure 16(b) shows the roll angle of the vehicle when it is traveling without constraints.

[0145] Figure 17 This is a flowchart of the method of the present invention. Detailed Implementation

[0146] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0147] Example:

[0148] See attached document Figure 1-1 As shown in Figure 6, a method for planning the safe trajectory of a multi-axle special vehicle considering obstacle avoidance stability includes:

[0149] Step 1: Establish a dynamic model of the five-axle special vehicle based on its four degrees of freedom (longitudinal, lateral, yaw, and roll), and verify the correctness of the dynamic model. Specifically:

[0150] Step 1.1: Establish a four-degree-of-freedom vehicle dynamics model in the vehicle coordinate system, and establish a five-axle vehicle kinematics model in the geodetic coordinate system;

[0151] Step 1.1.1: Considering the four degrees of freedom of a five-axle vehicle—longitudinal, lateral, yaw, and roll—establish a single-track model of the five-axle special vehicle in the vehicle coordinate system, such as... Figure 1 As shown. The four-degree-of-freedom vehicle dynamics equations are expressed as:

[0152] Longitudinal motion differential equation:

[0153]

[0154] Lateral motion differential equation:

[0155]

[0156] Differential equation of yaw motion:

[0157]

[0158] Differential equation of roll motion:

[0159]

[0160] Where m is the total vehicle mass; V x V y These represent the longitudinal velocity and lateral velocity of the vehicle along the X-axis and Y-axis in the vehicle coordinate system, respectively. Indicates the yaw rate of the vehicle; I z I is the yaw moment of inertia of the vehicle. x Let F be the vehicle's moment of inertia about its roll center. xi and F yi Let L represent the longitudinal and lateral forces exerted on the vehicle from the ground through the i-th (i = 1, 2, ..., 5) tire in the vehicle coordinate system; i (i = 1, 2, ..., 5) represents the distance from each axle of the vehicle to the center of gravity, θ represents the vehicle roll angle, h is the vertical distance from the vehicle's center of gravity to the ground, and a y Let be the lateral acceleration at the vehicle's center of gravity, expressed as Due to the presence of the suspension, the vehicle body is simultaneously subjected to anti-roll moment, using M R It indicates that its size is K and D represent the stiffness coefficient and damping coefficient of the vehicle body roll, respectively;

[0161] Step 1.1.2: Establish a five-axle vehicle kinematic model in the geodetic coordinate system:

[0162]

[0163] The above formula, The y-axis represents the vehicle's heading angle, and X and Y represent the vehicle's lateral and longitudinal positions in the geodetic coordinate system, respectively.

[0164] Step 1.2: Establish a tire model based on the Dugoff model;

[0165] Step 1.2.1: The tire force mapping relationship between the vehicle coordinate system and the tire coordinate system is as follows:

[0166]

[0167] Where, δ i F represents the steering angle of the i-th axis. li F ci This represents the longitudinal and lateral forces acting on the tire along the i-th axis in the tire coordinate system.

[0168] Step 1.2.2: Using the Dugoff tire model, establish the relationship between tire forces and vehicle dynamics:

[0169]

[0170] Among them, C li C ci Let α represent the longitudinal stiffness and lateral stiffness of the tire on the i-th axis, respectively. i Let S be the tire slip angle of the tire on the i-th axis. i Let F be the tire slip ratio of the tire on the i-th axis. μ is the ground adhesion coefficient, F zi Let λ be the longitudinal force of the tire on the i-th axis. i This is the tire force correction factor;

[0171] The tire force calculation based on equation (7) uses the tire slip angle and tire slip ratio as inputs, and the calculation method is as follows:

[0172] Tire slip angle:

[0173]

[0174] Tire slip ratio:

[0175]

[0176] The sign convention is that the first two axes are positive, and the last three axes are negative; ω i The wheel rotation speed, r w v is the tire rolling radius. li The longitudinal velocity of the wheel center is expressed as:

[0177]

[0178] Step 1.3: Establish the steering model of the five-axle special vehicle, specifically as follows:

[0179] The steering angle of the five-axle special vehicle is based on the Ackermann steering model, which means that all wheels have the same instantaneous steering center during steering, so that the wheels are approximately in a pure rolling state. The first, second, fourth, and fifth axles of the five-axle vehicle are steering axles, and their instantaneous centers are located on the extension line of the third axle. Figure 1 At this point, for the single-track model, the following relationship exists:

[0180]

[0181] and then:

[0182]

[0183] Step 1.4: A four-degree-of-freedom (DOF) vehicle model was built using Matlab / Simulink software, and a five-axle vehicle model was built using TruckSim software. The accuracy of the four-DOF vehicle model was verified by comparing the dynamic responses of the two models under the same given input conditions. Some parameters of the vehicle model are shown in Table 1.

