METHOD FOR DYNAMIC CONTROL OF FOUR-WHEEL DRIVE MOTOR VEHICLE AXLES, CONTROL MEANS, VEHICLE AND PROGRAM BASED ON SUCH A METHOD

Abstract

The invention relates to a method for controlling wheel sets (12, 17) of a powertrain of a motor vehicle having four drive wheels, comprising the following steps: determining:– a torque setpoint from the driver;– a distribution ratio of said torque setpoint between said sets (12, 17);– a derivative of said distribution ratio;– whether the motor vehicle is in sled mode;– torque gradient derivatives to be applied to said wheel sets (12, 17) depending on whether the vehicle is in sled mode or not; filtering said torque gradient derivatives to avoid oscillations on said sets (12, 17); and controlling the torques applied to said sets (12, 17) depending on the filtered torque gradient derivatives. The invention also relates to a program, a control means and a vehicle based on such a method. Figure 1

Description

Title of the invention: METHOD FOR DYNAMIC CONTROL OF FOUR-WHEEL DRIVE MOTOR VEHICLE AXLES, CONTROL MEANS, VEHICLE AND PROGRAM BASED ON SUCH A METHOD

[0001] The invention relates to the field of electric vehicles and hybrid vehicles, with four-wheel drive, having a front axle and a rear axle. The invention relates in particular to the dynamic control of the front and rear axles in various cases of use of these vehicles, in particular in "sled mode". The "sled mode" is defined for such a vehicle in that the front axle produces a positive torque in the direction of travel so as to move the vehicle forward, while the rear axle produces a negative torque in the opposite direction to the positive torque causing the vehicle to move backward, but the negative torque having a value lower than that of the positive torque, the vehicle moves forward.

[0002] The preventive comfort filter function is a function that allows the raw driver torque setpoint to be filtered in order to significantly reduce the oscillations of the powertrain when passing through mechanical clearances. It therefore has a preventive correction action on the oscillations generated by this passage of mechanical clearances. Thus, this makes it possible to significantly improve the longitudinal comfort of the vehicle and also to typify its longitudinal dynamics.

[0003] On a 4x4 type vehicle, the choice of architecture of the preventive approval filter function is very important since the location of the application of the distribution calculated by the operating points function can simplify or complicate the calculations of the preventive approval filter.

[0004] In the case of the continuous approval filter described in patent application FR2202362, the distribution is applied downstream of a global powertrain filter. In this case, the synchronization of the passage of games as well as the dynamics of the torques, is perfectly controlled.

[0005] However, in the case of choosing a preventive approval filter with an application of the distribution which is done upstream of the filter, some questions concerning the control of the torque dynamics on a variable distribution, may arise. For this reason, the applicant filed a patent application FR2304176 concerning the antagonistic control of the torque dynamics of the vehicle's trains on a variable distribution with the aim of having a single and unique torque dynamics while the vehicle has two trains. This makes it possible to achieve a variable distribution without the driver feeling the slightest loss of pleasure and performance.

[0006]

[0007] Figure 2 shows a preventive approval filter solution with a distribution and an antagonistic distribution control of the aforementioned patent application FR2304176. The torque of the powertrain, a distribution ratio ^ons and its derivative make it possible to determine a front torque setpoint gradient C^y and a rear torque setpoint gradient C1^'- The reference IVC in [Fig.2] concerns an overall torque control corresponding to the driver's acceleration. These gradients are filtered by respective positive approval filters FAV, FAR to obtain filtered instructions. BT designates the activation or not of the sled mode. The method according to the invention begins by detecting the sled mode on the 4x4 vehicle, that is to say that the front axle generates a positive torque in the direction of travel while the rear axle produces a lower negative torque forward so that the vehicle continues to move forward. This allows the rear axle to recharge the battery while driving.

[0008] However, the torque dynamics between the front and rear axles are very rarely the same. This different torque dynamic is felt from a user point of view in that the vehicle may accelerate or slow down unintentionally.

[0009] The applicant filed the aforementioned patent application FR2304176 concerning the antagonistic distribution control, however this invention can only be applied over a positive torque range. However, in sled mode, by definition, the front axle torque is positive and the rear axle torque is negative.

[0010] In this case, the equations defined in patent application FR2304176 concerning antagonistic distribution control are no longer valid and the invention proposes to add this specific case. Indeed, as the distribution factor is defined between 0 and 1, when there is a positive front axle torque and a negative rear axle torque, there is in this case a distribution greater than 1 which calls into question our equations.

[0011] [Fig.3] illustrates the problem that can be encountered on a 4x4 vehicle if the torque dynamics of each axle are not controlled when sled mode is activated.

