WO2024068272A1

WO2024068272A1 - System and method for calculating an adjusted current pulse for controlling an electromechanical actuator system

Info

Publication number: WO2024068272A1
Application number: PCT/EP2023/075094
Authority: WO
Inventors: Oliver Essbuechl-Srb
Original assignee: Magna powertrain gmbh & co kg
Priority date: 2022-09-28
Filing date: 2023-09-13
Publication date: 2024-04-04

Abstract

Method for controlling an electromechanical actuator system, wherein the actuator system serves to mechanically couple and uncouple two elements with an electric motor as the actuator, wherein the actuator is operatively connectedly coupled to a stepdown mechanism which provides a rotation of a stepdown gear up to 360° as a drive for an adjusting movement of an actuating element and further an adjusting movement of one of the elements to be coupled, wherein a differential rotation speed (setpoint rotation speed in relation to actual rotation speed of the elements to be coupled), the desired adjustment travel to a target position and the maximum available time for achieving the desired adjustment travel are used as the input variables to a control and evaluation unit, and wherein, depending on the input variables and an energy relationship which is determined by the actuator, a current pulse that is adequate in terms of inertia and rotation speed is calculated and wherein the adjustment travel is quickly covered by applying this calculated current pulse to the actuator and wherein the final adjustment to the target position of the actuating element is performed by means of a PI controller.

Description

System and method for calculating an adapted current pulse for

Control of an electromechanical actuator system

The invention relates to a system and a method for calculating an adapted current pulse for controlling an electromechanical actuator system.

Actuators form an essential part of electronic control systems in vehicles. Their task is to convert the electrical signals from the control unit into an action, for example an adjustment movement - linear or rotational.

Electromechanical actuators are used, for example, in the drive train of motor vehicles.

The actuator can be used in combination with corresponding sensors for clutch actuation of a claw clutch of a manual transmission, for actuation of the separating clutch in the hybrid transmission, as a parking lock actuator, actuation of the differential lock, and shift actuation in multi-gear axles.

In automated manual transmissions, the individual gears can be engaged by hydraulic or electric actuators, for example. A gear is usually engaged by at least one of the actuators moving a shift fork that is mechanically coupled to a sliding sleeve. To engage a gear, starting from a neutral position N, a positive connection is established between the sliding sleeve, which is connected in a rotationally fixed manner to a shaft, and a gearless gear that sits on this shaft by moving the shift fork and thus the sliding sleeve. In the positive connection, teeth of an internal toothing of the sliding sleeve are in contact with teeth of a gearless gear assigned to the gearless gear. clutch wheel is engaged. If the positive connection is established and the sliding sleeve has reached an end stop, the shift fork or the sliding sleeve is in the gear position.

To control the actuator with which the shift fork can be moved, a sensor is usually necessary to determine the position of the shift fork. With regard to the detection of the gear position of the gears of an individual manual transmission, in addition to the variation caused by the sensor, there are also mechanical inaccuracies, variations and games, such as. B. due to axial bearing play, shaft and shift fork deformation. In addition, in a large series there are also variations in the end stop between the individual gearboxes, which arise due to the tolerances of the individual components.

Conventional methods such as those in DE 10 2007 057 203 B4 for controlling an actuator regulate the position of one of the shift forks to a defined gear position. Due to the existing scatter, this can lead to problems. For example, if the actual gear position is before the gear position used as a basis for the control when engaging the gear, the desired end position will apparently not be reached. This may cause an unfounded error message. In contrast, the gear position targeted by the control can be ahead of the actual gear position. In this case, there is a risk that the control switches off too early and the sliding sleeve has not yet reached the position where the pull-in effect begins, so that the gear can jump out undesirably. According to the present invention, the shift fork is moved via the actuator from a first switching position, namely a neutral position, into at least a second switching position, namely a gear position, and vice versa, the position of the actuator being in the neutral position or in the gear position upon a switching request Basis of a stored mechanical play between the actuator and the shift fork and one is corrected by the sign generated by the state machine and assigned to the respective switching request.

The actuator is a position-controlled actuator, such as a hydraulic, electromotor, electromechanical or pneumatic actuator. The actuator is controlled or regulated via a control unit.