[0184] Table 1 Vehicle Model Parameters

[0185]

[0186] In the early stages, combined with real vehicle road tests, SPEEDBOX-INS was used to collect attitude data of the real vehicle under steering excitation. By comparing the TruckSim model used with the dynamic response of the real vehicle, the accuracy of the TruckSim model was verified.

[0187] The two models were simulated and verified under the excitation of steering wheel angle step input and sinusoidal input, with the vehicle speed set to 72km / h. The simulation results are shown in Figure 2.

[0188] It can be seen that, under the same given input excitation, the simplified four-degree-of-freedom dynamic model exhibits good consistency with the TruckSim model in terms of the sideslip angle response and yaw rate response, verifying the correctness of the proposed model. The next step of this invention will combine the established simplified dynamic model with consideration of the vehicle's own and surrounding vehicle dynamic states during driving to generate a safe and stable driving trajectory for the five-axle special vehicle.

[0189] Step 2: Based on the five-axle special vehicle dynamics model established in Step 1, and referring to the relative motion states of surrounding vehicles and special vehicles, a driving risk field model for the five-axle special vehicle is established based on the assumption of constant acceleration. Simultaneously, a dynamic constraint model for the five-axle special vehicle is established based on the vehicle's yaw stability and roll safety. Specifically:

[0190] Step 2.1.1: Based on the bivariate Gaussian distribution function, establish the predicted risk field for five-axle special vehicles:

[0191]

[0192] Among them, X ret Y ret σ represents the relative position between this vehicle and surrounding vehicles; Z is the collision tolerance coefficient. x σ y As a risk field moderating factor;

[0193] For the i-th vehicle surrounding this vehicle, the relative coordinates are defined as follows, based on the vehicle's relative direction of travel:

[0194]

[0195] Let (x0, y0) represent the heading angle between this vehicle and the i-th surrounding vehicle, and (x0, y0) be the coordinates of this vehicle in the geodetic coordinate system. i y i Let be the position of the i-th vehicle in the geodetic coordinate system;

[0196] Step 2.1.2: Considering that the difference in motion states between the obstacle vehicle and the vehicle itself will lead to different collision risks, and that the sign of the relative velocity or acceleration between the vehicles determines whether a collision will occur in the future (Table 2), the coefficient σ is defined as the risk field adjustment factor. The motion state of the vehicle at a certain moment is...

[0197] These represent the vehicle's position, speed, acceleration, and heading angle, respectively, with the motion states of surrounding vehicles as follows: The collision risk between vehicles varies according to their different states of motion, as shown in Table 2.

[0198] Table 2 Relationship between vehicle collision risk and vehicle motion state

[0199]

[0200] Taking into account the impact of the relative motion between vehicles on collision risk, the formula for calculating the risk field adjustment factor σ is as follows:

[0201]

[0202] In the formula, Here, Δx and Δy are constant coefficients, representing the differences in the X and Y directions between the vehicle and surrounding vehicles, respectively. x0 a y0 V represents the lateral and longitudinal accelerations of the vehicle. x0 V y0 C represents the lateral and longitudinal speeds of the vehicle. x C y Determined based on actual vehicle speed, using the following calculation formula:

[0203]

[0204] In the above formula, Indicates the maximum and minimum speed limits for the road;

[0205] For structured roads, in addition to dynamic obstacles such as surrounding vehicles, static obstacles include road boundaries and objects falling into the center of the road. Considering the motion state of the five-axle special vehicle itself, the static obstacle adjustment factor is defined in the same way as above, where both velocity and acceleration values ​​are 0. The road boundary is defined as:

[0206]

[0207] In the formula, Z is a constant coefficient, Y1 and Y2 are the shortest distances from the vehicle to the two sides of the lane boundary, respectively, and σ y The definition method is the same as that of equation (15); the generated risk field is shown in Figure 3.

