KR101228279B1 - finger gait planning method of robotic hands and finger gait planning apparatus of robotic hands - Google Patents

Abstract

The present invention proposes a gripping operation apparatus and method of a robot hand which enables the movement of a new gripping point using a static equilibrium condition to a finger of a held robot hand in a simpler operation than in the conventional method. The gripping operation method of the present robot hand generates a candidate set including a state candidate when each of the fingers other than the finger moved among the plurality of fingers of the robot hand holding the object is removed from the object, After calculating the supporting area, a finger to move from among the remaining fingers is selected based on the supporting area and moved in the direction of the target point. Compared to the existing method, the robot hand's finger can be moved to a new gripping point with simpler operation. Finger gait planning using supporting areas and stability margins to select fingers to move and move them to new gripping points is much simpler than conventional methods using complex algorithms.

Description

Finger gait planning method of robotic hands and finger gait planning apparatus of robotic hands

The present invention relates to a gripping operation apparatus and method of the robot hand, and more particularly to a gripping operation apparatus and method of the robot hand to move the finger of the robot hand held the object to a new gripping point.

When applied to an actual service robot, the robot hand may need not only simple gripping but also sophisticated manipulation. In particular, if the operating range of the finger of the robot hand is limited or the appropriate force cannot be applied to the object due to the gripping form, it is necessary to continue to work by changing to another gripping form. This behavior is called "regrasping" or "finger gait". The operation that causes such an operation to be performed is called "finger gait planning".

The gripping problem of the robot hand is a very old problem, but it is still difficult to grasp various objects that can be encountered in real life. The basic problem in gripping is the selection of the point to be gripped, and it should be possible to evaluate the stability for a particular gripping state.

Conventionally, a method of calculating a stable gripping using a force closure and an equilibrium relationship, and a method of calculating a suitable gripping area by sampling an interface of an object have been used for the problem of gripping a polyhedron with four fingers.

The force closure or equilibrium relation is a condition of whether gripping is possible given the gripping state (position of gripping points, frictional conditions). In other words, if a force closure or an equilibrium relationship is established, it can be regarded as being possible to hold. The sampling method is a method of finding out what holding state is possible for an object to be gripped based on this holding condition equation, which takes a long time because it examines all the sampled combinations.

As such, the conventional methods have a problem that the algorithm is very complicated and takes a long time to obtain the final result.

The present invention has been proposed to solve the above-mentioned conventional problems, and the robot which allows the movement of the finger of the held robot hand to a new gripping point using simpler equilibrium than in the conventional method. It is an object of the present invention to provide an apparatus and method for gripping a hand.

In order to achieve the above object, in the gripping operation apparatus for a robot hand according to a preferred embodiment of the present invention, each of the fingers other than the finger moved among the plurality of fingers of the robot hand holding the object is separated from the object. A controller configured to generate a candidate set including a state candidate of the second position, calculate a supporting region for the state candidate, and select a finger to move among the remaining fingers based on the supporting region; And a robot hand driver for moving the selected finger in the direction of the target point based on the drive signal from the controller.

The supporting area is the area calculated by projecting on the two-dimensional plane so that the area of the center of gravity of the object can maintain a static equilibrium.

When the controller selects a finger to move, the controller selects a state candidate with the center of gravity of the object in the supporting area, and selects the released finger that made the selected state candidate as the finger to move.

When a plurality of state candidates are selected, the controller selects a state candidate having the smallest stability margin among the plurality of state candidates.

The stability margin is the horizontal distance from the center of gravity of the object to the nearest boundary of the supporting area in the direction of the manipulation movement.

The robot hand drive unit moves the selected finger to the maximum within the movable range of the finger.

The robot hand holding operation method according to a preferred embodiment of the present invention comprises a candidate set including a state candidate when each of the fingers other than the finger moved among the plurality of fingers of the robot hand holding the object is removed from the object. Generating a; Calculating a supporting area for the status candidate; Selecting a finger to move among the remaining fingers based on the supporting area; And moving the selected finger in the direction of the target point.

The supporting area is the area calculated by projecting on the two-dimensional plane so that the area of the center of gravity of the object can maintain a static equilibrium.

Selecting a finger to move comprises: selecting a state candidate having a center of gravity of the object in a supporting area; And selecting, with the finger to move, the released finger that has made the selected state candidate.

The state candidate selection step selects a state candidate having the smallest stability margin among the plurality of state candidates when a plurality of state candidates are selected.

The stability margin is the horizontal distance from the center of gravity of the object to the nearest boundary of the supporting area in the direction of the manipulation movement.

The finger movement step moves as much as possible within the range of motion of the finger.