In the prior art, the not yet published DE 10 2021 204 859.5, the switching position of the shift fork is determined via at least one sensor, such as a Hall sensor, with a sensor signal from the sensor being processed in the state machine. The sensor is located on the actuator.

In order to save the costs of an additional position sensor, the translation between the pinion and the toothed shift fork must be designed in such a way that the entire axial travel of the shift fork can be covered by 360° within one revolution of the actuator motor, taking into account a wear reserve.

The communication between the commanding control unit of the vehicle and the attached control unit of the actuator runs via a vehicle bus.

The main requirements for an electromechanical actuator system are requirements regarding response times.

When connecting the claw clutch, a total switching time of 80ms should be sufficient.

When switching off, a total switching time of 100ms with a residual torque less than or equal to 20Nm is desired, with the dog clutch should be opened against a residual torque of 30Nm.

In order to achieve these specifications, PT1 filtering, i.e. a proportional transmission with a first-order delay, of the differential speed signal, obtained from the speed of the two axes to be connected, was carried out in order to change the cycle time of the vehicle bus, 10ms, to the internal one The cycle time of the computer in controlling the claw clutch is 2ms. The filtering is necessary because the gradient of the differential speed is used as a switching criterion.

The clutch is controlled via a PI controller. The controller integrated in a control loop acts on a controlled system in such a way that a variable to be controlled is adjusted to the level of the selected reference variable using negative feedback, regardless of disruptive influences.

There is no adjustment of the control strategy depending on the residual torque during decoupling.

The opening of the clutch against a residual torque is achieved by applying a constant current.

Resolving a detected blockage during switching is implemented via the integrating portion of the PI controller.

The resulting deficiencies or problems are mainly due to the very direct control and the resulting very narrow path conditions and the very high agility of the system.

With appropriate PT1 filtering there is an initial time loss in the range of 10 to 15ms. Furthermore, the current real differential speed does not match its internal signal value, which can lead to uncomfortable switching. If the filtering is too weak, the gradient condition can only be achieved when the differential speed has been fully adjusted. In order to achieve the required switching times, the P component of the controller would have to be chosen very high, which can lead to the controller swinging into the target range.

In the event of a blockage due to excessive residual torque near the tooth-to-tooth position, with minimal overlap, a sudden relief of load occurs when the system is switched off due to the applied current and the resulting elastically pre-stressed system as soon as the overlap area has been overcome. This creates a massive acceleration of the system towards the disengagement stop. Due to the small clearance of the claw clutch, the PID controller cannot react quickly enough to brake a switching sleeve and this results in a hard start against the stop, which very quickly leads to damage to the mechanics during continuous operation.

The actuator current required to open the system against residual torque is approximately proportional to the residual torque present. The actuator can only maintain the required 30A at 30Nm for a very short time for thermal reasons. Since the exact residual torque is not known in each case, only a compromise between the required current strength and the resulting tolerable duration can be set as a constant current. This can lead to unacceptably long design processes.

The current flow quickly leads to excessive thermal stress on the actuator if sufficient current is specified to clear blockages. This also happens when the PID controller is parameterized gently, without the blockage being able to be solved.

It is the object of the invention to provide a system and a method for controlling an electromechanical actuator system, the actuator system for mechanically coupling and decoupling two elements with one Electric motor serves as an actuator for which the problems of the existing systems and procedures are solved.

Description of the invention

The problem is solved with a method for controlling an electromechanical actuator system, wherein the actuator system serves for the mechanical coupling and decoupling of two elements with an electric motor as an actuator, wherein the actuator is operatively coupled to a reduction gear, which provides a rotation of a reduction gear up to 360° degrees as a drive for an adjustment movement of an actuator and further an adjustment movement of one of the elements to be coupled, wherein a differential speed (target speed to actual speed of the elements to be coupled), the desired adjustment path to a target position, the maximum time available to achieve the desired adjustment path are used as an input variable of a control and evaluation unit, and wherein, depending on the input variables and an energy relationship determined by the actuator, a current pulse that is adequate in terms of inertia and speed is calculated according to the formula E_mech=(J*cü ² ) /2 « R*l ² *At = E_el and wherein the adjustment path is determined by Applying this calculated current pulse to the actuator is quickly passed through and the final adjustment to the target position of the actuator is carried out via a PI controller.

