Actuator system
Patent Information
- Application Number
- PCT/EP2026/053335
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-02-20
- Filing Date
- 2026-02-09
- Publication Date
- 2026-08-27
Smart Images
Figure EP2026053335_27082026_PF_FP_ABST
Abstract
Description
[0001]
[0002] October 2, 2025
[0003] Actuator system and methods with energy reflection in an actuator system
[0004] The present invention relates to an actuator system, in particular for teleactuation, and a method with energy reflection in an actuator system.
[0005] In bilateral teleoperation, inputs from a first actuator are transmitted via a communication channel to a second actuator connected to the first. The first and second actuators can be controlled in a master-slave configuration, so that inputs applied to the first actuator by an operator, for example, are transmitted via the communication channel to the second actuator, which then executes the operator's inputs. The first actuator acts as the master, and the second as the slave. Not only are inputs, such as movement or force, transmitted from the first actuator to the second, but feedback from the second actuator, such as force or movement, is also transmitted back to the first actuator via the communication channel, thus ensuring force control through a force feedback system.This is intended to create sufficient system transparency so that the operator receives adequate information about the movement and interaction force performed by the second actuator. Such coupled actuator systems are particularly useful in robotics, for example in medical robotics for telesurgery or in aerospace engineering. This eliminates the need for the operator to be physically present at the point of execution. Instead, the operator's movements are detected by the first actuator and then transmitted via the communication channel to the second actuator, which executes the operator's movements, for example, to perform surgery on a patient or to remotely control an extraterrestrial robot.
[0006] However, real communication channels exhibit a certain latency or time delay, especially in long-distance teleoperations. This communication delay in bilateral teleoperation generates unwanted additional energy that can lead to instability.
[0007] The so-called "time-domain passivity approach" (TDPA) is known from the prior art, e.g., from KR 10 118 994 Bl, for maintaining the stability or passivity of the actuator system. The controller includes a first passivity controller that dampens the second actuator, thus adjusting its velocity. A second passivity controller dampens the first actuator, thus adjusting its force. A position controller is also included, which regulates the position coupling between the first and second actuators. This approach takes into account the directional dependence of the energy flow from the first to the second actuator and vice versa, meaning that no general energy storage is available and overcompensation is typically used to maintain passivity.This leads to large positional errors in the synchronization of the movement of the first actuator with the second actuator and vice versa (positional drift). Furthermore, overcompensation makes the actuator system energy-inefficient. Another problem within the context of TDPA is the reversal of energy flow direction. Such reversals can occur particularly during contact with the environment, for example, a wall contact, when there is a transition from the phase of increasing force transmitted from the first actuator to the second actuator to the phase of decreasing force transmitted from the first actuator to the second actuator. If the direction of energy flow changes, the time delay can lead to a sudden drop in force or excessive damping, which can result in force jumps and jitter, thus significantly reducing system transparency.
[0008] Based on the aforementioned disadvantages of the prior art, the present invention aims to provide an optimized actuator system and an optimized method for guaranteeing passivity, with a particular improvement in coupling behavior and / or performance. This objective is achieved by an actuator system according to claim 1 and a method with energy reflection in an actuator system according to claim 10.
[0009] The actuator system according to the invention, particularly for remote actuation, comprises a first actuator, especially for operation by a user. This actuator can, for example, be a master in a master-slave configuration. Furthermore, the actuator system comprises a second actuator, particularly for executing a movement by the user. This second actuator can, for example, be a slave in a master-slave configuration. Preferably, a second passivity controller is connected to the first actuator, and a first passivity controller is connected to the second actuator. The actuator system also includes a communication channel between the first actuator and the second actuator for transmitting the speed and / or force of the first actuator to the second actuator and vice versa.Within a force-feedback system, the speed and / or force of the second actuator is transmitted to the first actuator via the communication channel. This creates system transparency, allowing the user of the actuator system to preferably receive haptic feedback about the movement of the second actuator. Furthermore, a first control unit is provided, configured so that the energy input by the first actuator towards the communication channel can be measured as the target energy. This first control unit is also configured to determine a reference energy from a reflected component of the target energy and transmit this information back to the second passivity controller and / or the first actuator. Through this feedback of the energy information, the reference energy, as a reflected component of the target energy, is directly available to the first actuator.
