DRIVE UNIT ADOPTING ADMITTANCE CONTROL

Abstract

A drive unit 10A is configured to exert a driving force on an environment 50 in accordance with a target driving force command τ.sub.d, and includes a parameter storage device 30A, a force measuring instrument 35, an admittance model calculation device 31A, and a position control and driving device 33A. The parameter storage device 30A has stored therein dynamics parameters of first and second virtual objects affected by a virtual interactive force λ.sub.R. The force measuring instrument 35 is configured to output a measurement result for the driving force as a measured driving force value τ.sub.s. The admittance model calculation device 31A is configured to calculate and output a displacement of the first virtual object. The displacement is obtained by calculations based on the stored dynamics parameters, the target driving force command τ.sub.d, and the measured driving force value τ.sub.s. The position control and driving device 33A is configured to operate in accordance with a target position command. The force measuring instrument 35 is disposed between the position control and driving device 33A and the environment 50. The target position command corresponds to the first virtual object's displacement outputted by the admittance model calculation device 31A. The drive unit 10A achieves advantages of both high and low backdrivability.

Claims

1. A drive unit exerting a driving force on an environment in accordance with an externally provided target driving force command, the drive unit comprising: a parameter storage device having stored therein dynamics parameters of first and second virtual objects affected by a virtual interactive force; a force measuring instrument configured to output a measurement result for the driving force as a measured driving force value; an admittance model calculation device configured to calculate and output a displacement of the first virtual object, the displacement being obtained by calculations based on the first and second virtual objects' dynamics parameters stored in the parameter storage device, the target driving force command, and the measured driving force value; and a position control and driving device configured to operate in accordance with a target position command and thereby exert the driving force on the environment, wherein, the force measuring instrument is disposed between the position control and driving device and the environment, and the target position command corresponds to the first virtual object's displacement outputted by the admittance model calculation device.

2. The drive unit according to claim 1, wherein the position control and driving device is a single-axis device configured to operate in accordance with the target position command.

3. The drive unit according to claim 1, wherein the first and second virtual objects' dynamics parameters stored in the parameter storage device are externally modifiable.

4. The drive unit according to claim 1, wherein, the measured driving force value directly acts only on the first virtual object's dynamics in the calculations by the admittance model calculation device, the target driving force command directly acts only on the second virtual object's dynamics in the calculations by the admittance model calculation device, and the first and second virtual objects' dynamics interact with each other solely via the virtual interactive force outputted by a virtual interactive force model.

5. The drive unit according to claim 4, wherein, the parameter storage device further stores interaction parameters for the virtual interactive force model, and the interaction parameters include upper and lower limit values of the virtual interactive force.

6. The drive unit according to claim 4, wherein, the parameter storage device further stores interaction parameters for the virtual interactive force model, the virtual interactive force model has a function that converges a relative displacement or velocity between the first and second virtual objects to zero, and the interaction parameters include a parameter for the convergence.

7. The drive unit according to claim 5, wherein the virtual interactive force model's interaction parameters stored in the parameter storage device are externally modifiable.

8. A drive unit exerting a driving force on an environment in accordance with an externally provided target driving force command, the drive unit comprising: a parameter storage device having stored therein dynamics parameters of first and second virtual objects affected by a virtual interactive force; a force measuring instrument configured to output a measurement result for the driving force as a measured driving force value; an admittance model calculation device configured to calculate and output a velocity of the first virtual object, the velocity being obtained by calculations based on the first and second virtual objects' dynamics parameters stored in the parameter storage device, the target driving force command, and the measured driving force value; and a velocity control and driving device configured to operate in accordance with a target velocity command and thereby exert the driving force on the environment, wherein, the force measuring instrument is disposed between the velocity control and driving device and the environment, and the target velocity command corresponds to the first virtual object's velocity outputted by the admittance model calculation device.

9. The drive unit according to claim 8, wherein the velocity control and driving device is a single-axis device configured to operate in accordance with the target velocity command.

