METHOD FOR CONTROLLING A DRIVING ROBOT SIMULATING AGGRESSIVE DRIVING
The piloting process for a driving robot simulates aggressive driving maneuvers by defining zones and calculating dynamic speed setpoints, addressing the inability of existing robots to validate energy optimization in motorcycle groups, achieving compliant and adaptable testing.
Patent Information
- Application Number
- FR2023002158
- Authority / Receiving Office
- FR · FR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2023-03-08
- Publication Date
- 2025-01-31
- Estimated Expiration
- 2043-03-08
AI Technical Summary
Existing driving robots cannot automatically simulate aggressive driving maneuvers to validate the robustness of energy optimization calibration in motorcycle groups, failing to meet the requirements of aggressive driving tests.
A piloting process for a driving robot that includes loading gross speed instructions with maximum and minimum margins, determining acceleration, deceleration, and dead zones, defining a slippery function, and calculating dynamic speed setpoints to simulate aggressive driving, ensuring compliance with test requirements and energy optimization validation.
Enables autonomous, repeatable testing of motorcycle groups by simulating aggressive driving, ensuring quality monitoring and energy optimization validation without exceeding test tolerances, adaptable to various motorcycles and initial speeds.
Smart Images
Figure 00000010_0000 
Figure 00000010_0001 
Figure 00000011_0000
Abstract
Description
Title of the invention: METHOD FOR CONTROLLING A DRIVING ROBOT SIMULATING AGGRESSIVE DRIVING
[0001] The present invention relates to a method of piloting a driving robot simulating aggressive driving.
[0002] The invention finds a particularly advantageous application for the development of powertrains for internal combustion gasoline vehicles and / or hybrid type powertrains.
[0003] In a manner known per se, a driving robot used on a powertrain test bench is capable of autonomously and automatically monitoring the homologation cycles of a powertrain, such as a worldwide harmonized test procedure for light vehicles known as the WLTP cycle (for "Worldwide Light Vehicles Test Procedures" according to Anglo-Saxon terminology).
[0004] To this end, the driving robot includes a control unit and actuators capable of mechanically controlling the brake pedal, the accelerator pedal, and the gear shift lever of the powertrain. The driving robot can thus control the acceleration and deceleration of the powertrain in order to follow a set of speed commands.
[0005] However, existing driving robots are not capable of automatically simulating aggressive driving that maximizes acceleration and deceleration gradients in order to validate the robustness of an energy optimization calibration of a powertrain.
[0006] The invention aims to effectively remedy this drawback by proposing a method for controlling a driving robot for a motor vehicle powertrain comprising: - a step involving loading a template of raw speed instructions based on the time associated with a maximum gross margin and a minimum gross margin to be respected, - a step of calculating a new maximum margin lower than the maximum gross constraint margin and a new minimum margin higher than the minimum gross constraint margin, referred to respectively as "maximum constraint margin" and "minimum constraint margin", - a determination step, based on the template of raw speed instructions, of acceleration zones, deceleration zones, and dead zones, - a step of defining a sliding function that has an increasing segment for each acceleration zone, a decreasing segment for each deceleration zone, and a constant segment for each dead zone, and - a step of calculating a final dynamic speed setpoint as a function of the maximum gross stress margin, the minimum gross stress margin, the minimum stress margin, the maximum stress margin, and the sliding function.
[0007] The invention thus makes it possible to autonomously monitor speed commands simulating aggressive driving in order to validate the robustness of a powertrain energy optimization calibration by generating repeatable tests. The invention also provides a single setting, regardless of the powertrain being controlled and the initial speed command parameters, to generate a speed command simulating the aggressive driving of a driver. The invention makes it possible to obtain a quality of command monitoring compatible with the test requirements while avoiding exceeding the test tolerances.
[0008] According to one embodiment of the invention, the following are considered respectively: - an acceleration zone if the difference between a sample of rank n+1 and a sample of rank n of the raw speed setpoint template is greater than a threshold, - a deceleration zone if the difference between a sample of rank n+1 and a sample of rank n of the raw speed setpoint template is less than the inverse of the threshold, - a dead zone if an absolute value of a difference between a sample of rank n+1 and a sample of rank n is less than the threshold.
[0009] According to one embodiment of the invention, said method further comprises a filtering step consisting of removing dead zones surrounded by two identical zones consisting of two acceleration zones or two deceleration zones.
[0010] According to one embodiment of the invention, the sliding function is defined such that it is equal to: - 0 at the beginning of an acceleration zone, - 1 at the end of an acceleration zone, and - in a dead zone, a constant value equal to the value of the sliding function at the end of a previous zone.
