Unmanned vehicle safety obstacle avoidance control method under sideslip condition

Abstract

The invention discloses an unmanned vehicle safety obstacle avoidance control method under a sideslip condition, which decouples a vehicle safety obstacle avoidance problem under the sideslip condition into a safety obstacle avoidance trajectory planning problem and a drift control problem, and comprises the following steps: firstly, establishing a vehicle model, and performing experimental measurement on model parameters; designing a multi-layer controller comprising a planning layer, a rule layer and a motion control layer to realize a safety control target; the planning layer plans a safe obstacle avoidance track in real time through a method based on model prediction and obstacle avoidance function control; the rule layer optimizes the planning result in combination with the vehicle dynamic characteristics under the sideslip condition; the motion control layer realizes high-precision, high-real-time and high-robustness target state tracking control under the sideslip condition based on a feedforward feedback control method; through the provided controller, safe obstacle avoidance control under the vehicle sideslip condition can be realized, and a safe control strategy under the sideslip condition is provided, so that the safety boundary of the unmanned vehicle is widened, and the safety is improved.

Description

Technical Field

[0001] The present invention relates to the field of safe obstacle avoidance control for unmanned vehicles, and in particular to a real-time safe obstacle avoidance trajectory planning and control method in a side-slipping situation based on model predictive control, which is used to achieve safe obstacle avoidance operation under extreme conditions where an unmanned vehicle is side-slipping. Background Art

[0002] Active safety technology in vehicle emergency situations is an important research direction for the development of autonomous driving technology, such as automatic emergency braking system (AEB) and automatic emergency steering system (AES). After AEB detects that an obstacle is smaller than the safe braking distance, it will actively brake to ensure safety; and when facing a sudden obstacle and AEB cannot avoid a collision, AES can perform steering operations to achieve safe obstacle avoidance.

[0003] However, these existing active safety technologies limit the vehicle to a normal driving state, that is, the tire friction is not saturated and the vehicle does not slide. If the vehicle slides due to a decrease in the ground friction coefficient or improper steering operation at high speed, the existing active safety

[0004] The technology cannot guarantee safe obstacle avoidance control of the vehicle in the event of a side slip, that is, after a side slip occurs, the vehicle will not be able to safely avoid obstacles such as vehicles or pedestrians in front.

[0005] At present, there have been studies on the control and obstacle avoidance methods of unmanned vehicles in side-slip situations, but there are still obvious shortcomings, including:

[0006] 1. Some existing control methods can achieve vehicle stability control in a side-slip state, but cannot perform real-time obstacle avoidance planning control, and therefore cannot guarantee the vehicle's safe obstacle avoidance capability in a side-slip state.

[0007] 2. Some existing active drift obstacle avoidance methods can start from the normal driving state of the vehicle and adopt drift strategies to actively enter the side-slip situation to avoid obstacles, but they are not suitable for safe obstacle avoidance control when the vehicle is already in a side-slip situation. Summary of the invention

[0008] In order to solve the above problems existing in the prior art, the present invention discloses a safe obstacle avoidance control method for an unmanned vehicle in a side-slipping situation. The method designs a multi-layered controller, including a real-time planning layer, a rule layer, and a motion control layer. First, the planning layer plans the safe obstacle avoidance trajectory in real time based on the model prediction and control barrier function method; the rule layer optimizes the planning result; finally, the motion control layer tracks the target state based on the drift control method of feedforward feedback, realizes the safe obstacle avoidance capability of the vehicle in the side-slipping situation, and improves the safety of the unmanned vehicle in the extreme working condition of side-slipping.