10. The drive unit according to claim 8, wherein the first and second virtual objects' dynamics parameters stored in the parameter storage device are externally modifiable.

11. The drive unit according to claim 8, wherein, the measured driving force value directly acts only on the first virtual object's dynamics in the calculations by the admittance model calculation device, the target driving force command directly acts only on the second virtual object's dynamics in the calculations by the admittance model calculation device, and the first and second virtual objects' dynamics interact with each other solely via the virtual interactive force outputted by a virtual interactive force model.

12. The drive unit according to claim 11, wherein, the parameter storage device further stores interaction parameters for the virtual interactive force model, and the interaction parameters include upper and lower limit values of the virtual interactive force.

13. The drive unit according to claim 11, wherein, the parameter storage device further stores interaction parameters for the virtual interactive force model, the virtual interactive force model has a function that converges a relative displacement or velocity between the first and second virtual objects to zero, and the interaction parameters include a parameter for the convergence.

14. The drive unit according to claim 12, wherein the virtual interactive force model's interaction parameters stored in the parameter storage device are externally modifiable.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG. 1 is a schematic entity-relationship diagram of a drive unit according to a first embodiment of the present invention.

[0025] FIG. 2 is a block diagram of the drive unit according to the first embodiment of the present invention.

[0026] FIG. 3 is a diagram showing the relationship between first and second virtual objects in the present invention.

[0027] FIG. 4 is a schematic entity-relationship diagram of a drive unit according to a second embodiment of the present invention.

[0028] FIG. 5 is a block diagram of the drive unit according to the second embodiment of the present invention.

[0029] FIG. 6 is a schematic entity-relationship diagram of a drive unit according to a third embodiment of the present invention.

[0030] FIG. 7 is a block diagram of the drive unit according to the third embodiment of the present invention.

[0031] FIG. 8 is a schematic entity-relationship diagram of a drive unit according to a variant of the present invention.

[0032] FIG. 9 is a block diagram of the drive unit according to the variant of the present invention.

[0033] FIG. 10 is a block diagram of a conventional drive unit.

MODES FOR CARRYING OUT THE INVENTION

[0034] Hereinafter, some embodiments of a drive unit according to the present invention will be described with reference to the accompanying drawings.

First Embodiment

[0035] FIG. 1 shows a schematic entity-relationship diagram of a drive unit 10A according to a first embodiment of the present invention. The drive unit 10A according to the present embodiment is configured to exert a driving force on an environment in accordance with an externally provided target driving force command τ.sub.d, and includes a control portion 20A and a drive portion 21A.

[0036] The control portion 20A includes a microprocessor and accompanying volatile/non-volatile memory.

[0037] The drive portion 21A includes a driving device 41, which is a single-axis rotary motor, a position measuring instrument 42 configured to measure a displacement of the driving device 41 and output the measurement result as a measured position value, a transmission device 34, which is a reduction gear provided on an output shaft of the driving device 41, and a force measuring instrument 35, which is a torque sensor provided on a terminal end portion (terminal output shaft) of the transmission device 34. The force measuring instrument 35 is configured to measure the driving force that the drive unit 10A exerts on the environment, and output the measurement result as a measured driving force value τ.sub.s. The position of the driving device 41 is controlled in accordance with a driving current provided by the control portion 20A.

[0038] It should be noted that the driving device 41 may be any actuator or multi-axis robot. The transmission device 34 may be any power train (including a clutch, a transmission, a driveshaft, and a linkage) configured to transmit an internal driving force of the driving device 41 to the environment after converting the driving force into a suitable form. Moreover, in the case where the driving device 41 is of a translational type, the force measuring instrument 35 is preferably a force sensor.

[0039] FIG. 2 shows a block diagram of the drive unit 10A according to the present embodiment. As shown in the figure, the control portion 20A of the drive unit 10A includes a parameter storage device 30A, an admittance model calculation device 31A, an inverse model calculation device 32A, and a position controller 40, which along with the driving device 41 and the position measuring instrument 42, constitutes a position control and driving device 33A. The position control and driving device 33A operates in accordance with a target position command.

