An intelligent passenger car dynamic lane-changing trajectory planning method in a multi-vehicle complex traffic environment

Abstract

The application relates to an intelligent passenger car dynamic lane-changing trajectory planning method in a multi-vehicle complex traffic environment. In the dynamic traffic environment, a lane-changing candidate trajectory cluster is first generated; secondly, intelligent passenger car stability analysis is carried out according to different road conditions and other factors, the minimum lane-changing longitudinal distance of the lane-changing trajectory is determined, and the lane-changing trajectory with instability is removed; then, obstacle avoidance detection is carried out by considering the state change of surrounding vehicles, the minimum lane-changing longitudinal distance of the intelligent passenger car without collision with surrounding vehicles is obtained, and the lane-changing trajectory with insecurity is removed; finally, according to the safe lane-changing trajectory cluster of the real-time intelligent passenger car lane-changing trajectory longitudinal length, the target function of vehicle lane-changing efficiency and the like is considered, optimal lane-changing trajectory optimization is carried out, and the safe optimal lane-changing trajectory is obtained; and the lane-changing trajectory is re-planned and speed planning is carried out by considering the sudden change of the state of surrounding vehicles. The application establishes a more real driving scene, and the intelligent passenger car dynamic lane-changing trajectory planning is more in line with the actual traffic environment.

Description

Technical Field

[0001] This invention belongs to the technical field of intelligent bus trajectory planning, specifically referring to a dynamic lane-changing trajectory planning method for intelligent buses in complex multi-vehicle traffic environments. Background Technology

[0002] Domestic and international research on intelligent vehicle lane changing primarily focuses on passenger cars, especially four-wheeled independent vehicles, with limited research on intelligent buses. However, compared to passenger cars, buses are significantly heavier, carry more passengers, are four to five times longer, and are taller. Under the same lateral acceleration, they are prone to rollover accidents, resulting in catastrophic accidents with multiple fatalities and injuries, and causing substantial economic losses. Furthermore, prolonged driving time on buses can lead to driver fatigue, causing misjudgments of the surrounding environment and even operational errors. Therefore, research on intelligent buses is of paramount importance. Literature review reveals the following problems that need to be addressed in intelligent bus lane changing under real-world traffic conditions.

[0003] (1) In actual transportation, intelligent buses are always in a dynamic traffic environment. However, for lane-changing trajectory planning of intelligent buses, many previous studies have assumed that surrounding vehicles are stationary or moving at a constant speed, ignoring the dynamic changes in the speed or acceleration of surrounding vehicles, especially when changing lanes, the acceleration of surrounding vehicles changes suddenly. Therefore, lane-changing trajectory planning of intelligent buses needs to consider the influence of the surrounding dynamic traffic environment, and perform trajectory planning and replanning to obtain the safe and optimal lane-changing trajectory.

[0004] (2) Current intelligent vehicle lane-changing trajectory planning focuses more on vehicle yaw stability and less on vehicle rollover. However, intelligent buses are characterized by their length, height, and large passenger capacity, making them prone to rollover accidents. Therefore, intelligent bus trajectory planning needs to consider not only yaw stability but also roll stability to ensure vehicle stability. Therefore, this invention proposes a lane-changing trajectory replanning method that considers the stability of intelligent buses, including rollover. Summary of the Invention

[0005] To address the aforementioned shortcomings, this invention proposes a dynamic lane-changing trajectory planning method for intelligent buses in complex multi-vehicle traffic environments. This method combines information such as vehicle yaw stability and roll stability with the real-time motion status of surrounding traffic vehicles to dynamically plan the lane-changing trajectory of the intelligent bus in real time.

[0006] To achieve the above objectives, this invention proposes a dynamic lane-changing trajectory planning method for intelligent buses in complex multi-vehicle traffic environments. First, for dynamic traffic environments, a cluster of candidate unconstrained lane-changing trajectories for the intelligent bus is generated. Second, for different road conditions, stability analysis of vehicle yaw and roll is performed to determine the minimum longitudinal distance for stable lane changing for the intelligent bus, eliminating unstable lane-changing trajectories from the unconstrained trajectory cluster. Then, considering changes in the state of surrounding vehicles, obstacle avoidance detection is performed to obtain the maximum longitudinal length of the lane-changing trajectory for which the intelligent bus will not collide with surrounding vehicles, eliminating unsafe lane-changing trajectories that may collide. Finally, a multi-objective lane-changing trajectory performance function is set to obtain the optimal intelligent bus lane-changing trajectory among the aforementioned stable and collision-free trajectories. Simultaneously, considering sudden changes in the state of surrounding vehicles, lane-changing trajectory replanning and speed planning for the intelligent bus are performed.

[0007] Specifically, the steps include the following:

[0008] Step 1: Based on the location of the intelligent bus and the movement status of surrounding traffic vehicles, select a series of lane change endpoint candidate positions on the target lane, and use a cubic polynomial curve to connect the initial position of the vehicle and the endpoint candidate positions to generate an unconstrained lane change trajectory cluster.

[0009] y(x)=a0+a1x+a2x 2 +a3x 3

[0010] x and y(x) represent the longitudinal and lateral positions of the intelligent bus, respectively; a0, a1, a2, and a3 are parameters that satisfy:

[0011] y(x0)=a0+a1x0+a2x 2 0+a3x 3 0

[0012] y'(x0)=a1+2a2x0+3a3x 2 0

[0013] x0 and y(x0) are the initial longitudinal and lateral positions of the intelligent bus during lane change, respectively; y'(x0) is the tangent direction of the initial position of the intelligent bus.

