Automatic driving vehicle real-time escape decision-making method and system based on active safety

Abstract

The invention provides an automatic driving vehicle real-time escape decision-making method and system based on active safety. A collision risk real-time judgment model based on collision time and safety escape margin of a front vehicle and a rear vehicle is constructed; judging whether a collision risk exists between the current AEV and a rear conflict vehicle RCV of the current lane or not in real time through a collision risk real-time judgment model; and carrying out longitudinal escape or lane-changing escape based on the collision risk real-time judgment model. And preferentially executing a longitudinal acceleration escape strategy, starting an emergency lane changing escape mode when the longitudinal acceleration escape strategy is not met, and continuously performing behavior optimization on the decision-making system at a certain time interval in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous driving control, and particularly relates to a real-time escape decision-making method and system for an autonomous driving vehicle based on active safety. Background Art

[0002] With the rapid iteration of intelligent driving technology and the acceleration of the industrialization process, autonomous driving vehicles (Auto-escape Vehicles, AEVs) are gradually moving from closed test fields to open road environments.

[0003] When dealing with complex urban road conditions, although AEVs have technical advantages such as multi-modal environment perception, collaborative decision-making algorithms, and wire-controlled execution architectures, their dynamic risk avoidance capabilities in normal driving scenarios still face severe challenges. In particular, the challenges brought by dangerous traffic participants behind in the traffic flow have not been fully resolved: on the one hand, the behavioral uncertainties of heterogeneous traffic flow participants lead to a misjudgment rate of 17.3% in traditional collision prediction models, and the dangerous behaviors brought by some dangerous traffic participants from behind are not considered. When the dangerous traffic participants behind are very dangerous, the damage to AEVs and the entire traffic environment is inestimable.

[0004] Therefore, it is urgent to construct an intelligent risk avoidance system that integrates real-time risk situation assessment and multi-objective optimization control. By establishing a real-time collision risk determination model and a basic escape model for AEVs, the active safety defense capabilities of AEVs in normal driving scenarios can be upgraded, which has become a key technical path to break through the commercialization bottleneck of intelligent driving. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a real-time escape decision-making method for an autonomous driving vehicle based on active safety, including the following steps:

[0006] Step S1: Construct a real-time collision risk determination model based on the time to collision and safety escape margin between the front and rear vehicles.

[0007] Step S2: Use the real-time collision risk determination model to determine in real time whether there is a collision risk between the current autonomous driving vehicle AEV and the rear conflict vehicle RCV in the current lane. If there is, enter Step S3.

[0008] Step S3: Based on the real-time collision risk determination model, calculate the safe intrusion acceleration interval of AEV relative to the vehicle in front FV in the current lane and the safe escape acceleration interval of AEV relative to RCV. If there is an intersection between the two intervals, generate a linear longitudinal escape acceleration function based on the maximum speed limit and the acceleration interval, and drive AEV to accelerate and escape according to the linear longitudinal escape acceleration function. Otherwise, enter Step S4.

[0009] Step S4: Based on the real-time collision risk determination model, calculate the time to collision (TTC) with the vehicles in front of and behind the target lane for lane change and the safety escape margin to determine the feasibility of lane change. If the safety conditions are met, generate a linear lateral escape acceleration function based on the lane width and the initial lateral velocity, and drive the AEV to perform lane change.

[0010] Preferably, the expression of the real-time collision risk determination model is:

[0011]

[0012] In the formula, TTC is the time to collision between the vehicle behind and the vehicle in front; T safe represents the maximum time required for the autonomous driving vehicle to make a system response under normal conditions; SEM is the safety escape margin between the vehicle behind and the vehicle in front; D rv-fv represents the real-time distance between the vehicle behind and the vehicle in front during longitudinal driving.

[0013] Preferably, the calculation expression of the time to collision (TTC) between the vehicle behind and the vehicle in front is:

[0014]

[0015] In the formula, and are the distances traveled by the vehicle behind and the vehicle in front respectively until the two vehicles collide when driving in the current vehicle state; L fv is the actual vehicle length of the vehicle in front; v rv and v fv are the real-time vehicle speeds of the vehicle behind and the vehicle in front respectively; a in is the real-time intrusion acceleration of the vehicle behind.

[0016] Preferably, the calculation expression of the safety escape margin between the vehicle behind and the vehicle in front is:

[0017]

[0018] In the formula, and are the distances traveled by the vehicle behind and the vehicle in front respectively within the maximum reaction time of the autonomous driving vehicle; T safe is the maximum time required for the autonomous driving vehicle to make a system response under normal conditions; D safe is the limit safe following distance when the two vehicles are driving.

[0019] Preferably, in step S3, the linear longitudinal escape acceleration function is a piecewise linear function, and its expression is:

[0020]

[0021] In the formula, a AEV(y) (t) is the linear longitudinal escape acceleration of the current autonomous driving vehicle (AEV) changing with time; T is the time taken for the AEV to reach the maximum speed limit v max from the start of the longitudinal escape acceleration behavior; j y is the acceleration change rate when the AEV takes the longitudinal escape acceleration behavior; a (AEV)min represents the minimum longitudinal escape acceleration included in the acceleration interval; a (AEV)max represents the maximum longitudinal escape acceleration included in the acceleration interval; v AEV represents the real-time vehicle speed of the AEV.

[0022] Preferably, in step S4, the expression of the linear lateral escape acceleration function is:

[0023] a AEV(x) (t) = a AEV(x) (0) + j x t, t ∈ [0, TTC1];

[0024]

[0025] In the formula, a AEV(x) (t) is the linear lateral escape acceleration of the current autonomous driving vehicle (AEV) changing with time; j x is the acceleration change rate when the AEV takes the lateral escape acceleration behavior; a AEV(x) (0) is the initial lateral escape acceleration of the AEV; D is the current lane width; TTC1 represents the collision time.

