DE102014004115B3 - Control network with a haptic input device - Google Patents

Abstract

The invention relates to a control network comprising a master system (101) with a haptic input means (102), and a via a bidirectional communication channel (103) to the master system (101) connected to the slave system (104), wherein the master system (101) can generate a control variable sm (t) for driving the slave system (104) and can be transmitted to the slave system (104) by the slave system (104) for controlling the master system Based on the control variable sm (t) received from the slave system (104), a response variable fa (t) can be generated and transmitted to the master system, a first means (105) is provided, with which the control variable sm (t) and the response quantity fa (t) received from the master system is capable of determining a first energy quantity EP (t) indicative of energy flowing from the slave system (104) to the master system (101) and based on the control variable sm (t) and the response variable fa (t) has a second energy quantity EM ( t) indicating an energy flowing from the master system (101) in the direction of the slave system (104) and a second means (106) is provided, which in the case of EP (t)> EM ( t), an output of a warning to a user of the master system (101) is triggered by an output means (107), and in the case that EP (t)> EM (t) + Ethreshhold, the control network into one default security state switches, with Ethreshhold is given.

Description

The invention relates to a control network comprising a master system, and a slave system connected to the master system via a bi-directional communication channel. Furthermore, the invention relates to a method for controlling such a control network. The invention finds application in robotics, in particular telerobotics, telepresence, but also in all systems and uses of haptic input means which have a haptic feedback function.

Such control networks are known in the art as so-called "telepresence systems" or "virtual reality systems". Based on such control networks, an operator may, for example, interact with a remote real or simulated / virtual environment by manually manipulating an input means, in particular a haptic input means (such as a joystick) of the master system. Control signals, in particular position and / or speed signals and / or force signals are detected by the input means and sent to a remote slave system, for example with a robot manipulator. The slave system interacts with the real / virtual environment of the slave system, for example, by moving a real / virtual object based on the received control signals in the real / virtual environment.

In order to give the operator the haptic impression that is as real as possible when handling the remote object by the slave system, the slave system sends (feedback) response signals to the master system, which in particular indicate mechanical forces with which the environment acting on the slave system. These response signals are i.a. the input means, which conveys these forces to the operator by generating the response signals corresponding counter forces in an input of the operator.

For the following explanations, it is assumed that the reader is familiar with the terminology customary for the description of generic control networks and their meaning, the types of representation of such control networks and the mathematical methods for their description. An explanation of these principles is therefore omitted herein.

As is known, the passivation of such control networks is particularly a challenge when time-variable transmission delays occur in the communication between the master system and the slave system. Such time-variable transmission delays occur in particular in the case of communication over long distances (for example in the case of satellite links, radio links), in communications in / via media with slow signal propagation speeds (eg underwater acoustic links) or in the case of communication via systems which generate inherent delay (eg internet link) on.

In the prior art, in particular from the US 6,144,884 A. and the US Pat. No. 7,027,965 B2 , different approaches for Passivierung a generic control network are known.

The US 6,144,884 A. is based on the so-called "wave variables" approach, and proposes an energy filter coupled to the bidirectional communication channel that limits the total energy generated by the communication channel itself. The technical teaching is in the US 6,144,884 A. using a network with a position / speed-force architecture.

The US Pat. No. 7,027,965 B2 is based on the passivation of the network known as "Time Domain Passivity Control" or "TDPC". In this case, the so-called "passivity observer" determines the positive or negative energy amount which is dissipated or generated by the master system or the slave system. Furthermore, a so-called "passivity controller", in the case of the "Passivity Observer" a net energy generation, ie a negative energy contribution was determined, determined and applied an attenuation corresponding to the generated net energy. This results in the passivation of the control network, ie the communication channel, in which net energy can arise based on the time-varying transmission delays. For the term and the usual use of the term "energy" reference is made to the relevant literature. The technical teaching is in the US 6,144,884 A. using networks with a position / velocity force architecture or a force position / speed architecture. See also the article "Time-Domain Passivity Control of Haptic Interfaces" by B. Hannaford and JH Ryu, in IEEE Transactions on Robotics and Automation, Vol. 1, February 2002, pp. 1-10.