[0208] Figure 3(c) is a schematic diagram of the risk field distribution of surrounding vehicles in different motion states when a five-axle vehicle is traveling at a constant speed of 20 m / s. It can be seen that the differences in the speed and acceleration of surrounding vehicles lead to differences in the distribution of the collision risk field. For vehicle B, the vehicle speed is greater than that of the five-axle vehicle, so there is no collision risk when the five-axle vehicle is traveling behind vehicle B. However, when the five-axle vehicle is traveling in front of vehicle B, the collision risk increases as the distance decreases. The opposite is true for vehicle C.

[0209] Step 2.1.3: To establish a risk field model for the next n time steps, based on the assumption of constant acceleration and the current estimate of the other vehicle's motion state, the motion state of the other vehicle at the next time step is:

[0210]

[0211] Where T is the time interval between two motion moments, X i Y i Indicates the lateral and longitudinal positions of the i-th vehicle in the surrounding geodetic coordinate system.

[0212] If the acceleration is a constant value 'a' for a short period of time x a y The motion state predicted at time t for the next n times is as follows:

[0213]

[0214] The five-axle special vehicle dynamics constraint model in step 2 includes vehicle stability constraints and vehicle roll constraints;

[0215] The specific vehicle stability constraints are as follows:

[0216] Step 2.2.1: Vehicle stability constraints mainly consider the tire instability boundary. For example... Figure 4 As shown, this paper generates the vehicle stability envelope based on the yaw rate-center of mass sideslip angle phase plane.

[0217] Considering that the vehicle speed fluctuation is small during cruising, assuming the vehicle travels at a constant speed and ignoring the effect of roll, the vehicle dynamics equations simplify to two degrees of freedom:

[0218]

[0219] In the formula, β is the centroid sideslip angle, and its magnitude is V. y / Vx, when the vehicle is in a stable state, the rate of change of the sideslip angle is 0, and the maximum adhesion provided by the ground is given by the ground adhesion coefficient μ. The maximum stable yaw rate is expressed as:

[0220]

[0221] To characterize the vehicle's stable state and simplify the constraint model, the centroid sideslip angle constraint is combined with the tire force saturation condition of the non-steering axle, and equation (8) is used to obtain:

[0222]

[0223] The tire saturation sideslip angle is approximately calculated using the following formula:

[0224]

[0225] Among them, C α For tire lateral stiffness, F z The vertical force of the tires is given. Since the vehicle body and tire suspension of a five-axle vehicle form a statically indeterminate system, considering deformation compatibility conditions, the vertical force of the tires on the third axle is expressed as:

[0226]

[0227] In the formula, m b For the sprung mass of the vehicle, m w The unsprung mass of the vehicle is defined using the same sign convention as in steps 1, 2, and 3.

[0228] Step 2.2.2: Vehicle roll restraint

[0229] The vehicle roll constraint is based on the offset of the zero-moment point position. If there is a point on the ground that is offset by y relative to the vehicle's center of mass... zm If the sum of the vehicle's weight, inertial force, and the force exerted by the ground on the vehicle body relative to this point produces a zero lateral tilting moment, then the following relationship can be obtained:

[0230]

[0231] Because the body roll angle is small, based on the assumption of a small angle, then:

[0232]

[0233] By limiting the vehicle's center of gravity offset y zm To restrain vehicle roll:

[0234] -y ZM，max ≤y ZM ≤y ZM，max (27)

[0235] Ultimately, the constraints on vehicle driving stability can be expressed as:

[0236]

[0237] Preferably, the MPC trajectory planner design in step 3 is specifically as follows:

[0238] Step 3: Using the five-axle special vehicle driving risk field model established in Step 2 as soft constraints and the five-axle special vehicle dynamics constraint model as hard constraints, design an MPC trajectory planner to generate a safe driving trajectory.

[0239] Step 3.1.1: Record the vehicle dynamics model established in Step 1 as follows:

[0240]

[0241] In the formula, ξ represents the state variable of this trajectory planner system, which is... u represents the system control variable, which is [δ1, ω]. i ].

