• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The method comprises at least the following phases: a first phase in which a trajectory (33) to be generated is generated, the trajectory being cut into a series of sections whose starting point forms an intermediate position; a second phase in which, at the current intermediate position of said vehicle: a non-zero curvature is defined for each of the next n horizon prediction sections, said curvature gradually varying from one section to the next; ? it is predicted, before said vehicle engages a movement, whether said trajectory can be tracked on said prediction horizon, as a function of imposed constraints and estimated lateral and / or longitudinal slips; a third phase in which, if said trajectory can be followed, the steering angle of said front wheels (31) and the linear traction speed of said vehicle are controlled according to the state of said vehicle and the lateral sliding and / or or longitudinal to join the center of the axle of said rear wheels (32) on the path; if the trajectory can not be followed, a new alignment of said vehicle to said target position is performed and a new reference trajectory is generated according to the first phase.
    公开号:FR3072069A1
    申请号:FR1759456
    申请日:2017-10-10
    公开日:2019-04-12
    发明作者:Eric LUCET；Alain Micaelli；Francois-Xavier Russotto
    申请人:Commissariat a lEnergie Atomique CEA；Commissariat a lEnergie Atomique et aux Energies Alternatives CEA；
    IPC主号:
    专利说明:

    METHOD FOR AUTOMATICALLY DRIVING A VEHICLE, ESPECIALLY FROM A BUS IN A STORAGE CENTER, AND DEVICE IMPLEMENTING SUCH A METHOD
    The present invention relates to a method for automatically driving a vehicle under stress. It also relates to a device implementing such a method. It is particularly applicable for the automatic driving of buses inside storage centers.
    The use of buses, for urban public transport or long journeys, is strategic in the current and future transport economy. In this context, a problem to be resolved concerns the storage of buses in a closed center, this problem being all the more acute since the space allocated is generally limited.
    Thus, the storage centers, arranged on several floors, have cramped passage areas with limited parking spaces. This creates cases of a bus collision with the infrastructure, when it is not a collision between buses. As a result, the speed in the bus center is generally limited to 8 km / h. In addition to these accidents, the maneuvers necessary for parking the buses and removing the buses generate time losses for drivers caught up in their working time.
    There is therefore a need for automation of bus guidance inside a parking lot, at least to make maneuvers more reliable (avoid accidents) and to reduce wasted time in the parking lot.
    By way of example, such guidance should lead to the process described briefly below.
    A bus driver who begins his working day presents himself at the entrance to the bus center and, via a dedicated interface, selects the bus with which he wishes to work. The bus parked in the bus center pulls out of its location and follows a route to the exit of the bus center where it stops at a dedicated location, ready for use. Likewise at the end of the day, the driver who finishes his working day leaves the bus at a location provided at the entrance to the bus center. Then, the bus automatically follows a route to the parking space allocated to it.
    To achieve such guidance, it is necessary to overcome several obstacles and in particular in terms of accuracy in following paths which can have strong curvatures, in particular in a constrained environment such as a bus parking center for example.
    Patent application FR 1501414 discloses a method for automatically driving a vehicle but is not suitable for this type of guidance because it treats the tracking of sections of zero curvature. If we use this process to follow sections of non-zero curvature, that is to say arcs of circles, there is a static lateral error while following the path. This process is therefore not suitable for guiding buses in a constrained environment as specified above.
    A document by A. Micaelli and C. Samson, “Trajectory tracking for unicycletype and two-steering-wheels mobile robots >> Research Report RR-2097, INRIA, 1993, describes in particular a control law for non-zero curvature monitoring. The solution described in this document is not, however, suited to the problem of bus guidance in a constrained environment such as a parking center. Indeed, in this case, it is necessary to make a predictive guidance command that can take into account control and state constraints with regard to the environment. In particular, it is necessary to guarantee a small deviation of the bus from its reference trajectory over a given prediction horizon, all along its navigation route. It is also necessary to allow displacements at positive and negative speeds, without discontinuity during transitions between positive and negative speeds via a zero speed.
    An object of the invention is in particular to allow a bus or other vehicle to follow all types of constrained paths, with very good precision. To this end, the subject of the invention is a method of automatically driving a vehicle from a given first location to a given second location forming a target position, said vehicle being subjected to lateral and / or longitudinal slippage of the front wheels. and rear wheels during its movements, said method comprising at least the following phases:
    a first phase in which a trajectory to be followed is generated as a function of the state of said vehicle and of said target position, said state being defined by the current position and the orientation of said vehicle, said trajectory (33) being cut into a series of sections whose starting point forms an intermediate position;
    - a second phase in which, at the current intermediate position of said vehicle:
    o a curvature is defined for each of n next sections, said curvature varying from one section to the next according to a polynomial function of the curvature c rQ of the section at said current position and of the curvature c rn l of the n th next section;
    o it is predicted, before said vehicle commits a movement, if said trajectory can be followed over all of the next n sections, as a function of constraints imposed and of estimated lateral and / or longitudinal shifts;
    a third phase in which, if said trajectory can be followed, the steering angle of said front wheels and the linear speed of traction of said vehicle are controlled as a function of the state of said vehicle and of lateral and / or longitudinal shifts for rally the center of the axle of said rear wheels on said path;
    if the trajectory cannot be followed, a new alignment of said vehicle is carried out towards said target position and a new trajectory to be followed is generated according to the first phase.