[0040] The admittance model calculation device 31A is configured to perform time integration of simultaneous differential equations (1) to (3) and thereby simulate the motion of first and second virtual objects in accordance with the target driving force command τ.sub.d and the measured driving force value τ.sub.s.

First Virtual Object's Dynamics

[0041]
κMp″.sub.1+γDp′.sub.1=−τ.sub.s−λ.sub.R  (1)

Second Virtual Object's Dynamics

[0042]
(1−κ)Mp″.sub.2+(1−γ)D′p.sub.2=τ.sub.d+λ.sub.R  (2)

Virtual Interactive Force Model

[0043]
λ.sub.R=λ.sub.R(p″.sub.1,p′.sub.1,p.sub.1,p″.sub.2,p′.sub.2,p.sub.2)  (3)

where

[0044] M and κ are total virtual object inertia and an inertia distribution ratio (0<κ<1),

[0045] D and γ are total virtual object viscous friction and a viscous friction distribution ratio (0<γ<1),

[0046] p.sub.1, p′.sub.1, and p″.sub.1 are the first virtual object's displacement, velocity, and acceleration,

[0047] p.sub.2, p′.sub.2, and p″.sub.2 are the second virtual object's displacement, velocity, and acceleration, and

[0048] λ.sub.R is a virtual interactive force.

[0049] From the viewpoint of ensuring control stability and reducing the load on the mechanism of the drive portion 21A, the total virtual object inertia M is preferably set so as to match the actual total inertia of the mechanism. However, as in the case of model-based control, there is no need to take account of modeling errors, and if any modeling error occurs, there might be no problem so long as control stability is ensured.

[0050] On the other hand, the total virtual object viscous friction D does not have to match the actual total viscous friction of the mechanism of the drive portion 21A. However, viscous friction affects the ensuring of control stability and the improvement of positioning performance, and therefore needs to be properly set in view of overall balance.

[0051] The setting of the inertia distribution ratio κ and the viscous friction distribution ratio γ will be described later.

[0052] The parameter storage device 30A has the total virtual object inertia M, the inertia distribution ratio κ, the total virtual object viscous friction D, and the viscous friction distribution ratio γ stored as the first and second virtual objects' dynamics parameters. In the present embodiment, these parameters can be externally modified at any time.

[0053] It should be noted that instead of storing the parameters M, κ, D, and γ, the parameter storage device 30A may have first virtual object inertia M.sub.1 (=κM) and first virtual object viscous friction D.sub.1 (=γD) stored as the first virtual object's dynamics parameters and second virtual object inertia M.sub.2 (=(1−κ)M) and second virtual object viscous friction D.sub.2 (=(1−γ)D) stored as the second virtual object's dynamics parameters. It is preferred that these parameters can also be externally modified at any time.

[0054] FIG. 3 shows the relationship between the first and second virtual objects. As is apparent from the figure, the first and second virtual objects' dynamics are treated separately in calculations performed by the admittance model calculation device 31A, and the first and second virtual objects interact with each other solely via the virtual interactive force λ.sub.R outputted by the virtual interactive force model. Moreover, in the calculations performed by the admittance model calculation device 31A, the target driving force command τ.sub.d directly acts only on the second virtual object's dynamics, and the measured driving force value τ.sub.s directly acts only on the first virtual object's dynamics.

[0055] Referring back to FIG. 2, the inverse model calculation device 32A is configured to perform an inverse calculation and thereby convert the first virtual object's displacement p.sub.1 (specifically, the displacement at a terminal end of the drive unit 10A) calculated by the admittance model calculation device 31A into a target position command q.sub.d for the position control and driving device 33A.

[0056] More specifically, in the present embodiment in which the transmission device 34 is a reduction gear, the displacement p at the terminal end of the drive unit 10A and a displacement q of the position control and driving device 33A have a geometric relationship as given by equation (4), where n.sub.r is a reduction ratio of the reduction gear, and therefore the target position command q.sub.d for the position control and driving device 33A can be obtained using an inverse transmission device model as given by equation (5).