[0242] Step 3.1.2: The system in Step 3.1.1 is a nonlinear system. Linearize it at the current operating moment and define the system output as... The system is then represented as:

[0243]

[0244] In the formula, the matrix matrix Let d be a Jacobian matrix. t This is for linearization error;

[0245] Step 3.1.3: Based on the current time t, generate the driving trajectory for the next N time moments. Discretize the above system using the forward Euler method, and then represent the system as:

[0246]

[0247] In the formula, T represents the discrete time, and the linearization error after discretization is expressed as:

[0248]

[0249] Step 3.1.4: Replace the control quantity u(t) with the control increment Δu(t). By directly constraining the control increment, smooth trajectory planning of the vehicle can be achieved. If the control increment Δu(t) = u(t) - u(t-1), then equation (30) is transformed into:

[0250]

[0251] Equation (34) can be expressed as:

[0252]

[0253] If the prediction time domain is H p The control time domain is H c And H p >Hc Assuming at the current time t, given the control time domain [t, t+H] c -1] A series of control quantities Δu1, Δu2, ..., Δu M Outside the control time domain [t+H c ,t+H p -1], given a control quantity of 0, the prediction output in the prediction time domain is expressed as:

[0254]

[0255] According to equation (36), the optimal output trajectory is calculated with the current optimized series of input quantities as unknowns.

[0256] Step 3, the generation of the optimal driving trajectory, is specifically as follows:

[0257] When a five-axle special vehicle maintains its lane, the designed MPC planner acquires the vehicle's motion state and, based on predetermined time intervals, iteratively solves for the optimal trajectory that meets dynamic constraints. The process is as follows: Figure 5 As shown.

[0258] Step 3.2.1: Using the driving risk field model and dynamic constraint model from Step 2 as constraints, the planner iteratively solves the optimal problem J in the prediction time domain within a solution time T. The optimization problem J is defined as:

[0259]

[0260] In the formula, Tra represents the driving trajectory. ref For reference driving trajectory input, V x V is the longitudinal velocity of the vehicle. x ref For reference vehicle speed, P is the motion risk field generated in the prediction time domain based on the observation results of the motion state of other vehicles and the road boundary, calculated by equations (13)-(19);

[0261] Step 3.2.2: κ is the curvature of the vehicle's trajectory at a certain moment in the predicted time domain, calculated as follows:

[0262]

[0263] In the formula, X and Y are cubic spline fitting curves of the trajectory points output by the planner. Their boundary conditions are given by constraining the yaw angle state of the vehicle at the starting and ending points. Then, at time t, the output position of the planner is:

[0264]

[0265] Step 3.2.3: If the maximum permissible lateral acceleration of the vehicle is |a lat | maxThen, the vehicle speed is constrained as follows:

[0266]

[0267] In the formula, v sign The speed limit for the lane involves vehicle ride comfort and passenger comfort, and combined with the safety constraints of equation (27), the optimization problem is expressed as:

[0268]

[0269] In the prediction time domain H p Inside, the planner generates a continuous driving trajectory of the vehicle by solving the optimization problem (42), and the lower controller realizes the automatic driving of the vehicle on the structured road by tracking the driving trajectory in real time.

[0270] Simulation experiment verification

[0271] To verify the correctness of the method, a co-simulation platform was built using Matlab / Simulink and TruckSim software to test the trajectory generation method in straight road and continuous curve scenarios. The simulation scenarios included road boundary lines, surrounding vehicles, and static obstacles in the center of the road. The upper-level planner was implemented using the method proposed in this paper, and the lower-level PID controller was used for trajectory tracking. The planning verification principle is as follows: Figure 6 As shown.

[0272] The parameters for the trajectory planner are shown in Table 3. The vehicle model parameters are the same as those in Table 1.

[0273] Table 3 Planner Parameters

[0274]

[0275] Straight road driving conditions

[0276] If a five-axle special vehicle is cruising on a straight road, at time t, the number of surrounding vehicles is set to 3, and their relative positions are as follows: Figure 7 As shown, there is a static obstacle A_obs in front of the road.

[0277] The reference speeds for the five-axle special vehicles were set at 18 m / s and 20 m / s, respectively. Tests were conducted on the vehicles at both cruising speeds, and the motion states of each vehicle are shown in Table 4. The data format in the table is [Pi,,V]. i ,,a i ], i = A, B, C. The planned trajectories at different vehicle speeds are as follows: Figure 8 , Figure 9 As shown.