A further development of the technical teaching of US 6,144,884 A. is the article "Time Domain Passivity Control-based Telepresence with Time Delay" by Artigas, Jordi, Vilanova, J., Preusche, C., Hirzinger, G. in IEEE / RSJ International Conference on Intelligent Robotics and Systems, Beijing, China, 2006, p. 4205-4210. It proposes the use of two "passivity observers" and two "passivity controllers" in order to take account of the energy flow between the master system and the slave system in two directions (forwards / backwards) during passivation.

Further, prior art reference is made to the articles of Jee-Hwan Ryu, Jordi Artigas, and Cartsen Preusche: "A passive bilateral control scheme for a teleoperator with time-varying communication delay" in Mechatronics, Volume 20, October 2010, Issue 7, Pp. 812-823 and Artigas, Jordi; Jee-Hwan Ryu; Preusche, C .; Hirzinger, G. referred to "Network representation and passivity of delayed teleoperation systems" in the IEEE / RSJ International Conference on Intelligent Robots and Systems (IROS), 2011, pp. 177-183.

From the article by Beibei Han; Jee-Hwan Ryu; and Il-Kyun Jung: "FPGA based time domain Passivity Observer and Passivity Controllers" in IEEE / ASME International Conference on Advanced Intelligent Mechatronics, AIM 2009, 2009, pp. 433-438, is an implementation of a control structure for the remote control of a Actuator via a haptic input means, in which the slave controller additionally assumes the task of a passivity server and a passivity controller. The Passivity Observer monitors whether the control network stability criteria are met. If the energy balance is negative, the control variable at the slave is reduced by the passivity controller.

The object of the invention is to further develop a control network such that system instabilities have no detrimental, in particular harmful, effects on an operator of the control network or the input device.

The invention results from the features of the independent claims. Advantageous developments and refinements are the subject of the dependent claims.

Other features, applications and advantages of the invention will become apparent from the following description, as well as the explanation of embodiments of the invention, which are illustrated in the figures.

There is proposed herein a control network comprising a master system with a haptic input means, and a slave system connected to the master system via a bi-directional communication channel. The communication channel can be wired or wireless. In the present case, the haptic input means is also designed to output a haptic feedback signal on the basis of the response variable f _a (t) mentioned below. By the master system, a control variable s _m (t) for controlling the slave system can be generated and communicated to the slave system. The control variable s _m (t) is based in particular on a manual input of an operator into the haptic input means. On the other hand, a response variable f _a (t) can be generated by the slave system for controlling the master system on the basis of the control variable s _m (t) received by the slave system and transmitted to the master system. The control variable s _m (t) received by the slave system may be delayed in time or changed in comparison to the control variable s _m (t) emitted by the master system due to the transmission. The same applies analogously to the response variable f _a (t) emitted by the slave system. However, these effects are not further elaborated or considered here for the sake of simplicity

The control variable s _m (t) can be a flow quantity, in particular a first time derivative of a calculated or measured position or a calculated or measured speed, and the response variable a potential variable, in particular a calculated or measured force. The term "flow size" and "potential size" in the present context are known to the person skilled in the art, so that the present technical teaching can be applied or transferred to a large number of further flow quantities / potential variables known to a person skilled in the art. Alternatively, the control variable s _m (t) may be a potential variable, in particular a calculated or measured force, and the response variable f _a (t) a flow quantity, in particular a first time derivative of a calculated or measured position or a calculated or measured speed.