[0026] The present invention also provides a real-time escape decision-making system for autonomous driving vehicles based on active safety, including a collision risk determination module, a longitudinal acceleration escape module, and an emergency lane change escape module;

[0027] The collision risk determination module is used to calculate the collision time TTC and the safety escape margin SEM between the autonomous driving vehicle (AEV) and the rear conflicting vehicle (RCV), and when TTC ≤ the maximum reaction time of the system and the actual headway distance ≤ SEM, output a collision risk trigger signal;

[0028] The longitudinal acceleration escape module is used to, after receiving the collision risk trigger signal, calculate the safe intrusion acceleration interval of the AEV relative to the vehicle in front (FV) in the current lane and the safe escape acceleration interval of the AEV relative to the RCV. If there is an intersection between the two intervals, generate a linear longitudinal escape acceleration function based on the maximum speed limit and the acceleration interval, and drive the AEV to accelerate according to this function. Otherwise, transmit the collision risk trigger signal to the emergency lane change escape module;

[0029] The emergency lane-changing escape module calculates the lane-changing feasibility based on the TTC and SEM between the AEV and the vehicles in front and behind in the target lane. If the safety conditions are met, a linear lateral escape acceleration function based on the lane width and the lateral initial speed is generated, and the AEV is driven to change lanes.

[0030] Preferably, in the collision risk determination module, the TTC is calculated according to the following formula:

[0031]

[0032] In the formula, and are the distances traveled by the rear vehicle and the front vehicle respectively until the two vehicles collide when driving in the current vehicle state; L fv is the actual vehicle length of the front vehicle; v rv and v fv are the real-time vehicle speeds of the rear vehicle and the front vehicle respectively; a in is the real-time intrusion acceleration of the rear vehicle.

[0033] Preferably, in the collision risk determination module, the SEM is calculated according to the following formula:

[0034]

[0035] In the formula, and are the distances traveled by the rear vehicle and the front vehicle respectively within the maximum reaction time of the autonomous vehicle; T safe is the maximum time required for the autonomous vehicle to make a system reaction in the normal state; D safe is the limit safe following distance when the two vehicles are driving.

[0036] Preferably, in the longitudinal acceleration escape module, the linear longitudinal escape acceleration function is a piecewise linear function, and its expression is:

[0037]

[0038] In the formula, a AEV(y) (t) is the linear longitudinal escape acceleration of the current autonomous vehicle AEV changing with time; T is the time taken for the AEV to reach the maximum speed limit v max from the start of taking the longitudinal escape acceleration behavior; j y is the acceleration change rate taken when the AEV takes the longitudinal escape acceleration behavior; a (AEV)min represents the minimum longitudinal escape acceleration included in the acceleration interval; a (AEV)max represents the maximum longitudinal escape acceleration included in the acceleration interval; v AEV represents the real-time vehicle speed of the AEV.

[0039] The beneficial effects of the present invention at least include: The present invention provides a real-time escape decision-making system for autonomous driving vehicles based on active safety. By constructing a closed-loop control architecture of "perception - decision - execution", it realizes the dynamic safety response of an autonomous electric vehicle (AEV) to a rear vehicle (RCV). The system uses the time parameter T as the time reference unit to establish a real-time collision risk determination model to determine whether the RCV has a risk of colliding with the AEV. When a potential collision risk is detected, the system adopts a two-layer progressive escape strategy optimization mechanism: First, in the longitudinal control dimension, by establishing an acceleration constraint equation, the longitudinal optimal acceleration solution that meets the safety conditions is solved; then, from the lateral control dimension, based on the comfortable lane-changing mode, the lateral optimal acceleration solution of the AEV is calculated. The system preferentially executes the longitudinal acceleration escape strategy. When it is not satisfied, it starts the emergency lane-changing escape mode and continuously optimizes the behavior of the decision-making system at a certain time interval. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic flowchart of the method according to an embodiment of the present invention;

[0041] Figure 2 It is a schematic diagram of the vehicle classification signs according to an embodiment of the present invention;

[0042] Figure 3 It is a schematic diagram of the collision risk determination scenario of the real-time collision risk determination model according to an embodiment of the present invention;

[0043] Figure 4 It is a schematic diagram of the collision risk determination scenario for an autonomous driving vehicle according to an embodiment of the present invention;

[0044] Figure 5 It is a schematic diagram of the scenario for solving the intrusion acceleration according to an embodiment of the present invention;

[0045] Figure 6 It is a schematic diagram of the scenario for solving the escape acceleration according to an embodiment of the present invention;

[0046] Figure 7 It is a schematic diagram of the AEV longitudinal escape acceleration function according to an embodiment of the present invention;

[0047] Figure 8 It is a schematic diagram of the AEV emergency lane-changing scenario according to an embodiment of the present invention;

[0048] Figure 9 It is a graph of the AEV lateral escape acceleration function according to an embodiment of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] Combined with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present invention.

[0050] As Figure 1 shown, the embodiments of the present invention provide a real-time escape decision method for autonomous driving vehicles based on active safety, including the following steps:

[0051] Step S1: Construct a real-time determination model for collision risk based on the time to collision and safety escape margin between the front and rear vehicles.

[0052] The scenario of the present invention only studies the normal driving process of vehicles, and the vehicles under such heterogeneous traffic flow are classified as follows, as Figure 2 shown.

[0053] AEV is the automatic escape vehicle, that is, the autonomous driving vehicle studied in the invention;

[0054] RCV is the rear conflict vehicle, that is, the dangerous vehicle referred to in the invention;

[0055] FV, that is, the vehicle in front of the vehicle's driving road referred to in the invention;

[0056] ALFV, that is, the vehicle in front of the target lane of the vehicle when the invention performs a lane change operation;

[0057] ALRV, that is, the vehicle behind the target lane of the vehicle when the invention performs a lane change operation.