[0010] Furthermore, a controller is provided, wherein the controller is configured to regulate the damping of the first actuator, particularly via the second passivity controller, and / or the second actuator, particularly via the first passivity controller, depending on the reference energy. Preferably, energy is dissipated in the first and / or second passivity controller via the damping. This allows, in particular, an energy value that would be available at the controller for the second actuator to be continuously reflected back to the first actuator. Preferably, this energy information can be taken into account at the second passivity controller and / or first actuator. This makes energy available earlier on the side of the first actuator, which can exit in the event of a change in the direction of the energy flow at the first actuator. In particular, more energy can exit there than in an embodiment without feedback of the reflection energy information, e.g., in TDPA.This allows the force reduction or damping to be adjusted depending on the situation. This optimizes system transparency and increases the robustness of the actuator system to complex environmental energy behavior.
[0011] Preferably, the actuator system additionally comprises a second control unit connected to the controller, wherein the second control unit is configured to store information about the energy available in the controller and to reduce it by the determined value of the reference energy. To maintain passivity in the actuator system, the reference energy is subtracted from the available energy so that the reflection component of the target energy must not escape towards the second actuator. Preferably, the second control unit determines the energy E available in the controller. obs Qc) after
[0012]
[0013] E O bs(k) = (1 — ~ ^L2R(^)—
[0014] with the reflection component v to the sampling step k, where L2R denotes the energy flow from the direction of the first actuator to the second actuator, R2L the energy flow from the direction of the second actuator to the first actuator, and index 4 denotes the port of the controller facing the first actuator and index 5 the port facing the second actuator of the controller. Preferably, the energy available in the controller is energy information, in particular potential energy, and corresponds to the energy value that may exit the controller towards the second or first actuator. This ensures that the energy that should be available for the first actuator, in particular the second passive controller, on the operator side is not available on the robot side. In particular, the second control unit takes the reference energy into account for determining the available energy of the controller.This results in a discrepancy between the energy actually present in the controller and the energy available in the controller.
[0015] Preferably, the reflection component can assume values from the interval [0, 1], and particularly preferably values from the interval (0, 0.5). This allows the energy information transferred from the controller to the second passivity controller and / or first actuator to be adapted to the situation, while maintaining the passivity and stability of the actuator system at all times. In particular, the reflection component can assume the value zero. This allows, especially in the case of free movement, the entire energy measured as target energy of the first actuator to be transferred to the second actuator without reducing the available energy in the controller. In particular, the reflection component can assume the value one.This allows, particularly in the case of contact with the environment, such as an obstacle or a wall, the information about the total energy transferred from the first actuator towards the controller to be reflected back to the second passive controller and / or the first actuator, so that no available energy is present in the controller. Preferably, the reflection component can be a constant value. In this case, the reflection component does not change, and the reference energy depends only on the target energy. This simplifies the determination of the reference energy. Preferably, the reflection component v can be designed to depend on the positional deviation of the input device and the robot or the speed of the robot, in order to reflect energy only in the case of contact with the environment, e.g., contact with an obstacle or a wall. In this case, the positional deviation is measured, in particular, on the robot side.Due to the anticipatory reflection of the energy information, the reference energy is already available during the pressure phase of the wall contact at the first actuator and can exit towards the operator in the R2L direction as soon as the wall contact is released.
[0016] Preferably, the controller comprises a spring with adjustable spring stiffness for transmitting a velocity and / or force to control the second actuator. The second control unit is configured to adjust the spring stiffness based on an energy value of the actuator system. This energy value includes, in particular, the available energy in the controller and / or the actual energy in the controller, or a difference between the available energy and the actual energy in the controller. The available energy is, in particular, the available energy E. obs(k) of the actuator system. The difference between the available energy and the actual energy in the controller is, in particular, the energy value of the dissipated energy and, more specifically, the energy of a damping, preferably virtual, e.g., in a control circuit of the actuator system. It is particularly possible that the second control unit is configured to continuously acquire the energy value. The second control unit can, for example, be configured to measure the energy value. The spring stiffness, in particular the spring stiffness profile, is adjusted over the deflection of the spring. During the adjustment over the deflection, at least a partially continuous adjustment takes place.By adjusting the spring stiffness in the controller, the second control unit can ensure that only the available energy leaves the controller towards the second or first actuator, and not the actual energy. This results in a temporary adjustment of the force, eliminating the need for adjustments to speed and energy dissipation, as would be required in a first passive controller. This guarantees the passivity and thus the stability of the actuator system at all times and prevents additional position drift. Preferably, the additional energy supplied via the communication channel is dissipated by the first and / or second passive controller. It is particularly possible for the second control unit to be integrated with the first control unit, or for the first and / or second control unit to be an integral part of the controller.