The proposed control network furthermore comprises a first means with which a first energy quantity EP (t) can be determined on the basis of the control variable s _m (t) emitted by the master system and the response variable f _a (t) received by the master system. which indicates an energy that is transferred from the slave system to the master System flows, and with the basis of the control variable s _m (t) and the response variable f _a (t) a second energy quantity EM (t) can be determined, which indicates an energy flowing from the master system in the direction of the slave system ,

Finally, the proposed control network comprises a second means which, in the event that EP (t)> EM (t), triggers an issuance of a warning to a user of the master system by an output means, and that in case EP (t )> EM (t) + E _threshhold is to _switch the control network to a default security state, where E _threshhold is set and can be time constant or time varying depending on the application. The output of a warning by the output means is advantageously carried out optically, acoustically and / or haptically.

This means that the second means will issue a warning if more power is flowing to the master system from the slave system than flowing from the master system towards the slave system. If this is the case, then the control network is no longer in a guaranteed stable state, but is approaching an unstable state. If the (absolute) difference of the energy flowing in and out of the master system _exceeds the threshold E _threshhold , ie _{if there} is a potentially dangerous unstable state, the control network is switched by the second means to a predetermined safety state, which is advantageous limits or deactivates an activity of the haptic input device such that no response or movement of input means can be caused by a response signal f _a (t) received from the master system, which could (mechanically) damage an operator.

Which cause has also led to the fact that the control system has developed from a stable to an unstable state, is not essential in the present case and can therefore be left out. The proposed control network ensures in any case that the operator is protected in a potentially dangerous unstable network state by warning and, for example, a shutdown or partial shutdown of the control network.

Advantageously, the security state can be defined by the following features:
the response variable f _a (t) is set to zero: f _a (t) = 0, and / or the power supply of the haptic input device is switched off, and / or the mechanical drive means of the haptic input device are switched off, and / or the response variable f _a ( t) and / or the control variable s _m (t) is / are reduced, for example by f _a (t) '= f _a (t) / X _a and / or s _m (t)' = s _m (t) / X _m , where X _a and X _{m are} predetermined factors and f _a (t) 'is the reduced response magnitude and s _m (t)' is the reduced control magnitude, and / or movement of the haptic input means is mechanically blocked by a switchable mechanism. Other features and combinations of features of the control network for the protection of the operator which are obvious to the person skilled in the art are encompassed by the inventive idea.

The control network can be a telepresence system or a "virtual reality" system, in the first case an interaction between an operator (at the master system) and a remote real environment via a robot and in the second case an interaction between the operator and a user virtual environment takes place.

In a further development E _threshhold = E _{0 is} a constant energy value. The energy value E ₀ is advantageously selected depending on the desired security requirement and other user requirements. A low selected energy value E ₀ leads to a faster switching of the control network into the safety state, with the result that the control network never reaches a critical, unstable state.

In an advantageous development, the slave system has a third means, which determines an active energy input E ₁ (t) into the slave system, and transmits this to the second means. Furthermore, in this development, the means is set up and implemented in such a way that in the case that EP (t)> EM (t) + E is _threshhold (t), with E _threshold (t) = E ₀ + E ₁ (t), Control network switches to a predetermined safety state (see above), where E _{0 is} a predetermined constant energy value. An active energy input is present, for example, when the slave system (for example a robot arm) is actively moved by an operator and thus the energy E ₁ (t) is introduced. This energy E ₁ (t) is determined by the slave system and transmitted to the master system. Since it may be desired that this energy input E ₁ (t) is output as feedback via the haptic input means. This development thus makes it possible to take account of active energy sources in the slave system whose energy input E ₁ (t) should not lead to switching to the safety state.