[0278] Table 4 Vehicle Status under Straight-Ahead Driving Conditions

[0279]

[0280] exist Figure 8 and Figure 9 The upper figure shows the vehicle's trajectory planned using the method presented in this paper and the vehicle's final position after the planning process is completed. The lower figure shows the actual speed of the vehicle.

[0281] Figure 8 In the scenario, when the vehicle is traveling at 18 m / s, it begins to decelerate to reduce the risk of collision due to an obstacle Aobs ahead. After passing the obstacle (at a distance of 100 m), vehicle B ahead is traveling at a constant speed of 10 m / s, while vehicle A further away is traveling at the same constant speed. To avoid colliding with both vehicles A and B, the vehicle first moves away from vehicle B while maintaining its own stability. When the vehicle approaches vehicle A on its left front, it maintains a certain longitudinal distance from vehicle A and accelerates to overtake vehicle B (between 100 and 220 m). Then, at 270 m, the vehicle continues to accelerate, moving away from both vehicles A and B to avoid a collision. Since vehicle C maintains a constant forward speed of 20 m / s, it has no impact on the vehicle ahead.

[0282] right Figure 9 The vehicle is traveling forward at a constant speed of 25 m / s. Under this condition, since vehicle A is far away from the vehicle and is accelerating, and vehicle C's speed is greater than the vehicle's speed, the two vehicles have no impact on the vehicle's trajectory. After decelerating to reduce the risk of collision and bypassing obstacle Aobs, the vehicle overtakes vehicle B at 250 m and begins to accelerate until it reaches the reference speed. The spacetime diagram of the vehicle at different speeds is shown in Figure 10, and the phase trajectory of the center of gravity sideslip angle-yaw rate and the roll angle during the driving process are shown in Figure 11.

[0283] In Figure 10, the XY axes represent the vehicle's planar position, and the Z-axis represents time. This three-dimensional space shows the spatial positions of the vehicle and surrounding vehicles at different times. According to the trajectory spatiotemporal diagram shown in Figure 10, when the vehicle travels along the planned trajectory, dynamic obstacle avoidance can be achieved through path-velocity cooperative planning. Meanwhile, Figure 11 shows that planning the vehicle trajectory based on constraints on vehicle turning angle and wheel speed can ensure its driving stability and safety.

[0284] Curved driving conditions

[0285] To further verify the correctness of the trajectory generation method, a test was conducted considering a curve driving condition. The test setup involved three vehicles surrounding the five-axle special vehicle. The vehicle's speed was set to 20 m / s, and the initial speeds of the surrounding vehicles were set as follows: vehicle A 15 m / s, vehicle B 10 m / s, and vehicle C 13 m / s. The three vehicles changed speeds while driving on the road, with accelerations ranging from 0 to 2 m / s². 2 The relative positions and trajectories of the vehicles are as follows: Figure 12 As shown, the vehicle's spatiotemporal diagram is as follows: Figure 13 As shown.

[0286] Combination Figure 13 The vehicle's time-space diagram shows that when traveling on a curved road, the vehicle begins to turn at 100m. At this point, to ensure obstacle avoidance stability and safety, the vehicle speed is reduced. After entering the curve, the vehicle first avoids the slower-moving vehicle B. Since vehicle B is currently accelerating, the vehicle gradually accelerates at 120m until it overtakes vehicle B. Simultaneously, due to the large lateral distance between vehicle C and the vehicle, the vehicle slightly avoids vehicle C after approaching it and overtakes it. Afterward, the vehicle gradually decelerates to the reference speed and begins to follow vehicle A around 380m. The phase plane trajectory and roll angle changes of the vehicle during this process are shown in Figure 14.

[0287] Figure 14(a) shows that when the vehicle is driving on a curved road, its phase trajectory is always in the stable range during trajectory planning. In extreme cases, the vehicle can actively reduce its speed to ensure stability when avoiding obstacles, while effectively controlling the body roll (Figure 14(b)).

[0288] To further illustrate the effectiveness of the proposed method, driving trajectories without considering obstacle avoidance stability are generated under the same conditions, such as... Figure 15 As shown in Figure 16, the phase trajectory and roll angle changes of the vehicle during unconstrained trajectory driving are generated simultaneously.