[0012] The equations that define the antagonistic distribution control (demand co pending FR2304176 above) are as follows: Ccons ^cons >^cons -j-^cons x~xconsv^coiis tt-cous ^~scons / y7"COiis\ / —scons A y “ix XL” Ix XL L, - K XL +11 - ix IXL Or Ccons is the powertrain setpoint; ,cbns 'AV CAR ways is the derivative of the torque gradient to be applied to the front axle; is the derivative of the torque gradient to be applied to the rear axle ^cons is rat|Q I j-partitiQ!! defined as follows with Te a time said sampling): cons And f--œtis 1 CTAV ^COBS rAXmo ^TAV^-TAR Jq cons You

[0013]

[0014]

[0015]

[0016]

[0017]

[0018] The sampling time is the computation time of the function ^cbns or ^cons CAV CAR according to the case. For example, a time Te of 10ms means that every 10ms the function is calculated and therefore the value is updated. We note that the antagonistic distribution control which is defined by the equations above is not valid in this case of sled mode. Indeed, the distribution ratio is bounded between 0 and 1 and this is to robustify the antagonistic distribution control since otherwise the torque gradient would be greater than what the front axle actually does. However, in the case of sled mode the distribution ratio is greater than 1. Thus, the equation is not valid in this specific case since the torque gradients and therefore the torque balance will be erroneous. This is what is highlighted in the graphs in [Fig.3]. The torque gradient of the rear axle is 0 Nm / s. As the distribution ratio is limited to 1 while the torque gradient of the front axle is not limited. Here we can clearly see the problem generated by the limitation of the distribution ratio. An objective of the present invention is to remedy the defects of the prior art, and in particular to propose a solution for antagonistic control of a four-wheel drive electric or hybrid vehicle, in particular in sled mode, that is to say when the front axle produces a positive torque in the direction of travel while the rear axle produces a lower negative torque forward so that the vehicle continues to move forward. To achieve this objective, the invention provides a method for controlling a front wheel set and a rear wheel set of a powertrain of a motor vehicle having four drive wheels comprising the following steps: - determine a driver torque instruction; - determine a distribution ratio of the driver's torque instruction between the front wheel set and the rear wheel set; the method being characterized in that it comprises the following steps: - determine a derivative of the distribution ratio; - determine whether the motor vehicle is in sled mode; - determine a derivative of a torque gradient to be applied to the front wheel set and a derivative of a torque gradient to be applied to the rear wheel set, the derivative of the torque gradient to be applied to the front wheel set and the derivative of the torque gradient to be applied to the rear wheel set being a function of the driver's torque setpoint, the distribution ratio and the derivative of the distribution ratio determined in the case where the motor vehicle is in sled mode; and in the opposite case, the derivative of the torque gradient to be applied to the rear wheel set is rather a function of the derivative of the torque gradient to be applied to the front wheel set; - filtering the derivative of the torque gradient to be applied to the front wheel set and the derivative of the torque gradient to be applied to the rear wheel set to avoid oscillations on said front and rear wheel sets; - control a torque applied to the front wheel set according to the derivative of the torque gradient to be applied to the filtered front wheel set and a torque applied to the rear wheel set according to the derivative of the torque gradient to be applied to the filtered rear wheel set.

[0019] Advantageously, the invention makes the realization of the driver torque request robust even in sled mode. It makes it possible to secure the entire torque dynamics of each train when there is a change in distribution, an activation of the sled mode, and therefore the stability of the vehicle.

[0020] Furthermore, the invention contributes to the improvement of the long-term enjoyment service. vertical alignment of the vehicle. It improves the energy efficiency of the powertrain since the sled mode allows the battery to be recharged using the rear axle while driving without the driver noticing it.

[0021] According to a variant, the derivative of the torque gradient to be applied to the wheel set before is determined using the formula: Cons y-eons AV OR Ccons is the powertrain setpoint; ,cbns -AV icbns AR is the derivative of the torque gradient to be applied to the front axle; is the derivative of the torque gradient to be applied to the rear axle K*™* is a distribution ratio defined as follows: cons And Te is a calculation time of the function

[0022] This makes it possible to obtain an accurate value of the derivative of the torque gradient at apply to the front axle.