[0058] The AEV can obtain the vehicle trajectory information such as the headway distance information, vehicle speed, and vehicle acceleration between all vehicles in the road network and itself in real time through the roadside equipment laid in the road network. Unless otherwise specified, the unit of distance is m; v refers to the vehicle speed, the unit is m / s; the unit of time is s; the unit of acceleration is m / s 2 .

[0059] During the normal driving process of an autonomous vehicle, if a dangerous vehicle appears behind and approaches rapidly with a risk of rear-end collision, to ensure its own safety, the autonomous vehicle needs to avoid danger through active behaviors such as accelerating or changing lanes. Such behaviors are collectively referred to as vehicle escape behaviors. However, relying solely on qualitative risk perception is difficult to accurately trigger escape decisions. Therefore, it is necessary to construct a real-time collision risk determination model from two dimensions: time risk and space risk. This model quantifies and analyzes the dynamic interaction relationship between the rear vehicle and the front vehicle in the collision scenario when there is an autonomous vehicle, calculates the risk in real time, and thus determines whether the current scenario reaches the critical condition for triggering an escape decision. The collision risk determination scenarios are as shown in Figure 3 as shown.

[0060] According to the collision risk determination scenario and quantifying this collision risk from two perspectives of time and space, in terms of time, consider the collision time between the rear vehicle rv and the front vehicle fv, that is, when the two vehicles will collide when driving in the current vehicle state; in terms of space, consider the safety escape margin between rv and fv, that is, the actual head-to-head distance between the two vehicles that can still maintain the minimum safe distance when the fv has not made a reaction and the two vehicles are driving in the current vehicle state. Based on this, a collision risk determination model is established, and the determination formula is shown in Equation (1).

[0061]

[0062] In the formula: TTC is the collision time between the rv and fv vehicles, with the unit of s; T safe is the maximum time required for the autonomous vehicle to make a system reaction in the normal state, with the unit of s; SEM is the safety escape margin between rv and fv, with the unit of m; D rv-fv is the real-time actual distance between rv and fv during longitudinal driving, with the unit of m.

[0063] From the time perspective, since the autonomous vehicle needs a certain amount of time to react to decision-making behaviors, in order to prevent the fv from being collided by the rv in a collision-risk scenario, it is necessary to calculate from the time perspective when the front of the rv and the rear of the fv will collide on the longitudinal driving path between the vehicle itself and the rear vehicle. That is, according to Equation (1), it is necessary to calculate TTC to determine the collision risk, and based on this, a basic collision time calculation model is established.

[0064] Among them, the head distances traveled by rv and fv in the vehicle state starting from the determination moment within TTC are shown in Equations (2) and (3), and the equation regarding the distance difference between the distances traveled by rv and fv and the actual head-to-head distance between the two vehicles starting from the determination moment is Equation (4).

[0065]

[0066] In Formulas (2) to (4): are the distances traveled by rv and fv respectively until the two vehicles collide when driving in the current vehicle state, in m; L fv is the actual vehicle length of fv, in m; v rv and v fv are the real-time actual vehicle speeds of rv and fv respectively, in m / s; a in is the real-time intrusion acceleration of rv, in m / s 2 ; here, the TTC solution value takes its minimum positive solution value.

[0067] According to Formula (1), from a spatial perspective, it is necessary to calculate the safety escape margin SEM, so it is necessary to establish a safety escape margin calculation model. The safety escape margin is to consider that both rv and fv will travel a certain distance during the process of fv reacting to the decision-making behavior. In order to prevent the two vehicles from colliding during this process, it is necessary to calculate the initial headway distance that rv and fv can safely travel at the current vehicle driving state within the maximum reaction time of the autonomous vehicle on the premise of meeting the reserved limit safety following distance, so as to establish a safety escape margin calculation model.

[0068] The safety escape margin model should first calculate the distances traveled by rv and fv within the maximum reaction time of the autonomous vehicle, and the distance calculation formulas are Formulas (5) and (6).

[0069]

[0070] In Formulas (5) and (6): is the distance traveled by rv and fv within the maximum reaction time of the autonomous vehicle, in m; t safe is the maximum time required for the AEV to make a system reaction in the normal state, in s.

[0071] Then, according to the physical motion model, establish a safety escape margin calculation formula, and its calculation formula is Formula (7).

[0072]

[0073] In the formula: D safe is the limit safety following distance when the two vehicles are driving, which is the distance between the front of the rear vehicle and the rear of the front vehicle, in m.

[0074] Step S2: Use the collision risk real-time determination model to real-time determine whether there is a collision risk between the current autonomous vehicle AEV and the rear conflict vehicle RCV in the current lane. If so, enter Step S3.

[0075] Specifically, during the normal driving of an autonomous electric vehicle (AEV), if a risky vehicle (RCV) appears behind, the risk determination of the RCV and the AEV needs to be carried out through a real-time collision risk determination model. The collision risk determination scenario between the RCV and the AEV is shown in Figure 4 .

[0076] Calculate the collision time between the RCV and the AEV. The calculation formulas are Formulas (8) to (10).

[0077]

[0078] In the formulas: and are the distances traveled by the RCV and the AEV respectively until the two vehicles collide when driving in the current vehicle state, with the unit of m; L AEV is the actual vehicle length of the AEV, with the unit of m; v RCV and v AEV are the real-time actual vehicle speeds of the RCV and the AEV respectively, with the unit of m / s; a in is the real-time intrusion acceleration of the RCV, with the unit of m / s 2 ; here, the solution value of TTC1 is taken as its minimum positive solution value.

[0079] Calculate the safety escape margin between the RCV and the AEV. The calculation formulas are Formulas (11) to (13).