[0009] A safe obstacle avoidance control method for an unmanned vehicle in a side-slip situation, comprising:

[0010] Step 1: Get vehicle parameters and tire parameters:

[0011] The state variables of the vehicle dynamics model include the longitudinal velocity U x , sideslip angle β and heading angular velocity r; the control variables include the front wheel steering angle δ and the longitudinal driving force F xR , the model equation is shown as follows:

[0012]

[0013] The vehicle parameters in formula (1) include: vehicle mass m, moment of inertia I z , front and rear wheelbase a, b; F xF 、F xR Respectively represent the front and rear wheel longitudinal forces; F zF 、F zR represent the vertical loads on the front and rear wheels respectively;

[0014] Front wheel lateral force F yF and rear wheel lateral force F yR Calculated by the following magic tire formula:

[0015]

[0016] Where i = F, R represents the parameters of the front and rear wheels respectively, α i represents the tire slip angle, α icr Indicates the tire saturated slip angle; tire parameters include: tire model formula fitting coefficients B, C, friction coefficient μ, tire vertical load F zi ;

[0017] Step 2: According to the vehicle model, tire model, road constraints and obstacle constraints, the planning layer is designed based on the model prediction and control barrier function method to perform real-time obstacle avoidance trajectory planning, including:

[0018] The linearized prediction model is as follows:

[0019]

[0020] Among them, z is the state quantity of the prediction model, u is the control quantity of the prediction model; (X, Y) and, They are the vehicle’s position and heading information, (U x ,U y ,r) are the longitudinal velocity, lateral velocity and heading angular velocity information of the vehicle respectively, It is the position, heading, longitudinal velocity, lateral velocity and heading angular velocity information of the vehicle in the reference trajectory information;

[0021] The objective function is designed as follows:

[0022]

[0023] Where N represents the prediction time domain length, The t in the subscript indicates the time when the current optimization problem is solved, and t, t+1…, t+N indicate the next first, second, and Nth prediction time points that need to be calculated; u t+k,t and z t+k,t They represent the state quantity and the predicted control quantity taken at the kth prediction time node and after sub-time t, respectively. Q and R are the coefficient matrices of the state quantity error and the control quantity error weights.

[0024] The control barrier function constraint is designed as follows:

[0025] Δh(z t+k,t ,u t+k,t )+γh(z t+k,t )≥0 (5)

[0026] where h(z t+k,t ) is the designed control barrier function, Δh(z t+k,t ) is the rate of change of the barrier function, and γ is an adjustable parameter used to adjust the distance between the planned obstacle avoidance trajectory and the obstacle. The specific calculation formula of the function is as follows:

[0027]

[0028] Δh(z t+k,t ,u t+k，t )=h(z t+k+1,t )-h(z t+k,t ) (7)

[0029] Where (X t+k,t ,Y t+k,t ) is contained in the state quantity z t+k,t The vehicle position information in (X obs ,Y obs ) is the obstacle location information, R obs is a safe distance; and t+k+1,t =Az t+k,t +Bu t+k,t ;

[0030] Finally, the optimal control problem solved at each node is expressed as follows:

[0031]

[0032] Among them, z t+k+1,t =Az t+k,t +Bu t+k,t That is the discrete form of the prediction model; z t+k,t ∈z safe is the constraint that the vehicle position is within the road boundary; Δh(x t+k∣t ,u t+k∣t )+γh(x t+k∣t )≥0 is the control barrier function constraint, which is used to achieve safe obstacle avoidance; u min ≤u t+k,t ≤u max is the control quantity constraint, i.e., the boundary constraint of the vehicle speed and other information; Δu min ≤u t+k+1,t -u t+k,t ≤Δu max It is the control variable change rate constraint, that is, the boundary constraint of the vehicle's acceleration information;

[0033] By solving the optimal control problem, the final output of the planning layer is as follows:

[0034]

[0035] Among them, (U xd ,U yd ,r d ) is the desired vehicle state output, u t,t is the first control variable of the optimal control sequence to be solved, and the expected sideslip angle β d Calculated by the following formula:

[0036] β d =arctan(U yd / U xd ) (10)

[0037] Step 3: According to the vehicle dynamics characteristics in the side-slip state, design the rule layer to optimize the output results of the planning layer:

[0038] The change rate of the state quantity of the established vehicle dynamics model is set to zero, and the equilibrium dynamics equations for the side slip situation are obtained:

[0039]