[0014] The intelligent bus's final movement state after lane change is:

[0015] y(x f )=a0+a1x f +a2x 2 f +a3x 3 f

[0016] y'(x f )=a1+2a2xf +3a3x 2 f

[0017] x f y(x) f ) represent the longitudinal and lateral endpoints of the lane change for the intelligent bus; y'(x f ) represents the tangent direction at the destination of the intelligent bus; a0, a1, a2, and a3 are parameters in the lane change trajectory.

[0018] Step 2: Based on the speed of the intelligent bus and the road surface adhesion coefficient, and taking into account the saturation characteristics of the tires, a simulation model is used to obtain the minimum longitudinal distance x for lane changing without causing yaw or roll problems for the intelligent bus. fmin This results in a stable lane change trajectory;

[0019] (1) Vehicle yaw stability

[0020] Yaw velocity stability limits and center of mass sideslip angle stability limits:

[0021]

[0022]

[0023] Where, r s It is the vehicle's yaw rate, β s It is the vehicle's sideslip angle, F. yr F yf Let g and μ represent the lateral forces acting on the rear and front axles of the vehicle, respectively. g is the acceleration due to gravity, μ is the road adhesion coefficient, u is the vehicle speed, m is the vehicle mass, and C is the total mass. αr This indicates the rear axle lateral stiffness, where a and b represent the distances from the front and rear axles to the vehicle's center of gravity, respectively.

[0024] (2) Vehicle roll stability

[0025] Lateral load transfer rate

[0026] Where n is the number of vehicle axles, F Zl For the vertical load on the left side of the wheel, F Zr The vertical load on the right side of the wheel;

[0027] (3) Vehicle simulation

[0028] In the vehicle simulation model, set the longitudinal distance for lane changes at different vehicle speeds and different road surface adhesion coefficients;

[0029] The output yaw rate and sideslip angle are used to determine yaw stability in the phase plane formed by the yaw rate stability limit and the sideslip angle stability limit, thus obtaining the minimum longitudinal lane change distance x that satisfies the vehicle's yaw stability. f1 ;

[0030] Output the vertical loads on the left and right wheels, calculate the lateral load transfer rate (LTR) value, and determine the minimum longitudinal lane change distance (x) to prevent rollover. f2 ;

[0031] Minimum lane change longitudinal distance x fmin When the vehicle speed and road surface adhesion coefficient are the same, x f1 With x f2 The larger value in the range.

[0032] Step 3: Based on the minimum lane change longitudinal distance x fmin Collision detection is performed gradually from short to long to obtain the maximum longitudinal distance of the lane change without collision, thus obtaining a safe lane change trajectory.

[0033] In dynamic traffic environments, the most important aspect of lane changing is ensuring that the intelligent bus does not collide with other dynamic vehicles or surrounding obstacles during the lane change process. This invention treats the intelligent bus's outline as a rectangle, with the vehicle's heading angle representing the rectangle's orientation at that moment. A series of parameters, including the longitudinal distance x, are generated from a real-time planned reference lane change trajectory. f For the corresponding rectangle, when the longitudinal speed of the intelligent bus is known, the longitudinal distance x for lane changing is determined by real-time planning at time t. f The positions of each vertex of the determined rectangle are calculated as follows:

[0034]

[0035]

[0036] In the formula x 1...4 (t) and y 1...4 (t) represents the x and y coordinates of the four vertices of the intelligent bus outline rectangle, and l and w represent the length and width of the intelligent bus. Let x(t) and y(t) represent the heading angle of the intelligent bus, x(t) and y(t) represent the position of the center of mass of the intelligent bus, and T be the time interval.

[0037] Determining whether the intelligent bus and the traffic vehicle collide at time t becomes determining whether their outlines intersect. This invention constructs a function, Bounding Space and Hierarchies (BSH), for this determination. Based on this function, the intelligent bus and the traffic vehicle will not collide at time t when the following criteria are met:

[0038]

[0039]

[0040]

[0041] In the formula, i = 1, 2, 3, 4, representing 4 vehicles respectively, x K_n (t) and y K_n (t) represents the coordinates of the nth vertex of the intelligent bus, x T_in (t) and y T_in (t) represents the coordinates of the nth vertex of the i-th traffic vehicle profile. Based on the above derivation, the following obstacle avoidance detection formula during lane changes is obtained:

[0042]

[0043] Minimum lane change longitudinal distance x determined by stability fmin Initially, the length is increased to obtain a lane-changing trajectory, and the collision of the intelligent bus under this trajectory is detected. If a collision occurs, the trajectory of the previous length represents the maximum longitudinal distance for lane changing under this condition. Therefore, the safe lane-changing trajectory cluster for the intelligent bus extends from the longitudinal length of the trajectory determined by stability to the longitudinal length of the safe lane-changing trajectory determined by all surrounding vehicles. Within this range, the lane-changing trajectory cluster will not collide with surrounding vehicles, nor will it cause yaw or tilting instability during the lane-changing process of the intelligent bus.

[0044] Step 4: Consider the impact of lane change efficiency and passenger comfort on the safe lane change trajectory obtained in Step 3, construct an objective function, and select the optimal intelligent bus lane change trajectory.

[0045]

[0046] In the formula, J represents the cost function, a(x) f ) represents the lateral acceleration at the final position after the lane change, a(x) f ) max t(x) represents the maximum lateral acceleration within the safe trajectory cluster. f ) represents the lane change time, t(x) f ) max ω represents the maximum lane change time in the safe trajectory cluster, where ω is the weight value.