[0080]

[0081] In the formulas: and are the distances traveled by the RCV and the AEV respectively within the maximum reaction time of the AEV, with the unit of m; T safe is the maximum time required for the AEV to make a system reaction in the normal state, with the unit of s; D safe is the limit safety following distance between the two vehicles, which is the distance between the front of the rear vehicle and the rear of the front vehicle, with the unit of m.

[0082] According to the collision risk determination model, determine the current collision risk between the RCV and the AEV. The determination formula is shown in Formula (14).

[0083]

[0084] In the formula: TTC1 is the collision time between the RCV and the AEV, with the unit of s; T safe is the maximum time required for the AEV to make a system reaction in the normal state, with the unit of s; SEM1 is the safety escape margin between the RCV and the AEV, with the unit of m; D1 is the real-time actual distance between the RCV and the AEV during longitudinal driving, with the unit of m. When it is determined that there is a risk, go to step S3.

[0085] Step S3: Based on the real-time collision risk determination model, calculate the safe intrusion acceleration interval of the AEV relative to the vehicle FV in the current lane and the safe escape acceleration interval of the AEV relative to the RCV. If there is an intersection between the two intervals, generate a linear longitudinal escape acceleration function based on the maximum speed limit and the acceleration interval, and drive the AEV to accelerate and escape according to the linear longitudinal escape acceleration function; otherwise, enter Step S4.

[0086] Specifically, when it is determined that there is a collision risk between the RCV and the AEV, the AEV needs to perform relevant escape behaviors to ensure its own safety at this time. Since the AEV has good flexibility and a relatively fast reaction time during driving and can perform various different escape behaviors, this step will classify and establish the basic escape model of the AEV accordingly. Now, according to the type of escape behavior, this model is divided into a longitudinal acceleration escape model and an emergency lane change escape model.

[0087] This model is divided into two parts: a longitudinal acceleration feasibility determination model and the calculation of the AEV's longitudinal escape acceleration. If the speed of the AEV has reached the maximum speed limit v that can be achieved on the current road max , the AEV directly skips the longitudinal acceleration escape mode and enters the emergency lane change escape mode, that is, skips the longitudinal acceleration escape model and enters the emergency lane change escape model.

[0088] Since the AEV needs to consider the relationship between the FV and the RCV simultaneously during the longitudinal acceleration escape behavior to make the correct escape behavior, it is necessary to discuss the intrusion acceleration interval a in(AEV) calculated for the AEV relative to the FV under the condition of no collision risk and the escape acceleration interval a out(AEV) calculated for the AEV relative to the RCV under the condition of no collision risk.

[0089] That is, if there is an intersection between the acceleration intervals of a in(AEV) and a out(AEV) obtained, it is determined that the AEV can perform longitudinal acceleration escape with this intersection as the acceleration interval; if there is no intersection between the acceleration intervals of a in(AEV) and a out(AEV) obtained, it is determined that the AEV cannot perform longitudinal acceleration escape behavior, and the AEV then enters the emergency lane change escape process. The expression for its longitudinal acceleration feasibility determination is Equation (15).

[0090]

[0091] In the formula: a in(AEV) and a out(AEV) are the intrusion acceleration interval of the AEV relative to the FV and the safe escape acceleration interval of the AEV relative to the RCV, with the unit of m / s2 ; a (AEV)min 、a (AEV)max is a in(AEV) ∩a out(AEV) The minimum AEV longitudinal escape acceleration and the maximum AEV longitudinal escape acceleration included in the acceleration interval after, with the unit of m / s 2 .

[0092] The method for solving the intrusion acceleration interval of the longitudinal relationship between AEV and FV is as follows. In this step, before executing the longitudinal acceleration escape mode, AEV needs to determine whether it can accelerate relative to FV. The scenario is as Figure 5 shown.

[0093] That is, assume that AEV escapes with a specific intrusion acceleration relative to FV. In order to consider the principle that the driving state of other vehicles should not be changed when AEV makes relevant driving behavior changes, during this escape process, AEV and the preceding vehicle FV also need to meet the non-collision risk condition of the collision risk determination model with FV as the target vehicle. The conditional formula is shown in Equation (16).

[0094] TTC2≥T safe or D2≥SEM2 (16)

[0095] In the formula: TTC2 is the vehicle collision time between AEV and FV in this assumed state, with the unit of s; D2 is the real-time actual distance between AEV and FV during longitudinal driving, with the unit of m; SEM2 is the safe escape margin between AEV and FV in this assumed state, with the unit of m.

[0096] To satisfy Equation (16), AEV will calculate its own available acceleration interval on the premise of meeting the above conditions. Calculating the acceleration interval requires starting from the collision time between AEV and FV and the safe escape margin of AEV relative to FV.

[0097] First, start from the collision time between AEV and FV to establish the relationship function between the intrusion acceleration and time of AEV at the collision time. Its initial calculation formula refers to Equations (2) to (4), and the result is Equations (17) to (19).

[0098]

[0099] In the formula: and are the distances traveled by AEV and FV respectively until the two vehicles collide when driving in the current vehicle state, with the unit of m; TTC2 is the vehicle collision time between AEV and FV, with the unit of s; D2 is the real-time actual distance between AEV and FV during longitudinal driving, with the unit of m; L FV are the actual vehicle lengths of FV respectively, with the unit of m; vAEV , v FV are the real-time actual vehicle speeds of AEV and FV respectively, with the unit of m / s; a 1in is the real-time intrusion acceleration of AEV under the collision time condition, with the unit of m / s 2 .

[0100] Since in the above equations (17) and (18), both a 1in and TTC2 are unknowns, an equation is established according to equation (19). Let TTC2 be x and a 1in be y, and the equation is equation (20).