[0040] ② Solve the equilibrium dynamics equations to obtain the vehicle equilibrium state table DET under the side slip state:

[0041] In the speed range and steering angle range, given U x , δ is a known number; based on the three equations of formula (11), the other vehicle states β,r in the side slip equilibrium state are solved; thus, U in the side slip condition is obtained x,β,r relationship table DET; Based on the vehicle state relationship table in the side slip situation, the rule layer is designed to optimize the output of the planning layer:

[0042]

[0043] Among them (U xeq ,β eq ,r eq ) is the target vehicle state after rule layer optimization; κ in Eq. d is the U output according to the planning layer xd ,U yd ,r d The target curvature calculated is to optimize the vehicle state (U) by combining the sideslip relationship table DET while ensuring that the curvature of the planned target trajectory remains unchanged. xeq ,β eq ,r eq ), thereby ensuring safe obstacle avoidance in side-slip situations;

[0044] Step 4: Based on the vehicle dynamics model analysis and feedforward feedback control law, design the motion controller in the side slip situation, including:

[0045] ① Based on the idea of ​​controlling the sideslip angle based on the heading angular velocity, the desired heading angular velocity r is calculated by the feedback sideslip angle. des :

[0046] r des =r eq +K β e β (13)

[0047] Where K β is the sideslip angle feedback coefficient, e β =β-β eq is the sideslip angle error;

[0048] Next, the desired angular velocity is tracked by controlling the tire lateral force. The desired front wheel lateral force is calculated based on whether the front wheel lateral force is saturated. That is, it is unsaturated, which is divided into two situations: unsaturated front wheel lateral force and saturated front wheel lateral force:

[0049] ② When the front wheel lateral force is not saturated, the required front wheel lateral force is calculated by the following formula: and rear wheel drive

[0050]

[0051] in and K r are the longitudinal velocity and heading angular velocity feedback coefficients respectively; is the expected longitudinal force; is the expected lateral force; is the speed error, e r =rr des is the heading angular velocity error; k 1 =(a / I z )-(K β / U x ) and k 2 =(b / I z )+(K β / U x ) is an intermediate variable;

[0052] ③ After the front wheel lateral force is saturated, the required front wheel lateral force is calculated by the following formula: and rear wheel drive

[0053] F yF =μF zF

[0054]

[0055] ④ The obtained front wheel lateral force Converted into front wheel steering angle δ des , rear wheel drive Converted into rear wheel torque T xR :

[0056]

[0057] Where R ω is the effective rolling radius of the tire, i 0 is the transmission ratio.

[0058] Preferably, the tire parameters are obtained by driving the vehicle at a certain speed, gradually increasing the tire turning angle, recording the slip angle data and the tire force data, fitting the curve, and calculating.

[0059] Preferably, the vehicle's rotational inertia, mass, and wheelbase parameters are obtained through a swing test method or a direct measurement method.

[0060] The present invention has the following beneficial effects:

[0061] The present invention decouples the problem of safe obstacle avoidance of vehicles in side-slipping situations into a safe obstacle avoidance trajectory planning problem and a drift control problem. First, a vehicle model is established and the model parameters are experimentally measured. Next, a multi-layer controller including a planning layer, a rule layer and a motion control layer is designed to achieve the safety control goal. The planning layer plans the safe obstacle avoidance trajectory in real time by predicting and controlling the obstacle avoidance function based on the model. The rule layer optimizes the planning results in combination with the vehicle dynamics characteristics in side-slipping situations. The motion control layer achieves high-precision, high-real-time and high-robust target state tracking control in side-slipping situations based on the feedforward feedback control method. The proposed controller can realize safe obstacle avoidance control in side-slipping situations, provide a safety control strategy in side-slipping situations, thereby broadening the safety boundary of unmanned vehicles and improving safety. The algorithm can be applied to oil vehicles and electric vehicles, to various unmanned vehicle platforms such as rear-wheel drive and four-wheel drive, and to various scenarios according to different road constraints. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is a flow chart of a safe obstacle avoidance control method for an unmanned vehicle in a side-slip situation according to the present invention;

[0063] Figure 2 Schematic diagram of a three-degree-of-freedom monorail model of an unmanned vehicle in an embodiment of the present invention;

[0064] Figure 3 Schematic diagram of the safe obstacle avoidance result of an unmanned vehicle in a side-slip situation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The present invention is described in detail below with reference to the accompanying drawings and examples. This example is a rear-wheel-driven Ackerman steering unmanned vehicle that safely avoids obstacles when the vehicle in front is skidding. The control variables are the front wheel steering angle δ and the rear wheel torque T xR .