According to the present invention of such a configuration, it is possible to move the finger of the robot hand to a new gripping point by a simpler operation compared to the existing method.

Finger gait planning using supporting areas and stability margins to select fingers to move and move them to new gripping points is much simpler than conventional methods using complex algorithms.

1 is a block diagram showing the configuration of a gripping operation apparatus for a robot hand to which the present invention is employed.
2 is an example of the robot hand shown in FIG. 1.
FIG. 3 is a diagram for explaining a supporting region employed in describing an embodiment of the present invention.
4 is a comparative example of a supporting region calculated by a random sampling method and a linear programming method.
5 is a view for explaining a stability margin (stability margin) employed in the description of the embodiment of the present invention.
6 is a flowchart for explaining a gripping operation method of the robot hand according to an embodiment of the present invention.
7 is a view for explaining the movement of the robot hand finger to the final target gripping point in the initial gripping state.

Hereinafter, a gripping operation apparatus and method of a robot hand according to an embodiment of the present invention will be described with reference to the accompanying drawings. Prior to the detailed description of the present invention, terms and words used in the present specification and claims should not be construed as limited to ordinary or dictionary terms. Therefore, the embodiments described in this specification and the configurations shown in the drawings are merely the most preferred embodiments of the present invention and do not represent all the technical ideas of the present invention. Therefore, It is to be understood that equivalents and modifications are possible.

1 is a block diagram showing the configuration of a gripping operation apparatus for a robot hand to which the present invention is employed, and FIG. 2 is an example of the robot hand shown in FIG.

The gripping operation apparatus of the robot hand shown includes a controller 10, a robot hand driver 20, and a robot hand 30.

The controller 10 removes the remaining fingers from the object 40 based on the remaining fingers other than the finger moved among the plurality of fingers 30a, 30b, 30c, and 30d of the robot hand 30 holding the object 40. Create a candidate set including the state candidates in the case of missing. The controller 10 calculates a supporting region for the state candidate. The controller 10 selects a finger to move among the remaining fingers based on the supporting area. When selecting a finger to move, the controller 10 selects a state candidate having the center of gravity of the object 40 in the supporting area, and selects a finger to move in consideration of the stability margin for the selected state candidate. The stability margin depends on the size and location of the supporting area. Here, the supporting area is an area calculated by projecting on a two-dimensional plane so that the area of the center of gravity of the object 40 can maintain a static equilibrium. For a detailed description of the supporting region, refer to the description of FIG. 3 to be described later.

The controller 10 includes a finger gate calculator 12. The finger gate calculator 12 is responsible for a calculation to change the current gripping form of the finger of the robot hand 30. That is, the finger gate operator 12 performs the above-described operation for generating a candidate set, supporting region calculation, and selecting a finger to move.

The robot hand driver 20 moves the selected finger in the direction of the target point based on the drive signal from the controller 10. The robot hand drive unit 20 moves the selected finger to the maximum in the direction of the target point within the movable range of the finger.

FIG. 3 is a diagram for explaining a supporting region employed in describing an embodiment of the present invention.

When the robot hand 30 is assumed to hold an object (eg, 40), the force and moment act on the object 40 by fingers 30a, 30b, 30c, and 30d.

In addition, a force and a moment are generated due to the weight of the object 40 to be gripped.

Therefore, in order to maintain the gripped state of the robot hand 30, the sum of all acting forces and moments must be zero, and friction conditions must be satisfied because the friction between the robot hand 30 and the object 40 is considered. do.

The supporting region S is a projection of the set of centers of gravity c of the objects 40 satisfying all of these conditions in a plane perpendicular to the direction of gravity. That is, if the center of gravity c of the object 40 is present in the supporting area (S) it can be seen that the grip is stable.

This will be described again with the formula.

,

,

로 나타내고, 각 접촉지점에서의 물체 표면에 대한 노멀 벡터를

라고 표현하도록 한다. 또한

는 마찰계수를 나타낸다.Assume that the robot hand is holding the object as shown in FIG. In this case, where the finger is

,

,

Where the normal vector for the object surface at each

To be expressed. Also

Denotes the coefficient of friction.

는 마찰 콘(friction cone) 조건을 만족시켜야 한다.

,

,

는 절대 좌표계를 나타낸다. 중력은

방향으로 작용한다. 물체의 무게중심의 위치는 c로 나타내도록 한다. Force acting on each gripping point

Must satisfy the friction cone condition.

,

,

Represents an absolute coordinate system. Gravity is

Act in the direction. The position of the object's center of gravity is indicated by c.

In order to maintain the static grip state, the robot hand must satisfy the static equilibrium conditions as shown in Equations 1 to 3 below.

