Method for determining a speed profile of a motor vehicle with non-predefined accelerations

Abstract

The invention relates to a method for determining a speed profile to be followed by a vehicle, comprising the steps of: - acquiring event data comprising a distance to an event and a target speed (V3) of said vehicle in this event, - determining a speed profile to be followed as a function of time between an initial speed (V0) and this target speed in three successive different phases, respectively a first phase (Phase_1) in which jerk is set constant at a predetermined maximum jerk value so as to reach an optimal target acceleration value, a second phase (Phase_2) in which this optimal target acceleration value remains constant, and a third phase (Phase_3) in which the jerk is again set constant so as to reach a zero acceleration value at the end of this third phase, said optimal target acceleration value being such that the distance required to perform the three phases of said profile is equal to said distance to the event.

Description

[0001] This invention relates to the field of motor vehicles, and more particularly to the field of systems and devices for assisting in driving such vehicles. More specifically, this invention relates to a method for determining the speed curve to be followed by a motor vehicle.

[0002] Adaptive speed control systems for motor vehicles (known as the acronym ACC for "Adaptive Cruise Control") are known to be designed to continuously control the speed of a "master" vehicle based on a longitudinal control law. This law adjusts the speed of the "master" vehicle according to the driver's requested speed command and the presence of a vehicle ahead in the same lane (referred to as the "target vehicle"). This longitudinal control law can be used in various scenarios. Thus, the equipped vehicle can automatically adjust its speed to maintain a predetermined safe distance relative to the target vehicle. Once the lane ahead of the equipped vehicle is clear, its acceleration automatically increases to reach the speed commanded by the driver, thereby assisting the driver in completing the driving task. If a slower vehicle is detected in the lane ahead of the equipped vehicle, the speed is automatically adjusted to maintain a safe distance. The control law can also simultaneously check comfort criteria (boundary acceleration, boundary deceleration, and boundary jerk (i.e., the derivative of acceleration)).

[0003] Therefore, this control law allows the vehicle's longitudinal speed to be controlled based on the driver's defined speed command, and, where applicable, to decrease that speed if a target vehicle is detected ahead of the main vehicle. However, to predict deceleration and acceleration, this control does not consider background and semantic information that may be extracted from the road scene observed by the multi-sensor detection system mounted on the main vehicle, such as approaching a stop sign, approaching a roundabout, a curve, traffic congestion, or changes in speed limits. Therefore, the vehicle's behavior is not always adapted to its environment. However, to achieve a more robust control system (crucial for autonomous driving), background information needs to be considered.

[0004] Patent document US 2019106108 describes a control law that determines a speed profile of a motor vehicle along a planned trajectory in at least a partial response to: (i) an object is detected along the planned trajectory, (ii) a change in speed limits along the planned trajectory is determined, and (iii) a speed is selected by the user. The control law then enables the vehicle's speed along the planned trajectory to be managed based on the determined speed profile.

[0005] However, the published document remains very vague in its implementation. In particular, it uses a velocity curve based on acceleration curves calculated differently under certain conditions, but fails to explain how this curve is calculated to ensure driver comfort. Another issue raised is the lack of sufficient data to allow for system calibration.

[0006] More generally, there is another problem in the prior art, namely generating a speed curve that the speed controller should follow, which would allow for energy savings throughout the journey.

[0007] To overcome the aforementioned problems, one aspect of the present invention is a method for determining a speed curve to be followed by a motor vehicle, the method comprising the steps of: acquiring background information about the road environment of the vehicle via a multi-sensor system of the vehicle; extracting event data from the acquired background information, the event data including at least one distance from an event associated with the vehicle and a target speed of the vehicle in the event; providing a measured initial speed of the vehicle; determining a speed curve to be followed as a function of time in three consecutive distinct phases between the measured initial speed and the target speed, the phases being a first phase in which jerk is set constant at a predetermined maximum jerk value so as to reach an optimal target acceleration value at the end of the first phase, a second phase in which the optimal target acceleration value remains constant throughout the duration of the second phase, and a third phase in which jerk is again set constant so as to reach a zero acceleration value at the end of the third phase; the method comprising the step of: determining the optimal target acceleration value during the second phase such that the distance required to execute the three phases of the curve by applying the determined optimal target acceleration value is equal to the distance from the event.

