Improved control of torque request split between powertrain and braking system of a vehicle

Abstract

A motion control, MC, module (20) for controlling the movement of a vehicle (40), the MC module (20) being able to be coupled to an actuation module, A, (14) of the vehicle (40), comprising a braking system (14') and a powertrain (14") of the vehicle (40), the MC module (20) being configured to receive as input ego-motion, EM, control data (DEM,C) indicative of measured parameters of the vehicle (40) and a road to be travelled by the vehicle (40), and actuation feedback data (DA, Fdbk) generated by the A module (14), wherein the EM control data (DEM,C) comprise at least one road slope (9eg0) and a measured vehicle velocity (veg0) and the actuation feedback data (DA, Fdbk) being correlated with a feedback wheel torque signal (T wheel, Fdbk) indicative of an estimated torque value provided by the braking system (14') and powertrain (14"), the MC module (20) comprising a disturbance wheel torque observer compensator, DWTOC, module (26) configured to receive as input EM control data (DEM,C) and actuation feedback data (DA, Fdbk) and, as a function of such inputs, to output a corresponding observer wheel torque signal (T wheel, OBS) indicative of a torque contribution processed by an observer to correct actual torque drifts in the powertrain (14"), the MC module (20) being also configured to generate as output, as a function of the observer wheel torque signal (T wheel, OBS), a brake request torque signal (TBrake,Req) indicative of a torque command for the actuation of the braking system (14') in order to control the movement of the vehicle (40).

Description

[0001] IMPROVED CONTROL OF TORQUE REQUEST SPLIT BETWEEN POWERTRAIN

[0002] AND BRAKING SYSTEM OF A VEHICLE

[0003] CROSS-REFERENCE TO RELATED APPLICATIONS

[0004] This Patent Application claims priority from Italian Patent Application No. 102024000014749 filed on June 26, 2024, the entire disclosure of which is incorporated herein by reference.

[0005] TECHNICAL FIELD OF THE INVENTION

[0006] This invention relates generally to the automotive field, in particular to an improved control of torque request split between the powertrain and braking system of a vehicle. Specifically, it relates to a motion control module, a control system comprising the motion control module, a vehicle comprising the control system, a control method and corresponding software.

[0007] STATE OF THE ART

[0008] AS is well known, automated driving systems have seen a strong acceleration in terms of development in recent years, especially in the area of high-speed applications (e.g. “Adaptive Cruise Control” and “Lane Centering”). However, this is not the only area that research into autonomous driving has focused on.

[0009] In fact, there has recently been the development of driver assistance systems used at low speeds (e.g., at vehicle speeds below approximately 10 km / h). This is the case, for example, with the rear emergency braking functions that avoid hitting obstacles when manoeuvring to park, or the “Rear Cross Path” function that avoids collisions with passing vehicles when manoeuvring out of a parking space. In addition to assistance systems, autonomous parking functions have also been developed, such as “Remote Park Assist” and “Full Autonomous Parking”. These automated driving functionalities are based on very complex technologies, especially in terms of “motion control” aspects, that is, the management of the vehicle's actuators, where optimal control of the vehicle's longitudinal dynamics at low speeds specifically for parking manoeuvres is still problematic.

[0010] In general, control of the vehicle’s longitudinal dynamics at low speed can be classified as a SIMO (“Single Input Multi Output”) type system. This system must be able to manage the vehicle's speed and acceleration by interacting with two actuating variables on the powertrain and braking system. Managing these systems is not easy, especially at low speeds where the behaviour of the actuators is less linear than at higher speeds. Moreover, the type of implementation interface may vary between different original equipment manufacturers (OEMs). For example, in “drive-by-wire” systems, interfaces in terms of percentage accelerator pedal and brake pedal are permissible, although they are not optimal for the purpose as it is necessary to know the conversion logic of the pedal that translates the manual control into the final torque actually exerted. With the advent of ADAS systems, there has been a real transformation in this field, with the major suppliers providing increasingly integrated and optimised SMART actuation interfaces that allow, for example, torque requests to be made directly to the wheels (or, in the case of more conventional internal combustion engines, to the flywheel). Clearly, however, this complicates the control of the vehicle’s longitudinal dynamics at low speeds.

[0011] Another major challenge of low-speed control concerns the measurement and estimation of the fundamental vehicle quantities that govern longitudinal dynamics, in particular velocity and acceleration. Conventionally, vehicle speed is calculated from the number of phonic wheel counts of the vehicle and the measurement is asynchronous with respect to the speed of the wheel. Consequently, at very low speeds and especially in the stopping zone, it is difficult to obtain a reliable measurement from the phonic wheel system. The same applies to acceleration, normally measured by inertial units. The low sensitivity of the measurement, coupled with its noisiness, make it complex to implement a suitable feedback system, considering that trajectory planning involves stringent constraints in terms of maximum jerk values and maximum absolute acceleration.

[0012] It is therefore extremely complicated to design a control system capable of performing manoeuvres as comfortably as possible, due to the issues mentioned above. This becomes even more important at low speeds where acceleration peaks and vehicle jerks are perceived by the driver and passengers as pronounced.

[0013] Moreover, powertrains with automatic transmission with torque converter and internal combustion engine (ICE) are known to be affected by creeping. In fact, at low speed, the engine's “lock-up” converter is normally open and, therefore, the idle torque produced by the engine leads to slipping between the pump and turbine. In this situation, the torque seen on the transmission side is amplified according to a slip torque amplification characteristic that makes the vehicle “creep”. This means that, if it is necessary to maintain a very low vehicle speed, a combination of powertrain and braking torque is needed to achieve the correct balance of forces. As better described below, this effect is sometimes desired by drivers; however, it is currently difficult to control it reliably and accurately.