[0066] like Figure 1 As shown, a safe obstacle avoidance control method for an unmanned vehicle in a side-slip situation in this embodiment specifically includes:

[0067] Step 1: Obtain vehicle parameters and tire parameters;

[0068] In this embodiment, Figure 2 As shown, the vehicle dynamics model state includes the longitudinal speed U x , sideslip angle β and heading angular velocity r; the control variables include the front wheel steering angle δ and the longitudinal driving force F xR The model equation is shown below:

[0069]

[0070] The vehicle parameters in formula (1) include: vehicle mass m, moment of inertia I z , front and rear wheelbase a, b; F xF 、F xR Respectively represent the front and rear wheel longitudinal forces; F zF 、F zR Represent the vertical loads on the front and rear wheels respectively.

[0071] Front wheel lateral force F yF and rear wheel lateral force F yR Calculated by the following magic tire formula:

[0072]

[0073] Where i = F, R represents the parameters of the front and rear wheels respectively, α i represents the tire slip angle, α icr Indicates the tire saturated slip angle; tire parameters include: tire model formula fitting coefficients B, C, friction coefficient μ, tire vertical load F zi .

[0074] In specific implementation, the vehicle is driven at a certain speed, the tire angle is gradually increased, the slip angle data and tire force data are recorded, the curve is fitted, and the tire parameters are calculated. The vehicle's rotational inertia, mass, wheelbase and other parameters are obtained through the swing test method and direct measurement method.

[0075] Step 2: According to the vehicle model, tire model, road constraints and obstacle constraints, the planning layer is designed based on the model prediction and control barrier function method to perform real-time obstacle avoidance trajectory planning. Figure 3 As shown, the vehicle skids in a curve, and in the case of skidding, obstacles on the inside and outside of the curve must be avoided safely.

[0076] The linearized prediction model in the model prediction method is as follows:

[0077]

[0078] Among them, z is the state quantity of the prediction model, u is the control quantity of the prediction model; (X, Y) and, They are the vehicle’s position and heading information, (U x ,U y ,r) are the longitudinal velocity, lateral velocity and heading angular velocity information of the vehicle respectively, It refers to the position, heading, longitudinal velocity, lateral velocity and heading angular velocity information of the vehicle in the reference trajectory information.

[0079] The objective function of the model prediction method is designed as follows:

[0080]

[0081] Where N represents the prediction time domain length, The t in the subscript indicates the time when the current optimization problem is solved, and t, t+1…, t+N indicate the next first, second, and Nth prediction time points that need to be calculated. t+k,t and z t+k,t They respectively represent the state quantity and the predicted control quantity taken at sub-time t and the kth prediction time node thereafter, and Q and R are the coefficient matrices of the weights of the state quantity error and the control quantity error.

[0082] The control barrier function constraint is designed as follows:

[0083] Δh(z t+k,t ,u t+k,t )+γh(z t+k,t )≥0 (5)

[0084] where h(z t+k,t ) is the designed control barrier function, Δh(z t+k,t ) is the rate of change of the barrier function, and γ is an adjustable parameter used to adjust the distance between the planned obstacle avoidance trajectory and the obstacle. Through this inequality constraint, it can be ensured that the vehicle's state is within the safety boundary and will not collide with obstacles. The specific calculation formula of the function is as follows:

[0085]

[0086] Δh(z t+k,t ,u t+k,t )=h(z t+k+1,t )-h(z t+k,t ) (7)

[0087] Where (X t+k,t ,Y t+k,t ) is contained in the state quantity z t+k,t The vehicle position information in (X obs ,Y obs ) is the obstacle location information, R obs is a safe distance; and t+k+1,t =Az t+k,t +Bu t+k,t .