[0023]

[0024]

[0025]

[0026]

[0027]

[0028]

[0029]

[0030]

[0031] Alternatively, if the vehicle is not in sled mode, then the derivative of the torque gradient to be applied to the rear wheel set is determined using the formula: COfW.Ç -jyrCifnS sriCOHS j • AJi. — ~ । I 1 ” IX > Or Ccons is the powertrain setpoint; Cons AV ^cbns CAR is the derivative of the torque gradient to be applied to the front axle; is the derivative of the torque gradient to be applied to the rear axle is a distribution ratio defined as follows: '-.COBS LTAV .cons AV Jo r CJ' ; and v cons L c'w+cÀr“ J 0 K “ ïë Te is a computation time of the pcons function. 'AR This allows to obtain an accurate value of the torque gradient derivative to be applied to the rear axle in the specific case where the vehicle is not in sled mode. Alternatively, if the vehicle is in sled mode, then the derivative of the torque gradient to be applied to the rear wheel set is determined using the formula: Ccons / ^cônsx ar - - V av / Or C™yS CS( |a derivative of the torque gradient to be applied to the front axle; and pcons CS( |a derivative of the torque gradient to be applied to the rear axle. AR This allows to obtain an accurate value of the derivative of the torque gradient to be applied to the rear axle in the specific case where the vehicle is in sled mode. According to a variant, the step of filtering the derivative of the torque gradient to be applied to the front wheel set and the derivative of the torque gradient to be applied to the rear wheel set is carried out respectively by means of a first filtering means and a second filtering means. This allows the driver torque instruction to be filtered to best limit oscillations of the drive train caused by crossing mechanical clearances. Alternatively, the steps are implemented by a control means. This allows control to be carried out by the control means which is part of the powertrain. The invention further relates to a control means implementing the method of control according to the invention.

[0032] Another object of the invention relates to a motor vehicle implementing the control method according to the invention.

[0033] The invention further relates to a computer program comprising program code instructions for executing the steps of the control method according to the invention, when said program operates on a computer.

[0034] The invention will be further detailed by the description of non-limiting embodiments, and on the basis of the appended figures illustrating variants of the invention, in which: - [Fig.l] schematically illustrates a powertrain of a four-wheel drive motor vehicle, suitable for implementing the method according to the prior art and the method according to the invention; - [Fig.2] schematically illustrates a distribution of torque instructions suitable for implementing the method according to the prior art and the method according to the invention; - [Fig.3] schematically illustrates changes over time in antagonistic control parameters of the front and rear axles in the context of a method according to the prior art; and - [Fig.4] schematically illustrates changes over time in antagonistic control parameters of the front and rear axles in the context of a method according to the invention.

[0035] The applicant has designed a detection of the sled mode case in order to activate a specific strategy for antagonistic control in sled mode.

[0036] In particular, it is planned to add to the antagonistic control strategy of the co-pending patent application FR2304176, a strategy which makes it possible to detect the sled mode. This makes it possible to control the torque dynamics of each train in this specific case; and to respect the torque balance and thus realize the driver torque request and the request for recharging the battery by the rear train.

[0037] A vehicle suitable for the invention is illustrated in [Fig.l]. It comprises: - control means 11 comprising for example at least one processor and at least one memory storing software instructions for implementing the method according to the invention; - a hybrid type powertrain 10, comprising a front wheel set 12 comprising a thermal engine 13 and a first electric traction motor 14; and a rear wheel set 17 provided with a second electric traction motor 18 associated with a speed reducer 19 and a device 20 for coupling and decoupling the second electric traction motor 18 with the rear wheels of the powertrain 10; - a coupling and decoupling device 15 making it possible to selectively ensure the

[0038]

[0039]

[0040]

[0041]

[0042]

[0043]

[0044] coupling and decoupling of the thermal engine 13 and / or the first electric traction motor 14 with the front wheel set 12, the coupling and decoupling device 15 taking the form of a gearbox in order to be able to isolate the first electric traction motor 14 from the thermal engine 13 during a pure electric driving mode, a clutch 16 being interposed between the thermal engine 13 and the first electric traction motor 14. The coupling and decoupling device 20 may consist of a dog clutch device, a clutch or a gearbox. In the context of the invention, firstly it is detected that the powertrain 10 is in sled mode. The sign of the torque of the front axle and the rear axle is observed while checking whether the rear axle is declared below the clearances. The reference B generally designates a state. The determination of the sign can be done by the following formulas: BTAV_positive — SÎgtl (Cœns_TAV) ' BrAR_negative = sign (Ccons TAR); Bacvjraineau “ ®under_the_games & & BTAV_positive & ByAR_negative 5 If sign(CconsrAv) = “+” then BtAV_positive = h otherwise BrAV_positive - 0; If sign(CconsTAV) = “+” then BTAR_negative = 1, else BTAR .negative “0” WHERE Positive Btav denotes a positive state of the front axle; Negative Btar means a negative condition of the rear axle; Bacv sled denotes a sled mode state; and Bunder games denotes a state under games. Once the sled mode is detected, the specific torque dynamics control strategy of each train is applied. The front axle maintains its torque dynamics since the distribution ratio is saturated at 1 in sled mode. On the rear axle, we also take this torque dynamics value from the front axle where we apply the "-" sign to obtain an antagonistic torque dynamics, and thus that this is transparent for the user (no degradation of performance and pleasure). When ®acv_traineau 1 (Bsous_]es_jeux = 1» BTAV_posittf BTAR_negatif ~ 1)' then sled mode detected. SO cons •xeons -TAV $ con a ot is AV AR he = 1 And jçcans I Te 0.01 1VU P So the distribution rate for sled mode is 1.