[0014] Specifically, creeping is a well-known phenomenon in the powertrain field and mainly characterises the behaviour of thermal propulsion systems with automatic transmission with torque converter. In this vehicle configuration, releasing the brake pedal in drive mode causes the vehicle to move by dragging (creeping). After a certain transient period, which specifically depends on the characteristics of the powertrain, the vehicle will reach a stationary speed vCreep that is a function of the engine speed coengRPM, the gear ratio ig, the final transmission ratio id and the wheel R size, and which can be defined, for example, by the following expression:

[0015] This specific speed identifies a working point of the powertrain, also known as the “creeping region”. Normally the creeping region is between 0 and 10 km / h according to the different gear ratios given in the equation (1). In the case of an internal combustion engine, this operating area coincides with an “idle” control region, where the engine control unit is responsible for adjusting and tracking the minimum rpm target (DMiNengRPM to be maintained to overcome engine friction. The choice of target engine idle speed may depend on a multitude of factors, such as the temperature of the engine itself, current loads or anti-pollution systems, but in general, once the warm-up phase is complete, it remains almost constant (roughly around 700-800 rpm).

[0016] It is, therefore, clear that in order to maintain a speed below the creeping region (vCVcreep), it is necessary to balance the torque delivered by the powertrain by acting with the mechanical braking system. In fact, since creeping is the result of the coupling of a thermal propulsion mechanical component and a transmissive one, it is an effect that, although unavoidable, is favoured by motorists. Through the action of the brake pedal alone, it is in fact possible to reach a lower speed than in the creeping region, for example during parking and starting manoeuvres. This is why even in the most modem hybrid or “full electric” propulsion systems, creeping is an effect that is maintained and emulated (e.g. through the so-called “e-creeping feature”), even in the absence of specific technical reasons since an electric motor does not need to maintain a minimum idle speed.

[0017] From the point of view of a motion control module for automated driving, which is intended to handle low-speed manoeuvres, creeping can thus be exploited in the same way as by a human driver. However, this type of control is currently difficult to obtain.

[0018] OBJE T AND SUMMARY OF THE INVENTION

[0019] The Applicant noted that the solutions according to the prior art, although satisfactory in certain respects, can be improved.

[0020] The purpose of this invention, therefore, is to provide a solution that improves, at least in part, the solutions of the prior art. Specifically, the purpose of this invention is to provide a motion control module, a control system, and a control method, as claimed in the appended claims.

[0021] BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 shows a functional block diagram relating to a control system for controlling, in particular at low speeds, a vehicle according to one embodiment of this invention.

[0023] Figure 2 shows a functional block diagram relating to a motion control module of the control system in Figure 1 according to an embodiment of this invention.

[0024] Figure 3 shows a diagram of some of the forces acting on a vehicle during its use.

[0025] Figures 4-6 show examples of signals generated in use by the control system in Figure 1.

[0026] DESCRIPTION OF PREFERRED EMBODIMENTS OF THE INVENTION

[0027] This invention will now be described in detail with reference to the attached figures in order to allow a skilled person to implement it and use it. Various modifications to the described embodiments will be readily apparent to those skilled in the art and the general principles described may be applied to other embodiments and applications without however departing from the protective scope of this invention as defined in the attached claims. Therefore, this invention should not be regarded as limited to the embodiments described and illustrated herein but should be allowed the broadest protection scope consistent with the features described and claimed herein.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning commonly understood by one of ordinary skill in the art to which the invention belongs. In case of conflict, this specification, including the definitions provided, will control. Furthermore, the examples are provided for illustrative purposes only and as such should not be construed as limiting.

[0029] In particular, the block diagrams included in the attached figures and described below are not to be understood as a representation of the structural features, i.e. construction restrictions, but must be understood as a representation of functional features, i.e. intrinsic properties of the devices defined by the effects obtained, that is to say functional restrictions, which can be implemented in different ways, so as to protect the functionalities thereof (operational capability).

[0030] In order to facilitate the understanding of the embodiments described herein, reference will be made to some specific embodiments and a specific language will be used to describe the same. The terminology used herein is used for the purpose of describing particular embodiments only and is not intended to limit the scope of this invention.

[0031] In the following description, elements common to the different embodiments have been indicated with the same reference numbers.

[0032] Figure 1 shows a control system 10 for controlling, in particular at low speeds, a vehicle such as a motor vehicle.

[0033] In the following, “low speeds” mean speeds typically used to perform manoeuvres such as parking, for example speeds below about 10 km / h.

[0034] As can be seen, the control system 10 comprises a Motion Planning and Trajectory Definition (MPTD) module 12, a Motion Control (MC) module 20, an Actuation (A) module 14, a Vehicle Plant (VC) module 16 and an Ego-Motion (EM) module 18, which are operationally coupled together.

[0035] In particular, the MPTD module 12 is of a known type and in use provides a trajectory, typically expressed in terms of acceleration and speed, from input information such as a measurement of the distance to the “Closest Impact Obj ecf ’ (CIO) against which the vehicle might collide in the manoeuvre to be executed. The technology and criteria behind the MPTD module 12 are well known today, especially for fully autonomous parking applications. To make the manoeuvre as comfortable as possible, the MPTD module 12 can take certain constraints into account, for example a maximum and minimum speed Vmax,min, a maximum and minimum acceleration amax,min and a maximum and minimum jerk j max, min. For example, the MPTD module 12 incorporates convex optimisation methods (e.g. predictive control models based on optimisation, such as model predictive control strategies) for this purpose.

[0036] In general, the MPTD module 12 receives as input MPTD input data DMPTD and, based on these data, generates as output trajectory data Dtraj indicative of the planned trajectory for the vehicle.

[0037] MPTD input data DMPTD may comprise one or more of the following data: a CIO distance dcio indicative of the vehicle's distance from the CIO; the maximum and minimum speed Vmax,min ; the maximum and minimum acceleration amax,min ; and the maximum and minimum jerk j max, min. Potentially, the MPTD input data DMPTD may also comprise a rolling resistance p, indicative of the vehicle's rolling resistance on the road and thus related to the contact surface of the vehicle's tyres with the road; specifically, the rolling resistance p may be a fixed, predefined parameter of a known type. These data are obtained in a known way, for example by measurement using vehicle sensors, or by automatic or manual selection.