[0017] Preferably, the value of the reflection component can be varied during the transmission of the target energy. In particular, the value of the reflection component can be determined as a function of the deflection of the controller's spring and the robot's speed. Specifically, the reflection component can assume a maximum value above a certain spring deflection threshold to enable the return transmission of a maximum value of the reference energy, especially in the event of wall contact. Additionally, the robot's speed can preferably be taken into account by ensuring that the reflection component only reaches a maximum value when, in addition to the spring deflection threshold, the robot's speed falls below a certain limit.
[0018] Preferably, a minimum reflection component can be defined so that the reflected information about the reference energy is always non-zero. This allows for further reduction of force jumps and jitter, particularly with only slight delays in the communication channel and correspondingly short wall contact times of the robot. Preferably, the reflection component can also be determined as a function of the delay in the communication channel. This allows the magnitude of the reflection component to be adjusted according to the current situation. As a result, the reference energy, and thus the damping of the first and / or second actuator, can be adapted to the prevailing environmental conditions.
[0019] Preferably, in the actuator system, especially at the respective ports i, the power P l determined, after
[0020] P k) = v^k^F^k)
[0021] The sampling step k, with the velocity i?( / c) of the actuator and the force F'( / c) of the controller, is used. In particular, the powers Pi (k) are recorded at ports where power-related signals (velocity v and force F) can be observed.
[0022] Preferably, the direction of the power flow is determined by the sign of the power from the first actuator towards the second actuator P2RW and the power from the second actuator towards the first actuator ^. 2L (fc) from:
[0023] Pi r = f °' for PW ^ O
[0024] L2r1} lp k, forP i (fc)
[0025]
[0026] > 0,
[0027] and
[0028] pi rn = f °' for PW ^ O
[0029]
[0030] R2Ll J forP i (fc) < 0,
[0031] where in particular the sign of P l(k) depends on a sign convention defined in the controller.
[0032] Preferably, the sampling time T can be used. s the energy from the first actuator towards the second actuator E^^k) and the energy from the second actuator towards the first actuator E R l 2L (k) can be calculated by integration to:k
[0033] £Z,2R(^) = T s P[2R (D
[0034]
[0035] j=0
[0036] and
[0037] k
[0038] = Pfl2L (f)-
[0039]
[0040] j=0
[0041] The present invention further describes a method with energy reflection, in particular in an actuator system, comprising the following steps:
[0042] The energy supplied by the first actuator towards the second actuator between the communication channel and a controller is measured as the target energy, a reflection fraction v is determined to calculate a reference energy from the reflection fraction v and the target energy, the calculated value of the reference energy is transmitted back to the second passivity controller and / or first actuator, and the energy available in the controller is reduced by the value of the reference energy.
[0043] Preferably, the reflection component v is the value of a constant or a value depending on a deflection of a spring of the controller and a speed of a robot actuated by the actuator system.
[0044] Preferably, the spring stiffness in the controller is adjusted so that only the available energy leaves the controller.
[0045] Preferably, the energy reflection method is designed according to the characteristics of the actuator system.
[0046] They show:
[0047] Fig. 1 shows a schematic representation of an actuator system in the form of a port network.
[0048] Figure 1 shows a first actuator 10 and a second actuator 12 connected to each other via a communication channel 14. In particular, the first and second actuators can be in a master-slave configuration. The first actuator can be, in particular, a haptic input device for transmitting a movement initiated by a user. The second actuator can be, in particular, a robot configured to execute the user's movement.
[0049] Communication channel 14 can be a wired data transmission and / or a wireless data transmission. In particular, the transmission from the first actuator 10 to the second actuator 12 and vice versa can take place via the internet or another communication connection.
[0050] A movement of the first actuator (Al) 10, for example, applied to the first actuator 10 by a user, is then transmitted via communication channel (CC) 14 to the second actuator (A2) 12, which is then intended to perform the same movement with high positional accuracy. Conversely, however, particularly within the context of a force feedback system, forces and / or movements acting on the second actuator 12 are also to be transmitted via communication channel 14 to the first actuator 10. This creates system transparency, so that a user connected to the first actuator experiences forces acting on the second actuator 12, be it as haptic feedback, visual feedback, or the like.
[0051] As indicated by the dashed line 16 in the figure, communication channel 14 has a time delay. The transmission time from the first actuator to the second actuator is Tf, and the transmission time from the second actuator to the first actuator is Tb. Specifically, Tf and Tb can be the same, but they can also be different.