[0088] Finally, the optimal control problem solved by the model prediction algorithm at each node can be expressed as follows:

[0089]

[0090] Among them, z t+k+1,t =Az t+k,t +But+k,t This is the discrete form of the prediction model established above; t+k,t ∈z safe is the constraint that the vehicle position is within the road boundary; Δh(x t+k∣t ,u t+k∣t )+γh(x t+k∣t )≥0 is the control barrier function constraint, which is used to achieve safe obstacle avoidance; u min ≤u t+k,t ≤u max is the control quantity constraint, i.e., the boundary constraint of the vehicle speed and other information; Δu min ≤u t+k+1,t -u t+k,t ≤Δu max It is the constraint on the rate of change of the controlled variable, that is, the boundary constraint of the vehicle's acceleration information.

[0091] By solving the optimal control problem, the final output of the planning layer is as follows:

[0092]

[0093] Among them (U xd ,U yd ,r d ) is the desired vehicle state output, u t,t is the first control variable of the optimal control sequence to be solved, and the expected sideslip angle β d It can be calculated by the following formula:

[0094] β d =arctan(U yd / U xd ) (10)

[0095] Step 3: According to the vehicle dynamics characteristics in the side-slip state, design the rule layer to optimize the output results of the planning layer: ① Set the state quantity change rate of the established vehicle dynamics model to zero, and obtain the equilibrium dynamic equation group of the side-slip situation:

[0096]

[0097] ② Solve the equilibrium dynamics equations to obtain the vehicle equilibrium state table (DET) under the side slip state:

[0098] Since there are three equations and U in the equilibrium dynamics equations x ,β,r,δ,F xR Five unknowns, so in order to solve this set of equations. In the speed range and steering angle range, given U x , δ is a known number. Finally, based on the three equations and three unknowns, the other vehicle states β,r in the side slip equilibrium state can be solved. Thus, Ux ,β,r relationship table DET. ③ Based on the vehicle state relationship table in the side slip situation, the rule layer is designed to optimize the output of the planning layer:

[0099]

[0100] Among them (U xeq ,β eq ,r eq ) is the target vehicle state after rule layer optimization; κ in Eq. d is the U output according to the planning layer xd ,U yd ,r d The target curvature calculated is to optimize the vehicle state (U) by combining the sideslip relationship table DET while ensuring that the curvature of the planned target trajectory remains unchanged. xeq ,β eq ,r eq ), thereby ensuring safe obstacle avoidance in side-slip situations.

[0101] Step 4: Based on the vehicle dynamics model analysis and feedforward feedback control law, design the motion controller in the case of side slip. The design ideas of the motion controller are as follows:

[0102] ① Based on the idea of ​​controlling the sideslip angle based on the heading angular velocity, the desired heading angular velocity r is calculated by the feedback sideslip angle. des :

[0103] r des =r eq +K β e β (13)

[0104] Where K β is the sideslip angle feedback coefficient, e β =β-β eq is the side slip angle error. Next, the desired heading angular velocity is tracked by controlling the tire lateral force. The desired front wheel lateral force is calculated based on whether the front wheel lateral force is saturated. That is, it is unsaturated, which is divided into two cases ② and ③. ② If the front wheel lateral force is not saturated, the required front wheel lateral force is calculated by the following formula: and rear wheel drive

[0105]

[0106] in and K r are the longitudinal velocity and heading angular velocity feedback coefficients respectively; is the expected longitudinal force; is the expected lateral force; is the speed error, e r =rr des is the heading angular velocity error; k 1 =(a / I z )-(K β / U x ) and k 2 =(b / I z )+(K β / U x ) is an intermediate variable used to simplify the equation. ③ After the front wheel lateral force is saturated, the required front wheel lateral force is calculated by the following formula: and rear wheel drive

[0107] F yF =μF zF

[0108]

[0109] ④ The obtained front wheel lateral force Converted into front wheel steering angle δ des , rear wheel drive Converted into rear wheel torque T xR

[0110]

[0111] Where R ω is the effective rolling radius of the tire, i 0 is the transmission ratio.