[0008] Advantageously, the distance required to execute the three stages of the curve is calculated based on a set of equations implemented for calculating the velocity curve, the calculation steps of which include, for a set of fixed parameters including the initial velocity and initial acceleration of the vehicle at the start of the velocity curve, the target velocity in the event, and a predetermined maximum jerk value, and for non-fixed parameters including the optimal target acceleration to be achieved in the second stage of the curve:

[0009] - Calculate the duration of these phases and define the start and end times of these phases.

[0010] - Calculate the passing speed at the start and end times that define this second phase.

[0011] - Calculate the change in velocity over time for each of these stages, and

[0012] - Calculate the distance traveled at the start and end times of these defined phases.

[0013] Advantageously, the determination of the optimal target acceleration value is performed through iteration and a bisection method from a predetermined range bounded by the minimum and maximum acceleration values.

[0014] Advantageously, in each iteration, the distance required to generate the curve is calculated using an intermediate acceleration value, which is the centroid of the minimum and maximum acceleration values.

[0015] Advantageously, in the third stage, the jerk is set to be constant at the predetermined maximum jerk value in the first stage.

[0016] As a variant, in the third stage, the jerk is set to a predetermined maximum jerk value that is constant and different from the predetermined maximum jerk value in the first stage.

[0017] Advantageously, the predetermined maximum acceleration value depends on whether the speed curve is related to positive or negative acceleration of the vehicle.

[0018] Advantageously, the method includes the step of transmitting the speed curve as an instruction to an adaptive speed control system installed on the vehicle.

[0019] The present invention also relates to an apparatus intended for installation on a vehicle for implementing the above-described method, characterized in that the apparatus includes a multi-sensor system capable of acquiring background information about the road environment of the motor vehicle, a vehicle speed sensor, means for calculating the speed curve based on event data extracted from the acquired background information, and a control means for applying the calculated speed curve to a vehicle adaptive speed control system.

[0020] The present invention also relates to a motor vehicle including the equipment described above.

[0021] Other features and advantages of the invention will become apparent from the following description, given in a completely non-limiting manner with reference to the accompanying drawings, in which:

[0022] [ Figure 1 [Illustration] is a graph illustrating an example of a velocity curve as a function of time according to the invention, wherein acceleration depends on the distance from the event, and the velocity curve allows a vehicle to transition from an initial speed to the target speed to be reached in the event when the target speed is lower than its initial speed;

[0023] [ Figure 2 [This is a flowchart illustrating the bisection algorithm implemented to determine the acceleration required in the second phase of the velocity profile;]

[0024] [ Figure 3 The graph shows the relationship between the various distance values ​​required to generate the velocity curve and the jerk and acceleration values ​​selected for the initial and target velocity values.

[0025] [ Figure 4is a set of graphs showing the variation of the vehicle's speed with the speed curve, distance from the event, and vehicle acceleration.

[0026] The present invention is applicable to motor vehicles equipped with an adaptive speed control system and a multi-sensor perception system, which are capable of transmitting background information related to events in the road scene in front of the vehicle, such as approaching a roundabout, a bend, a traffic jam, a speed limit change, etc. The data collected by the on-vehicle sensors is sent to an electronic computer, which constructs a description of the environment near the vehicle and the configuration of the road scene through an environmental perception algorithm that analyzes these data. From this environment, the system is able to provide (distance, speed) pairs related to the detected event, including the distance D 事件 from the event and the speed V3 to be achieved in the event (referred to as the target speed). For example, the detected event can be a speed limit sign.

[0027] The vehicle also includes an on-vehicle speed sensor, which transmits information about its speed and its acceleration by processing its speed.

[0028] The speed curve described below is a speed curve determined based on the measured vehicle speed and is intended to be transmitted as an instruction to be followed to the vehicle's adaptive speed control system in order to predict the deceleration and acceleration when approaching an event. More specifically, the speed curve must allow the vehicle to transition from its initial speed to the target speed while complying with the vehicle dynamics constraints in terms of acceleration, which can be positive or negative (deceleration), depending on the type of the curve and the jerk (i.e., the derivative of the acceleration). These last constraints will enable the vehicle's adaptive speed control system to optimize its performance based on this curve.