[0038] The trajectory data Dtraj may comprise one or more of the following data: a reference acceleration aref; and a reference velocity vref. These data are obtained in a known way, for example through the convex optimisation methods mentioned above.

[0039] In addition, the MPTD module 12 can also receive as input control feedback data DMC dbk generated by the MC module 20, in order to generate trajectory data Dtraj also depending on the control feedback data DMC dbk.

[0040] For example, the control feedback data DMC,Fdbk can be used during optimisation for trajectory generation and may comprise one or more of the following: a maximum and minimum potential acceleration empirically estimated by the MC module 20, for example based on the maximum and minimum torque values deliverable by the actuators of the A module 14; and a maximum and minimum potential speed empirically estimated by the MC module 20, for example based on the maximum and minimum torque values deliverable by the actuators of the A module 14.

[0041] The MC module 20 is better described below, with reference to Figure 2.

[0042] In general, the MC module 20 allows the trajectory planned by the MPTD module 12 to be followed, providing an adequate request for braking and propulsive torque.

[0043] In use, the MC module 20 receives as input the trajectory data Dtraj generated by the MPTD module 12, EM control data DEM,C generated by the EM module 18 and actuation feedback data DA, Fdbk generated by the A module 14.

[0044] In particular, the EM control data DEM,C may comprise one or more of the following: a road slope 0ego indicative of the measured road slope; a measured vehicle speed veg0indicative of the measured vehicle speed; and a measured vehicle acceleration aeg0indicative of the measured vehicle acceleration.

[0045] In addition, the actuation feedback data DA, Fdbk may comprise one or more of the following data: a feedback wheel torque signal T wheel, Fdbk indicative of an estimated torque supplied to the wheels by actuators in the A module 14 (or signals TBrake,Fdbk and Tpwt dbk correlated to, and indicative of, the signal T wheel, Fdbk as further described below); a creeping feedback torque signal Tcreep dbk indicative of a torque provided as feedback by a powertrain 14" of the A module 14 (which normally corresponds to the minimum friction torque required to maintain the creeping effect) and, optionally, information regarding the availability of interfaces of the A module 14 and other useful feedback information, such as maximum and minimum saturation values of torque deliverable by the powertrain 14" (specifically, in the idle region the minimum torque generally coincides with the creeping / idle torque that is positive so as to avoid stalling of the powertrain 14").

[0046] On the basis of the input data and the criteria described in more detail below, the MC module 20 generates control data Dctri as output to control actuators of the A module 14.

[0047] The control data Dctri may comprise one or more of the following data: a brake control torque signal T Brake. Req indicative of a torque to be supplied to a braking system 14' of the A module 14; and a powertrain control torque signal Tpwt,Req indicative of a torque to be supplied to the powertrain 14" of the A module 14. Specifically, the sum of the brake control torque signal TBrake,Req and the powertrain control torque signal Tpwt,Req can correspond to a wheel torque signal TWheei indicative of an overall torque to be applied to the wheels of the vehicle.

[0048] In addition, based on the input data, the MC module 20 can also output the previously mentioned control feedback data DMC,Fdbk to be sent to the MPTD module 12.

[0049] The A module 14 (known and in particular comprising the braking system 14' and the powertrain 14") therefore receives the control data Dctri and controls the operation of the braking system 14' and the drive train 14" accordingly in a known way.

[0050] In use, the A module 14 can also output the wheel torque signal T wheel. The A module 14 can also output the previously mentioned actuation feedback data DA dbk, to be sent to the MC module 20.

[0051] The VC module 16, of a known type, is a modelling of the vehicle itself (that is, the remaining vehicle components). Specifically, the VC module 16 can comprise sensors of a known type to determine information indicative of the current state of the vehicle. Specifically, the VC module 16, controlled on the basis of the wheel torque signal T wheel, outputs VC data Dvc indicative of vehicle dynamics (e.g., inertial measurements made by inertial sensors, measurements related to the torque supplied to the wheels, wheel speed measurements, etc.).

[0052] The EM module 18, of a known type, aims to estimate the main vehicle states from the measurements provided by the vehicle's various sensors, such as odometry, filtered speed and measured acceleration.

[0053] For this purpose, the EM module 18 receives the VC data Dvc and, on the basis of this, generates output EM data that may comprise one or more of the following data: the road slope 0ego; the measured vehicle velocity veg0; the measured vehicle acceleration aeg0; and the instantaneous position xposof the vehicle.

[0054] For example, the output EM data can be divided into the EM control data DEM,C, previously described and received as input by the MC module 20, and the EM trajectory data DEM,I-

[0055] The EM trajectory data DEM,I may comprise one or more of the following: the measured vehicle velocity veg0; the measured vehicle acceleration aeg0; and the instantaneous position xposof the vehicle.

[0056] The EM trajectory data DEM,I can be received as input by the MPTD module 12, which can also generate the EM trajectory data Dtraj on the basis of the EM trajectory data DEM,I, in a known way.

[0057] Figure 2 shows the MC module 20 in detail.

[0058] The MC module 20 comprises a feedforward wheel torque compensator module (FWTC, optional) 22, a closed loop wheel torque compensator module (CLWTC, optional) 24, a disturbance wheel torque observer compensator module (DWTOC) 26 and a feedforward powertrain creeping control module (FPCC, optional) 28.

[0059] The (optional) FWTC module 22 receives FWTC input data as input, in particular the reference acceleration aref and the road slope 0ego, and based on these inputs generates a feedforward wheel torque signal T wheel, FFas output. As further described below, the signal T wheel, FF is indicative of a feedforward torque contribution that takes into account the reference acceleration aref as output from the MPTD module 12. This contribution is useful for controlling the braking system 14' during transient conditions and compensates for the major disturbances in vehicle dynamics, such as road slope 0ego.