[0011] According to one embodiment of the invention, said method comprises a step of calculating a raw dynamic speed setpoint from the following formula: Vdb = Mgbx_max * fg + (Mgbx_min * (l - fg)) - Vdb being the raw dynamic velocity setpoint, - Mgbx_max being the maximum gross constraint margin of a sample x, - Mgbx_min being the minimum gross constraint margin of a sample x, - fg being the sliding function.
[0012] According to one embodiment of the invention, the final dynamic speed setpoint is obtained by limiting the gross dynamic speed setpoint by the minimum constraint margin and the maximum constraint margin according to the following formula: Vd = max([min( [Vdb, Mgcx_max]), Mgcx_min]), - Vd being the final dynamic velocity setpoint, - Vdb being the raw dynamic velocity setpoint, - Mgcx_max being the maximum constrained margin of a sample x, - Mgcx_min being the minimum constrained margin of a sample x.
[0013] According to one embodiment of the invention, in dead zones, the final dynamic velocity retains its previous value.
[0014] According to one embodiment of the invention, during phases of stopping of the motor vehicle, the final dynamic speed is zero.
[0015] According to one embodiment of the invention, a transfer function applied to the maximum and minimum gross margins to calculate the maximum and minimum constrained margins is a parameterizable offset defined by a constraint value.
[0016] According to one embodiment of the invention, the maximum constrained margin is equal to the maximum gross margin less the constraint value and the minimum constrained margin is equal to the minimum gross margin plus the constraint value.
[0017] The invention will be better understood upon reading the following description and examining the accompanying figures. These figures are given only to illustrate, but in no way limit, the invention.
[0018] [Fig-1] schematically represents a powertrain test bench in which is installed a driving robot piloted according to the method according to the invention;
[0019] [Fig.2] is a graphical representation of a raw speed instruction template associated with maximum and minimum margins;
[0020] [Fig.3] illustrates the definition of the maximum and minimum constrained margins during the implementation of the process according to the invention;
[0021] [Fig.4] shows a template of raw speed setpoints as a function of time. to show acceleration zones, deceleration zones and dead zones, as well as a corresponding acceleration curve on which a velocity gradient threshold is represented;
[0022] [Fig.5] is a graphical representation illustrating the detection of the zones acceleration, deceleration zones and dead zones on a portion of a template of raw speed instructions;
[0023] [Fig.6] illustrates the removal of dead zones from [Fig.4] surrounded by two zones identical acceleration or deceleration;
[0024] [Fig.7] illustrates the development of a sliding function that varies according to the area acceleration zone, deceleration zone or dead zone;
[0025] [Fig.8] is a representation, on the same graph, of a template for instructions raw speed, maximum and minimum initial margins, constrained margins minimum and maximum, and dynamic speed commands obtained during the implementation of the process according to the invention during an acceleration phase;
[0026] [Fig.9] is a representation, on the same graph, of a template for instructions raw speeds, maximum and minimum initial margins, minimum and maximum constrained margins, and dynamic speed setpoints obtained during the implementation of the process according to the invention during a deceleration phase.
[0027] Figure 1 shows a test bench 10 of a powertrain 11 comprising a driving robot 12 controlled according to the method of the invention. The driving robot 12 comprises a control unit and actuators capable of mechanically controlling the brake pedal, the accelerator pedal, and the gear shift lever of the powertrain.
[0028] As explained below, the driving robot 12 is capable of developing a speed command from a template of raw speed commands which it will pass on to the powertrain 11 by imposing an acceleration Acc via the control of the brake pedal, the accelerator pedal and the gear shift lever.
[0029] The driving robot 12 communicates via a module 13 with a simulation environment 14, for example, of the dSpace-Simulink type (trade name). The module 13 manages, via the environment 14, a deceleration Dec_EE linked to a phase of electrical energy recovery. A drive system for the powertrain 15 manages a deceleration Dec_p linked to the road gradient.
[0030] The method of piloting the driving robot 12 according to the invention is described below with reference to figures 2 to 9.
[0031] A first step consists of loading a template of raw speed commands Cb as a function of time associated with a maximum gross margin Mgb_max and a minimum gross margin Mgb_min to be respected, as illustrated in [Fig. 2]. The template of raw speed commands Cb and the gross margins Mgb_max, Mgb_min can be stored in a memory of the control unit of the driving robot 12. The loading consists of transferring data representative of the template of raw speed commands Cb to this memory.
[0032] The raw speed setpoint template Cb is sampled according to a sampling period T_ech. The sampling period T_ech is for example 0.1s.