[0029] Now, the principle of the speed curve of the present invention will be demonstrated starting from an example of a decreasing curve as Figure 1 shown, which assumes zero initial acceleration and a constant initial vehicle speed (denoted as V0) at the start and zero acceleration and a constant target vehicle speed V3 at the end, where V3 < V0 because the speed curve here is decreasing.

[0030] Figure 1 The presented speed curve is defined according to the present invention as three consecutive phases:

[0031] - The first phase, denoted as Phase_1, ranges between the start time t0 and the end time t1 that define this phase, where the jerk value is set to a predetermined maximum jerk value, whose absolute value is denoted as J 最大值 so as to reach an optimal target acceleration value at time t1, whose absolute value is denoted as A 目标 ,,

[0032] - The second phase, denoted as Phase_2, is defined by the time between the start time t1 and the end time t2 of this phase. The target acceleration value reached at t1 remains constant throughout the entire duration of the second phase, after which the acceleration value becomes zero.

[0033] - The third stage, denoted as Phase_3, is defined within the range between the start time t2 and the end time t3 of this stage, wherein, preferably, the jerk value is again set to the predetermined maximum jerk value J. 最大值 So that the zero acceleration value is returned at time t3.

[0034] It is necessary to divide the speed curve according to the invention into three distinct consecutive phases in order to observe the constraints of vehicle dynamics on acceleration values ​​in phases Phase_1 and Phase_3.

[0035] Therefore, it has the following set of fixed parameters:

[0036] V0: The initial velocity when the velocity curve begins;

[0037] V3: The target speed to be achieved in the event;

[0038] A 初始 The acceleration of the vehicle at the start of this speed curve may not be zero.

[0039] J 最大值 : The maximum jerk value defined for this curve.

[0040] The speed curve is represented by two calculation functions actually used by the vehicle speed control system:

[0041] The command speed V to be followed varies with time (t). 曲线 :

[0042] V 曲线 (t)=f(t,V0V3,A 初始 J 最大值 )

[0043] The distance required to achieve the velocity curve is denoted as D. 曲线 .

[0044] For a given set of parameters, once the distance D from the target event sought to be predicted is derived from the vehicle's multi-sensor perception system... 事件 Become equal to D 曲线 This will trigger the following of the curve.

[0045] When the distance D from the event 事件Once the target velocity V3 for the event is known, the curve needs to be started, and its adherence must ensure that the behavior adapts to the target velocity required for the event. For this type of curve, the acceleration to be used is not a constraint. This involves generating a velocity curve with a non-predetermined acceleration. Therefore, it is necessary to determine the acceleration before calculating the velocity curve to be followed. The acceleration is determined using a bisection method.

[0046] The binary search method allows us to find the optimal target acceleration value A. 目标 , so that:

[0047]

[0048] The following section will detail the distance D required to calculate the velocity curve. 曲线 The calculation steps.

[0049] calculate Figure 1 The duration of Phase_1, Phase_2, and Phase_3 as defined in the text, as well as the start and end times t0, t1, t2, and t3 that define these phases.

[0050] Calculate the passing speeds V1 and V2 corresponding to the passing speeds at the start time t1 and end time t2 of the second phase Phase_2.

[0051] Calculate the behavior of velocity changes over time for each phase, Phase_1, Phase_2, and Phase_3.

[0052] Calculate the distance traveled as a function of time at each of the times t0, t1, t2, and t3 that define the different phases Phase_1, Phase_2, and Phase_3, denoted as X0, X1, X2, and X3, respectively. Therefore, the distance required to achieve this curve is D. 曲线 =X3.

[0053] The results of these velocity curve calculation steps will be described in detail below and further demonstrated.

[0054] To calculate the start and end times t0, t1, t2, and t3 of the defining phases Phase_1, Phase_2, and Phase_3, the following notation is used:

[0055] t0 = 0 (assumption)

[0056] t1 = T 01

[0057] t2=T 01 +T 12

[0058] t3 = T01 +T 12 +T 23

[0059] For T 01 T 12 and T 23 By integrating the velocity across all points, the corresponding durations of Phase_1, Phase_2, and Phase_3 are obtained:

[0060]

[0061]

[0062]

[0063] The speeds at each stage, i.e., the speeds at times t0, t1, t2, and t3, are denoted as V0, V1, V2, and V3, respectively, as follows:

[0064] V0, the initial velocity of the curve, is an imposed value that corresponds to the vehicle's current measured speed when the curve is calculated.