[0060] For example, the signal Twheei,FF helps to improve the control of the vehicle dynamics especially during vehicle stops, where the dead zone of the measured vehicle velocity signal veg0, for example from the phonic wheel, would lead to possible tracking discontinuities.

[0061] The (optional) CLWTC module 24 receives CLWTC input data as input, in particular the reference velocity vref and the measured vehicle velocity veg0, and based on these inputs generates a closed-loop wheel torque signal T wheel, CL as output.

[0062] As further described below, the signal T wheel, CL is indicative of a closed-loop torque contribution that takes into account the output reference velocity vrcf as target reference from the MPTD module 12.

[0063] The choice of using the reference velocity vref in feedback-loops arose to overcome some of the disadvantages described above. More specifically, due to the sensitivity problems of measuring acceleration at low vehicle speeds, a closed-loop control system based on this acceleration measurement would not be accurate enough to guarantee optimal tracking performance. Instead, setting the closed-loop control action against the reference velocity signal vref, which under stationary or semi-stationary conditions is a more stable and reliable measurement than acceleration, achieves this purpose.

[0064] Consequently, the combined action of the FWTC module 22 and the CLWTC module 24 makes it possible to consider both the high-frequency contribution provided by the FWTC module 22 and indicative of the feedforward control of acceleration, and the low-frequency contribution provided by the CLWTC module 24 and indicative of closed-loop velocity control. In general, this allows the coupling of an open-loop component that is based on the trajectory acceleration reference, and a closed-loop component that is based on the trajectory velocity reference; this allows the acceleration reference of the trajectory during the transient phase (component with a high-frequency information content) and the velocity reference during the steady-state phase (component with a low-frequency information content) to be exploited in a complementary manner. In addition, such control ensures that known limitations in measuring acceleration and velocity are overcome.

[0065] The DWTC module 26 receives as input DWTC input data, in particular the road slope 0ego, the measured vehicle velocity veg0and a feedback wheel torque signal T wheel, Fdbk described in more detail below, and on the basis of these inputs (that is, at least on the basis of the road slope 0ego and the measured vehicle velocity veg0) generates as output an observer wheel torque signal T wheel, OBS.

[0066] As further described below, the signal Twheei,OBS is indicative of a torque contribution made by an observer that is intended to correct for drifts (also referred to as deviations or variations) in the actual torque of the powertrain 14" supplied to the wheels, thereby improving the way in which the sharing of torque requests between the powertrain 14" and the braking system 14’ occurs. In particular, these drifts in actual torque correspond to deviations between the torque generated by the powertrain 14" (for example, defined on the basis of the actuation feedback data DA, Fdbk) and its target value (for example, defined on the basis of the EM control data DEM,C).

[0067] This is done in particular by implementing an asymptotic filter that relates measurements of vehicle dynamics (such as road slope 0egoand measured vehicle velocity Vego) to the signal T wheel, Fdbk that is indicative of torque measurements delivered to the wheels, for example from the braking system 14’ and the powertrain 14".

[0068] Next, a brake request torque signal TBrake,Req is generated from the signals Twheei,FF, T wheel, CL, Twheei,OBS and from a feedback powertrain torque signal Tpwt,Fdbk described in more detail below. In particular, the signals T wheel, FF, T wheel, CL and T wheel, OBS are added together and the signal Tpwt,Fdbk is subtracted from this sum to generate the signal TBrake,Req, for example via a first (optional) summing module 30 of the MC module 20. Specifically, the signal TBrake,Req is an actuation command for the braking system 14’.

[0069] The braking system 14’ then receives the signal TBrake,Req and is controlled accordingly.

[0070] Furthermore, the braking system 14', controlled via the signal TBrake.Req, also generates in use a brake feedback torque signal TBrake,Fdbk that is indicative of a feedback torque of the braking system 14', that is, a measured torque of the braking system 14' (e.g., acquired by sensors included in the braking system 14' and configured to detect information regarding the state of the actuators of the braking system 14') to be used as feedback.

[0071] The (optional) FPCC module 28 receives as input FPCC input data, in particular the road slope 0ego and the signal T creep, Fdbk (as well as, in a way not shown, possibly the maximum acceleration amax too, which is a manually selectable parameter, for example equal to approximately 0.3 m / s2; alternatively this parameter is predefined and therefore already stored in the FPCC module 28), and on the basis of these inputs generates a candidate powertrain torque signal Tpwt,cnd as output.

[0072] As described in more detail below, the signal Tpwt,cnd is indicative of a torque request to be possibly supplied to the powertrain 14", generated by exploiting the creeping effect.

[0073] Next, a powertrain request torque signal Tpwt,Req is generated from the signals Tpwt.Cnd and Tcreep dbk. In particular, the signals Tpwt,Cnd and T creep, Fdbk are compared with each other, for example by means of a comparison module 32 (optional) of the MC module 20, and the signal Tpwt,Req generated depends on the greater of the signals Tpwt,cnd and T creep, Fdbk (e.g., is equal to the greater of these two signals). Specifically, the signal Tpwt.Req is an actuation command for the powertrain 14”

[0074] The powertrain 14” thus receives the signal Tpwt,Req and is controlled accordingly.

[0075] In addition, the powertrain 14", controlled via the signal Tpwt,Req, generates both the signal Tcreep dbk and the signal Tpwt,Fdbk in use as feedback.

[0076] The signal T creep, Fdbk is then provided as feedback to the FPCC module 28, while the signal Tpwt,Fdbk is provided as input to the first summing module 30.

[0077] In particular, the signal Tpwt,Fdbk is indicative of a feedback torque of the powertrain 14", that is, a measured torque of the powertrain 14" (e.g., acquired by sensors included in the powertrain 14" and configured to detect information about the state of the actuators of the powertrain 14") to be used as feedback.