[0289] Figure 15 In the diagram, the blue trajectory represents the driving trajectory generated considering obstacle avoidance stability constraints, while the red trajectory represents the driving trajectory generated without considering obstacle avoidance stability constraints. Figure 15 It can be seen that the planned trajectory for this vehicle to achieve stable obstacle avoidance is relatively smooth. Compared to the trajectory considering obstacle avoidance stability, the red trajectory has a large curvature region, which means that the vehicle controller experiences a large rate of change in steering angle during trajectory tracking. Furthermore, comparing Figure 16, the vehicle's yaw rate of change also exceeds the stability region at this point, causing vehicle instability.

[0290] In summary:

[0291] Taking a five-axle special vehicle as an example, this invention proposes a safe trajectory planning method for multi-axle special vehicles that considers obstacle avoidance stability to address the stability and safety issues of multi-axle special vehicles during high-speed maneuvers.

[0292] (1) Taking a five-axle special vehicle as an example, a dynamic model considering four degrees of freedom of longitudinal, lateral, yaw and tilt is established, and the correctness of the model is verified by comparing the dynamic response of the TruckSim model and the established model.

[0293] (2) To address the obstacle avoidance stability problem of multi-axle special vehicles, stability constraints for multi-axle special vehicles are established based on the phase plane of the centroid side slip angle-yaw rate; and obstacle risk field for multi-axle special vehicles is established based on the vehicle motion state.

[0294] (3) In order to generate a safe driving trajectory for multi-axle special vehicles, a trajectory planner based on model predictive control was designed. Simulation experiments proved that the designed planner can meet the requirements of obstacle avoidance stability and driving safety of multi-axle special vehicles.

[0295] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for safe trajectory planning of a multi-axle special vehicle considering obstacle avoidance stability, characterized in that: Comprise: Step 1: based on five-axis special vehicle longitudinal, lateral, yaw and roll four degrees of freedom to establish five-axis special vehicle dynamics model, and the correctness of five-axis special vehicle dynamics model is verified; Step 2: based on the five-axis special vehicle dynamics model established in step 1, referring to the relative motion state of the surrounding vehicles and the special vehicle, based on the constant acceleration assumption, the five-axis special vehicle driving risk field model is established, and based on the vehicle yaw stability and roll safety, the five-axis special vehicle dynamics constraint model is established;The five-axis special vehicle driving risk field model of step 2 is specifically: Step 2.1.1: based on the bivariate Gaussian distribution function, the five-axis special vehicle predicted risk field is established: wherein, is the relative position between the host vehicle and the surrounding vehicle; Z is the collision adjustment coefficient, is the risk field adjustment factor; The relative coordinates are defined as follows with respect to the vehicle The relative coordinates are defined as follows with respect to the vehicle represent the angle between the heading of the vehicle and the heading of the surrounding vehicle, the angle between the heading of the dolly and the heading of the surrounding vehicle, the coordinates of the vehicle in the earth coordinate system, the coordinates of the surrounding vehicle in the earth coordinate system, the position of the dolly vehicle in the earth coordinate system; Step 2.1.2: Define the coefficient σ as the risk field adjustment factor, the motion state of the ego vehicle at a certain time as , respectively representing the vehicle position, speed, acceleration and heading angle, and the motion state of the surrounding vehicle as , The risk field adjustment factor σ calculation formula is: wherein, , and respectively represent the direction of the host vehicle and the surrounding vehicle direction difference, direction difference, are the lateral and longitudinal acceleration of the host vehicle, are the lateral and longitudinal speed of the host vehicle, determined according to the actual vehicle speed, using the following calculation formula: In the above formulae, denotes the maximum and minimum speed limit on the road; The road boundary is defined as: wherein is a constant coefficient, are the closest distances to the lane boundary on both sides of the vehicle, respectively, is defined in the same way as in equation (15); Step 2.1.3: to establish the risk field model of the next n time, based on the constant acceleration assumption, according to the current estimation of the other vehicle motion state, the other vehicle motion state at the next time is: where T is the time interval between two motion instants, Xi, yi represent the lateral and longitudinal position of the i-th vehicle in the earth coordinate system; If the acceleration is constant for a short time , at time, predict the future motion state at time ； Step 3: taking the five-axis special vehicle driving risk field model established in step 2 as the soft constraint, and the five-axis special vehicle dynamics constraint model as the hard constraint, the MPC trajectory planner is designed to generate a safe driving trajectory.