The invention further relates to a method for controlling a control network comprising a master system with a haptic input means, and a slave system connected to the master system via a bi-directional communication channel, wherein the master system controls a control variable s _m (t) generated to drive the slave system and to the slave system ( 104 ) is transmitted by the slave system ( 104 ) for controlling the master system on the basis of the slave system ( 104 ) Control quantity received s _m (t) is a response size f is _a generated (t) and transmitted to the master system, on the basis of the control variable s _m (t) and received from the master system response size f _a (t) a first energy size EP (t) is determined, which indicates an energy which is produced by the slave system ( 104 ) into the master system ( 101 ), and with which a second energy quantity EM (t) is determined based on the control variable s _m (t) and the response variable f _a (t), which indicates an energy which originates from the master system ( 101 ) in the direction of the slave system ( 104 ), and in case EP (t)> EM (t), a warning to a user of the master system ( 101 ), and in the case that EP (t)> EM (t) + E _threshhold , the control network is switched to a predetermined security state, where E _{threshhold is} given, and may be time constant or time varying depending on the application ,

Further developments of the proposed method and advantages result from a meaningful and analogous transmission of the statements made above to the control network.

The object of the invention is further achieved by a computer system having a data processing device, wherein the data processing device is configured such that a method as described above is performed on the data processing device.

In addition, the object of the invention is achieved by a digital storage medium with electronically readable control signals, wherein the control signals can interact with a programmable computer system so that a method as described above, is performed.

Furthermore, the object of the invention is achieved by a computer program product with program code stored on a machine-readable carrier for carrying out the method, as described above, when the program code is executed on a data processing device.

Finally, the invention relates to a computer program with program codes for carrying out the method, as described above, when the program runs on a data processing device. For this, the data processing device can be configured as any known from the prior art computer system.

Further advantages, features and details will become apparent from the following description in which - where appropriate, with reference to the drawings - at least one embodiment is described in detail. The same, similar and / or functionally identical parts are provided with the same reference numerals. Show it:

1 a greatly simplified construction of a control network in one embodiment, and

2 a schematic flow diagram of a method according to the invention.

1 shows a greatly simplified construction of a control network comprising a master system 101 with a haptic input device 102 , and via a bi-directional communication channel 103 with the master system 101 connected slave system 104 , Through the master system 101 is a control variable s _m (t) for controlling the slave system 104 producible and to the slave system 104 is transferable. Through the slave system 104 is for controlling the master system based on the slave system 104 received control variable s _m (t) a response variable f _a (t) can be generated and communicated to the master system. The control variable s _m (t) indicates a speed v (t) in the present case. The response variable f _a (t) indicates a force f (t) in the present case.

• – Steuergröße S_m(t) ist eine Position und Antwortgröße f_a(t) ist eine Kraft
• – Steuergröße S_m(t) ist eine Geschwindigkeit und Antwortgröße f_a(t) ist eine Kraft
• – Steuergröße S_m(t) ist eine Kraft und Antwortgröße f_a(t) ist eine Geschwindigkeit
• – Steuergröße S_m(t) ist eine Kraft und Antwortgröße f_a(t) ist eine Position

The following combination of control variable s _m (t) and response variable f _a (t) are conceivable:

• Control variable S _m (t) is a position and response variable f _a (t) is a force
• Control variable S _m (t) is a speed and response variable f _a (t) is a force
• Control variable S _m (t) is a force and response variable f _a (t) is a velocity
• Control variable S _m (t) is a force and response variable f _a (t) is a position

The control network here includes one in the master 101 arranged first means 105 in which, on the basis of the control variable s _m (t) and the response variable f _a (t) received by the master system, a first energy quantity EP (t) can be determined, which indicates an energy that originates from the slave system 104 into the master system 101 flows, and with the basis of the control variable s _m (t) and the response variable f _a (t) a second energy quantity EM (t) can be determined, which indicates an energy from the master system 101 in the direction of the slave system 104 flows.

Furthermore, the control network in the present case includes one in the master 101 arranged second means 106 which, in the event that EP (t)> EM (t), is an output of a warning to a user of the master system 101 by an output means 107 and, in the event that EP (t)> EM (t) + E _{threshhold, sets} the control network to a default security state, with E _threshhold set.