The method according to the invention enables a quick and safe decoupling and coupling of two elements of an actuator system that are to be coupled. Additional sensors are not required for this. As stated above, appropriate current pulses are used to accelerate and decelerate the actuator. Since the inertia of the system is practically only given by the rotor of the electric motor, in particular a brushless direct current motor (BLDC), and this is known, the necessary current pulse to achieve a desired target speed can be calculated based on the existing output speed of the actuator motor. After the current pulses have been executed accordingly, the system switches to the PI controller, which can then precisely adjust the final position. Adjustment via the PI controller is then not time-critical, since the system is already coupled or decoupled.

Due to the fact that there is no spring effect in the present system and practically all the inertia is caused by the rotor of the actuator, the energy relationship in the acceleration pulse mode can be defined as follows:

E_mech=(J*cü ² ) /2 « R*l ² *At = E_el.

From a desired change in speed, the necessary current pulse f (I, At) can be deduced.

If there is a corresponding sudden change in the speed specification An, a value for the current strength of the pulse for acceleration should be calculated according to the formula above. A current pulse is generated based on the difference between the target speed and the actual speed, taking into account that the pulse should be as short as possible, but discretized to 2ms, and must not exceed a parameterizable maximum current intensity.

As soon as the current pulse has been completed, the regular controller specification calculated in parallel is followed again. By varying the pulse duration (At), it should also be possible to maintain the pulse current for a little longer in order to achieve a similarly high level of dynamics with a lower maximum current. This means that the method can also be used in systems with limited maximum current.

As soon as the current pulse has been completed, the regular controller specification calculated in parallel should be followed again.

In a preferred embodiment of the invention, the actuator system comprises an electric motor as an actuator and a claw clutch, the method in this preferred embodiment for engaging and disengaging the claw clutch with the actuator and a reduction gear that enables a rotation of a reduction wheel up to 360 ° as a drive for a pivoting movement of a shift fork and further an axial movement of a shift sleeve, whereby a differential speed of the claw to the shaft is determined as an input variable, which should be within a predetermined range. If this is not the case, a prediction step is carried out in which this Differential speed signal is estimated, and is used for immediate initiation of the actuator, the axial path being quickly traversed by applying current pulses to the actuator, and the final adjustment to a target position of the shift fork via a PI controller.

The method available according to this further development leads to an improvement in function by starting the switching process more quickly, which results in up to 20% time savings with a direct configuration of “tooth to gap”.

The invention results in an improvement of the function by traveling through the axial path more quickly, namely the reduction of the characteristic travel time of the axial path from ~90ms to ~30ms. The quick connection causes an additional positive effect, as shift blockage due to torsional vibrations of the side shaft occurs significantly less often.

Without the method according to the invention, it would not be possible to implement the required parameters of the system with the given actuator system. A stronger actuator motor that would otherwise be necessary would have a significant cost disadvantage, as would a higher gear ratio due to the sensor that would then be required.

According to the invention, a prediction step is carried out on the basis of the respective weighted difference quotient of the differential speed and difference quotient of the signal thus obtained.

The initial change in the actuator's movement to follow a target speed change is carried out with a current pulse that depends on the actuator speed error and the armature inertia, provided the specified change exceeds a certain value. This current pulse is used to either accelerate or decelerate the actuator. The conventional control strategy is used again immediately after the current pulse.

A check for a switching blockage is carried out continuously, with the actuator being supplied with a current pulse to clear the blockage if there is a blockage.

An emergency braking function is available, which is activated if there is a risk of hitting a vehicle with a residual speed that is too high.

The emergency braking function leads to an improvement in wear and service life and to cost savings: Without the emergency braking function, excessive wear of the actuator double-plane occurs, which then eventually leads to breakage Without this emergency braking function, the actuator components would have to be much more massive and therefore more expensive.

The thermal load of the system is monitored and the risk of overheating is minimized by reducing the frequency and/or current of the current pulses.

Thermal monitoring improves wear and service life: The monitoring eliminates thermal overloads, which would otherwise deactivate the controller for the entire cycle. Regular overloads also have a negative effect on the wear and tear of the electronic switches, which reach the end of their service life more quickly if they are regularly overstressed.