[0047] Lateral acceleration a(x) at the final position f ) Calculated using the following equation:

[0048] a(x f )=u(x f ) 2 K(x f )

[0049] In the formula, u(x) f K(.) represents the speed of the intelligent bus at its final position, and K(.) represents the curvature function of the lane change trajectory. K(.) is calculated using the following equation:

[0050]

[0051] In the formula, y'(x f ) and y" (x f ( ) are the first and second derivatives of the lane change trajectory function, respectively;

[0052] y'(x f )=a1+2a2x f +3a3x 2 f

[0053] y"(x f )=2a2+6a3x f

[0054] Where a1, a2, and a3 are the parameters of the lane change trajectory function, x f This represents the longitudinal endpoint of the lane change trajectory.

[0055] The intelligent bus lane-changing trajectory running time t(x) f )for:

[0056]

[0057] In the formula, t c The intelligent bus has already changed lanes.

[0058] Step 5: During the lane change process, the intelligent bus adjusts its lane change trajectory and speed in real time. In case of an emergency, it abandons the lane change, returns to the original lane, and then replans until the intelligent bus reaches the target location.

[0059] When an intelligent bus is changing lanes, the status of surrounding vehicles may suddenly change drastically. Therefore, the lane-changing trajectory needs to be adjusted to ensure the intelligent bus avoids collisions. This invention adjusts the longitudinal length of the lane-changing trajectory curve at each moment to perform real-time trajectory adjustments, ensuring optimal planning of the intelligent bus's lane-changing trajectory. The lane-changing trajectory of the intelligent bus at each moment can be considered as a segment from the start of the lane change to the current trajectory; that is, the intelligent bus's lane-changing trajectory can be composed of three lane-changing trajectories of different longitudinal lengths. The specific algorithm is as follows: The intelligent bus lane-changing trajectory planning algorithm obtains the status of surrounding vehicles at intervals Δτ through sensing or vehicle-to-everything (V2X) communication. Combining this with the intelligent bus's position and speed, a cubic polynomial is used to plan the lane-changing trajectory in real time. This process is repeated until the intelligent bus reaches the target position, completing the lane change.

[0060] During lane-changing by an intelligent bus, if an emergency occurs in the surrounding dynamic environment, the intelligent bus should abandon the lane change and return to the original lane: when the vehicle in front in the target lane brakes suddenly, its deceleration increases dramatically, and the acceleration of the vehicle behind in the target lane increases sharply; when the longitudinal distance interval of the lane change is less than the minimum longitudinal distance x for lane change. fmin When the lane change is insufficient to ensure safe lane changing, the intelligent vehicle will abandon the lane change and plan a trajectory to return to the original lane.

[0061] When an intelligent vehicle performs a lane change operation, the longitudinal distance range of the lane change is less than the minimum longitudinal distance of the lane change x. fmin When the vehicle is about to leave, speed planning is performed to ensure the stability of the lane change trajectory. When speed planning is performed, the acceleration of the intelligent vehicle drops sharply. If the vehicle continues to travel at the planned speed, it will affect the comfort of the passengers. At this time, the lane change should be abandoned and a trajectory plan to return to the original lane should be performed.

[0062] When the longitudinal distance of the intelligent bus changing lanes is x f Less than the minimum lane change longitudinal distance x fmin At that time, speed planning is performed on the intelligent bus to reduce its speed, using a cubic polynomial model:

[0063] u x (t)=b0+b1t+b2t 2 +b3t 3

[0064] a x (t)=b1+2b2t+3b3t 2

[0065] In the formula, t, u x (t), a x (t) represents the current time, current speed, and current acceleration of the intelligent bus, respectively, where b0, b1, b2, and b3 are parameters.

[0066] The expressions for solving parameters b0, b1, b2, and b3 are as follows:

[0067]

[0068]

[0069]

[0070] Among them, u x0 Let a be the vehicle speed at the initial moment. x0 Let x be the vehicle's acceleration at the initial moment. f u is the longitudinal distance from the end of the lane change. limTo meet the critical vehicle speed requirements under boundary constraints such as driving safety and comfort, and to ensure occupant comfort, the maximum longitudinal acceleration of this vehicle was selected as the maximum acceptable longitudinal acceleration 'a' for occupant comfort. max The acceptable maximum longitudinal braking deceleration is a min To meet the comfort requirements during acceleration, we have:

[0071] a min ≤b1+2b2t+3b3t 2 ≤a max

[0072] When the critical terminal velocity u lim Once determined, the vehicle's speed planning depends only on t. f A variable, t f The value of should meet the above acceleration requirements.

[0073] To ensure that the intelligent bus does not collide with vehicles in front or behind in the target lane after changing lanes, and that the following vehicle also brakes when the preceding vehicle brakes suddenly, the minimum safe distances S1, S2, S3, and S4 must all be less than the actual distance between the two vehicles. The expression for the minimum safe distance is as follows:

[0074]

[0075]

[0076]

[0077]

[0078] Wherein, S1 is the minimum safe distance between the intelligent bus and the vehicle in front in the adjacent lane, S2 is the minimum safe distance between the intelligent bus and the vehicle behind in the adjacent lane, S3 is the minimum safe distance between the intelligent bus and the vehicle in front in its own lane, and S4 is the minimum safe distance between the intelligent bus and the vehicle behind in its own lane. V v1 represents the speed of the intelligent bus, v2 represents the speed of the vehicle in front in the adjacent lane, v3 represents the speed of the vehicle in front in this lane, v4 represents the speed of the vehicle in front in this lane, and a represents the speed of the vehicle behind. v τ represents the maximum braking deceleration of the intelligent bus, a1 represents the maximum braking deceleration of the vehicle in front in the adjacent lane, a2 represents the maximum braking deceleration of the vehicle behind in the adjacent lane, a3 represents the maximum braking deceleration of the vehicle in front in this lane, a4 represents the maximum braking deceleration of the vehicle behind in this lane, and τ V τ2 and τ4 represent the delay time in information interaction and control systems between the intelligent bus and surrounding vehicles, respectively, and the reaction time of the human driver. VV For the acceleration prepared for braking of the intelligent bus, a 22 and a 44The acceleration required to prepare for braking by vehicles in the adjacent lane and vehicles in the same lane.