    In a possible embodiment, at each of said intermediate positions between the position of said vehicle and said n th section, the curvature c r is defined by:
    = 2c ro + c rn _ 1 where1 <i <n-2.
    In another possible embodiment, at each of said intermediate positions between the position of said vehicle and said n th { Γ0η ς 0η the curvature c r is defined by
    n-1 i + c ro where 1 <i <n- 2.
    The law of control of said steering angle of the front and rear wheels is for example obtained according to a process of optimization of a constrained function where the variable is a vector u composed of the derivative with respect to the distance of said steering angle, regardless of the weather.
    Said control law is for example based on a kinematic model of said vehicle taking into account the distance between the wheel axis and the steering axis at the front and at the rear of said vehicle.
    A vector u is for example calculated for each section of said trajectory.
    Said constraint is for example a function of said constraints imposed depending on the size of said vehicle.
    The state of said vehicle and the slips are for example derived from an observation of variables independent of said trajectory, said variables being:
    - the average rotation speeds a) f and ω ν of the front and rear wheels;
    - the steering angles of said front and rear wheels a ^, a r ;
    - the derivatives with respect to time a, at r of said angles;
    - the position (x m , y m ) and the angle of a movable mark (m) linked to said vehicle with respect to a fixed mark (o).
    Said vehicle can in particular be a public transport bus, said bus being driven automatically inside a bus storage center.
    The invention also relates to a device for controlling the automatic driving of a vehicle from a given first location to a given second location forming a target position, said vehicle being subjected to lateral and / or longitudinal slippage of the front wheels. and rear wheels during its movements, characterized in that said device being capable of being loaded into said vehicle and being connected at least, via appropriate interfaces, to proprioceptive sensors, exteroceptive sensors and actuation motors for the direction and traction of said vehicle, it includes a computer implementing the method as described above.
    Other characteristics and advantages of the invention will become apparent with the aid of the description which follows, made with reference to the appended drawings which represent:
    - Figure 1, an illustration of the problem of guiding a bus inside a storage center, as an example of application of the invention;
    - Figure 2, an example of a functional architecture of vehicle control, used by the invention;
    - Figure 3, an example of kinematic model used by the invention.
    FIG. 1 illustrates the problem of guiding a bus 1 in a storage center 10. The route that the bus must follow, from the entrance to the center, to reach its parking space 3 is represented by a broken line 2. In this example, the bus must follow several curves with a large radius of curvature in a reduced space, in particular a spiral ascent or descent to go from one floor to another and the successive turns leading to the parking space. The reverse route is taken towards the center exit which also corresponds, in this example, to entry 4.
    FIG. 2 illustrates the trajectory control architecture used by the method according to the invention. This architecture comprises at least three blocks:
    - A first block 21 is intended for the command itself, it includes a controller built on the basis of a kinematic model of the bus 1 to be guided, this model will be described later;
    - A second block 22 formed by a slip observer based on a horizontal kinematic or dynamic 2D model, correcting the slip values taken into account in the control model;
    - The third block 23 estimates the slippages from the observations provided by the observer 22.
    This control architecture is the same as that described in patent application FR 1501414. In particular, the state observer 22 and the slip estimator 23 are the same. The kinematic model, described below, is simplified compared to that of document FR 1501414.
    A control device according to the invention therefore incorporates these three blocks 21, 22, 23, these blocks being functions performed by a computer implementing the different phases of the method according to the invention.
    The control device, attached to the bus, is connected via appropriate interfaces to:
    - An energy source ensuring its electrical supply;
    - A perception system which gives position and orientation information with respect to a docking station, this system being notably described in patent application FR 1455049, the docking station being for example the parking space 3 of Figure 1;
    - Motors for steering and traction of the bus that the controller controls in steering angle and in linear traction speed;
    - Proprioceptive sensors of the vehicle: encoders for measuring the angle of the front and rear steering axles, encoders for measuring the axle speed of the front and rear wheels and a distance sensor indicating the vertical position of the vehicle, this position being for example indicated by measuring the distance between a support and the ground.
    The various kinematic models used for control 21 and observation 22 are described below. The control algorithm is based on a linearized model.
    FIG. 3 presents the 2D kinematic model of the bus mentioned above, opposite a trajectory 33 constructed as a function of a predictive model and of the current position of the bus. This kinematic model is close to that described in the document A. Micaelli and C. Samson, "Trajectory tracking for unicycle-type and two-steering-wheels mobile robots" and adapted to the situation in Figure 3.