The problem is also solved with a system for engaging and disengaging a claw clutch with an actuator and a reduction gear, which provides a rotation of a reduction gear up to 360° as a drive for an axial movement of a shift fork.

A controller activates the actuator, whereby the controller receives the signals for clutch actuation via a vehicle bus.

Description of the characters

Fig. 1 shows a flow chart of the method based on a switching process of a dog clutch,

Fig. 2 shows the course of the differential speed and the gradient - comparison filter to prediction,

Fig. 3 shows a course of the switching sequence with pulse strategy,

Fig. 4 shows the course of the pulses,

Fig. 5 shows a flowchart of the thermal load. In Figure 1, the method 1 is shown as an entire process using an actuator system, with an electric motor as an actuator and a claw clutch.

The electric motor is preferably a brushless direct current motor (BLDC).

A switching request 2 is received via the vehicle bus to start.

In step 3, the controller checks whether the differential speed dnCluRaw between the claw and the shaft is within a specified range. If this is not the case, the controller jumps to prediction step 4 to determine the differential speed signal.

A prediction of the differential speed signal to a dnCluPred based on the respective weighted difference quotient of the differential speed dnCluRaw and the difference quotient of the signal obtained in this way is carried out if no signal update is available or if the measured differential speed dnCluRaw is not in the assigned range. The implementation in the controller is carried out by using the differences of the last two signal values of dnCluRaw and their differences.

The value obtained in this way replaces the dnCluPred value that lies outside the specified ranges in step 3.

If the value is now acceptable, the actuator of the dog clutch is activated with an acceleration pulse 5 depending on an actuator speed error, determined as the deviation between the target speed and the actual speed. The target speed 30 can be selected from different modes depending on the type of requirement.

In step 6, the controller checks whether a tooth-tooth detection threshold is exceeded. If this is the case, a brake pulse 7 is output depending on the actuator speed error.

In step 8, it is checked whether a tooth-tooth condition is detected and the actuator is blocked as a result. If a blockage is detected in step 8, an adjustment 9 is initiated on the actuator using the tooth-tooth current specification of the controller.

In checking step 10 it is determined whether a tooth overlap threshold has been exceeded. If so, an acceleration pulse 11 occurs.

In the next step 12, the target speed 30 is reached with adjustment, whereby the PI controller is used.

In step 14, the PI controller checks whether a switching blockage has been detected. If the answer is yes, a current pulse P1..Pn is again applied to release the blockage 15. If not, it is checked whether the overlap is sufficient, step 13.

If the determination is positive, a braking pulse 16 occurs and the target position is adjusted 17 with the PI controller.

The claw clutch is now correctly closed.

Appropriate current pulses are used to accelerate and decelerate the actuator. Since the inertia of the system is practically only given by the rotor of the BLDC and this is known, the necessary current pulse to achieve a desired target speed 30 can be calculated based on an existing output speed of the actuator motor. After the current pulses have been executed accordingly, the system switches to the PI controller, which can then precisely regulate the final position. The regulation via the PI controller is then not time-critical, since the system is already coupled or decoupled.

To address the problem of approaching the stop in the event of a blockage during deployment, an emergency braking function was implemented, which, when such a case is detected, specifies an adequate counter voltage within 6ms and thus slows down the system. The necessary high reaction speed is only possible by directly specifying the voltage that leads to current pulses.

The process is shown in Figure 1 on the right side for laying out the clutch.

Based on request 18 for designing the clutch, the controller checks whether the residual torque is below a threshold value. Only when the residual torque is small enough does an acceleration pulse occur in step 20. The target speed 30 is then adjusted in 21 using a PI controller.

In 22, a check is made to see whether there is a switching blockage, which is then released via current pulses P1..Pn in 23. It is also checked whether the overlap area 24 has been left; if not, the control process continues; if so, the controller checks in 25 whether the speed is too high. If this is the case, the emergency braking function 26 is activated, which then leads to the adjustment 28 of the target position. If the emergency braking function 26 is not activated, a moderate braking pulse 27 is emitted, which also leads to the adjustment 28 of the target position. The claw clutch is open.

