WO2014174181A1

WO2014174181A1 - Method for controlling the powershifting of an automatic transmission

Info

Publication number: WO2014174181A1
Application number: PCT/FR2014/050885
Authority: WO
Inventors: Grégory FLIPPE; Antoine Jean
Original assignee: Renault S.A.S
Priority date: 2013-04-23
Filing date: 2014-04-11
Publication date: 2014-10-30
Also published as: FR3004686A1; FR3004686B1

Abstract

A method for controlling the powershifting of an automatic transmission obtained by gradually switching the torque between two clutches each receiving a torque value to be transmitted to the wheels during shifting, characterised in that the actual result of these values is compared, during each shift, with predefined target trajectories, and in that these trajectories are optionally modified in anticipation of subsequent gear shifting on the basis of the observed variations, so as to bring the production of the torque values closer in time to the targets of same, during subsequent shifting carried out in similar conditions.

Description

METHOD FOR CHECKING THE TORQUE SPEEDS OF AN AUTOMATIC TRANSMISSION

The present invention relates to the control of automatic or automated transmissions.

It relates to a method of controlling the automatic torque transmission of an automatic transmission, obtained by progressive torque swing between two clutches each receiving a torque setpoint to be transmitted to the wheels during the passage.

This invention finds its application on any automatic transmission, or automated, where the gear changes are made by opening a first coupling means, and closing a second coupling means.

This is the case of automatic transmissions epicyclic train, in which two couplers (clutch or brake) are closed on each report, and shifts are made by opening one of the two couplers to replace it gradually with a third.

This is also the case of the robotised gearboxes, dual input clutch (DCT for "Dual Clutch Transmission"), in which each shift is the result of a rocker torque between the two clutches.

In these two applications, the gear changes without breaking torque, require simultaneous control of the opening of one of the clutches and the closing of the other clutch. The quality of the passages depends on the precision of the control of the clutches and the good realization of their torque setpoint. It is characterized by the time of change of ratio and by the jolts felt by the driver.

The direct measurement of the torque transmitted by each clutch at every moment, is difficult to achieve because of the size of the sensors to be implanted. The solution usually used to control the effort applied on the clutch discs, is to position a sensor on the clutch actuation system, for example a pressure sensor for the hydraulic actuation systems. The problem is then that the sum of the processing times of the information delivered by the sensor (for example filtering, analog / digital conversion, ...), the time of calculation of new setpoints, and the reaction time of the system of clutch actuation, is very important compared to the shift time. It is therefore necessary to use sensors and a very fast calculator, to correctly adjust the applied force.

A second aspect of the perceived quality of the gearshifts is their repeatability, because the driver is all the more satisfied, all the passages take place under the same conditions of comfort. For this, it is important that the instructions sent to the actuation systems are respected as faithfully as possible.

EP 0 414 547 discloses a method for learning the duration of the gear changes of an automatic transmission. This method only concerns the learning of the duration of the passage, on hydraulically controlled transmissions.

The present invention aims to perform, repeatedly, quality shifts, learning torque setpoints transmitted during the passage of time on most automatic or automated torque rocker transmissions.

For this purpose, it proposes that the achievement of the instructions is compared during each passage with predefined target trajectories, and that these trajectories are possibly modified in anticipation of subsequent shifts in accordance with the observed differences, so as to approximate in the time, the achievement of the target torque of their targets, during subsequent passages made under similar conditions.

Preferably, the trajectories are not modified according to the torque measurements made during the passage.

The present invention will be better understood on reading the following description of a nonlimiting embodiment thereof, with reference to the appended drawings, in which:

FIG. 1 illustrates the realization of the torque flip-flop during a passage,

FIG. 2 shows the effect of a difference in torque with the setpoint,

Figure 3 illustrates the result of the method, and Figure 4 shows the main steps. The automatic or dual-clutch transmissions perform the gearshift by torque rocker, that is to say by progressively transferring the torque delivered by the motor, a first clutch on a second clutch.

FIG. 1 illustrates the control of the under-torque shifts of an automatic transmission obtained by progressive rocking of torque between two clutches each receiving a torque set point C1, C2 to be transmitted to the wheels during the passage. The torque is transferred from a first clutch to a second clutch. Each clutch transmits C1 and C2 torque to the wheels, through a gear ratio RI and R2 (gear ratio obtained by an arrangement of the gears located in the downstream part of the transmission). The torque supplied to the wheels is the sum of the torques transmitted by the two clutches divided respectively by the transmission ratios of each downstream part of the transmission.

"Setpoint" and "setpoint2" torque instructions [in dashed lines on the diagram] are sent to the actuating systems at dates t ₀ i and t ₀ 2- Each actuating system has its own response time, ti and t ₂ : delay between the setpoint (recordal curves and setpoint2) its realization (curves C1, C2). A sensor placed on each actuating system, sends to the computer of the transmission information, which is a image of the torque transmitted by the clutch actuated. The processing times of this information by the sensor and the computer are t 'i and t' ₂ : delay between the realization of the setpoint and its measurement (curves Cmesl and Cmes2).