    For reasons of simplification, a single front wheel 31 represents the assembly of the two front wheels and their axle (axle). Likewise, a single rear wheel 32 represents the assembly of the two rear wheels and their axle (axle). A local mark (/) is linked to the front wheels and a local mark (r) is linked to the rear wheels.
    For reasons of clarity, the center of the local coordinate system (/) has not been represented on the trajectory 33, at the current point P. It is the same for the reference (r).
    The front wheels 31 have a speed vector Vf making an angle ff with respect to the steering axis of the wheels. This angle représentef represents the sliding of the wheels on the ground. The speed vector Vf makes an angle Pf with the axis 30 of the bus, equal to the direction angle otf corrected for the slip angle f. The bus is guided by the front wheels, according to the steering angle ctf. The angle Pf expresses the steering angle of the front wheels.
    In the case of the rear wheels 32, the steering angle a r is zero. There is always a sliding angle ô r , between the speed vector of the rear wheels V r and the steering axis 30.
    The resulting speed vector V of the bus makes an angle β relative to the axis 30 of the bus. This angle β expresses the slip resulting from the vehicle, in the absence of slip β = 0.
    The direction of the axis 30 of the vehicle is identified in an absolute coordinate system (o) by an angle 6 m . This angle 6 m also represents the angle between the mobile reference point (m) linked to the vehicle and the absolute reference point (o).
    The center of the front wheel axle 31 makes a difference jy, or lateral error, with the reference path 33. More specifically, by a process of optimization under stress, this difference must be reduced so that the center of the the axle of the wheels 81 reaches the trajectory at a point P where a tangential local coordinate system Cf makes an angle 6 C f with the movable coordinate system (o), this angle 6 C f being the angle of the reference trajectory at the point P. Likewise, the center of the axle of the rear wheels 32 must reach a point not shown in the trajectory 33.
    For the purpose of simplification, the kinematic model chosen for the synthesis of the command results from the following choice: the commanded reference is the reference (r) linked to the rear wheels 32, located at the center of the axle of these rear wheels. In other words, the command seeks to make this reference point converge on path 33.
    With regard to FIG. 3, we assume that the angles lesf, ô r are small, so that they are assimilated to a variation of the angles of direction oy, a r . In order to express the mathematical equations, we start from the kinematic torsor speed of the coordinate system (/) and the kinematic torsor speed of the coordinate system (r), denoted respectively T f and T r :
    4îi ~ f T / = V f t l x r -(1)0. θ.
    0 m being the speed of rotation of the mobile, and v r being respectively the projection of the speed vectors V f and V r on the axis ff X of the coordinate system (/) and on the axis f rx of the coordinate system (r).
    Furthermore, the kinematic torsors of the marks (/) and (r) relative to the body of the mobile and expressed in the reference (m) linked to the mobile, the bus in particular, are denoted T f / I11 and T r / m and thus formulated:
    day 1 Γ 0 T. · .... - 0 0 - T r y in - ------------------- 1 o o_______________________1
    (2}
    The torsors defined by equation (1) are not independent. The combination of relations (1) and (2) makes it possible to link and v r as well as to obtain an expression of the speed of rotation 0 m of the mobile, that is:
    T ra = m Adj (T; - T // m ) = 'Ad r T. f .
    (3) with:
    01x2 m Bj, r J • represents respectively the vectors mf and mr expressed in the coordinate system (m):
    • is the vector obtained by rotation of T of the vector m t /, r :
    • m Rf.; · Is a rotation matrix expressing respectively the orientation of the marks (f) and (r) in the mark (m).
    So it comes:
    From where :
     m~ l.lj: · .- V-rc-r-r,i ’* r - '·''rr 1 r * J ·The
    Leading to the speed of rotation according to the following relation:
    J T7-r '.' î7-r v ] '' r, 'CCX'J, ·
    After the evolution equations of the model, we deal below with the trajectory pursuit.
    For each of the references (f) and (r), the projection onto the trajectory and the associated reference point are considered, respectively the reference point (ci) and the reference point (c r ), not shown. Relative to these benchmarks, the evolution of benchmark (f) and of benchmark (r) is given by:
    'T, = T, -' .I T 'T. = t - <i T with:
    T Qixa 1 M R, i ( _ 0 '' J : (k K, t 4 ~ k. F 'I -mu - ",. Rr + Jp, J *'} s · 7 'j., · * / ·>
    . 0:
    in (c ( ) ri jc r ) rcbpcLTivcitient.
    , with “p, abst rnmligneb ut rt mrhuri's a-, socj nos à la trajectoire
    In particular at the axle of the front wheels:
    ' l -' R. !
    * 'i ‘P bf
    Dk:
    V.
    1 n. ·, - --îhi i] _} '- 1 f ÎJ (--- :: - 11 -.,. - <, i
    -, ·.> - - · '· / - J ‘.-B - :. ··. . ·· .-
    - in <:
    'T.