The use of current pulses P1..Pn means that a blockage can be resolved when it is detected. If the actuator is blocked during the switching process, the preloaded system is first relieved and then a maximum current pulse with IMax is specified in order to be able to use the inertia of the system to release the blockage.

The thermal load on the system is monitored throughout the entire pulse process, thus eliminating the risk of overload. Taken together, all of these individual measures result in a characteristic operating strategy, without which it would not be possible to control such a directly controlled claw clutch system while adhering to the specified parameters.

The prediction 4 of the differential speed signal dnCluPred is calculated as follows: dnCluPred(n)

> C dnClu(m), dnCu(m) current

(dnCluPred(n — 1) + fad * dnCluDiff(m) + fac2 * (dnCluDiff(m) — dnCluDiff(m — 1)), otherwise dnCluDiff(m)= dnClu(m)-dnClu(m-1 ) fad , fac2 freely selectable (parameter) m Running index for describing the bus clock n Running index for describing the model calculation clock

A distinction is made here as to whether a current differential speed for direct transfer, i.e. dnCluPred(n) = dnClu(m), is currently available on the vehicle bus (reception at a slower rate, about 10ms), or whether in the steps in between (controller calculates at a faster rate, about 2ms), where no current signal is available from the bus, the differential speed is predicted using dnCluPred(n) = dnCluPred(n-1 )+

The result can then be filtered. The gradient is calculated using the difference quotient of the filtered dnCluPred signal, namely dnCluPredFilt. Averaging can be carried out over 1 to 4 values. A comparison of the prediction method to conventional filtering based on an actually recorded synchronization process can be found in Figure 2 in the curves grdDnCluPredFilt and grdDnCluFilt in the upper third of the graph.

Due to the fact that there is no spring effect in the present system and practically all of the inertia is caused by the rotor of the actuator, the energy relationship in the acceleration pulse mode can be defined as follows:

E_mech=(J*cü ² ) /2 « R*l ² *At = E_el.

If the speed setting An changes abruptly, a value for the current intensity of the pulse for acceleration should be calculated according to the formula above. A current pulse is generated based on the difference between the target speed 30 and the actual speed 31, taking into account that the pulse should be as short as possible, but discretized to 2 ms, and must not exceed a parameterizable maximum current intensity.

As soon as the current pulse has been completed, the regular controller specification calculated in parallel is followed again.

Braking current pulse mode works analogously to the acceleration pulse, only with reversal of the signs for speeds and currents.

The schematic procedure for connection is shown graphically in Figure 3.

The triggering criterion for a pulse is a jump in the speed specification, ie the target speed 30 increases from an initial value around zero to around 1800 revolutions per minute. The actual speed 31 increases sharply at 0.03 and then remains in the tooth-tooth range 40 at 50 revolutions in the time window between 0.04 s and approximately 0.1 s. A rapid increase follows the decision that the target position of the axial displacement of the shift fork, here defined by the angle of rotation of the reduction wheel, is almost reached and the reduction in speed and the settling in the PI controller range 50.

The current pulse strength and duration are calculated from the speed error. This allows special, known positions to be taken into account, such as braking before a tooth-tooth position in the time period of 0.02 to 0.1 seconds to prevent a stop. This also allows a certain spread in the operating strategy, such as fast switching compared to optimized switching with low noise and vibration. At time 0.12s, the target position is almost reached and the remaining small deviations are controlled in the time window 0.12 to 0.18s via the PI controller.

The target angle 32 is set to an angle of approximately 300°. The actual angle 33 follows the target position via a position in the tooth-tooth range of 100°.

The entire coupling process is monitored by the controller. As soon as the two conditions actual angle 33 < threshold value and actual speed 31 > threshold value are met, the system immediately switches to emergency braking mode and the default voltage for the actuator is specified directly via a stored table depending on the actual speed 31.

The entire decoupling process is monitored; as soon as the two conditions actual current > threshold value and actual speed 31 < threshold value are met for a certain, parameterizable time, the pulsed opening is started. The schematic procedure is shown graphically in Figure 4. In the upper part of the graph, the curve of a target current Isoll is shown over time t. In the lower part, the actual angle <p Actual is plotted over time. Starting from the initial holding current Unit, the current in the pulse P1 increases up to a maximum pulse current IMax. The current IMax is held for a holding period T1. Over a reduction period T2, the current is reduced along a current reduction ramp Ides to a minimum pulse current IMin. During a relaxation period T3, the current is reduced to zero.