[Formula 1]

[Formula 2]

[Equation 3]

In particular, the friction cone condition can be expressed as a normal vector at the gripping point and a force vector acting on the gripping point, and can be expressed as in Equation 4 below.

[Formula 4]

Equations 1 and 2 denote force and moment equilibrium, and Equation 3 denotes constraint conditions of force vectors acting on each gripping point. We can find sets of force vectors and centers of gravity that satisfy these conditions. This means that the area of the center of gravity calculated in this way is an area capable of maintaining a static grip state.

On the other hand, Equation 4 expresses Equation 3 in more detail. Friction cones can be represented as cones around grip points when mathematically modeling friction, which is represented by Equation 4. This means that the force vector acting on the gripping point must be in the cone to maintain friction. This is also associated with the method of calculating the supporting area.

In the embodiment of the present invention, in calculating the supporting region, the region is calculated by projecting on a two-dimensional plane without considering the height of the center of gravity in the three-dimensional space, that is, the position in the e3 direction. Here, the calculation of the supporting area uses linear programming. In order to calculate the supporting area, linear programming is applied, and in order to apply the linear programming, Equation 4 needs to be linearized. When Equation 4 is linearized, it is expressed as a matrix. In other words, when the cone cone friction cone (linear friction cone) is linearized, it is expressed in the form of an approximation in the shape of a polygonal horn. 4 shows a comparative example of the supporting region calculated by the random sampling method and the linear programming method. 4 (a) shows the supporting region calculated by a conventional random sample. 4 (b) shows the supporting area calculated by the linear programming method in the embodiment of the present invention. As illustrated in FIG. 4, the random sampling method can find dense data points that meet the equilibrium conditions but still lacks filling when compared to linear programming. This means that the prior art method is superior to the random sampling method.

Typically, a force closure condition is used to assess the gripping state. In other words, if the grip can be maintained at any force and moment, the force closure is satisfied, and in this case, it can be regarded as a stable grip state.

Similarly, a supporting region of an embodiment of the present invention can be used to evaluate a phage. When the robot hand grips the object, it can be regarded as stable if the center of gravity c of the object is in the supporting region (S). That is, in the case where the robot hand grips the object in FIG. 3, in order to be gripped, the sum of all forces must be 0 and can be expressed by Equations 1 to 3 below. Equation 3 expresses friction with an object as a friction cone condition. By calculating the sets of the centers of gravity c of the objects satisfying the equations (1) to (3), the area S can be obtained, and the obtained area S is called a supporting area. In other words, applying linear programming with conditional equations (force equilibrium, frictional conditions) and cost functions (any straight line, with a slope of 0 to 360 degrees), one point of the support area boundary can be obtained. By applying this, the calculation is performed while changing the slope of the cost function to obtain an approximation of the entire area.

5 is a view for explaining the stability margin (stability margin) employed in the description of the embodiment of the present invention, an example of the stability margin for three contact grasp when moving the finger of the robot hand to the right; Shows.

The supporting region S means a position of a specific center of gravity that can maintain a static equilibrium with respect to a grip type of a specific finger. Using this concept, the concept of stability margin can be applied and quantitative comparisons can be made for each type of phage. In order to assess the quality of the swiveling with multiple fingers, a stability margin is applied. Of course, the embodiment of the present invention can perform the gripping operation of the robot hand by a simpler operation than the conventional method without applying a stability margin. However, if the stability margin is applied to find a more stable holding state, it is preferable to consider the stability margin as possible.

The stability margin can be defined in many forms, but in embodiments of the present invention the stability margin is defined by the supporting area (in the center of the projected object (2 of FIG. 5) (ie center of gravity; 6 in FIGS. It is preferable to define the horizontal distance L to the nearest boundary surface of S). Here, the stability margin may be said to be dependent on the size and position of the supporting region S, which is a function of the contact force and the contact position. Thus, the stability of the phage according to the phage form can be checked using the stability margin quantitatively obtained.

Next, a gripping operation method of the robot hand according to an embodiment of the present invention will be described with reference to the flowchart of FIG. 6. 7 is a view for explaining the movement of the robot hand finger to the final target gripping point in the initial gripping state. In the following description, it is assumed that the number of the fingers of the robot hand 30 is four.

First, the controller 10 determines whether the fingers 30a, 30b, 30c, and 30d of the robot hand 30 have moved to reach a target point. When the fingers 30a, 30b, 30c, and 30d of the robot hand 30 reach the target point as shown in Fig. 7 (see Fig. 63), they are terminated (Yes in S10).