    -’T: IL, 1
    - Ί ,. W.-; (·,. - n />-> =, · » f - i- ·, '; - f-inii ·.,; -' b I
    -'r., ·, Kibi.Lj · - V f -:, inii c.
    'k - • j ··' .: 1
    ...! ....-t ,; , 0, - - 11.; -, ·, - Γ ; 7 I, ”, ..i.
    : i. î ,. f -in i '/. - <i <.'s -. -η .: i - ·· ' P œ-T. „. -a <t ..
    l> D, ['
    ...... ,, ,, ,,,. T | · ·! Ad An
    - --y i
    - --- ki
    Γ <
    .i ,, | M ,, - -7 | ers · '. , · Üpj k 11 J
    Or in the coordinate system (Cf):
    , ( 17, - .-; Λ ·. V.. (. ·
    ί) '<7 d t. . <,.> <- II ·. , - · «! ! l
    By writing the relative speed f T f / cf in the coordinate system (c f ) without transport, that is to say by multiplying the speed of translation by the matrix f Rcf, we obtain:
    Similarly, we express f T f / cr in the coordinate system (c r ):
    The linear model is described below.
    The navigation path considered is a circle of curvature C r constant. The preceding models are then linearized around the equilibrium state:
    • 0 r =;
    • Vr / cr = 0 ΐ • = lc r = tan
    Considering the law of decomposition of a Taylor series function:
    t!
    it comes, for the tangent function:
    tan (.ï ') λ tan 3, - + (1 - tan 2 3 e ) (z - rij + ran (1 + tair 3 ,. j (.r - J c ) 2
    It follows that by asking:
    • 0 = 0r -;
    • w = υ Γ ;
    • y = Vr / c r · the linearization of the relations (6) leads to the following relations:
    __ 1 'COS y1' 7 l- <r, y 1 — C T // ÿ = v sin 0 u0 = (1 - c r y) Â r 0:
    = dc r + (1 + l ~ c ~ - 3c)] y - c r s v :
    = [lc r + (1 + Z 2 e 2 ) (ii / · - A,)] - Mr ·
    Then by considering y '= and by deleting the twice small terms like y (J3 f - y c ~) we obtain:
    if ff
    .)> - f- n f f c rlf
    Considering the state vectors:
    there
    where the exponent ‘>> expresses a derivation from the curvilinear abscissa;
    and x = y θ fif ~ Pc.
    we obtain the following linearized model:
    U 0
    not 1 (X e 2 0 ’R  0 1 0 1Ü 0 0 1 - 0 _ 0 • D-e; fll - - J
    px
    Al · ''> - Bu (8)
    In order to take into account strong accessibility constraints during navigation in the bus center, the method according to the invention uses a predictive command. It is a discrete predictive command which is based on the discrete model associated with relations (8). This predictive command optimizes a criterion depending on the predicted states and future commands on a spatial window of length n step S, if possible guaranteeing a certain number of constraints on the state (an admissible template) and the command (speed constraints of steering gear orientation).
    Given (8), the state matrices of the continuous system are:
    A c = AP 1 and B c = B
    The solution of the differential equation y '= A c y + B c u is y (s) = y o e AcS + e A < * s -hB c ιι (τ) άτ.
    For this, we assume A c and B c constant with respect to the curvilinear abscissa s. We keep this hypothesis for each small interval of distance S, the matrices A ck and B ck being a function of the curvilinear abscissa c rk and constant over a step S.
    By performing a discretization according to the curvilinear abscissa according to step S, the state matrices of the discrete system:
    Y k + i = A dk y k + B dk u k (k being between 1 and N, where N is the number of intervals) are given by:
    JA - r '> s - J Ij SA; + yAf ; · · + -ΓτΑ ,, -
    Ί B; , = e '*' f f- '> Τ άτ B ,, = p *' * s '' = ^ {A /, - h, |
    The input u k is assumed to be constant over the steps of distance S. If we consider displacements in forward and reverse, these calculations must be carried out for S> 0 and S <0.
    After this discretization step, we calculate predicted states. The predicted states are calculated starting from the knowledge of the current state y 0 and using the matrices Adk and B dk calculated previously. By performing a recurrence on n, we obtain:
    Iy i = A dg yo + Bd o uo y 2 = A dl A do yo + A dl B d (J "o + B dl " iy 3 = A d2 A dl A do y 0 + A d2 A dl B- do u 0 + A ^ B ^ Wi + B da , tt 2 etc.