[00001]$\begin{array}{cc}p=p\left(q\right)=\frac{q}{n_r}& \left(4\right)\end{array}$$\begin{array}{cc}{q}_d=p^{-1}\left(p_1\right)=n_r{p}_1& \left(5\right)\end{array}$

[0057] In other words, in the case where there is a geometric relationship (kinematics) as given by equation (6) between an input and an output of the transmission device 34, the inverse model calculation device 32A performs an inverse calculation (inverse kinematics) as given by equation (7).

p=p(q)  (6)

q=p.sup.−1(p)  (7)

[0058] The position controller 40, which constitutes a part of the position control and driving device 33A, is configured to control the position of the driving device 41 in accordance with the target position command. At the time of the position control, the position controller 40 references the measured position value outputted by the position measuring instrument 42.

[0059] In this manner, in the case of the drive unit 10A according to the present embodiment, the motion of the two virtual objects (i.e., the first and second virtual objects) linked only by the virtual interactive force λ.sub.R is simulated on the basis of driving force information (measured driving force value τ.sub.s) obtained by the force measuring instrument 35, and the position of the driving device 41 is controlled such that the driving device 41 follows the motion of the first virtual object. Therefore, properly setting the distribution ratios κ and γ allows the drive unit 10A to achieve advantages of both high and low backdrivability.

Second Embodiment

[0060] FIGS. 4 and 5 illustrate a drive unit 10B according to a second embodiment of the present invention. The drive unit 10B according to the present embodiment differs from the drive unit 10A in that the drive unit 10B includes a control portion 20B and a drive portion 21B in place of the control portion 20A and the drive portion 21A.

[0061] The control portion 20B has the same configuration as the control portion 20A except that the control portion 20B includes an inverse model calculation device 32B and a velocity controller 43 in place of the inverse model calculation device 32A and the position controller 40.

[0062] The drive portion 21B has the same configuration as the drive portion 21A except that the drive portion 21B includes a velocity measuring instrument 44 in place of the position measuring instrument 42.

[0063] The inverse model calculation device 32B is configured to perform an inverse calculation and thereby convert the first virtual object's velocity p′.sub.1 calculated by the admittance model calculation device 31A into a target velocity command for a velocity control and driving device 33B.

[0064] The velocity controller 43, which constitutes a part of the velocity control and driving device 33B, is configured to control the velocity of the driving device 41 in accordance with the target velocity command. At the time of the velocity control, the velocity controller 43 references a velocity (measured velocity value) of the driving device 41 as obtained by the velocity measuring instrument 44.

[0065] Properly setting the distribution ratios κ and γ allows the drive unit 10B according to the present embodiment to achieve advantages of both high and low backdrivability, as done in the first embodiment.

Third Embodiment

[0066] FIGS. 6 and 7 illustrate a drive unit 10C according to a third embodiment of the present invention. The drive unit 10C according to the present embodiment differs from the drive unit 10A in that the drive unit 10C includes a control portion 20C in place of the control portion 20A, but the drive unit 10C includes the same drive portion as the drive portion 21A included in the drive unit 10A.

[0067] The control portion 20C has the same configuration as the control portion 20A except that the control portion 20C includes a parameter storage device 30C and an admittance model calculation device 31C in place of the parameter storage device 30A and the admittance model calculation device 31A.

[0068] The admittance model calculation device 31C is configured to perform time integration of a system of simultaneous differential equations consisting of (1) and (2) above and (8) and (9) below and thereby simulate the motion of the first and second virtual objects in accordance with the target driving force command τ.sub.d and the measured driving force value τ.sub.s.