[0078] Subsequently, the signal Twheei,Fdbk is generated from the signals Tpwt dbk and TBrake dbk. In particular, the signals Tpwt dbk and TBrake,Fdbk are added together to generate the signal T wheel, Fdbk, for example via a second (optional) summing module 34 of the MC module 20. Specifically, the signal T wheel, Fdbk is a feedback for the DWTC module 26, which is generated taking into account the contributions of the braking system 14' and the powertrain 14".

[0079] It is, therefore, clear that the vehicle velocity regulation is primarily based on the request for braking torque due to the signal TBrake,Req. In fact, at the dynamic level, the braking system 14' is certainly a more responsive and more linear means of actuation than the powertrain 14", which at low speeds, especially in the case of vehicles with internal combustion engines, is affected by strong actuation delays and very slow dynamic responses that drastically reduce the phase margin and, consequently, the bandwidth of any feedback control system.

[0080] Therefore, in order for the control action to be handled through the braking torque request due to the signal TBrake,Req, the FPCC module 28 is present here, which provides the torque command identified by the signal Tpwt,Req. This allows for greater stationariness and takes into account the actual torque at low speeds, as well as any disturbances that may occur, such as the road slope and maximum acceleration demand. This adjusts the creeping point and means the vehicle is fully controllable in all working conditions.

[0081] The modules 22-28 shown in Figure 2 will now be described in more detail.

[0082] As seen above, the FWTC module 22 is intended to compensate for known disturbances that may occur and especially helps during transient phases of vehicle acceleration.

[0083] In particular, the FWTC module 22 is based on a simplified inverse model of the vehicle's longitudinal dynamics, which in terms of the Newton-Euler equilibrium equation can be seen as a linear combination of several forces acting at the vehicle's centre of gravity, as also shown in the free body diagram in Figure 3 and defined by the following mathematical expression: where Finteriai is an inertial force acting on the vehicle (shown in Figure 3 with reference 40), Ftraction is a traction force exerted by the vehicle 40 due to the A module 14, F slope is a gravitational force acting on the vehicle 40 and due to the road slope, Frou is a rolling force acting on the vehicle 40 and due to the frictional contact surface between the wheels of the vehicle and the road (specifically, due to the fact that the contact surface between the tyre and the asphalt is not point-like but rather it is a surface that alters the effective radius of the tyre), and Fair is a braking force acting on the vehicle 40 and due to the resistance exerted on the vehicle by the surrounding air.

[0084] Since this model is based on low-vehicle-speed scenarios, the contribution of air resistance can be neglected. Consequently, by expanding the other terms, the following mathematical expression is obtained:

[0085] M ■ a =Twheel'FF- M - g - sin(0) — M - g - p - cos(0) (3)

[0086] 7? where M is the vehicle mass, g the gravitational constant, R the tyre radius, 0 the road slope (that is, 9ego) a is the acceleration of the vehicle (that is, aeg0), p is the predefined rolling resistance parameter, and T wheel, FF is the torque at the wheel determined by the FWTC module 22.

[0087] In this case, it is assumed that all the traction force is transferred to the wheel and that no slipping occurs between the asphalt and tyre. For the rolling resistance model, only the tyre deformation effect due to the normal force acting on the tyre is taken into account; again, the other effects that depend on vehicle velocity are negligible. The vehicle mass and the rolling resistance coefficient can be considered as unchanging over time, although specific, known algorithms that allow real-time estimation of these parameters can similarly be used to improve the fit of this model. The road slope 9 can also be estimated by taking the derivative of the vehicle velocity and comparing it to the measured acceleration, derived for example from an inertial measurement unit. It has been verified that all these assumptions or considerations significantly reduce the complexity of calculation and modelling and have a completely negligible impact on modelling accuracy.

[0088] Imposing the reference acceleration aref in the above mathematical expression and resolving with respect to the term T wheel, tot, the following mathematical expression is obtained: where Twheei,FF is indicative of the feedforward action on wheel braking and, therefore, coincides with the signal T wheel, FF described above and generated as output of the FWTC module 22.

[0089] The CLWTC module 24 aims to minimise the error between the planned reference velocity vref and the measured vehicle velocity veg0; therefore, it implements the following control problem in the state of space (specifically, relating to a continuous-time, linear time-invariant dynamic system): min|vre / - vego\2

[0090] The state-space model in mathematical expressions (5) predicts two states, namely the vehicle velocity and the torque supplied to the wheel that is modelled here with a first- order transfer function with time constant 1 / T that is indicative of the actuator dynamics of the A module 14, considering a first-order transfer function with a unity gain pole (e.g., T is in the range of 5 to 10 rad / sec). In addition, Twheei CL\incorresponds to the system input, that is, the initial state of Twheei CL. This allows the possible dynamics of the braking system 14', given by its hydraulic part, to be properly considered and helps to stabilise the system. In addition, the system is guaranteed to be fully controllable since the rank of the matrix [B,AB] is equal to the number of states, and

[0091] B=©

[0092] It follows that an appropriate control command u that minimises the mathematical expressions (5) is: where K is the only degree of freedom in the mathematical expression (6) that defines the multiplicative coefficients Kvand KT, which can be calculated by solving a known Linear Quadric Regulator (LQR) problem. For example, the LQR problem aims to: asymptotically stabilise the closed-loop behaviour by requiring that the eigenvalues of the matrix (A-BK) have negative real parts, hence that (A-BK) is defined as negative; and minimise a quadratic cost function J in which the gain matrices Q and R of the LQR problem are appropriately selected to satisfy the desired performance (in particular, Q weighs the effect of the two states of speed and torque and, specifically, assigns a greater weight to speed as more precision in speed tracking is desired, while R acts generally on the aggressiveness of the control).

[0093] Nevertheless, as evident, other feedback loop control techniques can similarly be implemented by the CLWTC module 24. For example, other types of controllers such as the more conventional proportional-integral-derivative (PID) controllers or lead-lag compensators can be used.

[0094] The DWTC module 26, on the other hand, provides a correction to the implementation request for the braking system 14', with respect to the previously described contributions.