[0112] In summary, the motion controller in the side-slip situation first calculates the expected heading angular velocity through ①, and then through ② the front wheel lateral force. If it is found that the front wheel lateral force is saturated after calculation, the next step is to enter ③ to recalculate the lateral force, and then enter ④ to obtain the final control amount; if the calculated front wheel lateral force is not saturated, then directly enter ④ from ② to obtain the final control amount.

[0113] In this embodiment, when turning, the vehicle skids and encounters two stationary obstacle vehicles in front of it, one on the inner ring and the other on the outer ring of the road. At this time, through the above steps 2, 3, and 4, the result is as shown in the figure. Figure 3 As shown, the vehicle's safe obstacle avoidance control in the case of sideslip is realized, ensuring the vehicle's safety after sideslip occurs.

[0114] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A safe obstacle avoidance control method for an unmanned vehicle in a side-slip situation, characterized in that: include: Step 1: Get vehicle parameters and tire parameters: The state variables of the vehicle dynamics model include the longitudinal velocity U x , sideslip angle β and heading angular velocity r; the control variables include the front wheel steering angle δ and the longitudinal driving force F xR , the model equation is shown as follows: The vehicle parameters in formula (1) include: vehicle mass m, moment of inertia I z , front and rear wheelbase a, b; F xF 、F xR Respectively represent the front and rear wheel longitudinal forces; F zF 、F zR represent the vertical loads on the front and rear wheels respectively; Front wheel lateral force F yF and rear wheel lateral force F yR Calculated by the following magic tire formula: Where i = F, R represents the parameters of the front and rear wheels respectively, α i represents the tire slip angle, α icr Indicates the tire saturated slip angle; tire parameters include: tire model formula fitting coefficients B, C, friction coefficient μ, tire vertical load F zi ; Step 2: According to the vehicle model, tire model, road constraints and obstacle constraints, the planning layer is designed based on the model prediction and control barrier function method to perform real-time obstacle avoidance trajectory planning, including: The linearized prediction model is as follows: Among them, z is the state quantity of the prediction model, u is the control quantity of the prediction model; (X, Y) and, They are the vehicle’s position and heading information, (U x ,U y ,r) are the longitudinal velocity, lateral velocity and heading angular velocity information of the vehicle respectively, It is the position, heading, longitudinal velocity, lateral velocity and heading angular velocity information of the vehicle in the reference trajectory information; The objective function is designed as follows: Where N represents the prediction time domain length, The t in the subscript indicates the time when the current optimization problem is solved, and t, t+1…, t+N indicate the next first, second, and Nth prediction time points that need to be calculated; u t+k,t and z t+k,t They represent the state quantity and the predicted control quantity taken at the kth prediction time node and after sub-time t, respectively. Q and R are the coefficient matrices of the state quantity error and the control quantity error weights. The control barrier function constraint is designed as follows: Δh(z t+k,t ,u t+k,t )+γh(z t+k,t )≥0 (5) where h(z t+k,t ) is the designed control barrier function, Δh(z t+k,t ) is the rate of change of the barrier function, and γ is an adjustable parameter used to adjust the distance between the planned obstacle avoidance trajectory and the obstacle. The specific calculation formula of the function is as follows: △h(z t+k,t ,u t+k，t )=h(z t+k+1,t )-h(z t+k,t ) (7) Where (X t+k,t ,Y t+k,t ) is contained in the state quantity z t+k,t The vehicle position information in (X obs ,Y obs ) is the obstacle location information, R obs is a safe distance; and t+k+1,t =Az t+k,t +Bu t+k,t ; Finally, the optimal control problem solved at each node is expressed as follows: Among them, z t+k+1,t =Az t+k,t +Bu t+k,t That is the discrete form of the prediction model; z t+k,t ∈z safe is the constraint that the vehicle position is within the road