[0101]

[0102] In the formula: D2 - L FV This formula is always greater than 0, with the unit of m; since the magnitude relationship between v FV and v AEV depends on the actual driving scenario, the following is to define different situations to determine the real-time intrusion acceleration of AEV under the vehicle collision time condition. Through derivation and the solution of characteristic points, the real-time intrusion acceleration intervals of AEV in different states are obtained, as shown in equation (21).

[0103]

[0104] Then, starting from the safety escape margin of AEV relative to FV, an inequality discriminant about a 2in is established with D2 ≥ SEM2. The calculation formulas (22) to (24) of SEM2 in the discriminant are established with reference to formulas (5) to (7), and the solution result of the inequality discriminant is formula (25).

[0105]

[0106] In the formula: and are the distances traveled by AEV and FV within the maximum reaction time of AEV, with the unit of m; a 2in is the real-time intrusion acceleration under the condition of AEV safety escape margin, with the unit of m / s 2 .

[0107] When the intrusion acceleration intervals a 1in and a 2in of AEV that meet the collision time condition and the safety escape margin condition are obtained, in order to enable AEV to maintain a completely safe driving state with FV during the acceleration escape process, that is, to meet equation (16), the actual intrusion acceleration value adopted by AEV at this time should simultaneously meet the acceleration intervals under the two conditions, so there are equations (26) and (27).

[0108] ain(AEV) ∈a 1in ∪a 2in (26)

[0109] a in(AEV) ∈[0,a in(AEV)max (27)

[0110] In Equations (26) and (27): a in(AEV) is the AEV intrusion acceleration, and a in(AEV)max is the maximum intrusion acceleration that can be achieved under AEV intrusion conditions, with the unit of m / s 2 .

[0111] At this time, according to Equation (27), it can be known that when considering the relationship between AEV and FV, the maximum longitudinal escape acceleration corresponding to a in(AEV) can be obtained. Next, the relationship between RCV and AEV will be discussed.

[0112] The solution of the escape acceleration interval for the longitudinal relationship between RCV and AEV is as follows. When AEV performs a longitudinal acceleration escape mode, the acceleration adopted must meet the purpose of successful longitudinal acceleration escape before AEV accelerates to the maximum vehicle speed on the current road. The scenario is as Figure 6 shown.

[0113] To achieve this purpose, RCV and AEV should meet the following conditions, and the conditional equation is Equation (28).

[0114] TTC3≥T safe or D1≥SEM3 (28)

[0115] To satisfy Equation (28), it is also necessary to calculate the longitudinal escape acceleration interval of AEV from the perspectives of the collision time condition and the safe escape margin condition between RCV and AEV. First, starting from the collision time of RCV and AEV, TTC3≥T safe is used to establish the relationship function between the escape acceleration and time of AEV at the collision time. The initial calculation formula refers to Equations (2) to (4), and the results are Equations (29) to (31).

[0116]

[0117] In the formula: and are the distances traveled by RCV and AEV respectively until the two vehicles collide in the current vehicle state, with the unit of m; TTC3 is the collision time of RCV and AEV vehicles, with the unit of s; a 1out is the real-time escape acceleration of AEV under the collision time condition, with the unit of m / s 2 .

[0118] Since in Equations (29) and (30), both a 1out and TTC3 are unknowns, an equation is established according to Equation (31). Let TTC3 be x and a 1out be y, and the equation is Equation (32).

[0119]

[0120] In the formula: (D1 - L AEV ) This formula is always greater than 0, with the unit of m. From the former condition, it can be known that at this time, the AEV has passed the collision risk judgment model and already satisfies that when the acceleration of the AEV is 0, TTC3 < T safe , that is, the x corresponding to when y = 0 < T safe . Since the magnitude relationship between v AEV and v RCV depends on the actual driving scenario, the following is to define different situations to determine the real-time escape acceleration of the AEV under the vehicle collision time condition. Through derivation and the solution of characteristic points, the real-time escape acceleration intervals of the AEV in different states are obtained, as shown in Equation (33).

[0121]

[0122] Then, starting from the safety escape margin of the RCV relative to the AEV, an inequality discriminant about a 2out is established with D1 ≥ SEM3. The calculation formulas of SEM2 in the discriminant are Equations (34) to (36), and the reference is established from Equations (5) to (7). The solution result of the inequality discriminant is Equation (37).

[0123]

[0124] In the formula: and are the distances traveled by the RCV and the AEV within the maximum reaction time of the AEV, with the unit of m; a 2out is the real-time escape acceleration under the condition of the AEV safety escape margin, and a max is the maximum acceleration that the AEV can reach when driving on a normal road, with the unit of m / s 2 .

[0125] When the escape acceleration intervals a 1out and a 2out of the AEV that satisfy the collision time condition and the safety escape margin condition are obtained, in order to enable the AEV to successfully achieve the escape purpose during the acceleration escape process and satisfy Equation (28), the actual escape acceleration value of the AEV at this time should simultaneously satisfy the acceleration intervals under the two conditions, so there are Equations (38) and (39).

[0126] a out(AEV)∈a 1out ∪a 2out (38)

[0127] a out(AEV) ∈[a out(AEV)min ,a out(AEV)max (39)

[0128] In Equations (38) and (39): a out(AEV) is the AEV escape acceleration, a out(AEV)min is the minimum escape acceleration required under AEV escape conditions, a out(AEV)max is the maximum escape acceleration achievable under AEV escape conditions, with the unit of m / s 2 ;

[0129] Finally, calculate the longitudinal escape acceleration of the AEV. Since considering that using a single acceleration during the acceleration phase of the vehicle will greatly reduce the driver's comfort, and from Equation (15), it can be seen that when calculating [a (AEV)min ,a (AEV)max , it will be found that the AEV can adopt multiple accelerations for longitudinal acceleration behavior. Also, because when driving on relevant roads, the vehicle has a maximum speed limit v max , so a linear longitudinal escape acceleration a AEV(y) (t) of the AEV can be established based on this. It is a linear function of time that changes with time starting from when the AEV adopts longitudinal acceleration behavior. For this function, set a (AEV)min in Equation (15) as its intercept value and minimum value, and set a (AEV)max in Equation (15) as its function maximum value. Its function expression is Equation (40).