Hereinafter, as shown in Fig. 50 of Fig. 7 (indicative of the first gripping state), Fig. 63 (final target point) is shown while the object is held by the fingers 30a, 30b, 30c, and 30d of the robot hand 30. The process of transitioning to the holding state, such as the holding state). That is, a process of sequentially moving the fingers 30a, 30b, 30c, and 30d of the robot hand 30 one by one from left to right will be described.

At the starting point, the controller 10 generates a candidate set including a state candidate indicating a case in which one finger is removed from the object among the fingers 30a, 30b, 30c, and 30d of the robot hand 30 holding the object. do. Here, since there are four fingers that can be removed from the starting point, the state candidates generated at the starting point are shown in Fig. 51 to Fig. 54. That is, at the starting point, the state candidate when the finger 30a is removed (Fig. 51), the status candidate when the finger 30b is removed (Fig. 52), and the state candidate when the finger 30c is removed. (Fig. 53) and a candidate set including a state candidate (Fig. 54) when the finger 30d is removed (S12). Such a set of state candidates is called a candidate set, and the candidate set includes at least one state candidate.

Subsequently, the controller 10 calculates the supporting area for the situation of each state candidate (Fig. 51 to Fig. 54) (S14). Here, the calculation of the supporting area is replaced with the contents of FIGS. 3 and 4 described above.

Thereafter, the controller 10 selects and moves a finger to move based on the calculated supporting area (S16). That is, the controller 10 selects a state candidate having the center of gravity of the object in the calculated supporting area. This means that the controller 10 selects a state candidate that can be gripped based on the calculated supporting area. In the case of the state candidates (Fig. 53 and Fig. 54), the controller 10 is dropped because the supporting area does not exist. That is, the absence of a supporting area for a state candidate (Fig. 53, Fig. 54) means that there is no supporting area that satisfies a given gripping state and cannot be grasped. This is the same as if the center of gravity of the object is outside the supporting area.

As a result, the controller 10 selects a state candidate (Fig. 51, Fig. 52). Then, the controller 10 selects the state candidate (Fig. 51) which may move a finger to the right in consideration of the stability margin for the selected state candidate (Fig. 51, Fig. 52). Here, the method of considering the stability margin is to select the smallest value. As can be seen in FIG. 5, since the manipulation movement direction is taken into consideration, selecting a smaller value has greater degrees of freedom when considering the movement of the next finger. Thereby, a state candidate (Figure 51) having a smaller stability margin is selected from among the status candidate (Figure 51) and the status candidate (Figure 52). Subsequently, the controller 10 sends a command to the robot hand drive unit 20 to move the finger 30b (that is, the released finger) that has made the state candidate (Fig. 51). Accordingly, the robot hand driver 20 moves the finger 30b. At this time, the finger 30b moves to the maximum within the movable range in moving in the target point direction. This will look like Figure 55.

Thereafter, the controller 10 determines whether all the fingers have moved once (S18). As a result of determination, if all the fingers move once and reach the target point, the process ends.

On the contrary, if all the fingers have not moved once, the controller 10 performs the processes of S12 to S16 for the remaining fingers. This process will be described later.

Since the finger 30b is moved for the first time in the above description, the controller 10 subsequently moves the finger 30b moved among the fingers 30a, 30b, 30c, and 30d of the robot hand 30 holding the object. Except that a candidate set including a state candidate representing a case where one finger is removed from the object is generated. Here, in Fig. 55, since there are three fingers that can be removed, except for the finger 30b that has already moved, the state candidates generated are Fig. 56 to Fig. 58. That is, after one finger 30b is moved, the state candidate (figure 56) when the finger 30c is removed, the state candidate (figure 57) when the finger 30a is removed, and the finger ( A candidate set including a state candidate (Fig. 58) in the case of detaching 30d) is generated. Subsequently, the controller 10 calculates the supporting area for the situation of each state candidate (Fig. 56 to Fig. 58), and then selects and moves a finger to move based on the calculated supporting area. In other words, in the case of the state candidates (Fig. 57 and Fig. 58), the controller 10 drops out because the supporting area does not exist. That is, the absence of a supporting area for a state candidate (Fig. 57, Fig. 58) means that there is no supporting area that satisfies a given gripping state and cannot be grasped. This is the same as if the center of gravity of the object is outside the supporting area. As a result, the controller 10 selects only a state candidate (Fig. 56). Since only one status candidate has been selected here, the stability margin does not need to be considered. Then, the controller 10 sends a command to the robot hand driver 20 to move the finger 30c. Accordingly, the robot hand drive unit 20 moves the finger 30c. At this time, the finger 30c moves to the maximum within the movable range in moving in the target point direction. This will look like Figure 59.