Virtual Interactive Force Model

[0069] [00002]$\begin{array}{cc}{\lambda }_R=\{\begin{array}{cc}{R}_{\mathrm{upper}}& \left(R_{\mathrm{upper}}<{\lambda }_{\mathrm{RPD}}\right)\\ {\lambda }_{\mathrm{RPD}}& \left(R_{\mathrm{lower}}\le {\lambda }_{\mathrm{RPD}}\le R_{\mathrm{upper}}\right)\\ R_{\mathrm{lower}}& \left({\lambda }_{\mathrm{RPD}}<R_{\mathrm{lower}}\right)\end{array}& \left(8\right)\end{array}$$\begin{array}{cc}{\lambda }_{\mathrm{RPD}}={\lambda }_{\mathrm{RPD}}\left(p_1^{\prime },p_2^{\prime },p_1,p_2\right)=K_{\mathrm{RP}}\left(p_2-p_1\right)+K_{\mathrm{RD}}\left(p_2^{\prime }-p_1^{\prime }\right)& \left(9\right)\end{array}$

where

[0070] R.sub.upper is an upper limit value of the virtual interactive force,

[0071] R.sub.lower is a lower limit value of the virtual interactive force, and

[0072] K.sub.RP and K.sub.RD are gains for the virtual interactive force model.

[0073] The parameter storage device 30C has stored therein virtual interactive force parameters in addition to the first and second virtual objects' dynamics parameters, and the virtual interactive force parameters include the upper and lower limit values R.sub.upper and R.sub.lower of the virtual interactive force and the gains K.sub.RP and K.sub.RD for the virtual interactive force model. It is preferred that these parameters can also be externally modified at any time.

[0074] For the drive unit 10C according to the present embodiment, the virtual interactive force model includes two additional functions to be described below.

[0075] Described first is a “virtual torque limiter function”.

[0076] In the case where an external force from the environment 50 is greater than or equal to R.sub.lower but less than or equal to R.sub.upper, the virtual interactive force λ.sub.R is equal to λ.sub.RPD as given by equation (9). In this case, the first and second virtual objects can be regarded as moving as one in response to the external force. On the other hand, when the external force is greater than R.sub.upper or less than R.sub.lower, the first and second virtual objects move relative to each other. In other words, the first virtual object slides relative to the second virtual object.

[0077] In the present embodiment, as in the first embodiment, the inverse of the first virtual object's displacement p.sub.1 is used as the target position command for the position control and driving device 33A. Accordingly, when the first virtual object slides relative to the second virtual object, it can be perceived in the environment 50 as if the terminal output shaft of the drive unit 10C slides. Moreover, at this time, the actual torque that is applied to mechanical components of the drive portion 21A is limited.

[0078] In summary, this function is used to set the upper and lower limit values of the virtual interactive force λ.sub.R so as to limit the load on the mechanical components of the drive portion 21A (in particular, fragile components such as the transmission device 34 and the force measuring instrument 35) to a certain value or less and thereby prevent the mechanical components from being broken or damaged. That is, the drive unit 10C according to the present embodiment renders it possible to protect hardware in a software-like manner.

[0079] In the case where the driving device 41 is of a translational type, the actual force that is applied to the mechanical components of the drive portion 21A is limited. Accordingly, in such a case, the above function should be referred to as the “virtual force limiter”.

[0080] Described next is a “sliding deviation restoring function”.

[0081] Once the virtual torque limiter (virtual force limiter) is activated, the first and second virtual objects move relative to each other. The sliding deviation restoring function renders it possible to converge the relative displacement or velocity between the first and second virtual objects to zero. In other words, this function renders it possible to restore position and/or velocity deviations caused by sliding.

[0082] In the present embodiment, this function is realized using the parameter λ.sub.RPD as given by equation (9). Setting such a parameter allows the first and second virtual objects to be connected by a virtual spring-damper and thereby results in achieving the effects described above. The degree (or intensity) of convergence can be adjusted by the gains K.sub.RP and K.sub.RD.

[0083] The drive unit 10C according to the present embodiment can achieve advantages of both high and low backdrivability, as done in the first and second embodiments.