The slave system 104 For example, it may be a real or a virtual robot interacting with an environment, for example. The haptic input device 102 may be, for example, a joystick with one or more degrees of freedom

The coupling between master system 101 and slave system 104 must be stable to an operator of the master system 101 not by uncontrolled (response) forces from the input device 102 be created to endanger. In the present case, stability means that the amplitudes of the system variables s _m (t) and f _a (t) never rise above a predetermined safety limit.

The proposed control network enables its stability in real-time operation for all conceivable causes that could lead to instability of the control network, since the control network is switched to the security state as soon as the second means 106 a development trend towards instability is detected.

The energy quantities EP (t) and E _- (t) present in or out of the master system 101 Flow, can be easily determined on the basis of the power P, ie the product of the speed v (t) (= s _m (t)) and the force f (t) (= f _a (t)).

A positive sign of P indicates an inflow of power to the master system 101 at. A negative sign of P indicates an outflow of power from the master system 101 at. By integrating the line flows, one obtains two energy flows:

In this case, EP (0) or EM (0) is each an initially stored energy. Where P ₊ and P _-

A first case is that the operator by means of his manual input to the input means 102 the only power source of the control network is and therefore the master system and the rest of the network is passive. In particular, it is initially assumed that the slave system 104 no active energy input (for example by the environment) takes place. In this case, more energy flows from the master system to the slave system 104 as the master system 101 from the slave system 104 the control network is passive and therefore stable.

The control network becomes unstable when the energy flow to the master system 101 becomes larger than the energy drain from the master system 101 , In this case, a warning is issued. Furthermore, by the second means 106 the immediate switching of the control network into a safety state, if in addition the energy supply to the master system 101 is greater than a given energy value E ₀ . EP (t)> EM (t) -> Output of a warning. (5) EP (t)> EM (t) + E ₀ -> switching to the safety state (6)

A second case is that in addition to the slave system 104 an active energy input E ₁ (t) takes place, which is used as feedback (feedback) at the input means 102 should be reflected without the control network is immediately switched to the security state

In this case, the slave system 104 a third remedy 108 present that detects the energy input E ₁ (t) and to the second means 106 forwards. In this case, the second means 104 set up and executed so that as before the issue of a warning then takes place when the energy flow to the master system 101 becomes larger than the energy drain from the master system 101 , In contrast to the first case, switching to the safety state will only take place if the following applies: EP (t)> EM (t) + E ₀ + E ₁ (t) -> switching to the safety state (7)

2 shows a schematic flow diagram of a method according to the invention for controlling a control network comprising a master system 101 with a haptic input device 102 , and via a bi-directional communication channel 103 with the master system 101 connected slave system 104 , with the following steps.

In a first step 201 is through the master system 101 a control variable s _m (t) for controlling the slave system 104 generated and sent to the slave system 104 transmitted. In a second step 202 is through the slave system 104 for controlling the master system on the basis of the slave system 104 received control variable s _m (t) generates a response variable f _a (t) and to the master system 101 transmitted. In a third step 203 On the basis of the control variable s _m (t) and the response variable f _a (t) received by the master system, a first energy quantity EP (t) is determined, which indicates an energy that originates from the slave system 104 into the master system 101 flows, and on the basis of the control variable s _m (t) and the response variable f _a (t) a second energy quantity EM (t) is determined, which indicates an energy from the master system 101 in the direction of the slave system 104 flows. In a fourth step, in case EP (t)> EM (t), a warning is sent to a user of the master system 101 and if EP (t)> EM (t) + E _threshhold , the control network is switched to a default security state, with E _threshhold set.