[0130]

[0131] In the formula: a AEV(y) (t) is the linear longitudinal escape acceleration of the AEV that changes with time, with the unit of m / s 2 ; T is the time taken for the AEV to reach the speed v max from the start of adopting longitudinal escape acceleration behavior, with the unit of s; j y is the acceleration change rate adopted when the AEV adopts longitudinal escape acceleration behavior, with the unit of m / s 3 ; t starts from the initial time when the longitudinal escape acceleration behavior starts, with the unit of s; the function graph is as Figure 7 shown.

[0132] Obviously, this function expression of the linear longitudinal escape acceleration a AEV(y) (t) of the AEV that changes with time has two unknowns j yand T. As mentioned in the above expressions, due to the maximum vehicle speed v of the AEV max being limited, and when the AEV accelerates to a AEV(y) is a (AEV)max , at this time, equations can be established to solve for j and T, and the solution formula is formula (41).

[0133]

[0134] In the formula: v max is the maximum limited vehicle speed that the AEV can reach on the current road, with the unit of m / s.

[0135] By solving the above equation (41), j and T are obtained, and their expression is formula (42).

[0136]

[0137] According to the above solution results, the accurate linear longitudinal escape acceleration a AEV(y) (t) function expression of the AEV with respect to time can be obtained. At this time, the AEV can perform longitudinal escape behavior with this acceleration. During the longitudinal driving state where the AEV maintains this acceleration, relevant information about the current lane environment is updated in real time at a certain time interval, and the new linear longitudinal escape acceleration a AEV(y) (t) of the AEV with respect to time is continuously calculated. If the vehicle speed of the AEV reaches the maximum limited vehicle speed v max , the AEV enters the emergency lane-changing escape model.

[0138] Step S4: Based on the collision risk real-time determination model, calculate the collision time with the front and rear vehicles in the target lane for lane change and the safety escape margin to determine the feasibility of lane change. If the safety conditions are met, generate a linear lateral escape acceleration function based on the lane width and the initial lateral speed, and drive the AEV to perform lane change.

[0139] Specifically, when the AEV determines or calculates through the longitudinal acceleration escape model that longitudinal acceleration escape behavior cannot be performed, the AEV enters the emergency lane-changing escape model. In this model, due to the short lane-changing time, the AEV, ALRV, and ALFV are considered to be traveling at a constant speed longitudinally during the process from the start of preparing for lane change to the successful completion of lane change. This model is also divided into an emergency lane-changing escape feasibility determination model and the calculation of the AEV lateral escape acceleration.

[0140] Similarly, when the AEV enters the emergency lane-changing escape model, it is also necessary to determine the feasibility of emergency lane-changing escape. That is, when it is determined that the AEV successfully changes lanes to the adjacent lane in the current state, whether the TTC relationship and SEM relationship between the AEV and the ALRV and ALFV are in a safe state. The determination scenario is as Figure 8as shown

[0141] To ensure safe driving of the AEV from the start of preparing for lane change to successful lane change, the longitudinal spatio-temporal relationship between the AEV and the ALFV and the longitudinal spatio-temporal relationship between the ALRV and the AEV should both meet the safe driving state without collision risk. Therefore, there is the judgment formula (43).

[0142]

[0143] In the formula: TTC4 is the collision time between the AEV and the ALFV vehicle, TTC5 is the collision time between the ALRV and the AEV vehicle, and the unit is s; SEM4 is the safety escape margin between the AEV and the ALFV, SEM5 is the safety escape margin between the ALRV and the AEV, and the unit is m; D3 and D4 are the longitudinal distances between the AEV and the ALFV and the ALRV in the current lane after the AEV successfully changes lanes, and the unit is m.

[0144] Since the AEV, the ALRV, and the ALFV are all regarded as driving at a constant speed longitudinally during the process of the AEV preparing for lane change from the start to finally successfully changing lanes, that is, after the AEV changes to the adjacent lane, the AEV, the ALRV, and the ALFV still drive at the vehicle speeds before the judgment. Therefore, when performing the judgment calculation, the vehicle speeds at the current detection moment can be directly used for calculation. To ensure that the AEV can safely complete the lane change, the time for the AEV to change lanes from the start to the successful lane change should be equal to the collision time between the RCV and the AEV. First, calculate the relative relationship of the longitudinal positions after the AEV successfully changes lanes in the current state, and the calculation results are shown in formulas (44) and (45).

[0145] D3 = D ALFV -(v AEV -v ALFV )TTC1 (44)

[0146] D4 = D ALRV -(v ALRV -v AEV )TTC1 (45)

[0147] In formulas (44) and (45): D3 and D4 are the longitudinal distances between the AEV and the ALFV and the ALRV in the current lane after the AEV successfully changes lanes, and the unit is m; v ALFV and v ALRV are the vehicle speeds of the ALFV and the ALRV in the target lane for the AEV to change lanes, and the unit is m / s.

[0148] For the longitudinal spatio-temporal relationship between the AEV and the ALFV, referring to formulas (2) to (7), there are spatio-temporal relationship formulas (46) to (51).

[0149]

[0150] Where: and are the distances traveled by the AEV and the ALFV respectively when driving in the AEV lane-changing target lane in the current vehicle state until the two vehicles collide, with the unit of m; and are the distances traveled by the AEV and the ALFV within the maximum reaction time of the AEV, with the unit of m; TTC4 is the collision time between the AEV and the ALFV, with the unit of s; L ALFV are the actual vehicle lengths of the ALFV respectively, with the unit of m; SEM4 is the safety escape margin between the AEV and the ALFV, with the unit of m.