[0033] A second step consists of calculating a new maximum margin Mgc_max lower than the maximum gross stress margin Mgb_max and a new minimum margin Mgc_min higher than the minimum gross stress margin Mgb_min, referred to respectively as the "maximum stress margin" and "minimum stress margin", as illustrated in [Fig. 3]. The minimum stress margin Mgc_min and maximum stress margin Mgc_max are parameterizable so as to adapt the level of aggressive driving of the driver.
[0034] A transfer function Fct is applied to the gross margins Mgb_min, Mgb_max to calculate the constrained margins, i.e., Mgcx = Fct(Mgbx). Preferably, this transfer function Fct is a parameterizable offset defined by a constraint value Ctr.
[0035] The maximum constrained margin Mgc_max is equal to the maximum gross margin Mgb_max less the constraint value Ctr, i.e., Mgc_max = Mgb_max - Ctr. The minimum constrained margin Mgc_min is equal to the minimum gross margin Mgb_min plus the constraint value Ctr, i.e., Mgc_min = Mgb_min + Ctr.
[0036] A third step consists of determining the zones where the raw setpoint Cb is associated with an acceleration phase, a deceleration phase, or a dead zone. This calculation involves a velocity gradient Cbgrd to determine the type of zone (acceleration, deceleration, dead zone).
[0037] Following this step, a difference is compared between a sample of rank n+1 and a sample of rank n of the raw speed setpoint template Cb with a threshold Cb_grd.
[0038] We thus consider: - an acceleration zone Zacc if a difference between a sample of rank n+1 and a sample of rank n of the raw speed setpoint template Cb is greater than a threshold Cb_grd, i.e. Cb(n+l)-Cb(n)>Cbgrd - a deceleration zone Zdec if a difference between a sample of rank n+1 and a sample of rank n of the raw speed setpoint template Cb is less than the inverse of the threshold Cb_grd, i.e. Cb(n+1)-Cb(n)<-Cb_grd; - a dead zone Zm if an absolute value of a difference between a sample of rank n+1 and a sample of rank n is less than the threshold Cb_grd, i.e. ICb(n+l)-Cb(n)l<=Cbgrd.
[0039] Figure 4 illustrates the implementation of the Cbgrd threshold with respect to the Acc acceleration curve representing the differences between velocity samples divided by the time interval between these samples. Figure 5 illustrates the local detection of the different acceleration zones Zacc, Zdec, and Zm of the raw velocity setpoint template. The upper curve of Figure 4 illustrates an overall result of the detection of the different zones Zacc, Zdec, and Zm.
[0040] A fourth step consists of filtering the result obtained in the third step to remove dead zones Zm surrounded by two identical zones. Thus, a dead zone Zm located between two acceleration zones Zacc or a dead zone Zm located between two deceleration zones Zdec is removed. [Fig. 6] shows that the dead zones Zm have been removed at the arrows Fl, F2 and F3 compared to the initial detection of the different zones shown on the upper curve of [Fig. 4].
[0041] A fifth step consists of defining a sliding function fg which has an increasing segment Ter for each acceleration zone Zacc, a decreasing segment Tdec for each deceleration zone Zdec, and a constant segment Tco for each dead zone Zm. Preferably, the increasing segment Ter is defined by an increasing affine function, and the decreasing segment Tdec is defined by a decreasing affine function.
[0042] The sliding function fg is defined such that it is equal to: - 0 at the beginning of an acceleration zone Zacc, i.e. fg(start Zacc n°x)=0; - 1 at the end of an acceleration zone Zacc, i.e. fg(end Zacc n°x)=l; - in a dead zone Zm, a constant value equal to the value of the sliding function at the end of the previous zone, i.e. fg(Zm) = fg (Previous zone).
[0043] Figure 6 illustrates the definition of the sliding function as a function of the acceleration zones Zacc, the deceleration zones Zdec, and the dead zones Zm.
[0044] A sixth step consists of calculating a final dynamic speed setpoint Vd as a function of the maximum gross stress margin Mgb_max, the minimum gross stress margin Mgb_min, the minimum stress margin Mgc_min, the maximum stress margin Mgc_max, and the sliding function fg.
[0045] To this end, a raw dynamic velocity setpoint Vdb is first calculated using the following formula: Vdb= Mgbx_max*fg + (Mgbx_min*(l-fg)) - Vdb being the raw dynamic velocity setpoint, - Mgbx_max being the maximum gross constraint margin of a sample x, - Mgbx_min being the minimum gross constraint margin of a sample x, - fg being the sliding function.