    Or, in matrix form:
    ... M I ·· 4 = E - ............ A3 ... A3 yo + Brfo Osxi  >> l. ' <L .. . y ".. Ai w -i ..... A.;. .l Arf s _ t · · Ad t B ^ 0 ! · - Ad ^ Bdi ..] j i -
    m) and in a more synthetic way, we obtain the predicted state Y as a function of the current state y 0 and of the future command U either:
    Y = Ayp - BV
    111)
    The criterion to be optimized is a quadratic function of the predicted state Y and of the future command U. According to the invention, this criterion, noted crit, is expressed as follows:
    crit = | Y 1 <9 Y +1 U 1 iR U (12) where £ and are weighting matrices of dimension 3n x 3n and nxn respectively for the state and for the order. They are chosen from block-diagonal structure of generic elements / qQ3x3 and XrRixi where Yq and γ £, between 0 and 1, are forgetting factors, k being the rank in the diagonal and the matrix Q and R being positive defined adjustment matrices.
    The expression of the criterion, only as a function of the independent unknowns, that is to say as a function of U, is of the form:
    crit = | U £ (S ^ S + π) U + y 0 A l (13) from which we extract the constant depending only on y 0 .
    We give below the expression of the constraints on the state for a template given a priori. These constraints depend on the environment in which the guided bus operates, for example the structure of a storage center, as well as the dimensions of the bus. In a simplistic way, firstly, the bus is assimilated to a rectangle whose front and rear ends are respectively located at a point Df and at a point Dr of the direction axis of the nose gear along the longitudinal axis 30. The constraint is that these ends remain within a tolerance ô gap around the trajectory. This constraint can be written as follows:
    i | j / Dr- HÎllj ff lti $ e) l <^ gap | // + Df 6 » c ) | <5gep
    These relationships can still be expressed as follows, by linearizing and setting Θ = 0 m - 0 C :
    -D r 0
    D f O
    Γ i 1 X < _ ^ ap _
    (H)
    By asking :
    ~ D r 0 '
    Df 0
    -1 Dr 0 -1 ~ Df 0.
    and by setting d = ô gap 1 4x1 ;
    the constraint can be expressed by the following relation:
    Dy + d> Q 4x i (15)
    Extended to all the predicted state and expressed as a function of future orders, this constraint becomes:
    (16} with:
    • 5), the diagonal block matrix formed by n blocks D;
    Obtaining the control vector U is obtained by solving the following quadratic equation:
    min | u f (S'QBa K) U + y G UWU
    3 t * t 1 -ί- H-, 4y, i r. 1 ,. - I
    117!
    The first element β ' f = u 0 is then extracted from the vector U. The command is then given by the steering angle of the front wheels: (3 f = v f u 0 .
    We will now describe the implementation phases of the method according to the invention which apply for example the trajectory and control prediction described above. The steps of patent application FR 1501414 are repeated but in a different manner, in particular the trajectory is no longer cut along rectilinear sections but according to sections of non-zero curvature, advantageously using a fictitious curvature as a function of the real curvature of the trajectory to follow, in order to implement the command with a curvature with more progressive variation, as will be described later.
    The method according to the invention thus comprises the phases described below. These phases are repeated throughout the journey depending on the result of the prediction.
    In a first phase, the path 33 to be followed is generated to engage the bus towards its parking space 3, as a function of the state of the vehicle and of this parking space which is the target position. If we consider the example in figure 1, the trajectory begins at entrance 4 of the storage center. The method according to the invention also applies to extracting the bus from its parking space to take it to entrance 4 of the storage center. In this case, the target position is for example this input 4. The state of the bus according to which the trajectory is generated is the current position of the bus and its orientation (defined by the orientation of the axis 30 of the bus). The trajectory is cut into sections.
    In a second phase for each intermediate position of the bus (corresponding to the start of a section), a non-zero fictitious curvature is defined for a set of n following sections, forming the horizon on which the prediction is made. The curvature gradually varies from one section to the next. Examples of definition of curvatures will be given later. Then, before the bus initiates a movement, we predict at the start of each section whether the trajectory can be followed according to a part of the space constraints (for example those encountered on the route inside the storage center) or constraints of speed limitation of the actuators of the imposed wheels and on the other hand of the estimated lateral and / or longitudinal slippages.
    In a third phase, if the prediction indicates that the trajectory can be followed, the steering angle of the front wheels 31 and the linear speed of traction of the bus are controlled as a function of the state of the bus and of the lateral shifts and / or longitudinal, to join the center of the axle of the rear wheels 32 on the path 33. The command applied is for example that described above in relation to the kinematic model chosen for the bus.
    The second and third phases are repeated as long as the trajectory can be followed. If the result of the prediction is that the trajectory cannot be followed, a new alignment of the bus is carried out towards the target position and a new trajectory is generated according to the first phase.
    In this journey along the trajectory 33, the bus can move forward or reverse.
    We now describe the nature of the trajectory sections on which the prediction is made.
    When the kinematic model of the bus is linearized around a section of curvature c r , we assume a priori a nominal front steering angle configured for the corresponding curvature. There is, therefore, a steep angle jump in the prediction. If this jump is too large, it is not feasible in a real situation.
    According to the invention, a solution to this problem is to artificially modify the value of the curvature c r over the entire prediction horizon n, so as to obtain a progressive variation at the change of section. The number n indicates the number of sections to cover on the prediction horizon.