[0095] In fact, the actuators of the A module 14 provide useful feedback regarding the instantaneous torque delivered to the vehicle. In most cases, these are only estimates and not actual measurements made by means of dynamometric sensors, in order to reduce the overall cost. For example, the braking torque for the braking system 14' can be expressed as a function of the master cylinder pressure and, in general, the parameters that characterise the braking system 14', such as the size of the pad contact surfaces and others. In the case of the powertrain 14", on the other hand, there can be multiple solutions depending on the type. For example, in the case of an ICE engine, the estimated torque is normally given at the vehicle's flywheel, is calculated as a function of various parameters such as manifold pressure, engine rpm, and amount of fuel injected, and is particularly affected by uncertainty as a function of the vehicle's current load; the torque at the wheels is then determined by the transmission parameters such as final reduction and gear ratios; furthermore, in the case of torque converters, the amplification factor must be included, another element that can generally lead to uncertainty in estimating torque at the wheels. In the case of electric motors, on the other hand, the estimate is more reliable as the relationship between current and torque is certainly more linear and deterministic than in the case of an ICE engine; however, even in this case, uncertainties may arise, especially at low speeds where non-linearity effects are greater.

[0096] In order to account and correct for these possible uncertainties in this model, the mathematical expression (3) can be revised by adding a term T wheel, OBS.

[0097] To obtain the uncertainty torque term indicated by the signal T wheel, OBS, the DWTC module 26 implements an asymptotic filter.

[0098] Specifically, by integration of the mathematical expression (7), the asymptotic filter outputs an estimated velocity vest that is then compared with the measured vehicle velocity Vego and scaled by a gain P. In particular, the gain P is a predefined value that, for example, can be appropriately calibrated in a known way to minimise the error between estimated velocity vest and measured vehicle velocity veg0, in order to manage the stability and convergence of the asymptotic filter. For example, the gain P can be calculated using optimisation methods (e.g. based on kalman filters) that aim to solve the trade-offbetween dynamic measurement estimation performance and measurement stability. Merely by way of non-limiting example, the gain P can be between approximately 500 and about 2000.

[0099] In terms of the equation, the asymptotic filter can be implemented as described in the mathematical expressions (8a) and (8b):

[0100] Consequently, although the reliability of the estimate made by the modules 22 and 24 is affected by the chosen model and its parameters, as well as any additional uncertainties, this does not reduce the overall accuracy of the control on the braking system 14' as these uncertainties are corrected by the signal T wheel, OBS generated by the DWTC module 26. Therefore, the presence of the DWTC module 26 gives this closed- loop control system the ability to achieve a reference steady-state speed with zero steadystate error.

[0101] With regard to the control performed by the FPCC module 28, it is instead based on the creeping effect described in more detail above.

[0102] In particular, the creeping effect is exploited in the FPCC module 28 to ensure that the control implemented by the MC module 20 always operates in a zone where it is possible to act with the braking system 14' to regulate vehicle velocity. As described above, this provides a braking system 14' dynamic that allows for cleaner and more efficient control action. This is achieved by forcing the system so that the creeping, that is, the continuous dragging, is maintained under all conditions at low vehicle velocity. For example, in the case of significant road disturbances, such as positive road slopes, creeping alone would not be sufficient to keep the vehicle dragging, resulting in its stalling. Instead, here, through the signal torque control interface Tpwt,Req supplied to the powertrain 14", it is possible to shift the slipping point so that creeping is always guaranteed.

[0103] To do this, the feedforward control law defined, for example, by the following mathematical expression can be implemented: in which the signal Tpwt,Req is equal to the maximum between the slow torque supplied as feedback from the powertrain 14" and indicated by the signal T creep, Fdbk (this normally corresponds to the minimum friction torque required to maintain minimum creeping) and the torque request that takes into account the maximum acceleration request amax and the road slope 0 (that is, 0ego).

[0104] In other words, the FPCC module 28 generates the signal Tpwt,cnd that may be equal to M ■ R ■ (amax+ g ■ sin(0) + g ■ p ■ cos(0) ), so that the comparison module 32 selects the maximum between this signal and the signal Tcreep,Fdbk in order to generate the Signal Tpwt,Req.

[0105] It follows that the final braking request at the wheels, defined by the signal TBrake,Req, is equal to the sum / sub traction of all the various contributions described above, as better indicated in the following mathematical expression: wherein the term Tpwt dbk represents the estimated torque at the wheels that is provided as feedback from the powertrain 14" as a result of the request Tpwt,Req, as indicated in the mathematical expression (9), and which is used to counterbalance the effect of creeping in the control action.

[0106] Therefore, considering that the torque related to uncertainties (essentially associated with the creeping contribution due to Tpwt Fdbk) is offset with T wheel, OBS, the total theoretical torque delivered to the wheel is basically given by the sum of T wheel, FF and TWheei,CL according to the following mathematical expression:

[0107] Consequently, the MC module 20 implements in use a control method (not shown) for controlling the motion of the vehicle.

[0108] The steps of the control method are quite obvious in light of the previous description of the MC module 20 and are, therefore, not described in detail again here.

[0109] In addition, corresponding software (also not shown) for controlling vehicle motion is stored in the MC module 20 and is designed in such a way that, when executed, the MC module 20 is configured to perform this control method.

[0110] It was verified that the control system implemented via the MC module 20 allows optimal control of the vehicle, especially during low-speed manoeuvres.

[0111] To this end, Figures 4-6 refer to results obtained by implementing this control on a car with an internal combustion engine and torque converter transmission. Specifically, in order to interact with the vehicle's actuators, some modifications were made to the vehicle's electrical system to bypass the ADAS control unit. More specifically, the interfaces used are those for the Adaptive Cruise Control (ACC) system, which concern: an engine flywheel torque channel for the powertrain 14" and a deceleration channel for the braking system 14', which is also used for collision mitigation (AEB functions). The latter was used to modify the software of the brake control unit (BSM) to provide an additional brake torque and wheel interface more suitable for the present low-speed control manoeuvres.