[0079] This invention proposes a dynamic lane-changing trajectory planning method for intelligent buses in complex multi-vehicle traffic environments. It considers whether the lane-changing trajectory satisfies the vehicle's lateral and longitudinal dynamic stability, whether the lane-changing trajectory can be completed safely and comfortably in complex external traffic conditions, and factors such as the intelligent bus's dynamic speed planning during lane changing and the safe following distance after lane changing. Previous studies have generally rarely addressed the dynamic characteristics of vehicles in lane-changing trajectories, the speed changes of intelligent buses during lane changing, and the safe following distance after lane changing. Therefore, this invention uses these conditions as constraints to establish a more realistic simulated driving scenario, making the simulation process and results more consistent with the dynamic lane-changing process of intelligent buses in actual traffic environments.

[0080] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0081] 1. This invention considers whether the lane-changing trajectory of an intelligent bus satisfies the vehicle's yaw and roll stability, thereby deriving the minimum stable longitudinal distance x for lane changing under different vehicle speeds u and different road surface adhesion coefficients μ. fmin .

[0082] 2. This invention treats the vehicle outline as a rectangle, calculates the coordinates of the rectangle's vertices, determines whether the lane-changing trajectory is safe, and thus determines whether the vertices of the intelligent bus rectangle intersect with the vertices of surrounding traffic vehicles. The BSH function is used to determine the safety of the lane-changing trajectory more quickly and efficiently.

[0083] 3. This invention establishes a dynamic speed planning model for intelligent buses during lane changes, where the longitudinal distance of the lane change is x. f Less than the minimum lane change longitudinal distance x fmin At the same time, speed planning is performed for intelligent buses to ensure the stability of intelligent buses when changing lanes.

[0084] 4. This invention takes into account the safe following distance after lane change, so as to avoid collisions between the intelligent bus and vehicles in front and behind in the target lane after lane change.

[0085] 5. By establishing a complex and dynamic traffic environment and considering the real-time movement of surrounding traffic vehicles, this invention provides a more realistic, reliable, and universal simulation environment for experimental simulation. Attached Figure Description

[0086] Figure 1 This is a schematic diagram of the present invention.

[0087] Figure 2 This is a schematic diagram of a complex multi-vehicle traffic environment.

[0088] Figure 3A schematic diagram for generating unconstrained trajectory clusters.

[0089] Figure 4 This is a schematic diagram of the vehicle yaw dynamics model.

[0090] Figure 5 x is the minimum longitudinal lane change distance in this embodiment of the invention. fmin A three-dimensional graph relating vehicle speed u and road surface adhesion coefficient μ.

[0091] Figure 6 This is a schematic diagram illustrating the safety assessment principle of the lane change obstacle avoidance detection algorithm.

[0092] Figure 7 This is a diagram illustrating the safe following distance.

[0093] Figure 8 This is a schematic diagram of the replanning of dynamic lane change trajectories.

[0094] Figure 9 A schematic diagram of the trajectory planning for an intelligent bus to abandon lane changes and turn around.

[0095] Figure 10 This is a diagram illustrating the lane change planning trajectory process according to an embodiment of the present invention. Detailed Implementation

[0096] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0097] Example

[0098] This invention provides a dynamic trajectory planning method for intelligent buses in complex multi-vehicle traffic environments, the principle of which is as follows: Figure 1 As shown, firstly, after receiving a lane-change command, the intelligent bus begins planning its lane-change trajectory. Based on the intelligent bus's position and the motion state of surrounding vehicles, a series of candidate lane-change endpoint positions are selected on the target lane. A cubic polynomial curve is used to connect the initial position and the candidate endpoint positions to generate an unconstrained lane-change trajectory cluster. Secondly, based on the intelligent bus's speed and road surface adhesion coefficient, and taking into account the tire's saturation characteristics, a simulation model is used to obtain the minimum longitudinal lane-change distance x that prevents the intelligent bus from yawing and tilting. fmin This yields a stable lane-change trajectory; the lane-change trajectory obstacle avoidance detection module detects the minimum longitudinal lane-change distance x. fminThe system proceeds by performing collision detection from shortest to longest possible lane-change trajectory length to obtain the maximum lane-change longitudinal distance without collision, thus achieving a safe lane-change trajectory. The trajectory planning module sets multiple objective functions, including lane-change efficiency, and optimizes to obtain the optimal lane-change longitudinal length within the determined minimum and maximum lane-change longitudinal length range. The trajectory generation module generates a reference lane-change trajectory for the intelligent bus based on the longitudinal length. The lane-change abort module plans a lane-change abort trajectory when a sudden change in the surrounding vehicle's state causes the intelligent bus to exceed its stable range, and the original vehicle allows the intelligent bus to abort the lane change and return to its original lane. The speed planning module plans the intelligent bus's lane-change speed to continue if a lane-change abort would pose a danger to the original vehicle. Simultaneously, the trajectory planning process progresses in real-time based on the surrounding vehicle status, real-time road conditions, and the vehicle's own status, rolling forward at a time interval of Δτ until the intelligent bus changes to the target lane, at which point the intelligent bus trajectory planning ends.

[0099] The specific steps are as follows:

[0100] Step 1: Establish a complex dynamic traffic environment and generate an unconstrained lane change trajectory cluster.