With each pulse of the current, the angle of rotation increases by a delta amount A<p until a target angle <pSoll is reached.

The sequence of the blocking pulse process with thermal monitoring is shown in Figure 5 as a flow chart.

A current pulse of maximum strength Imax is generated in response to a current pulse to release the blockage 15. If the thermal load is above a threshold value TSchw, the actuator is regulated down to a holding current llnt for cooling. From this state, the regulation takes place to the minimum current IMin of the pulse. After the time period T2, a check is made as to whether the maximum number of pulses has been reached; if not, the maximum current IMax is generated again.

If no thermal overload is detected, the current pulse oscillates back and forth between IMax and IMin.

Claims

Expectations

1. Method for controlling an electromechanical actuator system, wherein the actuator system serves for the mechanical coupling and decoupling of two elements with an electric motor as an actuator, wherein the actuator is operatively coupled to a reduction gear, which provides a rotation of a reduction gear up to 360° as a drive for an adjustment movement of an actuator and further an adjustment movement of one of the elements to be coupled, wherein a differential speed (target speed to actual speed of the elements to be coupled), the desired adjustment path to a target position, the maximum time available to achieve the desired adjustment path are used as an input variable of a control and evaluation unit, and wherein depending on the input variables and an energy relationship determined by the actuator, according to the formula

E_mech=(J*cü ² ) /2 « R*l ² *At = E_el a current pulse that is adequate for inertia and speed is calculated and the adjustment path is quickly traversed by applying this calculated current pulse to the actuator and the final adjustment to the target position of the Actuator via a PI controller.

2. Method for controlling an electromechanical actuator system, wherein the actuator system is used for mechanically coupling and decoupling two elements with an electric motor as an actuator, the elements to be mechanically coupled and decoupled being elements of a claw clutch as a sliding sleeve/switching sleeve and gearless wheel, and wherein the adjustment movement is a Pivoting movement of a shift fork, which is coupled to the shift sleeve for axial displacement, whereby the differential speed (dnCluRaw) is determined as the speed difference between the claw and the shaft, which should be within a predetermined range. If this is not the case, a Prediction step (4) is carried out in which a differential speed signal (dnCluPre) is estimated and is used for immediate initiation of the actuator, the axial path being quickly traversed by applying an inertia- and speed-adequate current pulse to the actuator, and the final adjustment to a target position the switching fork is done via a PI controller.

3. Method (1) according to claim 2, characterized in that the prediction step (4) takes place on the basis of the respective weighted difference quotient of the difference speed dnCluRaw and the difference quotient of the signal thus obtained.

4. Method according to claim 1 or 2, characterized in that the respective initial change in movement of the actuator in order to follow a desired speed change (30) is carried out with a current pulse which is dependent on the actuator speed error and on the armature inertia.

5. The method according to claim 2, characterized in that a check for a tooth-tooth blockage position is carried out, with the actuator regulating a holding current should there be a blockage.

6. Method according to claim 5, characterized in that after checking whether the blockage has been removed, an acceleration pulse is again applied as a current pulse until the target position is almost reached.

7. The method according to claim 2, characterized in that an emergency braking function is present, which is activated when there is a risk of hitting a residual speed that is too high.

8. Method for monitoring switching blockages while engaging and disengaging a claw clutch and in particular for applying current pulses (P1 ..Pn) to resolve these blockages.

9. Method according to claim 8, characterized in that the thermal load of the system is monitored and the risk of overheating is minimized by reducing the frequency and/or the current of the current pulses (P1..Pn).

10. System for coupling and decoupling two elements to be mechanically connected with an electric motor as an actuator, in particular for inserting and disengaging a claw clutch with an actuator and a reduction gear, which allows a rotation of a reduction wheel up to 360 ° as a drive for an axial movement of one Shift sleeve provides, the method according to claims 1 to 9 being used.

11. System according to claim 10, characterized in that a controller causes the actuation of the actuator, the controller receiving the signals for clutch actuation via a vehicle bus.