[0112] Therefore, the control logic was implemented within a rapid prototyping control unit (e.g. dSPACE Micro Autobox II unit), placed in parallel with the standard production ADAS unit. The control logic was developed using a purely “model-based” approach through the Matlab-Simulink environment, which was then integrated into the prototyping control unit using the code generation and integration tools provided directly by dSPACE. The control run time was approximately 10ms. In addition, for validation and initial control tests, a model based on vehicle dynamics and actuators was developed in accordance with the known development practices relating to the model-based design approach.

[0113] The tests were performed by equipping the system with two different trajectory types. The first, shown in Figure 4. a, represents a synthetic reference obtained from constant acceleration steps with infinite jerk at ± 0.3 m / s2that in terms of velocity produces a trapezoidal profile with a stationary zone at 1 m / s. To ensure repeatability, the profile is run three times in a row. The second, in Figure 4.b, is a recorded trajectory generated by a real MPTD module during a parking manoeuvre and has a duration of 100s. In this case, it is worth remembering that the trajectory tracking by the MC module 20 was performed exclusively in an open loop, following the previously recorded trajectory within a fixed “timestamp”. This allows the manoeuvre to be repeated in several scenarios and the performance to be evaluated according to various control calibration parameters. In this regard, both trajectories in Figures 4. a and 4.b were run under three different road slope conditions: flat, uphill and downhill. This was done primarily to assess the soundness of the control with respect to road slope disturbances, which in the case of longitudinal dynamics, is one of the key elements of performance degradation.

[0114] As far as the main Key Performance Indicators (KPIs) are concerned, these can be classified according to objective and subjective indicators. In this case, objective ones have been considered, as they are more easily measurable. These refer to the canonical indicators for assessing the performance of a closed-loop system: stability, performance monitoring (in transient and steady state) and robustness.

[0115] Figure 5 shows the results obtained by plotting the first reference trajectory in the three different road slope scenarios: flat, downhill (-8%) and uphill (+8%). Each of these three sub-figures 5.a-5.c is structured as follows: the first subplot shows the acceleration reference used in the feedforward control; the second subplot shows the speed reference together with the actual measurement; the third subplot shows the propulsive and braking torques, together with the minimum torque required to maintain creeping; finally, the last subplot shows the percentage of road slope.

[0116] In terms of stability, the control performance in all three scenarios is very good, and this is due to the specific control strategy that ensures a high phase margin. The few oscillations present are mostly due to the unevenness of the road surface; these oscillations are, however, effectively reduced by the control action performed. Tracking performance is also remarkable during both transient and stationary phases. The overshoot is less than 5% and the rise and stabilisation time for the speed ramp profile is less than 0.1 s. Moreover, the steady-state error, barring measurement disturbances, can be considered zero in the steady-state region and less than 0.05 m / s near the ramp region. Similar conclusions can also be extended to the case of a negative or positive slope. No deviations with respect to the case defining the nominal system performance are evident. In other words, the control demonstrates robust and repeatable performance and this too can be attributed to the phase margin consideration made earlier.

[0117] The control action determined by the actuation commands of the powertrain 14" and the braking system 14' is also effective. In this regard, it is interesting to point out that, in the case of a negative slope, the torque required from the powertrain 14" matches the minimum creeping torque perfectly, and the braking torque controls attempt to balance the thrust given by gravity and the creeping forces at the same time. On the uphill slope, the demand on the powertrain 14” is greater, as it is necessary to move the creeping point higher up to allow proper modulation via the braking torque control. Even in the case of a flat road, the torque command of the powertrain 14" is slightly higher than the creeping torque, to reinforce the stability behaviour of the control (as the creeping zone is influenced by the variations given by the idle control of the internal combustion engine). Therefore, this facilitates the control balance within the braking system 14’ control and linearises the torque distribution behaviour despite energy consumption. Nevertheless, a parking manoeuvre is normally completed in a short time, so the impact in this case is completely negligible.

[0118] Figure 6 shows the results obtained from tracking the second trajectory under consideration. Similarly to what was seen before and with reference to Figure 6. a, the control behaves stably under all circumstances, even in the lowest speed range around 0.2 m / s (a significant result given the strong non-linearity of the powertrain 14” at such low speeds). Again, the visible oscillations can be attributed to the roughness of the road surface; these oscillations are effectively controlled even when there is less demand for speed. Moreover, the control reaches the different reference steady-state regions with zero error and without significant delays during the transient phases. The same conclusions also apply in the case of the descent in Figure 6.b and the ascent in Figure 6.c. Once again, the control behaviour proves to be robust and repeatable.

[0119] The advantages enabled by the discovery according to this invention will be apparent from an examination of the features thereof. In particular, the MC module 20 enables an innovative approach to vehicle actuation control that aims to manage low-speed parking manoeuvres. This is possible by optimally managing the various actuation requirements, both on the powertrain 14" side and on the braking system 14' side, while exploiting the effect of creeping.

[0120] In addition, the control performs well and its behaviour is stable and robust even with respect to significant road disturbances.

[0121] Lastly, it is clear that modifications and variations may be made to the invention described and illustrated herein without departing from the scope of this invention, as set forth in the claims. For example, the different embodiments described can be combined to provide additional solutions.