[0065]

[0066]

[0067] V3 is the target speed that is expected to be achieved by following the speed curve, and it is imposed just like the initial speed.

[0068] The equation for the velocity curve v(t) at each stage is determined as a function of the elapsed time, as shown below:

[0069] Phase_1, 0≤t≤t1:

[0070] Phase_2, t1≤t≤t2: V(t)=V1-A 目标 ×(t-t1)

[0071] Phase_3, t2<=t<=t3:

[0072] The distance traveled as a function of time is calculated as follows:

[0073] X0 = 0 (assumption)

[0074]

[0075]

[0076]

[0077] The distance required to achieve this curve is D. 曲线 =X3. Therefore, when the distance to the event is equal to X3, following the velocity curve will be triggered by sending the velocity command V(t) as defined above.

[0078] These results will now be demonstrated.

[0079] For phase Phase_1, for t such that t0≤t≤t1:

[0080] The acceleration J of the curve in this first stage 01 (t) is set to constant, and its value is: J 01 (t)=-J 最大值

[0081] Due to the initial acceleration A 初始 The acceleration as a function of time in this first stage is zero (denoted as A). 01 The value of (t) is:

[0082] A 01 (t)=A 初始 +∫0 t J(t)dt

[0083] A 01 (t)=-J 最大值 ×t

[0084] Then, the change in velocity over time in this first stage can be derived, denoted as V. 01 (t):

[0085] V 01 (t)=V0+∫0 t A 01 (t)dt

[0086]

[0087] Therefore, the travel distance X, as a function of time, in the first phase (Phase 1) 01 (t) is:

[0088] X 01 (t)=X0+∫0 t V 01 (t)dt

[0089]

[0090] For the second phase, Phase_2, for t such that t1≤t≤t2:

[0091] The acceleration as a function of time in this second stage (denoted as A) 12 (t) is constant during this stage, that is:

[0092] A 12 (t)=-A 目标

[0093] Therefore, the change in velocity over time in this second stage can be derived, expressed as V. 12 (t):

[0094]

[0095] V 12 (t)=V1-A 目标 ×(t-t1)

[0096] Therefore, the travel distance X in this second stage is a function of time. 12 (t) is:

[0097]

[0098]

[0099] Finally, for the third phase, Phase_3, for t such that t2≤t≤t3:

[0100] The acceleration J of the curve in this first stage 23 (t) is set to be constant, and its value is:

[0101] J 23 (t)=J 最大值

[0102] Due to the final acceleration A in the third stage 最终 The value is zero, therefore the acceleration as a function of time during this stage (denoted as A) is zero. 23 The value of (t) is:

[0103]

[0104] A 23 (t)=(t-t3)×J 最大值

[0105] Then, the change in velocity over time in this third stage can be derived, denoted as V. 23 (t):

[0106]

[0107]

[0108] Therefore, the distance traveled X as a function of time during this stage 23 (t) is:

[0109]

[0110]

[0111] For the corresponding stage duration T 01 T 12 and T 23 And the expressions that define the start and end times of these phases, recalling the use of the following notation:

[0112] t0 = 0 (assumption)

[0113] t1 = T 01

[0114] t2=T 01 +T 12

[0115] t3 = T 01 +T 12 +T 23

[0116] The duration T of Phase 1 01 The value is:

[0117] T 01 =t1-t0=t1

[0118] However, the acceleration is continuous between Phase 1 and Phase 2, that is:

[0119] A 01 (t1)=A 12 (t1)

[0120] It is equivalent to:

[0121] -J 最大值 ×t1=-A 目标

[0122] therefore:

[0123] Therefore, the value of velocity V1 reached at time t1 (corresponding to the end of the first phase and the beginning of the second phase) is:

[0124] V1 = V 01 (t1)=V 01 (T 01 )

[0125] Right now,

[0126] Now let's consider the duration T of Phase 3. 23 Initially, its value is:

[0127] T 23 =t3-t2

[0128] However, just as between the first and second phases, the acceleration is continuous between Phase 2 and Phase 3, that is:

[0129] A 23 (t2)=A 12 (t2)

[0130] It is equivalent to:

[0131] (t2-t3)×J 最大值 =-T 23 ×J 最大值 =-A 目标

[0132] therefore:

[0133] Therefore, the value of velocity V2 reached at time t2 (corresponding to the end of the second stage and the beginning of the third stage) is:

[0134] V2 = V 23 (t2)=V 23 (T 23 )

[0135] therefore:

[0136] Now let's consider the duration T of Phase 2 in the second stage. 12 Initially, its value is:

[0137] T 12 =t2-t1

[0138] However, the velocity at time t2 at the end of the second stage is denoted as V2 and can be written as:

[0139] V 12 (t2)=V2

[0140] It is equivalent to:

[0141] V1-A 目标 ×(t2-t1)=V1-A 目标 ×T 12 =V2

[0142] therefore:

[0143] By replacing the previously developed expressions V1, V2, T 01 and T 23 , the following equation is derived therefrom:

[0144]

[0145] From the demonstration of the calculation results of the speed curve, it can be clearly seen that although the speed curve is defined as three stages as a function of time, its implementation is still simple because it only requires simple mathematical operations (addition, multiplication, division) and simple logical checks. Therefore, the required computing power is limited.

[0146] Now referring to a more general speed curve, that is, a speed curve that can increase or decrease, and its initial acceleration value may not be zero.

[0147] Therefore, the starting assumption used here is an initial acceleration A that may not be zero 初始 and a constant initial vehicle speed V0, and a zero acceleration and a constant target vehicle speed V3 are used as the ending assumption, where V3 < V0 or V3 > V0, depending on whether the speed curve is decreasing or increasing.

[0148] [[ID=2)) As in the previous example, according to the present invention, the speed curve is always defined by the following three consecutive stages:

[0149] - The first stage Phase_1, where the jerk value is set to the maximum jerk value J 最大值 , in order to reach the optimal target acceleration value (the absolute value of which is denoted as A 目标 ), and this optimal target acceleration value depends on the distance from the detected event,

[0150] - The second stage Phase_2, where the acceleration is maintained at the optimal target acceleration value,

[0151] - The third stage, at the maximum jerk value, to return to the zero acceleration value.

[0152] In the same way as above, the distance required to achieve the speed curve is denoted as D 曲线 . Therefore, when the distance D from the event is reached 曲线 , the curve will be triggered.

[0153] Since the parameters A 初始 and J 最大值 are absolute values, the following variables s and s1 are introduced to reflect the relative acceleration value and the jerk value.

[0154] The variable s is defined as follows:

[0155]

[0156] Therefore, if the curve is increasing, i.e., V3 > V0, then the value of s will be 1, while if the curve is decreasing, then the value of s will be -1. Furthermore, in the first phase (Phase 1) of the velocity curve, the acceleration will change from A... 初始 Transform into s×A 目标 In the second phase, Phase 2, the acceleration will remain at s×A 目标 Furthermore, in the third phase, Phase 3, the acceleration will increase from s×A 目标 It becomes 0 m / s².

[0157] In addition, variable s1 is defined as follows:

[0158]

[0159] Therefore, this variable represents the direction of acceleration change in the first phase, Phase_1, which starts from A. 初始 Transform into s×A 目标 Therefore, the value of the jerk in this stage is s1×J. 最大值 .

[0160] As explained in the previous example, the steps to calculate the velocity curve are as follows:

[0161] The calculations define the duration of Phases 1, 2, and 3, as well as the start and end times t0, t1, t2, and t3 for these phases.

[0162] Calculate the passing speeds V1 and V2 corresponding to the passing speeds at the start time t1 and end time t2 of the second phase Phase_2.

[0163] Calculate the behavior of velocity changes over time for each phase, Phase_1, Phase_2, and Phase_3.

[0164] Calculate the distance traveled as a function of time for each of the times t0, t1, t2, and t3 that define the different phases Phase_1, Phase_2, and Phase_3, denoted as X0, X1, X2, and X3, respectively.

[0165] The table below summarizes the changes in jerk, acceleration, velocity, and distance traveled at each stage of a typical velocity curve:

[0166] [Table 1]

[0167]

[0168] By following the same procedure as in the previous examples regarding the decreasing velocity curve and the case of zero initial acceleration, the parameters of the curve and the velocity are defined as follows:

[0169] The start and end times of Phase 1 are represented as follows:

[0170] t0 = 0 (assumption)

[0171] t1 = T 01

[0172] The corresponding durations of Phase 1 and Phase 3 are represented as follows:

[0173]

[0174]

[0175] The throughput speeds for these stages are expressed as follows:

[0176] V0 is the forced initial velocity of the curve.