[0045] The formula Qcons _ ^cons x ^eons gcons ^'«« can be used as the equation for the front axle torque dynamics in sled mode (because the front axle has a positive torque range).

[0046] The formula — -Kcons x P^'+^l - x ““ cannot be used as an equation for the torque dynamics of the rear axle in sled mode since a negative torque is required.

[0047] Thus, in the case of sled mode B«cv_sled — 1, the torque dynamics of the rear axle is: Cons f AR “ "V~AVj

[0048] With an addition of this last formula to the antagonistic distribution control strategy the method is robust and adapted to all life situations of the powertrain.

[0049] On the graphs of [Fig.4], we can see that this strategy for the antagonistic control of the torques during the sled mode solves the problem mentioned previously.

[0050] The maximum and minimum gradients of the powertrain are determined from calibratable tables (the same maps as for the antagonistic distribution control strategy). This makes it possible to always calibrate on the dynamics of the slowest axis so that the antagonistic dynamics are the same and the driver does not perceive it when driving.

[0051] Variations in vehicle speed and even feelings of discomfort for the user are no longer present.

[0052] The process of calculating the antagonistic distribution control with the integration of the addition for the sled mode is done according to the scheme of [Fig.2].

Claims

Claims

1. Method for controlling a front wheel set (12) and a rear wheel set (17) of a powertrain of a motor vehicle having four drive wheels comprising the following steps: - determining a torque setpoint from the driver; - determine a distribution ratio of the driver's torque setpoint between the front wheel set (12) and the rear wheel set (17); the method being characterized in that it comprises the following steps: - determining a derivative of the distribution ratio; - determine whether the motor vehicle is in sled mode; - determining a torque gradient to be applied to the front wheel set (12) and a torque gradient to be applied to the rear wheel set (17); - filtering the torque gradient to be applied to the front wheel set (12) and the torque gradient to be applied to the rear wheel set (17) to avoid oscillations on said front (12) and rear (17) wheel sets - controlling a torque applied to the front wheel set (12) as a function of the filtered torque gradient to be applied to the front wheel set (12) and a torque applied to the rear wheel set (17) as a function of the filtered torque gradient to be applied to the rear wheel set (17), characterized in that the torque gradient to be applied to the front wheel set (12) is determined by means of the formula: éof,sxCcens । ^corts . v Aï / WHERE x-tcons ^œns '—AV '-AR is the powertrain setpoint; is the torque gradient to be applied to the front axle; is the torque gradient to be applied to the rear axle; is a distribution ratio defined as follows: cons ^AV pcons ^cons LAV +CAR he • D0 K cons l Vw^-ar Jq Te is a calculation time of the function in that if the vehicle is not in sled mode, then the gradient of torque to be applied to the rear wheel set (17) is determined by means of the formula: v-cons _ ^cons {1 rrcafis\ ^ccms . G-” Æ XG “F | 1 - £V 1 X G.- ' where Te is also a calculation time of the ^cons function, and in that if the vehicle is in sled mode, then the torque gradient to be applied to the rear wheel set (17) is determined by means of the formula: ^cbns / ^cbnsx CAR “ ' (CAV )'

2. Control method according to claim 1, characterized in that the step of filtering the torque gradient to be applied to the front wheel set (12) and the torque gradient to be applied to the rear wheel set (17) is carried out respectively by means of a first filtering means (FAV), and a second filtering means (FAR).

3. Control method according to any one of claims 1 to 2, characterized in that the steps are implemented by a control means (11).

4. Control means (11) implementing the control method according to claim 3.

5. Motor vehicle implementing the control method according to any one of claims 1 to 3.

6. A computer program comprising program code instructions for carrying out the steps of the control method according to any one of claims 1 to 3, when said program is running on a computer.

Images