[0101] This embodiment establishes a complex dynamic traffic environment, such as... Figure 2 As shown, the research object is an intelligent bus, labeled (V). During the modeling process, the motion states of the four surrounding vehicles that may affect the intelligent bus's lane-changing trajectory should be considered simultaneously. These are: the vehicle in front in the adjacent lane (1), the vehicle behind in the adjacent lane (2), the vehicle in front in this lane (3), and the vehicle behind in this lane (4). When the intelligent bus receives a lane-changing command, based on the intelligent bus's position and the motion states of the surrounding traffic vehicles, a cubic curve is used to generate a cluster of unconstrained lane-changing trajectories, such as... Figure 3 As shown.

[0102] y(x)=a0+a1x+a2x 2 +a3x 3

[0103] x and y(x) represent the longitudinal and lateral positions of the intelligent bus, respectively; a0, a1, a2, and a3 are parameters that satisfy:

[0104] y(x0)=a0+a1x0+a2x 2 0+a3x 3 0

[0105] y'(x0)=a1+2a2x0+3a3x 2 0

[0106] x0 and y(x0) are the initial longitudinal and lateral positions of the intelligent bus during lane change, respectively; y'(x0) is the tangent direction of the initial position of the intelligent bus.

[0107] The intelligent bus's final movement state after lane change is:

[0108] y(x f )=a0+a1x f +a2x 2 f +a3x 3 f

[0109] y'(x f )=a1+2a2x f +3a3x 2 f

[0110] x f y(x) f ) represent the longitudinal and lateral endpoints of the lane change for the intelligent bus; y'(x f ) represents the tangent direction at the destination of the intelligent bus; a0, a1, a2, and a3 are parameters in the lane change trajectory.

[0111] Step 2: Establish vehicle yaw and roll stability models, and derive the minimum longitudinal distance x for lane changes under different vehicle speeds u and different road surface adhesion coefficients μ. fmin This results in a stable lane change trajectory.

[0112] (1) Vehicle yaw stability

[0113] Yaw velocity stability limits and center of mass sideslip angle stability limits:

[0114]

[0115]

[0116] Where, r s It is the vehicle's yaw rate, β s It is the vehicle's sideslip angle, F. yr F yf Let g and μ represent the lateral forces acting on the rear and front axles of the vehicle, respectively. g is the acceleration due to gravity, μ is the road adhesion coefficient, u is the vehicle speed, m is the vehicle mass, and C is the total mass. αr This indicates the rear axle lateral stiffness, where a and b represent the distances from the front and rear axles to the vehicle's center of gravity, respectively.

[0117] like Figure 4 As shown, different lane change longitudinal distances are set in the vehicle simulation model for different vehicle speeds and road adhesion coefficients; the yaw rate stability limit and the center of gravity sideslip angle stability limit are obtained respectively. The phase plane used to judge the lateral stability of the vehicle is obtained by combining these two limits. When the output yaw rate and center of gravity sideslip angle are within the phase plane, the lane change longitudinal distance x is...f1 To ensure the vehicle's lateral stability.

[0118] (2) Vehicle roll stability

[0119] Lateral load transfer rate

[0120] Where n is the number of vehicle axles, F Zl For the vertical load on the left side of the wheel, F Zr The vertical load on the right side of the wheel; in the simulated vehicle model, the longitudinal distance x for lane changes is mainly set for different vehicle speeds u and different road surface adhesion coefficients μ. f Using experimental data on lateral acceleration and vertical load on each wheel as output, the minimum longitudinal lane change distance x to prevent rollover is determined. f2 .

[0121] Minimum lane change longitudinal distance x fmin When the vehicle speed and road surface adhesion coefficient are the same, x f1 With x f2 The larger value in the range.

[0122] Minimum lane change longitudinal distance x for intelligent buses fmin A three-dimensional graph showing vehicle speed u and road surface adhesion coefficient μ is shown below. Figure 5 As shown, when the longitudinal distance x of the lane change f Greater than x fmin This means that the lane-changing trajectory is assumed to satisfy the vehicle's yaw and roll stability. Through... Figure 5 The 3D graph shown is fitted with a surface to determine the minimum longitudinal lane change distance x. fmin The functional expression relating vehicle speed u and road adhesion coefficient μ is:

[0123] x fmin (u,μ)=58.58-0.3892u-139.1μ+0.01839u 2 -1.669μm +261.5μm 2 -5.158e -5 u 3 -0.003438u 2 μ+1.09uμ 2 -147.9μ 3

[0124] In this embodiment, the speed u of the intelligent bus is set to 90 km / h. -1 The road surface adhesion coefficient μ is 0.8, according to the above function expression and Figure 5 The minimum lane change longitudinal distance x is obtained. fmin It is 35m. When the longitudinal distance x of the lane change trajectory f When the distance is greater than 35m, the lane change trajectory satisfies the vehicle's yaw and roll stability.

[0125] Step 3: Obtain the optimal lane change trajectory

[0126] Through such Figure 6 and 7 The obstacle avoidance detection algorithm during lane changes and the safe following model after lane changes, as shown, yield a cluster of safe lane change trajectories. The objective function is then optimized to obtain the optimal lane change trajectory from this cluster. This invention treats the vehicle's outline as a rectangle, with the vehicle's heading angle representing the rectangle's orientation at that moment. Therefore, a series of trajectories with longitudinal distance x are generated from the real-time planned reference lane change trajectory. f The corresponding rectangle. When the longitudinal speed of the intelligent bus is known, the longitudinal distance x for lane changing is planned in real time at time t. f The positions of the vertices of the defined rectangle are calculated as follows:

[0127]

[0128]

[0129] In the formula x 1...4 (t) and y 1...4 (t) represents the x and y coordinates of the four vertices of the intelligent bus outline rectangle, and l and w represent the length and width of the intelligent bus. Let x(t) represent the current heading angle of the intelligent bus, x(t) and y(t) represent the centroid positions of the intelligent bus, and T be the time interval. Surrounding traffic vehicles are calculated using a similar method. Therefore, determining whether the intelligent bus and the traffic vehicle collide at time t becomes determining whether their outlines intersect. The longitudinal distance x corresponding to the lane change for each trajectory in the stable trajectory cluster is then used. f The BSH function is used for rigorous testing. A trajectory that satisfies the following criteria is considered a safe reference trajectory during lane changes; otherwise, it is considered an unsafe trajectory. Therefore, the safe lane-change trajectory cluster for intelligent buses extends from the longitudinal length of the trajectory determined by stability to the longitudinal length of the safe lane-change trajectory determined by all surrounding vehicles. Lane-change trajectory clusters within this range will not collide with surrounding vehicles, nor will they cause yaw or tilting instability during the intelligent bus's lane-change process.