[0084] The effects of the present embodiment will be described in more detail using an example where the inertia distribution ratio κ is set to a relatively low value of 0.1 so that the inertias of the first and second virtual objects are 0.1 M and 0.9 M, respectively.

[0085] When the virtual torque limiter is not activated, the first and second virtual objects move as one, and therefore it can be perceived in the environment 50 as if the inertia at the terminal output shaft of the drive unit 10C is M (=0.1 M+0.9 M). Therefore, in the case of the drive unit 10C according to the present embodiment, matching the value M to the actual inertia of the drive portion 21A prevents an excess load from being applied to the driving device 41 while achieving more stable position control. That is, it is rendered possible to achieve advantages of low backdrivability.

[0086] On the other hand, when the virtual torque limiter is activated, the inertia at the terminal output shaft as viewed from the environment 50 can be lowered to the minimum of 0.1 M. Therefore, the drive unit 10C according to the present embodiment can protect the mechanical components of the drive portion 21A against overload. That is, it is rendered possible to achieve an advantage of high backdrivability.

[0087] In the present invention, the total virtual object inertia M and the total virtual object viscous friction D are distributed to the first and second virtual objects at the distribution ratios κ and γ, as described earlier. The distribution ratios κ and γ affect the behavior of the first and second virtual objects where the virtual torque limiter is activated. As the distribution ratios κ and γ decrease, overload protection performance increases. However, it should be noted that if the distribution ratios κ and γ are excessively decreased, position control in the position control and driving device 33A might become unstable when the virtual torque limiter is activated.

[0088] <Variants>

While the first through third embodiments of the drive unit according to the present invention have been described above, the present invention is not limited to the configurations of these embodiments.

[0089] For example, a drive unit 10D according to a variant of the present invention may include a control portion 20D without an inverse model calculation portion and a drive portion 21D without a transmission device (see FIGS. 8 and 9). The drive unit 10D can be considered equivalent to the drive unit 10C according to the third embodiment without the inverse model calculation device 32A and the transmission device 34. In the case of the drive unit 10D, the first virtual object's displacement p.sub.1 calculated by the admittance model calculation device 31C is used as the target position command for the position control and driving device 33A.

[0090] It should be understood that the drive unit 10A according to the first embodiment or the drive unit 10B according to the second embodiment can also be provided without the inverse model calculation portion and the transmission device.

[0091] Furthermore, the drive unit 10A according to the first embodiment, the drive unit 10B according to the second embodiment, or the drive unit 10D according to the variant may additionally have the virtual torque limiter function and the sliding deviation restoring function.

[0092] Furthermore, in the case where the driving device 41 is of a type that does not require feedback of the measured position value (for example, a stepping motor), the position measuring instrument 42 can be omitted from the position control and driving device 33A. Similarly, the velocity measuring instrument 44 can be omitted from the velocity control and driving device 33B.

[0093] Furthermore, the first virtual object's dynamics are not limited to those given by equation (1) and may be defined by equations including terms related to coulomb friction (static and dynamic friction), motion range limits, etc. Similarly, the second virtual object's dynamics are not limited to those given by equation (2).

[0094] Furthermore, the virtual interactive force models are not limited to those given by equations (3) and (8).

INDUSTRIAL APPLICABILITY

[0095] The present invention is advantageous particularly in applications where the force of a drive portion with high power and a high reduction ratio is flexibly controlled in a highly safe and adaptable manner.

DESCRIPTION OF THE REFERENCE CHARACTERS

[0096] 10A, 10B, 10C, 10D drive unit [0097] 20A, 20B, 20C, 20D control portion [0098] 21A, 21B, 21D drive portion [0099] 30A, 30C parameter storage device [0100] 31A, 31C admittance model calculation device [0101] 32A, 32B inverse model calculation device [0102] 33A position control and driving device [0103] 33B velocity control and driving device [0104] 34 transmission device [0105] 35 force measuring instrument [0106] 40 position controller [0107] 41 driving device [0108] 42 position measuring instrument [0109] 43 velocity controller [0110] 44 velocity measuring instrument [0111] 50 environment