[0177]

[0178]

[0179] V3 imposes a target speed.

[0180] The duration T of Phase 2 in the second stage 12 The definition is as follows:

[0181]

[0182] The times t2 and t3 for Phase 3 are defined as follows:

[0183] t2=T 01 +T 12

[0184] t3 = T 01 +T 12 +T 23

[0185] The equation for the velocity curve v(t) at each stage is determined as a function of the elapsed time, as shown below:

[0186] Phase_1, 0≤t≤t1:

[0187] Phase_2, t1≤t≤t2: V(t)=V1+A 目标 ×s×(t-t1)

[0188] Phase_3, t2<=t<=t3:

[0189] The distance traveled as a function of time is calculated as follows:

[0190] X0 = 0 (assumption)

[0191]

[0192]

[0193]

[0194] The distance required to achieve this curve is D. 曲线 =X3.

[0195] As a variant, the jerk values ​​set in the first and third stages of the velocity curve may differ.

[0196] As a further variation, it can be specified that different acceleration values ​​depend on whether the curve is related to acceleration or deceleration.

[0197] The acceleration A required for the second phase of the curve, Phase 2 目标 By according to Figure 2 The presented binary search algorithm determines the distance D required to reach the curve. 曲线 As a rule, this distance is calculated using the equation above and should be equal to the distance to the event.

[0198] Advantageously, the speed curve is therefore based on the following set of fixed parameters (including the vehicle's initial speed V0 at the start of the speed curve, the speed V3 to be reached during the event, and the vehicle's initial acceleration A at the start of the speed curve). 初始 (Predetermined maximum jerk value) and unpredetermined parameters (in this case, the target acceleration A to be achieved in the second stage of the curve). 目标 Generated by [the method]. According to a specific feature of the invention, the final parameter adopts the optimal value determined when the bisection search is completed, the optimal acceleration value being selected within a bounded range, for example, the distance required to achieve the curve for this set of fixed parameters, and the optimal acceleration value being equal to the distance D from the event. 事件 .

[0199] Now refer to Figure 2 A more detailed description of the binary search algorithm is provided. The goal is to determine the minimum acceleration value 'a'. 最小 With the maximum acceleration value a 最大 The optimal acceleration value a is selected within the predetermined range. 最佳This makes the distance required to realize the curve equal to the distance D from the event. 事件 It should be noted that the distance required to achieve the curve decreases as acceleration increases.

[0200] Therefore, in the first initialization step E0, these minimum and maximum acceleration values ​​within the defined range are set, namely:

[0201] a 最小 =a min_init

[0202] a 最大 =a max_init

[0203] In step E1, the distance Da is first calculated using the above equation. 最小 This corresponds to the acceleration value required in the second stage of the curve as the minimum acceleration value a. 最小 The distance required to achieve the velocity curve under certain conditions; secondly, calculate the distance Da. 最大 This corresponds to the acceleration value required in the second stage of the curve as the maximum acceleration value a. 最大 The distance required to achieve the velocity curve under certain conditions.

[0204] Then test step E2 is performed, the purpose of which is to check whether the optimal acceleration value can be determined. In other words, the distance D from the location of the event is checked. 事件 Is it at the maximum acceleration value a respectively? 最大 and minimum acceleration value a 最小 The calculated distance between the curves, i.e.:

[0205]

[0206] If the test fails, the algorithm terminates at step E20, which means that it is impossible to find an acceleration value that makes the distance required to realize the curve equal to the distance from the event.

[0207] Conversely, if the test passes, the algorithm continues to execute step E3, where the index NB is iterated in a loop. 迭代 It was initialized to 0.

[0208] Next, in each iteration, in step E4, the intermediate value 'a' within the range bounded by the predefined minimum and maximum acceleration values ​​is first determined. 重心 The intermediate value is two values ​​a. 最小 and a 最大 When assigning the centroids with the corresponding coefficients δ and (1-δ):

[0209] a 重心 =δ×a 最小 +(1-δ)×a最大

[0210] Next, in step E5, still based on the above equation, the value a is... 重心 Given the acceleration value required for the second stage of the curve, calculate the distance D corresponding to the distance required to achieve the velocity curve. 重心 .