Subsequently, the controller 10 indicates a case in which one finger is removed from an object except for the fingers 30b and 30c moved among the fingers 30a, 30b, 30c, and 30d of the robot hand 30 holding the object. Create a candidate set that includes a status candidate. Here, in Fig. 59, there are two detachable fingers except the fingers 30b and 30c which have already been moved, so the state candidates generated are Fig. 60 and Fig. 61. In other words, after the two fingers 30b and 30c are moved, the state candidate (Fig. 60) when the finger 30d is removed and the state candidate (Fig. 61) when the finger 30a is removed. A candidate set for inclusion is generated. Subsequently, the controller 10 calculates the support area for the situation of each state candidate (Fig. 60, Fig. 61), and then selects and moves a finger to move based on the calculated support area. In other words, in the case of the state candidate (Fig. 61), the controller 10 drops out because there is no supporting area. That is, the absence of a supporting region for the state candidate (Figure 61) means that there is no supporting region that satisfies the given holding state and cannot be grasped. This is the same as if the center of gravity of the object is outside the supporting area. As a result, the controller 10 selects a state candidate (Fig. 60). Since only one status candidate has been selected here, the stability margin does not need to be considered. Then, the controller 10 sends a command to the robot hand drive unit 20 to move the finger 30d. Thus, the robot hand drive unit 20 moves the finger 30d. At this time, the finger 30d moves to the maximum within the movable range in the movement to the target point direction. This will look like Figure 62.

After that, the controller 10 controls the robot hand driver 20 to move the finger 30a that has not moved yet to the right. That is, the controller 10 determines whether the finger 30a can be moved and moves the finger 30a to the right if it can be moved. This will look like Figure 63. Of course, it is also possible to perform candidate set generation, supporting region calculation, stability margin consideration, and the like for the finger 30a that is moved last, but this is not necessary because it is a waste of time.

In this way, when all the fingers move to reach the target point once, it ends.

In the above-described embodiment of the present invention, the case in which the fingers 30a to 30d of the robot hand 30 are horizontally moved from left to right has been described. .

The present invention described above can also be embodied as computer readable code on a computer readable recording medium. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like, which are also implemented in the form of a carrier wave (for example, transmission over the Internet). It also includes. The computer readable recording medium may also be distributed over a networked computer system so that computer readable code can be stored and executed in a distributed manner.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. You must see.

10 controller 12 finger gate calculator
20: robot hand drive unit 30: robot hand
30a to 30d: finger 40: object

Claims (12)

Generating a candidate set including a state candidate when each of the fingers other than the finger moved among the plurality of fingers of the robot hand holding the object is removed from the object;
Calculating a supporting area for the status candidate;
Selecting a finger to move among the remaining fingers based on the supporting area; And
And moving the selected finger in a direction of a target point. Claim 1
And the supporting area is an area calculated by projecting on a two-dimensional plane so that the area of the center of gravity of the object can maintain a static equilibrium. The method according to claim 1,
The finger selection step to move,
Selecting a state candidate with a center of gravity of the object within the supporting area; And
And selecting the removed finger that has made the selected state candidate as the moving finger. The method according to claim 3,
And selecting a state candidate having a smallest stability margin among the plurality of state candidates when a plurality of state candidates are selected. The method of claim 4,
And the stability margin is a horizontal distance from the center of gravity of the object to the closest boundary surface of the supporting area in the direction of the manipulation movement. The method according to claim 1,
The finger movement step is a robot hand holding operation method characterized in that to move to the maximum within the movable range of the finger. Generating a candidate set including a state candidate when each of the fingers except the finger moved out of the plurality of fingers of the robot hand holding the object is removed from the object, calculating a supporting area for the state candidate, A controller for selecting a finger to move among the remaining fingers based on the supporting area; And
And a robot hand driver for moving the selected finger in a direction of a target point based on a drive signal from the controller. Claim 7
And the supporting area is an area calculated by projecting on a two-dimensional plane so that the area of the center of gravity of the object can maintain a static equilibrium. The method of claim 7,
The controller selects a state candidate having the center of gravity of the object in the supporting area when selecting the finger to move, and selects the released finger that has made the selected state candidate as the moving finger. Holding device. The method according to claim 9,
And when the plurality of state candidates are selected, the controller selects a state candidate having the smallest stability margin among the plurality of state candidates. The method of claim 10,
And the stability margin is a horizontal distance from the center of gravity of the object to the closest boundary surface of the supporting area in the direction of the manipulation movement. The method of claim 7,
And the robot hand driving unit moves the selected finger to the maximum within the movable range of the finger.

Images