[0052] Due to the delay 16 of the communication channel 14, the actuator system could become unstable. To ensure the stability or passivity of the actuator system, a controller is provided. The controller has a second passivity controller (PC2) 18, which dampens the first actuator 10, in particular by changing the force. Furthermore, a second passivity controller (PC2) 20 is provided, which dampens the second actuator 12, in particular by changing the target speed. The damping of the first actuator 10 and the second actuator 12 is thus achieved by the passivity controllers 18 and 20, ensuring the passivity and therefore the stability of the actuator system at all times. In addition, the controller 22 is configured to regulate position coupling between the first actuator 10 and the second actuator 12.
[0053] Between the first passive controller 20 and the controller 22, a first control unit (CU1) 28 is connected to the actuator system. The first control unit 28 is configured to measure the energy supplied by the first actuator 10 towards the communication channel 14 as the setpoint energy. Furthermore, the first control unit 28 is configured to determine a reference energy via a reflection component v of the setpoint energy and to transmit the information about the reference energy to the second passive controller 18, which allows this reference energy to be output to the first actuator 10.
[0054] A second control unit (CU2) 24 is connected to the controller 22. The second control unit 24 is configured to store information about the energy available in the controller 22 and to reduce the determined value of the reference energy accordingly. Specifically, the second control unit manages and balances the system's energy, recording how much energy was supplied via the first actuator 10 and / or the second actuator 12, and ensures that this information is forwarded to the second passivity controller 18 and taken into account when determining the available energy in the controller 22. In the figure, the second control unit 24 is depicted as a separate element. However, the second control unit 24 can also be an integral part of the controller 22.Alternatively, the second control unit 24 can also be designed together with the first control unit 28, or the first 28 and / or second 24 control unit can be an integral part of the controller 22.
[0055] The controller 22 determines the power Pt(k~) = v at the respective ports where power-related signals (velocity v and force F) can be observed. i k')F i k') measured at port i in time step k. The direction of power flow can be determined from the sign of the power:
[0056] for P l (k) < 0
[0057] p / 2R (fc) = (1)
[0058]
[0059] L2RV J lP l (k), for P l k) > 0,
[0060] and
[0061] 0, for P l (k) > 0
[0062] PR2LW = (2)
[0063]
[0064] -P%k), for Pl k) < 0.
[0065] According to arrows 26, "L2R" denotes the power Pi or energy Ei from the first actuator towards the second actuator, and "R2L" denotes the power Pi or energy Ei from the second actuator towards the first actuator. Furthermore, i denotes the respective port between the individual elements of the actuator system, such that i = l, 5.
[0066] It should be noted that the sign of the flow direction depends on the sign convention in controller 22. With the sampling time T s The energies can be calculated by integration:
[0067] k
[0068] E L ' 2B m = (3)
[0069]
[0070] j=0
[0071] and
[0072] k
[0073] (4)
[0074]
[0075] j=0
[0076] According to the present invention, the information about a reflection component 30 (v) of the energy E^ is 2R k) from the first actuator 10 towards the second actuator 12 and back to the first actuator 10. The value of the remaining reflection component 32 (1 - v) of the energy E R2R(k) The energy transmitted from the first actuator 10 towards the second actuator 12 is transferred to the control unit 24 for consideration in determining the available energy in the controller 22. Preferably, the reflection component 30 can assume a value from the interval [0, 1]. Particularly preferably, the reflection component 30 can assume a value from the interval (0, 0.5). This allows the energy transmitted from the first actuator 10 to the second actuator 12 to be adapted depending on the situation. In particular, the reflection component 30 can assume the value zero, for example, when transmitting a free movement from the first actuator 10 to the second actuator 12. In this case, the information about the total energy E^ transmitted from the first actuator 10 towards the second actuator 12 is 2Rk is transmitted to the second control unit 24 for the execution of the commanded movement. In particular, the value of the reflection component 30 can be one. This allows, especially in the case of contact with an obstacle, e.g., a wall contact, the information about the total energy transmitted from the first actuator 10 to the controller 22 to be reflected back to the second passive controller and / or first actuator 18 and / or first actuator 10 even before the contact with the obstacle is broken. As a result, the passive controller 18 (PCI) does not have to attenuate the force feedback from the second actuator 12 to the first actuator 10. This improves system transparency in the event of changes in the energy flow direction.
[0077] In control unit 24, the available energy of the controller is determined according to E
[0078]
[0079] O bs(k) = (1 — V )^L2R (^) + —^L2R (X) — ^R2L (^) (5) with the reflection component v to the sampling step k, where L2R denotes the energy flow from the direction of the first actuator 10 to the second actuator 12, R2L denotes the energy flow from the direction of the second actuator 12 to the first actuator 10 and index 4 denotes the port of the controller facing the first actuator 10 and index 5 denotes the port of the controller facing the second actuator 12.