    The bus being at the current time at the start of a section of curvature c rQ and having to reach at its horizon n a section of curvature% _ ι; at all intermediate positions the curvature c r is defined by one of the following polynomials:
    where 1 <i <n-2
    2c 4 c
    - c r . = ——, this first polynomial being a first efficient approximation 1 3, in order to anticipate a third of the curvature at the end of the prediction horizon;
    c 4 - c ' c rt = rn ~ l ± r ° i + c r0> this second polynomial allowing a linear progression between c ro and% _ 1 .
    The intermediate positions defined below correspond to the departures of the successive sections of the trajectory.
    Other choices are possible, including more complex polynomials of higher degrees. Depending on the case, the relationship between the value of the prediction horizon nS and the size of the sections of the trajectory should be considered. Indeed, in the case where the prediction horizon nS is small compared to the size of the sections, this method ensuring a redefinition of each of the intermediate curvatures c r is the most relevant, insofar as it is comparable to a virtual subdivision of them. On the other hand, in the case where the prediction horizon nS becomes significant with respect to the size of the sections, a virtual variation of the values of curvatures is not always necessary. If however such a variation is necessary, it must be ensured that any new value of virtual curvature does not distort the trajectory, always remaining between the real value and the following different real value.
    As regards the characterization of the trajectory, this is defined at least in a geometric form parametrized x (u) by a parameter u.
    We assume that the function x (.) Is C 3 by pieces. This expression makes it possible to deduce all the desired characteristics of the trajectory, in terms of tangent, normal, curvature and variation of curvature, that is:
    nr = t ± .r rkl (x '„.%) llvt ^ X ^ .X ,, | Χ ,, ΙΙ - 3 {x ,, x u 'lx „j where:
    G) u> G) u and Wu respectively define a derivation along u in the order one, two and three;
    - and (a. b) = a ^ b.
    In the case where u varies like the curvilinear abscissa s, we have | x ' M | = 1, and equations (18) are simplified by:
    = dot ιχ (. xl · = det (x ' s .x') (19)
    Conventionally, the bus trajectory is defined as a succession of segments and arcs of circles, the position of the bus on this trajectory varying according to a curvilinear abscissa s. A possible problem is that the control law put in place tends in priority to keep a zero error on a section covered, including at the end of this section, even in the case of a large variation in the value of the curvature at the following section. This induces errors at the time of the section change, notably linked to the limitations of the axle actuators. In order to limit this phenomenon, depending on the size of the sections and the variation in their curvature, any pair of two successive sections having a significant variation in curvature is subdivided into several sections of smaller sizes in order to have a minimal variation in the curvature.
    Among other advantages, the method according to the invention allows a bus to follow all types of trajectories or paths, with very good accuracy on all types of section geometry, including transitions between sections.
    The invention has been described for driving a bus, it can of course be applied to 5 other types of vehicles.
    权利要求:
    Claims (10)
    [1" id="c-fr-0001]
    1. Method for automatically driving a vehicle from a first given location (4, 3) to a second given location (3, 4) forming a target position, said vehicle being subjected to lateral and / or longitudinal sliding of the wheels front (31) and rear wheels (32) during its movements, characterized in that it comprises at least the following phases:
    - A first phase in which a trajectory (33) is generated to be followed as a function of the state of said vehicle and of said target position, said state being defined by the current position and the orientation of said vehicle, said trajectory (33) being cut into a series of sections whose starting point forms an intermediate position;
    - a second phase in which, at the current intermediate position of said vehicle:
    o a curvature is defined for each of n next sections, said curvature varying from one section to the next according to a polynomial function of the curvature c ro of the section at said current position and of the curvature c rn l of the n th next section;
    o it is predicted, before said vehicle commits a movement, if said trajectory can be followed over all of the next n sections, as a function of constraints imposed and of estimated lateral and / or longitudinal shifts;
    - A third phase in which, if said trajectory can be followed, the steering angle of said front wheels (31) and the linear speed of traction of said vehicle are controlled as a function of the state of said vehicle and of lateral sliding and / or longitudinal to join the center of the axle of said rear wheels (32) on said path;
    if the trajectory cannot be followed, a new alignment of said vehicle is carried out towards said target position and a new trajectory to be followed is generated according to the first phase.
    [2" id="c-fr-0002]
    2. Method according to claim 1, characterized in that at each of said intermediate positions between the position of said vehicle and said n th section, the curvature c r is defined by:
    = 2% + c rn _ 1 where 1 <j < n _2.
    [3" id="c-fr-0003]
    3. Method according to claim 1, characterized in that at each of said intermediate positions between the position of said vehicle and said n th section is defined the curvature c r by _ 2c rn-l + c r0 j _ | _ where 1 <i <n - 2.
    r i n-1 r ° '
    [4" id="c-fr-0004]
    4. Method according to any one of the preceding claims, characterized in that the law of control of said steering angle of the front and rear wheels is obtained according to a process of optimization of a function under stress where the variable is a vector u composed of the derivative with respect to the distance β ' of said steering angle, regardless of time.