[0130]

[0131]

[0132]

[0133] In the formula, i = 1, 2, 3, 4, representing 4 vehicles respectively, x K_n (t) and y K_n (t) represents the coordinates of the nth vertex of the intelligent bus, x T_in (t) and yT_in (t) represents the coordinates of the nth vertex of the i-th traffic vehicle profile. Based on the above derivation, the following obstacle avoidance detection formula during lane changes is obtained:

[0134]

[0135] Minimum lane change longitudinal distance x determined by stability fmin Initially, the length is increased to obtain a lane-changing trajectory, and it is detected whether the intelligent bus collides under this trajectory; if a collision occurs, the trajectory of the previous length is the maximum longitudinal distance of lane changing under this working condition.

[0136] In this embodiment, a minimum safe distance model is used to ensure that the intelligent bus does not collide with vehicles in front or behind in the target lane after changing lanes. If the vehicle in front brakes suddenly, the vehicle behind also brakes. For the two vehicles to avoid a collision, the minimum safe distances S1, S2, S3, and S4 must all be less than the actual distance between the two vehicles. The expression for the minimum safe distance is as follows:

[0137]

[0138]

[0139]

[0140]

[0141] Wherein, S1 is the minimum safe distance between the intelligent bus and the vehicle in front in the adjacent lane, S2 is the minimum safe distance between the intelligent bus and the vehicle behind in the adjacent lane, S3 is the minimum safe distance between the intelligent bus and the vehicle in front in its own lane, and S4 is the minimum safe distance between the intelligent bus and the vehicle behind in its own lane. V v1 represents the speed of the intelligent bus, v2 represents the speed of the vehicle in front in the adjacent lane, v3 represents the speed of the vehicle in front in this lane, v4 represents the speed of the vehicle in front in this lane, and a represents the speed of the vehicle behind. v τ represents the maximum braking deceleration of the intelligent bus, a1 represents the maximum braking deceleration of the vehicle in front in the adjacent lane, a2 represents the maximum braking deceleration of the vehicle behind in the adjacent lane, a3 represents the maximum braking deceleration of the vehicle in front in this lane, a4 represents the maximum braking deceleration of the vehicle behind in this lane, and τ V τ2 and τ4 represent the delay time in information interaction and control systems between the intelligent bus and surrounding vehicles, respectively, and the reaction time of the human driver. VV For the acceleration prepared for braking of the intelligent bus, a 22 and a 44 The acceleration required to prepare for braking by vehicles in the adjacent lane and vehicles in the same lane.

[0142] The above method is used to eliminate lane-changing trajectories that do not meet the minimum safe distance requirement during the lane-changing process, thus forming a safe lane-changing trajectory cluster. The objective function value of each trajectory in the safe trajectory cluster is calculated, and the lane-changing trajectory corresponding to the minimum value is selected as the optimal lane-changing trajectory for the current moment.

[0143] The objective function considers the impact of lane change efficiency and passenger comfort on the lane change trajectory:

[0144]

[0145] In the formula, J represents the cost function, a(x) f ) represents the lateral acceleration at the final position after the lane change, a(x) f ) max t(x) represents the maximum lateral acceleration within the safe trajectory cluster. f ) represents the lane change time, t(x) f ) max ω represents the maximum lane change time in the safe trajectory cluster, where ω is the weight value.

[0146] Lateral acceleration a(x) at the final position f ) Calculated using the following equation:

[0147] a(x f )=u(x f ) 2 K(x f )

[0148] In the formula, u(x) f K(.) represents the speed of the intelligent bus at its final position, and K(.) represents the curvature function of the lane change trajectory. K(.) is calculated using the following equation:

[0149]

[0150] In the formula, y'(x f ) and y" (x f ( ) are the first and second derivatives of the lane change trajectory function, respectively;

[0151] The intelligent bus lane-changing trajectory running time t(x) f )for:

[0152]

[0153] In the formula, t c The intelligent bus has already changed lanes.

[0154] Step 4: Obtain the trajectory and then plan the lane change trajectory

[0155] In this embodiment, due to the sudden change in speed of surrounding vehicles (1)(2)(3)(4) during the lane change process, the intelligent bus may collide with surrounding vehicles if the original trajectory is continued. To ensure safety, the trajectory needs to be replanned. By outputting the vehicle's lateral position, longitudinal position, and heading angle at the current moment through the vehicle system, steps one, two, and three above are repeatedly called to obtain the replanned optimal lane change trajectory, such as... Figure 8-9 As shown.

[0156] Figure 10 The entire lane-changing process of the intelligent bus in this embodiment under the given environmental conditions is described. Figure 10 It can be seen that at any given moment, the intelligent bus does not overlap with surrounding vehicles. Based on this, the proposed trajectory planning algorithm can adjust the lane-changing reference trajectory in real time according to the dynamic state of surrounding vehicles, thus realizing dynamic planning of lane-changing trajectories.