[0046] The final dynamic speed setpoint (Vd) is obtained by limiting the gross dynamic speed setpoint (Vdb) by the minimum constraint margin (Mgc_min) and the maximum constraint margin (Mgc_max) according to the following formula: Vd = max([min( [Vdb, Mgcx_max]), Mgcx_min]), - Vd being the final dynamic velocity setpoint, - Vdb being the raw dynamic velocity setpoint, - Mgcx_max being the maximum constrained margin of a sample x, - Mgcx_min being the minimum constrained margin of a sample x.
[0047] In the dead zones Zm, the final dynamic velocity Vd retains its previous value.
[0048] During phases of stopping of the motor vehicle, the final dynamic speed Vd is zero.
Claims
Claims
1. Method for controlling a robot for driving a motor vehicle powertrain, characterized in that said method comprises: - a step of loading a template of raw speed instructions (Cb) as a function of time associated with a maximum gross margin (Mgb_max) and a minimum gross margin (Mgb_min) to be respected, - a step of calculating a new maximum margin (Mgc_max) less than the maximum gross constraint margin (Mgb_max) and a new minimum margin (Mgc_min) greater than the minimum gross constraint margin (Mgb_min) called respectively "maximum constraint margin" and "minimum constraint margin", - a step of determining, on the template of raw speed instructions (Cb), acceleration zones, deceleration zones, and dead zones (Zm), - a step of defining a sliding function (fg) which has an increasing section for each acceleration zone (Zacc),a decreasing section for each deceleration zone (Zdec), and a constant section for each dead zone (Zm), and - a step of calculating a final dynamic speed setpoint (Vd) as a function of the maximum gross stress margin (Mgb_max), the minimum gross stress margin (Mgb_min), the minimum stress margin (Mgc_min), the maximum stress margin (Mgc_max), and the sliding function (fg).,
2. Method according to claim 1, characterized in that the following are considered respectively: - an acceleration zone (Zacc) if a difference between a sample of rank n+1 and a sample of rank n of the raw speed setpoint template (Cb) is greater than a threshold (Cb_grd), - a deceleration zone (Zdec) if a difference between a sample of rank n+1 and a sample of rank n of the raw speed setpoint template (Cb) is less than the inverse of the threshold (Cb_grd), - a dead zone (Zm) if an absolute value of a difference between a sample of rank n+1 and a sample of rank n is less than the threshold (Cb_grd).
3. Method according to claim 1 or 2, characterized in that it further comprises a filtering step consisting of removing the dead zones (Zm) surrounded by two identical zones consisting of two zones acceleration (Zacc) or two deceleration zones (Zdec).
4. Method according to any one of claims 1 to 3, characterized in that the sliding function (fg) is defined in such a way that it is equal to: - 0 at the start of an acceleration zone (Zacc), - 1 at the end of an acceleration zone (Zacc), and - in a dead zone (Zm), a constant value equal to the value of the sliding function at the end of a previous zone.
5. Method according to any one of claims 1 to 4, characterized in that it comprises a step of calculating a raw dynamic speed setpoint (Vdb) from the following formula: Vdb = Mgbx_max * fg + (Mgbx_min * (l-fg)) - Vdb being the raw dynamic speed setpoint, - Mgbx_max being the maximum raw stress margin of a sample x, - Mgbx_min being the minimum raw stress margin of a sample x, - fg being the sliding function.
6. Method according to claim 5, characterized in that the final dynamic speed setpoint (Vd) is obtained by limiting the raw dynamic speed setpoint (Vdb) by the minimum constrained margin (Mgc_min) and the maximum constrained margin (Mgc_max) according to the following formula: Vd = max([min( [Vdb, Mgcx_max]), Mgcx_min]), - Vd being the final dynamic speed setpoint, - Vdb being the raw dynamic speed setpoint, - Mgcx_max being the maximum constrained margin of a sample x, - Mgbx_min being the minimum constrained margin of a sample x.
7. Method according to any one of claims 1 to 6, characterized in that in the dead zones (Zm), the final dynamic speed (Vd) retains its previous value.
8. Method according to claim 6 or 7, characterized in that, during stopping phases of the motor vehicle, the final dynamic speed (Vd) is zero.
9. Method according to any one of claims 1 to 8, characterized in that a transfer function applied to the maximum and minimum gross margins (Mgb_max, Mgb_min) to calculate the maximum and minimum constrained margins (Mgc_max, Mgc_min) is a parameterizable offset defined by a constraint value (Ctr).
10. Method according to claim 9, characterized in that the maximum constrained margin (Mgc_max) is equal to the maximum gross margin (Mgb_max) minus the constraint value (Ctr) and the minimum constrained margin (Mgc_min) is equal to the minimum gross margin (Mgb_min) plus the constraint value (Ctr).