    [5" id="c-fr-0005]
    5. Method according to claim 4, characterized in that said control law is based on a kinematic model of said vehicle taking into account the difference (df, d r ) between the axis of the wheels (B) and the axis of steering at the front and rear of said vehicle.
    [6" id="c-fr-0006]
    6. Method according to any one of claims 4 or 5, characterized in that a vector u is calculated for each section of said trajectory.
    [7" id="c-fr-0007]
    7. Method according to any one of claims 4 to 6, characterized in that said constraint is a function of said imposed constraints depending on the size of said vehicle.
    [8" id="c-fr-0008]
    8. Method according to any one of the preceding claims, characterized in that the state of said vehicle and the slippages result from an observation (22) of variables independent of said trajectory, said variables being:
    - the average rotation speeds a) f and ω ν of the front and rear wheels;
    - the steering angles of said front and rear wheels ccf, a r ;
    - the derivatives with respect to time a, at r of said angles;
    - the position (x m , y m ) and the angle (6 m ) of a mobile reference (m) linked to said vehicle with respect to a fixed reference (o).
    [9" id="c-fr-0009]
    9. Method according to any one of the preceding claims, characterized in that said vehicle is a public transport bus, said bus being driven automatically inside a bus storage center.
    [10" id="c-fr-0010]
    10. Device for controlling the automatic driving of a vehicle from a first given location (4, 3) to a second given location (3, 4) forming a target position, said vehicle being subjected to lateral sliding and / or longitudinal of the front wheels (31) and the rear wheels (32) during its movements, characterized in that said device being able to be installed in said vehicle and to be connected at least, via appropriate interfaces, to proprioceptive sensors, exteroceptive sensors and motors for actuating the direction and traction of said vehicle, it comprises a computer implementing the method according to any one of the preceding claims.
    类似技术:
    公开号 | 公开日 | 专利标题
    FR3072069B1|2019-09-20|METHOD FOR AUTOMATICALLY DRIVING A VEHICLE, IN PARTICULAR A BUS IN A STORAGE CENTER, AND DEVICE IMPLEMENTING SAID METHOD
    US10399562B2|2019-09-03|Park assist with tire radius consideration
    EP2870052B1|2016-07-20|Improved method for determining the absolute angular position of the steering wheel of a motor vehicle
    FR2924831A1|2009-06-12|METHOD AND DEVICE FOR GENERATING A LACET SPEED ORDER FOR A GROUNDING AIRCRAFT
    FR2924832A1|2009-06-12|METHOD AND APPARATUS FOR GENERATING A CONTROLLED SPEED FOR AN AIRCRAFT RUNNING ON THE GROUND
    FR3063701A1|2018-09-14|METHOD AND SYSTEM FOR AIDING THE PARKING OF A MOTOR VEHICLE
    WO2014167255A1|2014-10-16|Parallel‑parking strategy using a continuously curving path
    WO2000039650A1|2000-07-06|Method for horizontal routing of an aircraft between two waypoints
    EP3115280B1|2018-04-18|Automatic driving method for inserting and removing a vehicle in a docking station, and monitoring device implementing such a method
    EP2470971B1|2014-11-05|Method and apparatus for guiding a vehicle in respect to overhead wires
    FR3096947A1|2020-12-11|A method of following the spiral trajectory of a vehicle under constraints
    FR2925710A1|2009-06-26|Flight line modifying method for e.g. cargo aircraft, involves calculating coordinates of turning point from flight distance in plane of follower aircraft, and implementing trajectory section in flight line to obtain modified flight line
    FR3045171A1|2017-06-16|METHOD FOR MONITORING AN ELECTROMECHANICAL ACTUATION SYSTEM
    EP3174775B1|2019-09-04|Method for regenerating a path for an automatic parking maneuver
    EP3730365A1|2020-10-28|Method and system for assisting the parking of a motor vehicle comprising a module for controlling any discontinuity in the remaining distance to be travelled
    EP2788246B1|2016-03-09|Parking assistance method for a motor vehicle
    EP3730385B1|2021-06-02|Method and device for detecting and reducing any speed deviations of a motor vehicle integrated in a parking assistance system
    CN111016891B|2022-03-22|Method and device for controlling vehicle driving route tracking
    EP3589533A1|2020-01-08|Device for controlling the trajectory of a vehicle
    EP1910154A2|2008-04-16|Method for real-time parking assistance for a motor vehicle, and corresponding device
    WO2021001112A1|2021-01-07|Method for determining an avoidance path of a motor vehicle