2. The multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability according to claim 1, characterized in that: The step 1 comprises: Step 1.1: a four degree of freedom vehicle body dynamics model is established in the vehicle coordinate system, and a five-axis vehicle kinematics model is established in the earth coordinate system; Step 1.2: a tire model is established based on Dugoff model; Step 1.3: a steering model of five-axis special vehicle is established; Step 1.4: a four degree of freedom vehicle model is established by using Matlab / Simulink software, and a five-axis vehicle model is established by using TruckSim software, and the accuracy of the four degree of freedom vehicle model is verified by comparing the dynamic response of the two models under the same given input condition.

3. The multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability according to claim 1, characterized in that: The step 1.1 comprises: Step 1.1.1: a single track model of five-axis special vehicle is established in the vehicle coordinate system, and the four degree of freedom vehicle body dynamics equation is represented as: Longitudinal motion differential equation: Lateral motion differential equation: Yaw motion differential equation: Roll motion differential equation: wherein, is the total vehicle mass; denote the longitudinal and lateral velocities of the vehicle along the x-axis and y-axis in the vehicle coordinate system, respectively; denote the longitudinal and lateral velocities of the vehicle along the x-axis and y-axis in the vehicle coordinate system, respectively; denotes the yaw angular velocity of the vehicle; is the yaw moment of inertia of the vehicle; is the roll moment of inertia of the vehicle about the roll center, and denote the longitudinal and lateral forces of the ground acting on the vehicle through the ith tire in the vehicle coordinate system, where i = 1, 2,..., 5; i denote the longitudinal and lateral forces of the ground acting on the vehicle through the ith tire in the vehicle coordinate system, where i = 1, 2,..., 5; denotes the roll angle of the vehicle body, is the vertical distance from the vehicle center of mass to the ground, is the lateral acceleration at the vehicle center of mass, denoted as ; due to the presence of the suspension, the vehicle body is also subjected to an anti-roll moment, denoted as , whose magnitude is , denote the stiffness and damping coefficients of the vehicle body roll, respectively;​ Step 1.1.2: a five-axis vehicle kinematics model is established in the earth coordinate system: The above formula, denotes the vehicle heading angle, respectively the lateral and longitudinal position of the vehicle in the earth coordinate system.