[0211] Then in step E6, the distance value D is... 重心 With D 事件 Compare. If the distance D from the event... 事件 Greater than or equal to use a 重心 If the distance required for the calculated curve is determined, then in step E7, the maximum acceleration value 'a' required for the curve is set. 最大 and distance Da 最大 , using a 最大 Perform the following calculations:

[0212] a 最大 =a 重心

[0213] Da 最大 =D 重心

[0214] Otherwise, in step E8, set the minimum acceleration value a required for the curve. 最小 and distance Da 最小 , using a 最小 Perform the following calculations:

[0215] a 最小 =a 重心

[0216] Da 最小 =D 重心

[0217] Then, in step E9, it is checked whether the maximum number of iterations for the loop has been reached. If not, the loop iteration index NB is incremented in step E30. 迭代 ,Right now:

[0218] NB 迭代 =NB 迭代 +1

[0219] And then the loop returns to step E4.

[0220] If the maximum number of iterations is reached in step E9, the algorithm provides the optimal acceleration value 'a' in step E10. 最佳 As a result, the following is achieved:

[0221] a 最佳 =a 最小

[0222] Therefore, in the range [a min_init ,a max_init Determining the optimal acceleration value within the curve is a binary search process. In each iteration, the distance required to achieve the curve with an intermediate acceleration value is calculated, which is one of two values, 'a'. 最小 and a 最大 The centroid is assigned when the corresponding coefficients δ and (1-δ) are applied.

[0223] Because the jerk is set to the value J 最大值 Therefore, this method can obtain the optimal acceleration A. 目标 =a 最佳 ∈[a min_init ,a max_init ], so that:

[0224] D 曲线 (a 最佳 ) = D 事件

[0225] In other words, the optimal acceleration value a 最佳 This makes the distance required to realize the curve, calculated using the value determined by the bisection method, equal to the distance to the event.

[0226] This method for determining the optimal acceleration value in the second stage of the curve is particularly advantageous. First, the required computational load is predictable. Specifically, convergence is ensured within a fixed number of iterations. Second, the method ensures good convergence. Therefore, for a fixed number of iterations N and coefficient δ = 0.5, the result a... 最佳 The accuracy is Magnitude.

[0227] Specific exemplary applications of the velocity curves calculated according to the present invention are used in the following situations:

[0228] J 最大值 =0.6m / s 3

[0229] A 初始 =0m / s 2

[0230] V0 = 22.2 m / s = 80 km / h

[0231] V3 = 10m / s = 36km / h

[0232] a min_init =0.5ms -2 ,a max_init =3.4ms -2

[0233] For D 事件=180m, the following results were obtained:

[0234] a 最佳 =2.1ms -2

[0235] Figure 3 Show D 曲线 The various values ​​used are related to the jerk and acceleration values ​​selected for the settings of V0 and V3. Here, since the jerk value is fixed, the scan is performed by accelerating along arrow F.

[0236] Therefore, in this case, by using the value J 最大值 Fixed at 0.6 m / s 3 To obtain the predetermined maximum jerk value, a scanning technique can be used to predetermine it in a different way. Specifically, the binary search determination step will then not use methods such as... Figure 3 Instead of the acceleration scan along a straight line with a constant y, a scan along an axis such as x = y is used, and then the acceleration value J is applied. 最大值 The optimal acceleration A will be predetermined based on its relationship with acceleration. 目标 Value and jerk J 最大值 The value will be obtained in the bisection step in such a way that the distance D_profil required to achieve the three stages of the curve by applying the determined optimal target acceleration value and the acceleration value is equal to the distance D from the event. 事件 .

[0237] Based on the above principles, the velocity curve P related to these constraints is calculated as follows: Figure 4 The first graph is shown. Figure 4 The second graph shows the distance D from the event. 事件 The curve shows how the event changes over time. As can be seen in the first graph, when event D is reached... 事件 When V = 0, the speed V measured by the vehicle, generated by the control law of longitudinal vehicle control following the curve, does indeed reach the target speed required in the event, namely V3.