boundary; Δh(x t+k∣t ,u t+k∣t )+γh(x t+k∣t )≥0 is the control barrier function constraint, which is used to achieve safe obstacle avoidance; u min ≤u t+k,t ≤u max is the control quantity constraint, i.e., the boundary constraint of the vehicle speed and other information; Δu min ≤u t+k+1,t -u t+k,t ≤Δu max It is the control variable change rate constraint, that is, the boundary constraint of the vehicle's acceleration information; By solving the optimal control problem, the final output of the planning layer is as follows: Among them, (U xd ,U yd ,r d ) is the desired vehicle state output, u t,t is the first control variable of the optimal control sequence to be solved, and the expected sideslip angle β d Calculated by the following formula: β d =arctan(U yd / U xd ) (10) Step 3: According to the vehicle dynamics characteristics in the side-slip state, design the rule layer to optimize the output results of the planning layer: The change rate of the state quantity of the established vehicle dynamics model is set to zero, and the equilibrium dynamics equations for the side slip situation are obtained: ② Solve the equilibrium dynamics equations to obtain the vehicle equilibrium state table DET under the side slip state: In the speed range and steering angle range, given U x , δ is a known number; based on the three equations of formula (11), the other vehicle states β,r in the side slip equilibrium state are solved; thus, U in the side slip condition is obtained x ,β,r relationship table DET; Based on the vehicle state relationship table in the side slip situation, the rule layer is designed to optimize the output of the planning layer: Among them (U xeq ,β eq ,r eq ) is the target vehicle state after rule layer optimization; κ in Eq. d is the U output according to the planning layer xd ,U yd ,r d The target curvature calculated is to optimize the vehicle state (U) by combining the sideslip relationship table DET while ensuring that the curvature of the planned target trajectory remains unchanged. xeq ,β eq ,r eq ), thereby ensuring safe obstacle avoidance in side-slip situations; Step 4: Based on the vehicle dynamics model analysis and feedforward feedback control law, design the motion controller in the side slip situation, including: ① Based on the idea of ​​controlling the sideslip angle based on the heading angular velocity, the desired heading angular velocity r is calculated by the feedback sideslip angle. des : r des =r eq +K β e β (13) Where K β is the sideslip angle feedback coefficient, e β =β-β eq is the sideslip angle error; Next, the desired angular velocity is tracked by controlling the tire lateral force. The desired front wheel lateral force is calculated based on whether the front wheel lateral force is saturated. That is, it is unsaturated, which is divided into two situations: unsaturated front wheel lateral force and saturated front wheel lateral force: ② When the front wheel lateral force is not saturated, the required front wheel lateral force is calculated by the following formula: and rear wheel drive in and K r are the longitudinal velocity and heading angular velocity feedback coefficients respectively; is the expected longitudinal force; is the expected lateral force; is the speed error, e r =rr des is the heading angular velocity error; k1=(a / I z )-(K β / U x ) and k2=(b / I z )+(K β / U x ) is an intermediate variable; ③ After the front wheel lateral force is saturated, the required front wheel lateral force is calculated by the following formula: and rear wheel drive F yF =μF zF ④ The obtained front wheel lateral force Converted into front wheel steering angle δ des , rear wheel drive Converted into rear wheel torque T x R: Where R ω is the effective rolling radius of the tire, and i0 is the transmission ratio.

2. The method for controlling an unmanned vehicle to avoid obstacles safely in a side-slipping situation as claimed in claim 1, characterized in that: By driving the vehicle at a certain speed, gradually increasing the tire turning angle, recording the slip angle data and tire force data, fitting the curve, and calculating the tire parameters.

3. The method for controlling an unmanned vehicle to avoid obstacles safely in a side-slipping situation as claimed in claim 1, characterized in that: The vehicle's moment of inertia, mass and wheelbase parameters are obtained through swing test method and direct measurement method.

Images