    WO2021084096A1|2021-05-06|System and method for managing the position of an autonomous vehicle
    FR3110129A1|2021-11-19|Trajectory calculation module, associated trajectory control device and method
    CN111016891A|2020-04-17|Method and device for controlling vehicle driving route tracking
    FR2983449A1|2013-06-07|Method for assisting to e.g. angled parking of electric motor vehicle in parking location, involves calculating distance corresponding to distance to be traveled during operation in case in which vehicle does not have angular shift
    同族专利:
    公开号 | 公开日
    WO2019072449A1|2019-04-18|
    FR3072069B1|2019-09-20|
    US20200298878A1|2020-09-24|
    US11254330B2|2022-02-22|
    EP3694767A1|2020-08-19|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US20150353082A1|2014-06-05|2015-12-10|Carnegie Mellon University|Unified motion planning algorithm for autonomous driving vehicle in obstacle avoidance maneuver|
    FR3038279A1|2015-07-03|2017-01-06|Commissariat Energie Atomique|AUTOMATIC DRIVING METHOD FOR THE INSERTION AND EXTRACTION OF A VEHICLE IN A RECEIVER STATION, AND A CONTROLLING DEVICE USING SUCH A METHOD|
    WO2017077223A1|2015-11-03|2017-05-11|Newtl |Device to assist with manoeuvres for parking alongside a platform|
    CN106681327A|2017-01-11|2017-05-17|中南大学|Method and system for intelligent driving horizontal and vertical decoupling control of great inertia electric motor coach|CN110228465A|2019-05-31|2019-09-13|深圳市航盛电子股份有限公司|A kind of semiautomatic parking system path planing method|
    CN112286049A|2019-07-27|2021-01-29|华为技术有限公司|Motion trajectory prediction method and device|
    CN112428991A|2020-11-30|2021-03-02|北京车和家信息技术有限公司|Vehicle control method, device, medium, equipment and vehicle|
    CN112874511A|2021-02-07|2021-06-01|奇瑞汽车股份有限公司|Automatic driving control method and device for automobile and computer storage medium|FR4306M|1965-01-11|
    CH463106A|1965-11-23|1968-09-30|Ciba Geigy|Hardenable, filler-containing epoxy resin mixture|
    US8170739B2|2008-06-20|2012-05-01|GM Global Technology Operations LLC|Path generation algorithm for automated lane centering and lane changing control system|
    US20150143913A1|2012-01-19|2015-05-28|Purdue Research Foundation|Multi-modal sensing for vehicle|
    FR3021938B1|2014-06-04|2016-05-27|Commissariat Energie Atomique|PARKING ASSIST DEVICE AND VEHICLE EQUIPPED WITH SUCH A DEVICE.|
    WO2017220705A1|2016-06-24|2017-12-28|Jaguar Land Rover Limited|Control system for a vehicle|
    JP6748619B2|2017-09-20|2020-09-02|日立オートモティブシステムズ株式会社|Vehicle control device, vehicle control method, and vehicle control system|
    US10684622B2|2017-11-22|2020-06-16|Uatc, Llc|Vehicle dynamics monitor for autonomous vehicle|FR3096947A1|2019-06-04|2020-12-11|Commissariat A L'energie Atomique Et Aux Energies Alternatives|A method of following the spiral trajectory of a vehicle under constraints|
    CN110789530B|2019-11-19|2021-04-09|中国科学院深圳先进技术研究院|Four-wheel independent steering-independent driving vehicle trajectory tracking method and system|
    CN111016886B|2019-12-19|2021-07-30|合达信科技集团有限公司|Automatic parking path planning method based on B spline theory|
    法律状态:
    2018-10-30| PLFP| Fee payment|Year of fee payment: 2 |
    2019-04-12| PLSC| Publication of the preliminary search report|Effective date: 20190412 |
    2019-10-31| PLFP| Fee payment|Year of fee payment: 3 |
    2020-10-30| PLFP| Fee payment|Year of fee payment: 4 |
    2021-10-29| PLFP| Fee payment|Year of fee payment: 5 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1759456A|FR3072069B1|2017-10-10|2017-10-10|METHOD FOR AUTOMATICALLY DRIVING A VEHICLE, IN PARTICULAR A BUS IN A STORAGE CENTER, AND DEVICE IMPLEMENTING SAID METHOD|
    FR1759456|2017-10-10|FR1759456A| FR3072069B1|2017-10-10|2017-10-10|METHOD FOR AUTOMATICALLY DRIVING A VEHICLE, IN PARTICULAR A BUS IN A STORAGE CENTER, AND DEVICE IMPLEMENTING SAID METHOD|
    US16/755,150| US11254330B2|2017-10-10|2018-08-27|Method for automatically driving a vehicle under constraint, in particular a bus in a storage facility, and device implementing such a method|
    PCT/EP2018/072949| WO2019072449A1|2017-10-10|2018-08-27|Method for automatically driving a vehicle under constraint, in particular a bus in a storage facility, and device implementing such a method|
    EP18756256.6A| EP3694767A1|2017-10-10|2018-08-27|Method for automatically driving a vehicle under constraint, in particular a bus in a storage facility, and device implementing such a method|
    [返回顶部]