4. The multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability according to claim 3, characterized in that: The step 1.2 comprises: Step 1.2.1: the tire force mapping relationship in the vehicle coordinate system and the tire coordinate system is: wherein, denotes the steering angle of the i-th axis, denotes the longitudinal and lateral forces experienced by the i-th axis tire in the tire coordinate system; Step 1.2.2: the Dugoff tire model is adopted to establish the relationship between tire force and vehicle dynamics: （7） wherein, respectively represent the longitudinal stiffness and lateral stiffness of the axis tire, is the tire slip angle of the axis tire, is the tire slip ratio of the axis tire, is the ground adhesion coefficient, is the tire longitudinal force of the axis tire, λ i is a tire force correction coefficient; Based on the tire force calculation formula (7), the tire side slip angle and tire slip ratio are input, and the calculation method is as follows: Tire side slip angle: Tire slip ratio: Wherein, the sign convention is that the first two axes are positive and the last three axes are negative; denotes the wheel rotational speed, is the tire rolling radius, is the wheel center longitudinal speed, denoted by: 。 5. The multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability according to claim 4, characterized in that: Step 1.3 specifically comprises: The steering angle of five-axis special vehicle is based on Ackerman steering model, the first, second, fourth and fifth axes of five-axis vehicle are steering axes, and the instantaneous center is located on the extension line of the third axis, for single track model, the following relationship is established: Further: 。 6. The multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability according to claim 5, characterized in that: The five-axis special vehicle dynamics constraint model of step 2 comprises vehicle stability constraint and vehicle roll constraint; The vehicle stability constraint specifically comprises: Step 2.2.1: the vehicle body dynamics equation is simplified to two degrees of freedom: wherein is the side slip angle of the center of mass, and is the rate of change of the side slip angle of the center of mass, and is the maximum lateral acceleration of the vehicle, and is the maximum lateral acceleration of the vehicle, and is the maximum lateral acceleration of the vehicle, and is the maximum lateral acceleration of the vehicle, and is the maximum lateral acceleration of the vehicle, and is the maximum lateral acceleration of the vehicle, and is the maximum lateral acceleration of the vehicle, and is is the ground adhesion coefficient, For the constraint of the center of mass side slip angle, combined with the saturation condition of the tire force of the non-steering axle, combined with formula (8), the following formula is obtained: where the tire saturation side slip angle is approximated by wherein, is the tire lateral stiffness, is the tire vertical force, is the ground adhesion coefficient, since the five-axle vehicle body and tire suspension constitute an over-determined system, the third axle tire vertical force is expressed as: wherein is the sprung mass of the vehicle, is the unsprung mass of the vehicle, the sign convention is the same as in step 1.2.3; Step 2.2.2: Vehicle roll constraint The vehicle roll constraint is based on the shift of the zero moment point (ZMP) if there is a point on the ground that is shifted relative to the vehicle's center of mass and the sum of the vehicle roll moments of the vehicle's gravity, inertial forces and the ground's force on the vehicle body relative to this point is zero, then the following relationship is obtained: Due to the small roll angle of the vehicle body, based on the small angle assumption, it is further obtained that By limiting the vehicle center of mass offset , the vehicle roll is constrained: Finally, the vehicle ride stability constraint can be expressed as 。 7. The multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability according to claim 6, characterized in that: The MPC trajectory planner design of step 3 is specifically as follows: Step 3.1.1: The vehicle dynamics model established in step 1 is recorded as In the formula represents the state quantity of the trajectory planner system, and is u represents the system control quantity, and is ; Step 3.1.2: The trajectory planner system of step 3.1.1 is a nonlinear system, which is linearized at the current time instant and the system output quantity is defined as The trajectory planner system is then represented as: In the formula, the matrix ,matrix For Jacobi matrix, This is for linearization error; Step 3.1.3: Based on the current time t, the ride trajectory of the future N time is generated, the above system is discretized by using the forward Euler method, and then the system is expressed as wherein , , , represents a discrete time length, and the linearization error after discretization is represented as: Step 3.1.4: The control increment Δu(t) is used to replace the control amount u(t), and the smooth trajectory planning of the vehicle can be realized by directly constraining the control increment, if the control increment Δu(t)= u(t)- u(t-1), then formula (30) is changed to Formula (34) is expressed as If the prediction time domain is H p The control time domain is H c And H p >H c Assuming at the current time t, given the control time domain [t, t+H] c -1] A series of control quantities Outside the control time domain Given a control variable of 0, the predicted output in the prediction time domain is expressed as: According to formula (36), the optimal output trajectory is calculated with the current optimized series of input quantities as unknown quantities.

8. The multi-axle special vehicle safety trajectory planning method considering obstacle avoidance stability according to claim 7, characterized in that: The optimal ride trajectory generation of step 3 is specifically as follows: Step 3.2.1: With the ride risk field model and the dynamics constraint model in step 2 as constraints, the planner solves the optimal problem J in the prediction time domain in a loop within the solution time T, and the optimal problem J is defined as wherein is the driving trajectory, is the reference driving trajectory input, is the vehicle longitudinal speed, is the reference vehicle speed, is the motion risk field generated according to the observation results of the motion state of the other vehicle and the road boundary in the prediction time domain, calculated by equations (13)-(19). Step 3.2.2: To predict the curvature of the vehicle trajectory at a certain time in the time domain, the calculation method is: In the formula, X and Y are the cubic spline fitting curves of the planner output trajectory points, the boundary conditions are given by constraining the vehicle yaw angle state of the starting point and the ending point, and then at time t, the planner output position is Wherein, the parameters in the formula are the fitting coefficients of the cubic spline curve, which have no physical meaning; Step 3.2.3: If the vehicle allowable maximum lateral acceleration is then constrain the vehicle speed to: where For the lane speed limit, which is related to the vehicle ride comfort and the passenger comfort, and combined with the safety constraint of equation (27), the optimization problem is expressed as In the prediction horizon H p The planner generates a continuous driving trajectory of the vehicle by solving an optimization problem (42) in the prediction horizon H, and the lower-level controller realizes the automatic driving of the vehicle on the structured road by tracking the driving trajectory in real time.

Images