A gearshift is "of good quality", if the torque to the wheels does not undergo a sudden change during the transition. However, the computer is informed of the actual achievement of the torque setpoints on the clutches, that dates t ₀ i + ti + t 'i and to2 + t ₂ + t' ₂ . As a result, any correction made in real time to one or other of the instructions, requires a double processing time. It is effective only at dates t ₀ i + 2x (t i + t 'i) and t ₀ 2 + 2x (t ₂ + t' ₂ ) · During this time, the torque at the wheels does not have the value desired, and the quality of passage is degraded.

In order not to be penalized by the information processing delay, the method provides that the corrections are not made to the torque setpoints during shifting. But the information from the sensors are recorded and compared at the end of the passage with the signals sent by the sensors. The calibrations can then be updated, so that during the following passages made under similar conditions, the trajectories of the pairs C1 and C2 converge progressively towards the desired trajectories.

Figure 2 shows the impact of a difference in torque with the setpoint, on the torque to the wheels. On the graph, the Cl couple falls faster than the target (curve C1 below the "target 1" curve). The C2 pair increases more slowly than the target (curve C2 below the "target 2" curve). It is also observed that the torque to the wheels is disturbed at the beginning and end of the passage. The calculator analyzes these shifts, and modifies the calibrations. During the following passage in similar conditions, the couples C1 and C2 are close to the targets 1 and 2, as in Figure 3. There is then a torque curve to the wheels more progressive pace, and a better quality of passage . The same principle can be applied to more complex variations between the target curve and the actual transmitted torque. Without departing from the scope of the invention, it is possible to apply to the setpoint, other evolutions than a simple time shift, within the limits of the dynamics of the actuating system and the clutch.

The general method, as illustrated in Figure 4, can be summarized as follows. At each start (or order) of change of gear [initialisation], the calculation unit of the powertrain concerned (engine and / or transmission computer) analyzes its operating conditions (engine speed, torque, gear engaged, etc.) [Step 1]. The set of predefined trajectories to cover all possible operating conditions can be grouped together in a multi-dimensional matrix called calibration.

In step 2, the computer searches in calibration tables, instructions and targets corresponding to the operating conditions identified. These instructions are sent to the actuating systems (clutches 1 and 2) [Step 3A]. In parallel [Step 3B], the information from the sensors on the actuating systems is recorded. The instructions and information received are compared [Step 3B]. If there are no differences [Step 4], the instructions are maintained for the following similar passages. Otherwise [Step 5], the setpoints of the actuators are changed in the calibration. As indicated above, the trajectories are not modified as a function of the torque measurements made during the passage. The beginning (t ₀ i, t ₀ 2> of the opening or closing of the clutches can be delayed or anticipated, to approach the predefined trajectory.

Torques transmitted by the clutches during shifts are measured by sensors associated with their actuation systems. The implantation of a torque sensor on the clutches may present particular difficulties. Therefore, sensors associated with their actuating systems are preferably used. Such sensors give an image of the actual torque applied, by measuring another magnitude representing the state of the actuating system during gearshifting, actuator displacement, actuation force, etc.

In summary, the achievement of the torque setpoints is compared during each run, with predefined target trajectories stored in the calculator. These trajectories are possibly modified in anticipation of subsequent shifts, depending on the differences noted.

With this learning, the quality of passage converges to a stable situation, because the torque in each clutch approximates the target curves at each pass. Indeed, the learning advantageously takes into account not only the last pass made, but also a succession of shifts made in the same conditions, to adjust the setpoint and progress each time the quality thereof. The behavior of the actuation system becomes predictable as its response to the setpoints stabilizes. The quality of passage reaches the expected repeatability. Moreover, any slow evolution of the actuation system (increase of the wear, variation of the friction, etc.) is automatically taken into account by the learning, to be compensated.

Claims

RE SELL ICAI ON S

1. A method for controlling the under-torque transitions of an automatic transmission obtained by progressive torque flip-flop between two clutches each receiving a torque setpoint to be transmitted to the wheels during the passage, according to which the performance of these instructions is compared during each passage with predefined target trajectories, and these trajectories are possibly modified in anticipation of subsequent shifts as a function of the observed discrepancies, so as to bring the achievement of the torque setpoints of their targets closer together during subsequent passes made in similar conditions, characterized in that the beginning (t ₀ i, t ₀ 2> of the opening or closing of the clutches can be delayed or anticipated to approach the predefined path.

2. Control method according to claim 1, characterized in that the couples transmitted by the clutches during the speed passes are measured by sensors associated with their actuating systems.

3. Control method according to claim 1 or 2, characterized in that, at each beginning of the passage, the calculation unit concerned of the powertrain analyzes the operating conditions of the passage to search in calibration tables, the trajectories corresponding to the operating conditions identified, and send the corresponding instructions to the operating systems of the two clutches.

4. Control method according to claim 3, characterized in that the information from the sensors associated with these actuating systems, are recorded at each pass to be compared to the instructions.

5. Control method according to claim 4, characterized in that the instructions are maintained for the following similar passages if there are no differences between the received information and it, while in the opposite case, they are modified in the calibration.

6. Control method according to one of the preceding claims, characterized in that the learning takes into account a succession of shifts made under the same conditions, to adjust the setpoint.