[0080] The controller 22 has a spring with adjustable spring stiffness. The second control unit 24 can preferably manage the system's energy levels, i.e., it can record how much energy was supplied to the controller 22 by L2R and by R2L. Furthermore, the second control unit ensures, in particular, that information about the energy available in the controller is determined, which may differ, for example, from the energy actually present in the controller due to the reflection component of the target energy.
Claims
Claims 1. Actuator system, especially for teleactuating, with a first actuator (10), in particular for operation by a user, a second actuator (12), in particular for executing a movement of the user, a communication channel (14) between the first actuator (10) and the second actuator (12) for transmitting the speed and / or force of the first actuator (10) to the second actuator (12) and vice versa, a first control unit (28), wherein the first control unit (28) is configured such that the energy supplied by the first actuator (10) towards the communication channel (14) can be measured as the target energy, wherein the first control unit (28) is configured to determine a reference energy via a reflection component (30) v of the target energy and to transmit the information about the reference energy back to the second passivity controller (18) and a controller (22), wherein the controller (22) is configured to control the damping of the first actuator (10) and / or the second actuator (12) depending on the reference energy.
2. Actuator system according to claim 1, characterized in that the actuator system additionally comprises a second control unit (24) connected to the controller (22), wherein the second control unit (24) is configured to store the information about the energy available in the controller (22) and to reduce it by the determined value of the reference energy.
3. Actuator system according to claim 2, characterized in that the second control unit (24) uses the energy E available in the controller (22). O bsW determines according E O bs(k) = (1 — V )^L2R (^) + — ^L2R (X) — ^R2L (^) with the reflection component (30) v to the sampling step k, where L2R denotes the energy flow from the direction of the first actuator (10) to the second actuator (12), R2L denotes the energy flow from the direction of the second actuator (12) to the first actuator (10) and index 4 denotes the port of the controller (22) facing the first actuator (10) and index 5 denotes the port of the controller (22) facing the second actuator (12).
4. Actuator system according to one of the preceding claims, characterized in that the reflection component (30) v assumes values from the interval [0; 1], particularly preferably values from the interval (0, 0.5].
5. Actuator system according to one of the preceding claims, characterized in that the reflection component (30) v is a constant value.
6. Actuator system according to one of claims 2 to 5, characterized in that the controller (22) comprises a spring with adjustable spring stiffness for transmitting a speed and / or force for controlling the second actuator (12), wherein the second control unit (24) is designed to adjust the spring stiffness based on the available energy E O bs in the controller (22), and where the spring stiffness is adjusted across the range of motion during a deflection of the spring.
7. Actuator system according to claim 6, wherein the reflection component (30) v is varied, in particular depending on the deflection of the spring of the controller (22) and / or the speed of the robot.
8. Actuator system according to one of the preceding claims, characterized in that the power P l The determination at port i is based on P l (k) = v i (k')F i (k') to the sampling step k, at speed v l (k of the actuator and the force F k) of the controller (22).
9. Actuator system according to one of the preceding claims, characterized in that the direction of the power flow is determined by the sign of the power from the first actuator (10) in the direction of the second actuator (12) P L 1 2RW and the power from the second actuator (12) towards the first actuator (10) This results from: pi r = f °' for PW ^ O L2r1} lp k, forP i (fc) > 0, and pi rn = f °' for PW ^ O R2 / Ä} lP\k), forP i (fc) < 0, wherein in particular the sign of P k) depends on a sign convention defined in the controller (22) and wherein with the sampling time T sthe energy from the first actuator (10) towards the second actuator (12) E L l 2R (k) and the energy from the second actuator (12) towards the first actuator (10) E R l 2L (k) can be calculated by integration to: k £Z,2R(^) = T s P[ 2R (J') j=0 and k £R2L(^) = T s PR2L (f) j=0 10. Energy reflection method, in particular in an actuator system according to one of the preceding claims, comprising the steps: Detection of the energy introduced by the first actuator (10) towards the second actuator (12) between the communication channel (14) and a controller (22) as the target energy, determination of a reflection component (30, 32) v to calculate a reference energy from the reflection component (30, 32) v and the target energy, transmission of the calculated value of the reference energy back to the first actuator (10), in particular the second passivity controller (18), Reduction of the energy available in the controller (22) by the value of the reference energy.
11. Method according to claim 10, wherein the reflection component (30, 32) v is the value of a predefined constant or assumes a value depending on a deflection of a spring of the controller (22) and a speed of a robot actuated by the actuator system.
12. Method according to one of claims 10 or 11, wherein the spring stiffness in the controller (22) is adjusted so that only the available energy leaves the controller (22).