[0151] For the longitudinal spatio-temporal relationship between the ALRV and the AEV, referring to Equations (2) to (7), there are spatio-temporal relationship equations (52) to (57).

[0152]

[0153] Where: and are the distances traveled by the ALRV and the AEV respectively when driving in the AEV lane-changing target lane in the current vehicle state until the two vehicles collide, with the unit of m; and are the distances traveled by the ALRV and the AEV within the maximum reaction time of the AEV, with the unit of m; TTC5 is the collision time between the ALRV and the AEV, with the unit of s; SEM5 is the safety escape margin between the ALRV and the AEV, with the unit of m.

[0154] When it is determined by the emergency lane-changing escape feasibility determination model that the AEV can perform the emergency lane-changing escape behavior at this time, in order to consider the driver's comfort during the lane-changing process, it is necessary to calculate the lateral escape acceleration of the AEV. First, determine the lane-changing time of the AEV. When making the determination by the emergency lane-changing escape feasibility determination model, it is mentioned that the AEV should complete the lane change when the collision time between the RCV and the AEV arrives. Then, determine the lane-changing distance of the AEV. Its lane-changing distance should be the current lane width. Obviously, the lateral speed of the AEV should also be zero at the end. Based on this, a linear lateral escape acceleration a AEV(x) (t) of the AEV is established. It is a linear function of time that changes with time starting from the time when the AEV takes the lateral acceleration behavior, and its function expression is Equation (58).

[0155] a AEV(x) (t) = a AEV(x) (0) + j x t, t ∈ [0, TTC1] (58)

[0156] Where: a AEV(x)(t) is the linear lateral escape acceleration of the AEV varying with time, with the unit of m / s 2 ; j x is the acceleration change rate adopted when the AEV takes the lateral escape acceleration behavior, with the unit of m / s 3 ; t takes the start time of the lateral escape acceleration behavior as the initial time, with the unit of s; a AEV(x) (0) is the initial lateral escape acceleration of the AEV, with the unit of m / s 2 ; The function image is as Figure 9 shown, and the positive acceleration direction is the lane-changing direction of the AEV.

[0157] From the above, the lane-changing distance of the AEV should be the current lane width, and when t takes TTC1, the lateral escape acceleration of the AEV is the opposite of the initial lateral acceleration of the AEV. Thus, an equation can be established to solve for j x and a AEV(x) (0), and the solution formula is Formula (59).

[0158]

[0159] In the formula: d is the current lane width, with the unit of m.

[0160] Solving j x and a AEV(x) (0) from the above equation, and their expression is Formula (60).

[0161]

[0162] According to the above solution results, the accurate linear lateral escape acceleration a AEV(x) (t) function expression of the AEV varying with time can be obtained. At this time, the AEV can perform an emergency lane-changing escape behavior with this lateral acceleration until the AEV successfully changes to the target lane. After the lane change is successful, the AEV continues to update the vehicle information of the current lane in real time on the new lane.

[0163] The present invention also provides a real-time escape decision-making system for an autonomous driving vehicle based on active safety, including a collision risk determination module, a longitudinal acceleration escape module, and an emergency lane-changing escape module;

[0164] The collision risk determination module is used to calculate the time to collision TTC and the safety escape margin SEM between the autonomous driving vehicle AEV and the rear conflicting vehicle RCV, and when TTC ≤ the maximum reaction time of the system and the actual headway distance ≤ SEM, output a collision risk trigger signal;

[0165] The longitudinal acceleration escape module is used to calculate the safe intrusion acceleration interval of the AEV relative to the front vehicle FV in the current lane and the safe escape acceleration interval of the AEV relative to the RCV after receiving the collision risk trigger signal. If there is an intersection between the two intervals, a linear longitudinal escape acceleration function based on the maximum speed limit and the acceleration interval is generated, and the AEV is driven to accelerate according to this function. Otherwise, the collision risk trigger signal is transmitted to the emergency lane change escape module;

[0166] The emergency lane change escape module calculates the lane change feasibility based on the TTC and SEM between the AEV and the front and rear vehicles in the target lane. If the safety conditions are met, a linear lateral escape acceleration function based on the lane width and the lateral initial speed is generated, and the AEV is driven to change lanes.

[0167] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. Only the preferred embodiments of the present invention are expressed. The description is relatively specific and detailed, but it should not be construed as a limitation to the scope of the present invention. As long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0168] It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the appended claims.

Claims

1. A real-time escape decision-making method for autonomous driving vehicles based on active safety, characterized in that: It includes the following steps: Step S1: Construct a real-time collision risk determination model based on the collision time between the front and rear vehicles and the safety escape margin. Step S2: Use the real-time collision risk determination model to determine in real time whether there is a collision risk between the current autonomous driving vehicle (AEV) and the rear conflict vehicle (RCV) in the current lane. If there is a collision risk, proceed to Step S3. Step S3: Based on the real-time collision risk determination model, calculate the safe intrusion acceleration interval of the AEV relative to the vehicle in front (FV) in the current lane and the safe escape acceleration interval of the AEV relative to the RCV. If there is an intersection between the two intervals, generate a linear longitudinal escape acceleration function based on the maximum speed limit and the acceleration interval, and drive the AEV to accelerate and escape according to the linear longitudinal escape acceleration function. Otherwise, proceed to Step S4. Step S4: Based on the real-time collision risk determination model, calculate the collision time and safety escape margin with the front and rear vehicles in the target lane for lane change to determine the feasibility of lane change. If the safety conditions are met, generate a linear lateral escape acceleration function based on the lane width and the initial lateral speed, and drive the AEV to perform a lane change.