Claims

CLAIMS1. A motion control, MC, module (20) for controlling the movement of a vehicle (40), the MC module (20) being couplable to an actuation module, A, (14) of the vehicle (40), comprising a braking system (14’) and a powertrain (14”) of the vehicle (40), the MC module (20) being configured to receive as input ego -motion, EM, control data (DEM,C) indicative of measured parameters of the vehicle (40) and a road to be travelled by the vehicle (40), and actuation feedback data (DAjdbk) generated by the module A (14), wherein the EM control data (DEM,C) comprise at least a road slope (Oego) and a measured vehicle velocity (veg0) and the actuation feedback data (DAjdbk) being correlated with a feedback wheel torque signal (Twheei,Fdbk) indicative of an estimated torque value provided by the braking system (14’) and powertrain (14”), the MC module (20) comprising a disturbance wheel torque observer compensator, DWTOC, module (26) configured to receive as input EM control data (DEM,C) and actuation feedback data (DAjdbk) and, as a function of such inputs, to output a corresponding observer wheel torque signal (Twheei,OBs) indicative of a torque contribution processed by an observer to correct actual drifts in the powertrain (14"), the MC module (20) being also configured to generate as output, as a function of the observer wheel torque signal (Twheei.oiis)- a brake request torque signal (TBrake,Req) indicative of a torque command for the actuation of the braking system (14') in order to control the movement of the vehicle (40).

2. The MC module according to claim 1, wherein the observer of the DWTOC module (26) is configured to implement an asymptotic filter configured to compare EM control data (DEM,C) and implementation feedback data (DA dbk) with each other.

3. The MC module according to claim 2, wherein the asymptotic filter is configured to determine an estimated velocity (vest) from the EM control data DEM,C), compare the estimated velocity (vest) with the measured velocity of the vehicle (veg0) and scale the result of said comparison by a gain factor (P) of a predefined type.

4. The MC module according to claim 3, wherein the gain factor (P) is determined in such a way as to minimise an error between the estimated velocity (vest) and the measured velocity of the vehicle (veg0).

5. The MC module according to claim 3 or 4, wherein the asymptotic filter is defined by the mathematical expressions:where M is a mass of the vehicle (40), g is the gravitational constant, R is a tyre radius of the vehicle (40), 9 is said road slope (0ego), p is a rolling resistance of the vehicle, T wheel, OBS is the observer wheel torque signal (T wheel, OBS), Twheei,Fdbk is the feedback wheel torque signal (T wheel, Fdbk), P is the gain factor (P), veg0is said measured velocity of the vehicle (veg0) and vest is said estimated velocity (vest).

6. The MC module according to any one of the preceding claims, further comprising a first summing module (30) coupled to the DWTC module (26) and configured to receive the observer wheel torque signal (Twheei,OBs), a feedforward wheel torque signal (T wheel, FF), a closed-loop wheel torque signal (T wheel, CL) and a feedback powertrain torque signal (Tpwt,Fdbk) and, as a function of these inputs, generate the corresponding brake request torque signal (TBrake,Req) as output, wherein the feedforward wheel torque signal (T wheel, FF) is indicative of a feedforward torque contribution determined using a reference acceleration (aref) of the vehicle (40), the closed-loop wheel torque signal (Twheei,ct) is indicative of a closed-loop torque contribution determined using a reference velocity (vrcf) of the vehicle (40), and the feedback powertrain torque signal (Tpwt,Fdbk) is indicative of a measured feedback torque provided by the powertrain (14").

7. The MC module according to claim 6, wherein the first summing module (30) is configured to sum the feedforward wheel torque signal (T wheel, FF), the closed loop wheel torque signal (Twheetc ) and the observer wheel torque signal (T wheel, OBS) and subtract the feedback powertrain torque signal (Tpwt dbk) in order to generate the brake request torque signal (TBrake,Req)-8. The MC module according to any one of the preceding claims, wherein the actuation feedback data (DA, Fdbk) comprise a brake feedback torque signal (TBrake,Fdbk) and a feedback powertrain torque signal (Tpwt,Fdbk), the MC module (20) further comprising a second summing module (34) coupled to the DWTC module (26) and configured to receive the brake feedback torque signal (TBrake,Fdbk) from the braking system (14') and the feedback powertrain torque signal (Tp„t.i dbk) from the powertrain (14") and, as a function of these inputs, generate the corresponding wheel feedback torque signal (Twheel,Fdbk) aS Output, wherein the brake feedback torque signal (TBrake,Fdbk) is indicative of a measuredvalue of feedback torque of the braking system (14') and the feedback powertrain torque signal (Tpwt dbk) is indicative of a measured value of feedback torque of the powertrain (14'), and wherein the wheel feedback torque signal (TwheetFdbk) depends on the sum of the brake feedback torque signal (TBrake,Fdbk) and the feedback powertrain torque signal (Tpwt,Fdbk).

9. A control system (10) for controlling a vehicle (40), the control system (10) comprising a motion control, MC, module (20), according to any one of the preceding claims, and said actuation module, A, (14), wherein the braking system (14') of the module A (14) is configured to be controlled via the brake request torque signal (TBrake,Req) generated by the MC module (20).

10. A control method for controlling the movement of a vehicle (40), executable by a motion control, MC, module, (20) coupled to an actuator module, A, (14) of the vehicle (40), which comprises a braking system (14') and a powertrain (14") of the vehicle (40), the MC module (20) being configured to receive as input ego -motion, EM, control data (DEM,C) indicative of measured parameters of the vehicle (40) and a road to be travelled by the vehicle (40), and actuation feedback data (DA,Fdbk) generated by the module A (14), wherein the EM control data (DEM,C) comprise at least a road slope (Oego) and a measured vehicle velocity (veg0) and the actuation feedback data (DA,Fdbk) being correlated with a feedback wheel torque signal (Twheei,Fdbk) indicative of an estimated torque value provided by the braking system (14’) and powertrain (14”), the control method comprising the steps of:- by a disturbance wheel torque observer compensator DWTOC, module, (26) of the MC module (20), receiving as input EM control data (DEM,C) and actuation feedback data (DAjdbk) and, as a function of said inputs, generating as output a corresponding observer wheel torque signal (Twheei,OBs) indicative of a torque contribution processed by an observer to correct actual torque drifts in the powertrain (14”); and- generating as output, as a function of the observer wheel torque signal (T wheel, OBS), a corresponding brake request torque signal (TBrake,Req) indicative of a torque command for actuating the braking system (14') in order to control the movement of the vehicle (40).

Images