[0157] After the intelligent bus initiates a lane change, the minimum longitudinal length of the lane change trajectory determined by the intelligent bus's stability is interpolated into a 3D MAP, yielding a minimum longitudinal length of 26m. Due to the dynamic changes in the acceleration of surrounding vehicles, the intelligent bus's lane change trajectory may need to be replanned. The optimal trajectory length at the start of the lane change is 91m; however, as the state of surrounding vehicles changes, the optimal trajectory length becomes 89m. Therefore, the proposed trajectory planning algorithm can adjust the intelligent bus's lane change trajectory according to changes in the dynamic traffic environment to ensure safe and optimal lane changes.

[0158] By performing vehicle stability analysis on the planned trajectory curve, it can be found that the sideslip angle and yaw rate of the planned trajectory curve are located at the center of the phase plane of the sideslip angle and yaw rate, which does not exceed the stability range. Therefore, the proposed trajectory planning algorithm guarantees the vehicle's yaw stability. The lateral acceleration is found to be -0.8 m / s². 2 up to 1.1 m / s 2 The roll angle varies smoothly between -0.2° and 0.3°. The load transfer rate (LTR) is in the range of [-1, 1]. Therefore, the proposed trajectory planning algorithm guarantees vehicle roll stability.

[0159] It will be readily understood by those skilled in the art that the above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the invention should be included within the protection scope of the present invention.

Claims

1. A method for dynamic lane-changing trajectory planning of intelligent buses in complex multi-vehicle traffic environments, characterized in that: First, for dynamic traffic environments, a cluster of unconstrained lane-changing trajectories for the intelligent bus is generated. Second, for different road conditions, stability analysis of vehicle yaw and roll is performed to determine the minimum longitudinal distance for stable lane changing of the intelligent bus, and unstable lane-changing trajectories are eliminated from the unconstrained lane-changing trajectory cluster. Then, considering changes in the state of surrounding vehicles, obstacle avoidance detection is performed to obtain the maximum longitudinal length of the lane-changing trajectory of the intelligent bus without colliding with surrounding vehicles, and unsafe lane-changing trajectories that may collide are eliminated. Finally, a multi-objective lane-changing trajectory performance function is set to obtain the optimal intelligent bus lane-changing trajectory among the above stable and collision-free trajectories. At the same time, considering sudden changes in the state of surrounding vehicles, lane-changing trajectory replanning and speed planning for the intelligent bus are performed. Specifically, the steps include the following: Step 1: Based on the location of the intelligent bus and the movement status of surrounding traffic vehicles, select a series of lane change endpoint candidate positions on the target lane, and use a cubic polynomial curve to connect the initial position of the vehicle and the endpoint candidate positions to generate an unconstrained lane change trajectory cluster. Step 2: Based on the speed of the intelligent bus and the road surface adhesion coefficient, and taking into account the saturation characteristics of the tires, a simulation model is used to obtain the minimum longitudinal distance for lane changing without causing yaw or tilting issues in the intelligent bus. This results in a stable lane change trajectory; Step 3: Based on the minimum lane change longitudinal distance Collision detection is performed gradually from short to long to obtain the maximum longitudinal distance of the lane change without collision, thus obtaining a safe lane change trajectory. Step 4 considers the impact of lane change efficiency and passenger comfort on the safe lane change trajectory obtained in Step 3, constructs an objective function, and selects the optimal intelligent bus lane change trajectory. Step 5: During the lane change process of the intelligent bus, its lane change trajectory and speed are adjusted in real time. In case of emergency, the lane change is abandoned, the vehicle returns to the original lane, and the lane is replanned until the intelligent bus reaches the target location. When an intelligent vehicle performs a lane change operation, the lane change longitudinal distance range is less than the minimum lane change longitudinal distance. When the vehicle is about to leave, speed planning is performed to ensure the stability of the lane change trajectory. When speed planning is performed, the acceleration of the intelligent vehicle drops sharply. If the vehicle continues to travel at the planned speed, it will affect the comfort of the passengers. At this time, the lane change should be abandoned and a trajectory plan to return to the original lane should be performed.

2. The method for dynamic lane-changing trajectory planning of intelligent buses in complex multi-vehicle traffic environments according to claim 1, characterized in that: The specific content of Step 1 is as follows: , These refer to the longitudinal and lateral positions of the intelligent bus, respectively. , , , Let be a parameter, and satisfy: , These represent the initial longitudinal and lateral positions of the intelligent bus during lane changing; The tangent direction is the initial position of the intelligent bus; The intelligent bus's final movement state after lane change is: , These are the longitudinal and lateral endpoints of the intelligent bus lane change; The tangent direction at the destination of the intelligent bus; , , , These are the parameters in the lane change trajectory.

3. The method for dynamic lane-changing trajectory planning of intelligent buses in complex multi-vehicle traffic environments according to claim 1, characterized in that: The specific content of Step 2 is as follows: (1) Vehicle yaw stability Yaw velocity stability limits and center of mass sideslip angle stability limits: in, It is the vehicle's yaw rate. It is the vehicle's sideslip angle. , These represent the lateral forces acting on the rear axle and the front axle of the vehicle, respectively. It is the acceleration due to gravity. The road surface adhesion coefficient, For vehicle speed, It's the overall vehicle quality. Indicates the rear axle lateral stiffness. , These represent the distances from the front axle and rear axle to the vehicle's center of gravity, respectively. (2) Vehicle roll stability Lateral load transfer rate in, For the number of vehicle axles, The load is the vertical load on the left side of the wheel. The vertical load on the right side of the wheel; (3) Vehicle simulation In the vehicle simulation model, set the longitudinal distance for lane changes at different vehicle speeds and different road surface adhesion coefficients; The output yaw rate and sideslip angle are used to determine yaw stability in the phase plane formed by the yaw rate stability limit and the sideslip angle stability limit, thus obtaining the minimum longitudinal lane change distance that satisfies vehicle yaw stability. ; Output the vertical loads on the left and right wheels, calculate the lateral load transfer rate (LTR) value, and determine the minimum longitudinal lane change distance to prevent rollover. ; Minimum lane change longitudinal distance When taking the same vehicle speed and the same road surface adhesion coefficient, and The larger value in the range.