[0238] exist Figure 4 The third graph shows the acceleration A measured for the vehicle. 测量 The curve shows how the vehicle's acceleration changes over time. It can be seen that the vehicle's acceleration is limited to 1.6 m / s². 2 It is far below the limit defined in the computational constraints (a max_init =3.4ms -2 ).

[0239] Therefore, the speed curve calculated according to the invention and transmitted as an instruction to the vehicle's adaptive speed control system appears to allow the vehicle to automatically reduce its speed behavior to gradually reach the target speed (36 km / h in the example). This allows the vehicle's speed to adapt to background elements detected by the vehicle's sensors. In particular, this allows the desired speed to be reached at the desired distance, i.e., only at the arrival event.

[0240] As seen above, the implementation of the velocity curve requires only limited computing power due to the need for simple mathematical operations. Furthermore, it does not require preliminary measurements of the predetermined path. In other words, it can be generalized to receive events ({distance, velocity}) anytime and in any way.

Claims

1. A method for determining a speed profile to be followed by a motor vehicle, the method comprising the steps of: acquiring background information about the road environment of the vehicle via a multi-sensor system of the vehicle; extracting event data from the acquired contextual information, the event data comprising at least one distance (D 事件 ) from an event related to the vehicle and a target speed (V3) of the vehicle in the event; providing a measured initial speed (V0) of the vehicle; determining a speed profile (P) to be followed as a function of time between the measured initial speed and the target speed in three successive different phases, respectively a first phase (Phase_1) in which jerk is set constant at a predetermined maximum jerk value so as to reach a best target acceleration value (A_target) at the end of the first phase, a second phase (Phase_2) in which the best target acceleration value remains constant throughout the duration of the second phase, and a third phase (Phase_3) in which the jerk is again set constant so as to reach a zero acceleration value at the end of the third phase; the method comprises the steps of: determining said optimal target acceleration value (A 目标 ) during the second phase such that the distance (D 曲线 ) required to perform the three phases of the curve by applying said determined optimal target acceleration value is equal to said distance (D 事件 ) from the event, the determination of said optimal target acceleration value during the second phase being performed by iteration and dichotomy from a predetermined range bounded by a minimum acceleration value, calculated from an initial acceleration value, and a maximum acceleration value; and transmitting the speed profile as an instruction to an adaptive speed control system installed on the vehicle, the adaptive speed control system controlling the vehicle in accordance with the speed profile.

2. The method of claim 1, wherein, The distance required to perform the three phases of the curve is calculated according to a set of equations implemented for calculating this speed curve, whose calculation steps include, for a set of fixed parameters including the initial speed (V0) and initial acceleration (A 初始 ) of the vehicle at the start of the speed curve, the target speed (V3) in the event and a predetermined maximum jerk value, and for non-fixed parameters including the optimal target acceleration (A 目标 ) to be reached at the second phase of the curve: - calculating the duration (T 01 ,T 12 ,T 23 ) of these phases and delimiting the start and end times of these phases, - calculating a passing speed (V1, V2) delimiting a start time and an end time of the second phase, - calculating a change in speed over time for each of the phases, and - calculating a distance of travel delimiting a start time and an end time of the phases.

3. The method of claim 1 or 2, wherein, In each iteration, the distance required to produce the profile is calculated using an intermediate acceleration value which is the barycenter of the two minimum and maximum acceleration values.

4. The method of claim 1 or 2, wherein, In the third phase, the jerk is set constant at the predetermined maximum jerk value of the first phase.

5. The method of claim 1 or 2, wherein, In the third phase, the jerk is set constant at a predetermined maximum jerk value different from the predetermined maximum jerk value of the first phase.

6. The method of claim 1 or 2, wherein, The predetermined maximum jerk value is different depending on whether the speed profile relates to positive or negative acceleration of the vehicle.

7. A device intended to be installed on a motor vehicle for implementing the method according to any one of the preceding claims, characterized in that, The device comprises a multi-sensor system capable of acquiring background information about the road environment of a motor vehicle, a vehicle speed sensor, means for calculating the speed profile from event data extracted from the acquired background information, and control means for applying the calculated speed profile to a vehicle adaptive speed control system.

8. Motor vehicle, characterized in that The motor vehicle comprises a device as claimed in claim 7.

Images