LIST OF REFERENCE NUMBERS

101

Master system

102

input means

103

communication channel

104

Slave system

105

first means

106

second means

107

output means

108

third means

Claims (10)

Control network comprising a master system ( 101 ) with a haptic input means ( 102 ), and via a bi-directional communication channel ( 103 ) with the master system ( 101 ) connected slave system ( 104 ), whereby - by the master system ( 101 ) a control variable s _m (t) for controlling the slave system ( 104 ) and to the slave system ( 104 ), - by the slave system ( 104 ) for controlling the master system on the basis of the slave system ( 104 received control variable s _m (t) a response variable f _a (t) can be generated and transmitted to the master system, - a first means ( 105 ) is present, with the basis of the control quantity s _m (t) and the received response from the master system response variable f _a (t) a first energy quantity EP (t) can be determined, which indicates an energy that from the slave system ( 104 ) into the master system ( 101 ), and with which on the basis of the control variable s _m (t) and the response variable f _a (t), a second energy quantity EM (t) can be determined, which indicates an energy that originates from the master system ( 101 ) in the direction of the slave system ( 104 ), and - a second means ( 106 ), which, in the event that EP (t)> EM (t), is an output of a warning to a user of the master system ( 101 ) by an output means ( 107 ), and that in case EP (t)> EM (t) + E _threshhold is to _switch the control network to a default security state, with E _threshhold set. A control network according to claim 1, wherein the security state is defined as follows: - the response quantity f _a (t) is set to zero: f _a (t) = 0, and / or - the haptic input means power supply is turned off, and / or - the mechanical drive means of the haptic input means are switched off, and / or - the response variable f _a (t) and / or the control variable s _m (t) is / are reduced, for example by f _a (t) '= f _a (t) / Xa and / or s _m (t) '= s _m (t) / Xm, where Xa and Xm are predetermined factors and f _a (t)' the reduced response size and s _m (t) 'the reduced control quantity are and / or - a movement of the haptic input means is mechanically blocked by a switchable mechanism. Control network according to claim 1 or 2, in which the control variable s _m (t) is a flow quantity, in particular a first time derivative of a calculated or measured position or a calculated or measured speed, and the response variable is a potential variable, in particular a calculated or measured force is. Control network according to claim 1 or 2, in which the control variable s _m (t) is a potential variable, in particular a calculated or measured force, and the response variable f _a (t) is a flow variable, in particular a first time derivative of a calculated or measured one Positions or a calculated or measured speed. A control network according to any one of claims 1 to 4, wherein the control network is a telepresence system. Control network according to one of Claims 1 to 5, in which the slave system ( 104 ) is a robot or a virtual reality system. Control network according to one of claims 1 to 6, wherein the output of a warning by the output means is optical, acoustic and / or haptic. Control network according to one of claims 1 to 7, _wherein E _threshhold = E _{0 is} a constant energy value. Control network according to one of Claims 1 to 7, in which the slave system ( 104 ) a third remedy ( 108 ) having an active energy input E ₁ (t) into the slave system ( 104 ) and send it to the second 106 ), and the second means is set up and executed in such a way that in case EP (t)> EM (t) + E is _threshhold (t), with E _threshold (t) = E ₀ + E ₁ (t), the control network switches to a predetermined safety state, where E _{0 is} a predetermined constant energy value. Method for controlling a control network according to the preceding claims, comprising a master system ( 101 ) with a haptic input means ( 102 ), and via a bi-directional communication channel ( 103 ) with the master system ( 101 ) connected slave system ( 104 ), in which - by the master system ( 101 ) a control variable s _m (t) for controlling the slave system ( 104 ) and sent to the slave system ( 104 ) - by the slave system ( 104 ) for controlling the master system on the basis of the slave system ( 104 ) Control quantity received s _m (t) is a response size f is _a generated (t) and transmitted to the master system, - on the basis of the control variable s _m t) and received from the master system response size f _a (t) (a first Energy quantity EP (t) is determined, which indicates an energy which is generated by the slave system ( 104 ) into the master system ( 101 ), and with which a second energy quantity EM (t) is determined based on the control variable s _m (t) and the response variable f _a (t), which indicates an energy which originates from the master system ( 101 ) in the direction of the slave system ( 104 ) and, in the case that EP (t) is> EM (t), a warning to a user of the master system ( 101 ), and in case EP (t)> EM (t) + E _threshhold , the control network is switched to a predetermined security state, with E _threshhold given.

Images