2. The real-time escape decision-making method for an autonomous driving vehicle based on active safety according to claim 1, wherein: The expression of the real-time collision risk determination model is: In the formula, TTC is the collision time between the rear vehicle and the front vehicle. T safe represents the maximum time required for an autonomous vehicle to make a system response in a normal state; SEM is the safety escape margin between the vehicle behind and the vehicle in front; D rv-fv represents the real-time distance between the vehicle behind and the vehicle in front during longitudinal driving.

3. The real-time escape decision-making method for an autonomous driving vehicle based on active safety according to claim 2, wherein: The calculation expression of the collision time TTC between the rear vehicle and the front vehicle is: Wherein, and are respectively the distances traveled by the rear vehicle and the front vehicle until the two vehicles collide when driving in the current vehicle state; L fv is the actual vehicle length of the front vehicle; v rv and v fv are respectively the real-time vehicle speeds of the rear vehicle and the front vehicle; a in is the real-time intrusion acceleration of the rear vehicle.

4. A real-time escape decision-making method for an autonomous driving vehicle based on active safety according to claim 2, characterized in that: The calculation expression of the safety escape margin between the rear vehicle and the front vehicle is: In the formula, and are the distances traveled by the vehicle behind and the vehicle in front respectively within the maximum response time of the autonomous vehicle; T safe is the maximum time required for the autonomous vehicle to make a system response under normal conditions; D safe is the limit safe following distance when the two vehicles are traveling.

5. The real-time escape decision-making method for an autonomous driving vehicle based on active safety according to claim 1, characterized in that: In Step S3, the linear longitudinal escape acceleration function is a piecewise linear function, and its expression is: where a AEV(y) (t) is the linear longitudinal escape acceleration of the current autonomous emergency vehicle (AEV) varying with time; T is the time taken for the AEV to reach the maximum speed limit v max from the start of the longitudinal escape acceleration behavior; j y is the acceleration change rate adopted when the AEV takes the longitudinal escape acceleration behavior; a (AEV)min represents the minimum longitudinal escape acceleration included in the acceleration interval; a (AEV)max represents the maximum longitudinal escape acceleration included in the acceleration interval; v AEV represents the real-time vehicle speed of the AEV.

6. The real-time escape decision-making method for an autonomous driving vehicle based on active safety according to claim 1, wherein: In Step S4, the expression of the linear lateral escape acceleration function is: a AEV(x) (t) = a AEV(x) (0) + j x t, t ∈ [0, TTC1]; where a AEV(x) (t) is the linear lateral escape acceleration of the current autonomous emergency vehicle (AEV) varying with time; j x is the acceleration change rate when the AEV takes a lateral escape acceleration behavior; a AEV(x) (0) is the initial lateral escape acceleration of the AEV; D is the current lane width; TTC1 represents the time to collision.

7. A real-time escape decision-making system for autonomous driving vehicles based on active safety, characterized in that: It includes a collision risk determination module, a longitudinal acceleration escape module, and an emergency lane change escape module. The collision risk determination module is used to calculate the collision time TTC and the safety escape margin SEM between the autonomous driving vehicle AEV and the rear conflict vehicle RCV, and output a collision risk trigger signal when TTC ≤ the maximum system response time and the actual headway distance ≤ SEM. The longitudinal acceleration escape module is used to calculate the safe intrusion acceleration interval of the AEV relative to the vehicle in front (FV) in the current lane and the safe escape acceleration interval of the AEV relative to the RCV after receiving the collision risk trigger signal. If there is an intersection between the two intervals, generate a linear longitudinal escape acceleration function based on the maximum speed limit and the acceleration interval, and drive the AEV to accelerate according to this function. Otherwise, transmit the collision risk trigger signal to the emergency lane change escape module. The emergency lane change escape module calculates the feasibility of lane change based on the TTC and SEM between the AEV and the front and rear vehicles in the target lane. If the safety conditions are met, generate a linear lateral escape acceleration function based on the lane width and the initial lateral speed, and drive the AEV to perform a lane change.

8. The real-time escape decision-making system for an autonomous driving vehicle based on active safety according to claim 7, characterized in that: In the collision risk determination module, TTC is calculated according to the following formula: Wherein, and are respectively the distances traveled by the vehicle behind and the vehicle in front until the two vehicles collide when driving in the current vehicle state; L fv is the actual vehicle length of the vehicle in front; v rv and v fv are respectively the real-time vehicle speeds of the vehicle behind and the vehicle in front; a in is the real-time intrusion acceleration of the vehicle behind.

9. The real-time escape decision-making system for an autonomous driving vehicle based on active safety according to claim 7, characterized in that: In the collision risk determination module, SEM is calculated according to the following formula: In the formula, and are the distances traveled by the vehicle behind and the vehicle in front within the maximum reaction time of the autonomous vehicle respectively; T safe is the maximum time required for the autonomous vehicle to make a system reaction under normal conditions; D safe is the limit safe following distance between the two vehicles when driving.

10. The real-time escape decision-making system for an autonomous driving vehicle based on active safety according to claim 7, characterized in that: In the longitudinal acceleration escape module, the linear longitudinal escape acceleration function is a piecewise linear function, and its expression is: Where a AEV(y) (t) is the linear longitudinal escape acceleration of the current autonomous emergency vehicle (AEV) changing with time; T is the time taken for the AEV to reach the maximum speed limit v max from the start of the longitudinal escape acceleration behavior; j y is the acceleration change rate adopted when the AEV takes the longitudinal escape acceleration behavior; a (AEV)min represents the minimum longitudinal escape acceleration included in the acceleration interval; a (AEV)max represents the maximum longitudinal escape acceleration included in the acceleration interval; v AEV represents the real-time vehicle speed of the AEV.

Images