4. The method for dynamic lane-changing trajectory planning of intelligent buses in complex multi-vehicle traffic environments according to claim 1, characterized in that: The specific content of Step 3 is as follows: The outline of the intelligent bus is considered as a rectangle, and the heading angle of the vehicle is the orientation of the rectangle at any given time. A series of lane-change longitudinal distances are generated from the reference lane-change trajectory planned in real time. The corresponding rectangle, when the longitudinal speed of the intelligent bus is known, in t Real-time lane change longitudinal distance The positions of each vertex of the determined rectangle are calculated as follows: In the formula and The coordinates of the four vertices of the rectangle representing the outline of the intelligent bus x and y The value, and This indicates the length and width of the intelligent bus. Indicates the heading angle of the intelligent bus. and Indicates the center of gravity position of the intelligent bus. For time intervals; exist Determining whether a collision occurs between an intelligent bus and a regular vehicle becomes determining whether their outlines intersect. Specifically, this is based on the BSH function, and the intelligent bus and the regular vehicle are considered to collide when the following criteria are met. Never collide: In the formula, Each refers to one of the four vehicles. and The first intelligent bus n vertex coordinates and It is the first The first outline of the traffic vehicle n The coordinates of the vertices are used to derive the following obstacle avoidance detection formula during lane changes: Minimum lane change longitudinal distance determined by stability Initially, the length is increased to obtain a lane-changing trajectory, and it is detected whether the intelligent bus collides under this trajectory; if a collision occurs, the trajectory of the previous length is the maximum longitudinal distance of lane changing under this working condition.

5. The method for dynamic lane-changing trajectory planning of intelligent buses in complex multi-vehicle traffic environments according to claim 1, characterized in that: The objective function for Step 4 is: In the formula, Represents the cost function. This represents the lateral acceleration at the final position after the lane change. This represents the maximum lateral acceleration within the safe trajectory cluster. Indicates the time of lane change. This represents the maximum lane change time within the safe trajectory cluster. These are weight values; Lateral acceleration at final position Calculate using the following equation: In the formula, This indicates the speed of the intelligent bus at its final location. The curvature function representing the trajectory of a lane change. Calculate using the following equation: In the formula, and These are the first and second derivatives of the lane change trajectory function, respectively; Intelligent bus lane change trajectory running time for: In the formula, The intelligent bus has already changed lanes.

6. The method for dynamic lane-changing trajectory planning of intelligent buses in complex multi-vehicle traffic environments according to claim 1, characterized in that: Every now and then The lane-changing trajectory is updated in real time and continuously until the intelligent bus reaches the target lane-changing position. During lane-changing by an intelligent bus, if an emergency occurs in the surrounding dynamic environment, the intelligent bus should abandon the lane change and return to the original lane: when the vehicle in front in the target lane brakes suddenly, its deceleration increases dramatically, and the acceleration of the vehicle behind in the target lane increases sharply; or when the longitudinal distance of the lane change is less than the minimum longitudinal distance of the lane change. When the lane change is insufficient to ensure safe lane changing for the vehicle, the intelligent vehicle will abandon the lane change and plan a trajectory to return to the original lane.

7. The method for dynamic lane-changing trajectory planning of intelligent buses in complex multi-vehicle traffic environments according to claim 6, characterized in that: When the intelligent bus changes lanes longitudinal distance Less than the minimum lane change longitudinal distance At that time, speed planning is performed on the intelligent bus to reduce its speed, using a cubic polynomial model: In the formula, , , Let $\mathbf$ represent the current time, $\mathbf$ the current speed, and $\mathbf$ the current acceleration of the intelligent bus, respectively. , , , For parameters.

8. The method for dynamic lane-changing trajectory planning of intelligent buses in complex multi-vehicle traffic environments according to claim 1, characterized in that: The minimum safe distance model is used to ensure that the intelligent bus does not collide with vehicles in front or behind in the target lane after changing lanes. If the vehicle in front brakes suddenly, and the vehicle behind also brakes, the minimum safe distance must be met to prevent a collision. , , , Both must be less than the actual distance between the two vehicles in front and behind. The expression for the minimum safe distance is as follows: in, The minimum safe distance between the intelligent bus and the vehicle in front in the adjacent lane. The minimum safe distance between the intelligent bus and the vehicle behind in the adjacent lane. The minimum safe distance between the intelligent bus and the vehicle in front in this lane. The minimum safe distance between the intelligent bus and the vehicle behind it in this lane. For the speed of the intelligent bus, The speed of the vehicle in front in the adjacent lane. This represents the speed of the vehicle following in the adjacent lane. The speed of the vehicle ahead in this lane. This represents the speed of the vehicle following in this lane. This is the maximum braking deceleration of the intelligent bus. The maximum braking deceleration of the vehicle in front in the adjacent lane. This is the maximum braking deceleration of the vehicle following in the adjacent lane. This is the maximum braking deceleration of the vehicle ahead in this lane. This is the maximum braking deceleration of the vehicle following in this lane. The delay in information exchange and control systems between the intelligent bus and surrounding vehicles is due to the time required for these systems to function. and For human driver reaction time Acceleration for intelligent buses during braking and The acceleration required to prepare for braking by vehicles in the adjacent lane and vehicles in the same lane.

Images